I now have the microSD and EEPROM working on Blade #2. · Only the Flash to go, but that can wait for the moment... more ZiCog for me
I used the code from OBC's website http://www.warrantyvoid.us/tiki-index.php?page=Objects·SD_trainer.zip with changes to the microSD I/O pins, and 115200 baud rate to the PC. The routine TestMountSD did not appear to work (hung) but the write to and read from the microSD (1GB) worked and I could read the microSD file on a PC. The code required an EEPROM to be present, which I hadn't fitted until now. I have posted this modified code for those interested.
For those who may not be following, I have the SRAM working with the ZiCog emulator for both 64KB code space and a 960KB ram disk (see ZiCog thread).
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔ Links to other interesting threads:
The TribladeProp has arrived in Indiana in the midwest United States. It took only 7days!! I will start assembly in ·in the morning. Almost can't wait but I need to take it slow and reread all the posts and info you and others have
been sharing. Keep up the "chatter"!!!
This is a fantastic project. Every day almost there seems to be some cool progress.
I've got the board but I've been holding off buying parts till all the testing is done. It seems like this is going to work though. The thread seems to have split into two threads and I've slightly lost track of the parts and all the links. Eg some links go to the website http://bluemagic.biz/clusodocs/tribladeprop_a1d.jpg but then there are links that go back to these threads and it is all a bit confusing. I'm sure I saw a parts list somewhere. But then it seems some parts are not necessary? eg on that website above is a photo and two flash chips have not been installed and some ram chips are installed and some are not. Does the terminal prop need a ram chip? And does the middle prop need two or can it run with just one? Sorry about getting confused - I'm sure all the answers are in one of the theads. I did find this on the website "If there is sufficient demand, I may stock a kit of components.". Can I put my vote in for that please!
James: I am sorry for any confusion. I have been heavily into getting ZiCog running. Also it's been coming so fast and furios I hav not had enough time to combine all the info in one place.
Above (page 3 of this thread) are two posts answering some questions and assembly suggestions/options. I think this is duplicated on my website www.bluemagic.biz/cluso.htm There are also some extra photos and if you click on them you can also zoom in to get better detail. However, the thread always has the latest info (discussion) but I also post a copy of materials such as Schematics, Parts Lists, Errata, etc on my website.
So far, I haven't fitted any Flash parts and only two days ago fitted (pushed in, not soldered) 1 eeprom which was required for the SD (microSD) driver of Rokiki's to work. Everything else has been fitted to my TriBladeProp except the connectors at the left hand edge because I don't know what I am going to add to them and therefore what pins to fit (multiple options here).
Blade #1 (the Terminal) does not require the SRAM or 74HC573. It is for later as I see getting hires graphics without SRAM will be difficult. If the SRAM is used, there are not enough pins for both the VGA/TV, keyboard and mouse so headers are provided to jump them by cable to Blade #3. I have fitted all the pin headers so that I can use shunts to swap between TV and VGA. I have tested the VGA, TV, and SRAM but not the keyboard and mouse.
Blade #2 (the Brain) does not require both SRAMs but I am recommending the 74HC573 be fitted as it also has outputs to enable the SRAMs, microSD, Flash and reset the other two Props (untested). If you fit the 74HC573 then you must do the modification to the SRAM as detailed on my website (RevA1). If you fit the microSD and/or Flash you also need the 74HC573 and the mod. For the ZiCog I am using the SRAM for the 64KB ram/eprom emulation and the remaining 960KB as a CPM ram disk. The microSd is another group (7-8 or more) of 8MB disk drives. I am trying to host the drives under FAT but don't yet know if this will be fast enough.
Blade #3 (the I/O) does nothing yet, except it works. Ultimately it will talk to the PC and other I/O expansion.
The ZiCog thread contains code and details relevant for getting ZiCog working on the TriBladeProp SBC.
As for kits, at this time there just isn't the demand. Freight is an extra·50%-100% to my·parts costs.
I am often on msn and recently on OBC's Propeller IRC. email me and I'll give you my msn address.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔ Links to other interesting threads:
Wow cluso99, you are going ahead in leaps and bounds. The N8VEM came to life 8 months ago but only now are we even starting to talk about FAT on a SD card.
Thanks for the description. It helps narrow down which bits are the essential bits. Blade 1 and 3 can probably wait for the moment - we know the terminal will work as I've got Vince's Pocketerm running fine (I think that was OBC's code).
So blade 2 is where the action is. Ram drive and SD drive. And all the intricacies of getting the custom bios working etc. I'll look into getting the bits for blade 2, as I need some of those bits anyway for the N8VEM. Are you at the point where it can boot to an A> prompt and then get the full 64k of ram available?
James said...
Are you at the point where it can boot to an A> prompt and then get the full 64k of ram available?
Was there about 2 weeks ago. I can do a DIR on the 960KB Ram Disk. And yes, other than verify the VGA and TV work I have left the Terminal alone.
The Ram Disk·took too long to load from the PC so I started on the microSD a couple of days ago. I now have blank images in the FAT file of 8MB drives· (only using the first 128 bytes in a·sector so the files are really 32MB) and a 960K CPM image (actually it is 1MB and contains the Z80 64K plus the 960KB CPM image together in 512 byte sectors so I can load quick as it will be preloaded during bootup). I posted them in .rar format on ZiCog last night. I am hoping the sectors within the files·are contiguous so I can directly read/write into them. That way·I can keep FAT16 meaning we can also boot prop code or whatever, by having PropDos sitting in a layer above.
FYI: I am using the onboard PropPlug emulator to load the code - saves multiple parts. The FT232RL (smt) wasn't too hard to solder with some solder paste and solder wick, even with my aging eyesight. The microSd smt socket was a breeze.
I was going to ask you a couple of questions about CPM last week regarding file sizes and 128/137 byte sectors, etc. I found most of what I needed and will emulate the extra 3+6 bytes for 137 byte sectors if I need to. I wrote a VB6 program to build the 64KB file from the cpm files. I also have an 8MB CPM file which just has the sector number as the first 6? characters followed by "*". A good test program to see that I am reading what I expect.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔ Links to other interesting threads:
I'm still deep inside the coding for a wireless mesh, with auto discovery of new nodes and auto update of software using xmodem. Also working on ethernet to serial devices, so that a mesh can interface directly to the internet without needing a PC. We had a test link going recently from Canada to Australia so the concept is sound. Soon we should have a BBS system running. The aim ultimately is to be able to log into a home automation network from anywhere on the web, turn on things, sense things etc, but without needing to leave a PC on. Ultimately, nodes should be low power (?battery or solar even) but smart enough to think for themselves, recompile new updates of software etc.
One of the features of wireless is that everyone is sharing the same bandwidth. So you can't type "DIR" into a wireless mesh, because 10 boards will reply at once, and then each one spits out its listing and the whole thing goes epileptic. There are software solutions, essentially involving disabling echo within CP/M and random delays before responding. So sometimes you need to tweak CP/M itself. On a N8VEM this involves making a new eprom, so the prop solution could end up being quicker too. I've also been working on a hardware model of the wireless system, essentially a bunch of RS232 connections all connected together at a single point (but with the correct logic gates as you can't physically just connect them all together). I've also discovered you need more than one RS232 port - in fact I need one for the PockeTerm, one to a PC, one to the wireless transceiver, one to an analog I/O and relay board, and one to a watchdog reset circuit (using a picaxe chip). So that is a 1:5 RS232 router that accepts control signals at 1200 baud, but which can switch at any baud rate. I've just sent that off to be made tonight.
Getting back to what you are doing, ok the N8VEM has CP/M sitting in 16K of eprom and loads that into the top of 64k of ram, and then runs it. That version of CP/M has enough smarts to work out all the drives. I have a vb.net program that can build these eprom images with many different options (12 at present) using simple check boxes. But you still have to then download actual programs at 38400 baud to the drives.
Your model sounds better. Just to clarify, are you just taking a 1meg .bin image off the altair SIMH or whatever, and just emulating that as a total package? IE, CPM plus a whole pile of files. If so, that makes a lot of sense, because it would be much faster to package up and download.
If that is what is going on, I have some vb.net software that can link to the altair simh. It can link in two ways - either by saving a batch file that does a whole lot of things on startup and then starting the simh and running that batch file. Eg - load program x, compile program x, save program x, write program x back to the PC directory. So a compile of C or BASIC becomes a very fast single step operation. The other way to interface to the simh is to run the simh and attach it to a telnet terminal. And then from vb.net, hook into that telnet terminal with a winsock .dll This is done locally, but it probably could be done to another PC on the internet too. This is useful for moving groups of files from a PC directory into a 1 meg (or 8meg) .bin simh image. So you can copy all the .COM files with one keypress. Or all the Wordstar files. Or Basic files. It makes the whole process of creating disk images so much quicker when it is just a checkbox process.
The altair simh only lets you have 8 drives A to H. CP/M can only go to 8meg. But that means a SD card could hold drives A-Z with 8 meg in each =26*8. Ok, maybe a bit less because drive A would be the ram drive.
I'd better keep a close eye on this thread - all very exciting stuff!
Very interesting... I have had a quick look a few times at the N8VEM forum. There are two things I would have liked on my board - sound and ethernet. Unfortunately, there is no room and they can be done externally. I saw a part recently that had the whole ethernet in the RJ45 connector - IC as well, serial or SPI out - do you think I can find it again! Like an ENJxxx in a socket.
Your RS232 board doesn't surprise me. Did you think about or are you using a prop for this, or just real gates?
My VB6 program is no where as sophisticated as yours. But it allows me to build the 64KB RAM/EPROM from files and I just added the CPM disk image (converting on the fly stripping the 137 byte sectors into 128 bytes) as well and combined it into a single 1MB file using the whole 512 byte sectors since it is being preloaded by spin/pasm into the SRAM. So it is as you said above. Heater just made me a 960KB drive with the files on it. As I said above, I created a blank file and a file with sectors filled with the sector number in each sector, each 8MB (ie. 32MB) and just copied them, renaming them DRIVE__A.CPM thru to H, and then copied to the microSD which uses FAT16. My board displays these files with code I wrote using the Femto SD Trainer code as the base. I also have an SD adapter and one of my laptops has an SD slot so I can read and write directly to the card
I think if can get the drives to be files within the FAT16 system, then it will be easy to copy any CPM files required to the SD via FAT and have a spin section that will read them and copy them onto one of the drive files. Then there will be no slow xmodem to worry about. Well, other than your wireless exploits. I am thinking of putting the z80 block move instruction into the driver and heaters ZiCog can just call the other cog directly to do this.
By renaming FAT files, it would be easy to swap hard drives, but I guess there wont be much requiring 8 x 8MB drives anyway.
While writing this I have been running a walking bit code (spin/pasm) through my SRAM to test its integrity - looks good.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔ Links to other interesting threads:
I also thought I would share my reasoning for the choice of microSD... see the attached photo. The top left is a USB connector with an 8GB microSD inserted (you can see the "8"). The top right is an adapter to take a microSD and convert to the SD format (you can see the slight recess at the bottom where·the microSD card to plugs into and is completely enclosed). The bottom is obviously the microSD card.
There is virtually no difference in price. For the time being, only use 2GB or less. The only issue is that SPI which is what we use to access the card is not mandatory for microSD. There have been reports that some microSD's do not support SPI although I have not found any.
The choice of miniUSB is·because it is used by most cameras and phones, so users are very likely to have a cable(s) already.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔ Links to other interesting threads:
Dr_A. "The altair simh only lets you have 8 drives A to H."
This is not true[noparse]:)[/noparse] The SIMH AltairZ80 simulator suports 8 floppy drives of 1MB (A to H) and 4 hard drives of 8MB (I to L). The number of each is configurable in an include file when you rebuild the BIOS.
Adding hard drive simulation to PropAltair/Zicog has been on my todo list for a long while.
Now, SIMH has a config file which is read at start up which specifies which disk image file on the host file system (in our case FAT16) is used as which CP/M drive.
It has lines like:
attach dsk1 cpm2.dsk
It would be great to use a similar system on TriBlade, no copying of disk images around required.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
For me, the past is not over yet.
I have just posted the latest BOM, suppliers and part numbers and costs (in tab seperated format suitable for import into Excel) to the first post of this thread.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔ Links to other interesting threads:
Here is my latest ZiCog code on Blade #2 of the TriBladeProp. I thought it more relevant to post here than the ZiCog thread.
I have attached the zipped source files. The ZiCog Z80 code is not heaters latest (I am still using v005). There is an issue I need to fix by updating the TBP (TriBladeProp) driver. Currently I am resetting the cog to overcome this until I get time to fix (add a simple routine). All software to compile is in the demo zip file including the homespun compiler and the batch file to compile. There are 3 or 4 warnings about missing '#' in the jmp/ret instructions but this is in fact correct.
I have also attached the zipped disk files for the microSD card. It is a large file. Expand it first on your pc and then copy DRIVE__C.DSK to files·A through H. This is a blank CPM2.2 disk which can be used for 1MB drives A...H. Expand DRIVE_1MB.RAM. This file is used to load the CPM boot information into the RAM (sram) by the spin program. DRVCPM_2.DSK is the disk file containing all the CPM2 files, formatted as a 1MB drive. It is best to leave this untouched and just copy it over the DRIVE__A.DSK you have made above. This way anything you do will overwrite/add to the A CPM drive and you will still have an original. And CPM will boot to the A drive.
Once these files are ready, format your microSD card with FAT16 and preferably 32KB clusters. Then copy these files one by one. They will then be contiguous (using successive clusters) because we access them using physical addresses once the starting point is found under FAT16.
SURVEY is the CPM program that shows the disk drives and memory map. It will not display drives until you have done a DIR command.
As I haven't included the CPM3 disk file, edit the spin demo file to change the H drive to use DRIVE__H.DSK and recompile.
This version only uses Blade #2 and the PropPlug connected to your PC and PropTerminal.
I edit and save using PropTool (dont forget to save), then go to a dos box and run z.bat from the source folder (set the path etc - ensure there are no blanks in the path/file names). Check the compile produced an output binary file with no errors (other than the warnings mentioned above). Go back to PropTool and load the binary file to the prop ram. The switch to PropTerminal. You will have about 3 seconds to do this before the software starts on the prop. If you miss this, just type "y" to load the ram and continue.
If you have any questions just ask. I am a bit tied up for a couple of days (in case I am slow to respond).
Enjoy
PS. Later I will publish the entire code you require to boot from an SD card without the PC.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔ Links to other interesting threads:
With my parts finally on order for my TriBlade, I can't help but think there are more possibilities here than just running Z80 emulation. [noparse]:)[/noparse]
Have you written a generic object for the SRAM configuration yet?
Aghh! OBC What are you saying? Don't divert Cluso's attention. This is going so well.
A megabyte of RAM for code and data, 2G file store, built in display, multiple I/Os, a pile of unused 32 bit CPUs. No I can't think of any other possible use for this[noparse]:)[/noparse]
I think this is driven by the fact that if you want to use 64K of RAM (or more) for running editors, compilers whatever on the Prop this emulation is the only possible way to do it at the moment.
Things could/will change dramatically when we get C compilers that can execute code from external memory or perhaps better for most Prop heads a Spin compiler/interpreter combo that works from external RAM !
Anyway, what about a 6809 emulation[noparse]:)[/noparse]
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
For me, the past is not over yet.
What about the rest of us that don't care about CPM and other emulation? I second OBC need for a general RAM driver. CPM is just a sideline hobby. Boring...
I have published a TriBlade driver for SRAM access·· Pass it a hub address, sram address and a length and wait for it to be done.
The SD card driver is the Femto drivers with a published fix
I have said many times, emulators are NOT my only purpose of the TriBlade. It is just my focus. However, everything I have written is with general purpose use in mind, just like heaters ZiCog is deliberately NOT tied to the TriBlade.
We eagerly await the Prop spin/pasm or c compiler running on the prop.
Meanwhile we can run/compile other code under CPM, like Turbo Pascal, MBasic, Fortran, ADA, etc, etc, etc. And we can run SuperCalc (not tried yet), etc, etc, etc.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔ Links to other interesting threads:
Well, I went to the lab today to solder up one of my boards, and whew! It takes a while. I worked on it for about 6 hours straight, and I'm still not done. A couple of questions:
Does the text on the resistor network go under the small square on the outline?
For the micro SD card holder, are all four pads in back used as electrical connections?
For the pairs of capacitors in and near C16, C26, C33, what values do we use? Also, those are for overclocking with a 6 MHz crystal, right?
Does the switch go in any particular orientation?
Blade two, what are the U24A and U23A for as opposed to the versions without A?
What are the various holes for, notably the LK1, LK2, LK3, ...?
I suppose that the Blade 2 SRAM pins most definitely cannot touch the PCB (the pins you mention in your mod)?
Edit: another question: I don't have 240 ohm resistors or any 1K1 ohm resistors. Is it okay to substitute with something close (perhaps 250 and 1K respectively)?
1. The resistor network is soldered on the underside of the pcb and the usual "dot" indicating the common connection on the part goes to the square box end.
2. None of the pads at the back of the microSD connector are used in my design. They just provide stability for the connector in this application. Normally these back pads could be used for write protect and card inserted functions.
3. C16/17, C26/27 & C33/34 are for possible use in overclocking. No values are specified although I think 0.1uF and a 1uF (tantalum - so careful of polarity "+" closest to the vga connector) might be necessary from Sapheia's information. If required, they mount on the underside of the pcb. A 6MHz crystal hopefully should not require these.
4. There is no orientation for the switch. If the switch is in the left position·the power will be off, right it will be on.
5. U23A and U24A are for 2MB SRAMs which replace the U23 and U24 512KB SRAMs. They are EXPENSIVE and nothing uses them at present.
6. Links:
LK1 & LK2 allow Blade #1 & #2 to connect their P30/31 serial pins together. You cannot use the PropPlug to connect to these Blades when this is active. Beware that P30 connects to P30, and P31 connects to P31, so the terminal function on Blade #1 will need it's serial pins reversed in software,·otherwise there will be a bus conflict which may cause damage to the prop chips. Alternately use a small cable on these headers and cross them over.
LK3, LK4 & LK5 are designed to connect Blade #2 to Blade #3 together via the serial bus mentioned in LK1 & LK2. It can route the PropPlug to Blade #3 as well, including the reset line. See the schematic. Note these shunts can be all horizontal or all vertical for different pins
7. The modification to the SRAMs U23 and U24: These pins must not touch the pcb. If you are experienced, you may alternately drill the connection but·this is not wise. You may bend then cut the pin shorter on each SRAM to ensure it does not touch the pcb, but long enough to solder the wire to. Put the 2 wires under the ICs to hide most of the modification and keep the wires from getting caught. This is the professional way.
8. Just to reiterate... U12, U22, U32 & U25 must be soldered to the underside of the pcb as the pinout is reversed for this. They will not function if soldered to the top of the pcb !!!
9. U26 label should read FLASH, not EEPROM.
10. Q91 is inserted differently to the label. (see the errata A1)
11. For now, I suggest you do not fit U13 and U15 (SRAM and Latch for Blade #1) until such time as you have a requirement for it or someone produces a requirement for it. There are other possible uses for these pins if left unsoldered - such as allowing VGA AND TV concurrently with a cable, or more I/O.
12. I have not fitted (permanently) any of the EEPROMs as it is my intention (when time allows) to have Blades #1 and #3 code loaded by Blade #2 when in standalone (SBC = Single Board Computer) operation. The USB serial connection to the PC will go to Blade #3. All Blades will communicate between them via a common·high speed serial communications on a 1 or 2 wire bus (TBD).
13. I have not fitted the FLASH although I understand Mike Green is doing this.
14. You will see from the schematics, the LEDs are uncommitted, so you can wire them where you like. The pad will light the LED if it is logic high. I used a blue LED as LED2. I cannot recall which of the others were red and green. I only held wires to check the props were functioning.
Good luck and I await to hear you are running... If you have any questions, please ask.··
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔ Links to other interesting threads:
One alternative on SRAM Pins is to cut conections around holes.
U23 have only conections on solder side.
U24 have on both sides ... On Component side it have conection to one line on PCB.
It is that I modified it ...
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔ Nothing is impossible, there are only different degrees of difficulty. For every stupid question there is at least one intelligent answer. Don't guess - ask instead. If you don't ask you won't know. If your gonna construct something, make it·as simple as·possible yet as versatile as posible.
1)I've been socketing all the chips, and I'm wondering about the U23/U24 business. If I get a socket to fit the extra four pins, how easily can I switch between the two versions of SRAM? My guess is that I'll have to do some resoldering...
2) So for the microSD, is it okay if I bridge the four connections on the U23 side?
3) Will the overclocking caps hurt me if I only have a 5Mhz crystal?
4)A question that I edited in above at the same time you were posting: "Edit: another question: I don't have 240 ohm resistors or any 1K1 ohm resistors. Is it okay to substitute with something close (perhaps 250 and 1K respectively)?"
Thanks for telling be about the tantalum capacitor polarity. Of the three on the board (right!?) I've only gotten to C91 and it's in backwards...
Extra capasitors have little efect on 5-6MHz crystal.
It give sxtra stablity on PLL system in Propeler chip most with frequensies 14-15MHZ x PLL8
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔ Nothing is impossible, there are only different degrees of difficulty. For every stupid question there is at least one intelligent answer. Don't guess - ask instead. If you don't ask you won't know. If your gonna construct something, make it·as simple as·possible yet as versatile as posible.
1. Yes you can use a 36 pin socket and you could remove the modified pin (-OE)·from the socket first. Both the 512KB and 2MB SRAMs are pin compatible, so no resoldering wil be required.
2. No problems if these pins are bridged as long as they are not shorting to the power plane underneath. Just meter the connection to verify this. (I presume you are not referring to the holes (vias) on the left side of the microSD connector near U26 - these cannot be shorted.)
3. No they will not cause any issues if installed. They just provide better decoupling and storage reservoir (in laymans terms). You can always solder these later by bending the legs slightly and soldering on the underside. Unless you want to significantly overclock, don't bother with them.
4. I am unsure. It will affect the colors not sure about the actual display.
5. C91 needs to be corrected. C11 (electrolytic) and C12 & C13 all require to be·correctly (polarity)·inserted.
6. Since you are asking about polarity, I presume you have fitted the diode D4 correctly - the small block is the white bar end·on the diode.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔ Links to other interesting threads:
If you have to use an alternative resistor value on the VGA or composite I would err on the higher side to prevent any clipping/"overloading". I have always used 270 Ohms and 1K2 if the right ones were left at work: again!
I have joined in, now. I have a "Blade 2" with eeprom, SD and two 512KB rams. I have made it with the /OE mods with latch.
I originaly tried your ram test (with blue led on P24) which did not work, which i guess is due to the mods???. Next I tried the Test201, with this I get one screen full of number dumps. The numbers start wit 10,11,12 ...get up to 7f turn to 77s for some more rows and thenstart at ff count down to 80 and the 77s come back until i) complete gets a row of its own. Reading the code I think that it should test all the pages and the i) should be a number indicating that all 1024 pages have been vetted
Am I using the wrong test prog or have I got another puzzle waiting, whoopee.
I got one of my TriBladeProps assembled now and can boot CPM from your posts
I assume the files on the disk are the cbios that is being used? I want to look at using 128 byte sectors and packing 4 per SD sector to be more efficient and want to look at CPM 3 long term to get to 512MB hardrives
Have you enabled using console going to Blade 1 with VGA and keyboard ? I am interested in testing that out next.
Toby: Great news. At various times I use sequences such as $77, $33 etc to prove if something has correctly overwritten the data. So it could just be me publishing a debug version of the code (or it could be that the page was not written at all). However, usually my result that it is correct will test everything.
Yoda: I packed 4 x 128 byte sectors into the DRIVE_1M.xxx file. It has the CPM boot info (Z80 ram) in the first 64KB followed by the 960KB CPM2.2 Ram Disk (drive A). However, I decided not to pack the other drives as 4 x 128 bytes and just use the first 128 bytes as the data for each sector. I wasn't sure of the overheads used in the card itself to perform the write to the flash, and I was not concerned (at this point anyway) about available space on my 1GB and 2GB cards. You may notice, that while the files are 32MB (for an 8MB drive), only a 1MB drive is actually used. On the agenda is the 2 or 4 HD drives which can be 8MB each. Changing the BIOS will of course increase this, but later.
I did have the console working to Blade #1, but not satisfactory to me, and not at present. I went back to getting CPM3 running. As soon as I get time (shortly) I will get the bank switching for CPM3 running. Then I can go back to the Terminal in Blade #1. I think Mike is working on this anyway.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔ Links to other interesting threads:
Comments
I used the code from OBC's website http://www.warrantyvoid.us/tiki-index.php?page=Objects·SD_trainer.zip with changes to the microSD I/O pins, and 115200 baud rate to the PC. The routine TestMountSD did not appear to work (hung) but the write to and read from the microSD (1GB) worked and I could read the microSD file on a PC. The code required an EEPROM to be present, which I hadn't fitted until now. I have posted this modified code for those interested.
For those who may not be following, I have the SRAM working with the ZiCog emulator for both 64KB code space and a 960KB ram disk (see ZiCog thread).
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Links to other interesting threads:
· Home of the MultiBladeProps: TriBladeProp, SixBladeProp, website (Multiple propeller pcbs)
· Single Board Computer:·3 Propeller ICs·and a·TriBladeProp board (ZiCog Z80 Emulator)
· Prop Tools under Development or Completed (Index)
· Emulators: Micros eg Altair, and Terminals eg VT100 (Index)
· Search the Propeller forums (via Google)
My cruising website is: ·www.bluemagic.biz·· MultiBladeProp is: www.bluemagic.biz/cluso.htm
My boards arrived today
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Aka: CosmicBob
Post Edited (Bob Lawrence (VE1RLL)) : 4/2/2009 2:36:52 AM GMT
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Links to other interesting threads:
· Home of the MultiBladeProps: TriBladeProp, SixBladeProp, website (Multiple propeller pcbs)
· Single Board Computer:·3 Propeller ICs·and a·TriBladeProp board (ZiCog Z80 Emulator)
· Prop Tools under Development or Completed (Index)
· Emulators: Micros eg Altair, and Terminals eg VT100 (Index)
· Search the Propeller forums (via Google)
My cruising website is: ·www.bluemagic.biz·· MultiBladeProp is: www.bluemagic.biz/cluso.htm
The TribladeProp has arrived in Indiana in the midwest United States. It took only 7days!! I will start assembly in
·in the morning. Almost can't wait but I need to take it slow and reread all the posts and info you and others have
been sharing. Keep up the "chatter"!!!
TrapperBob
I've got the board but I've been holding off buying parts till all the testing is done. It seems like this is going to work though. The thread seems to have split into two threads and I've slightly lost track of the parts and all the links. Eg some links go to the website http://bluemagic.biz/clusodocs/tribladeprop_a1d.jpg but then there are links that go back to these threads and it is all a bit confusing. I'm sure I saw a parts list somewhere. But then it seems some parts are not necessary? eg on that website above is a photo and two flash chips have not been installed and some ram chips are installed and some are not. Does the terminal prop need a ram chip? And does the middle prop need two or can it run with just one? Sorry about getting confused - I'm sure all the answers are in one of the theads. I did find this on the website "If there is sufficient demand, I may stock a kit of components.". Can I put my vote in for that please!
Above (page 3 of this thread) are two posts answering some questions and assembly suggestions/options. I think this is duplicated on my website www.bluemagic.biz/cluso.htm There are also some extra photos and if you click on them you can also zoom in to get better detail. However, the thread always has the latest info (discussion) but I also post a copy of materials such as Schematics, Parts Lists, Errata, etc on my website.
So far, I haven't fitted any Flash parts and only two days ago fitted (pushed in, not soldered) 1 eeprom which was required for the SD (microSD) driver of Rokiki's to work. Everything else has been fitted to my TriBladeProp except the connectors at the left hand edge because I don't know what I am going to add to them and therefore what pins to fit (multiple options here).
Blade #1 (the Terminal) does not require the SRAM or 74HC573. It is for later as I see getting hires graphics without SRAM will be difficult. If the SRAM is used, there are not enough pins for both the VGA/TV, keyboard and mouse so headers are provided to jump them by cable to Blade #3. I have fitted all the pin headers so that I can use shunts to swap between TV and VGA. I have tested the VGA, TV, and SRAM but not the keyboard and mouse.
Blade #2 (the Brain) does not require both SRAMs but I am recommending the 74HC573 be fitted as it also has outputs to enable the SRAMs, microSD, Flash and reset the other two Props (untested). If you fit the 74HC573 then you must do the modification to the SRAM as detailed on my website (RevA1). If you fit the microSD and/or Flash you also need the 74HC573 and the mod. For the ZiCog I am using the SRAM for the 64KB ram/eprom emulation and the remaining 960KB as a CPM ram disk. The microSd is another group (7-8 or more) of 8MB disk drives. I am trying to host the drives under FAT but don't yet know if this will be fast enough.
Blade #3 (the I/O) does nothing yet, except it works. Ultimately it will talk to the PC and other I/O expansion.
The ZiCog thread contains code and details relevant for getting ZiCog working on the TriBladeProp SBC.
As for kits, at this time there just isn't the demand. Freight is an extra·50%-100% to my·parts costs.
I am often on msn and recently on OBC's Propeller IRC. email me and I'll give you my msn address.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Links to other interesting threads:
· Home of the MultiBladeProps: TriBladeProp, SixBladeProp, website (Multiple propeller pcbs)
· Single Board Computer:·3 Propeller ICs·and a·TriBladeProp board (ZiCog Z80 Emulator)
· Prop Tools under Development or Completed (Index)
· Emulators: Micros eg Altair, and Terminals eg VT100 (Index)
· Search the Propeller forums (via Google)
My cruising website is: ·www.bluemagic.biz·· MultiBladeProp is: www.bluemagic.biz/cluso.htm
Post Edited (Cluso99) : 4/2/2009 9:29:18 AM GMT
Thanks for the description. It helps narrow down which bits are the essential bits. Blade 1 and 3 can probably wait for the moment - we know the terminal will work as I've got Vince's Pocketerm running fine (I think that was OBC's code).
So blade 2 is where the action is. Ram drive and SD drive. And all the intricacies of getting the custom bios working etc. I'll look into getting the bits for blade 2, as I need some of those bits anyway for the N8VEM. Are you at the point where it can boot to an A> prompt and then get the full 64k of ram available?
The Ram Disk·took too long to load from the PC so I started on the microSD a couple of days ago. I now have blank images in the FAT file of 8MB drives· (only using the first 128 bytes in a·sector so the files are really 32MB) and a 960K CPM image (actually it is 1MB and contains the Z80 64K plus the 960KB CPM image together in 512 byte sectors so I can load quick as it will be preloaded during bootup). I posted them in .rar format on ZiCog last night. I am hoping the sectors within the files·are contiguous so I can directly read/write into them. That way·I can keep FAT16 meaning we can also boot prop code or whatever, by having PropDos sitting in a layer above.
FYI: I am using the onboard PropPlug emulator to load the code - saves multiple parts. The FT232RL (smt) wasn't too hard to solder with some solder paste and solder wick, even with my aging eyesight. The microSd smt socket was a breeze.
I was going to ask you a couple of questions about CPM last week regarding file sizes and 128/137 byte sectors, etc. I found most of what I needed and will emulate the extra 3+6 bytes for 137 byte sectors if I need to. I wrote a VB6 program to build the 64KB file from the cpm files. I also have an 8MB CPM file which just has the sector number as the first 6? characters followed by "*". A good test program to see that I am reading what I expect.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Links to other interesting threads:
· Home of the MultiBladeProps: TriBladeProp, SixBladeProp, website (Multiple propeller pcbs)
· Single Board Computer:·3 Propeller ICs·and a·TriBladeProp board (ZiCog Z80 Emulator)
· Prop Tools under Development or Completed (Index)
· Emulators: Micros eg Altair, and Terminals eg VT100 (Index)
· Search the Propeller forums (via Google)
My cruising website is: ·www.bluemagic.biz·· MultiBladeProp is: www.bluemagic.biz/cluso.htm
I'm still deep inside the coding for a wireless mesh, with auto discovery of new nodes and auto update of software using xmodem. Also working on ethernet to serial devices, so that a mesh can interface directly to the internet without needing a PC. We had a test link going recently from Canada to Australia so the concept is sound. Soon we should have a BBS system running. The aim ultimately is to be able to log into a home automation network from anywhere on the web, turn on things, sense things etc, but without needing to leave a PC on. Ultimately, nodes should be low power (?battery or solar even) but smart enough to think for themselves, recompile new updates of software etc.
One of the features of wireless is that everyone is sharing the same bandwidth. So you can't type "DIR" into a wireless mesh, because 10 boards will reply at once, and then each one spits out its listing and the whole thing goes epileptic. There are software solutions, essentially involving disabling echo within CP/M and random delays before responding. So sometimes you need to tweak CP/M itself. On a N8VEM this involves making a new eprom, so the prop solution could end up being quicker too. I've also been working on a hardware model of the wireless system, essentially a bunch of RS232 connections all connected together at a single point (but with the correct logic gates as you can't physically just connect them all together). I've also discovered you need more than one RS232 port - in fact I need one for the PockeTerm, one to a PC, one to the wireless transceiver, one to an analog I/O and relay board, and one to a watchdog reset circuit (using a picaxe chip). So that is a 1:5 RS232 router that accepts control signals at 1200 baud, but which can switch at any baud rate. I've just sent that off to be made tonight.
Getting back to what you are doing, ok the N8VEM has CP/M sitting in 16K of eprom and loads that into the top of 64k of ram, and then runs it. That version of CP/M has enough smarts to work out all the drives. I have a vb.net program that can build these eprom images with many different options (12 at present) using simple check boxes. But you still have to then download actual programs at 38400 baud to the drives.
Your model sounds better. Just to clarify, are you just taking a 1meg .bin image off the altair SIMH or whatever, and just emulating that as a total package? IE, CPM plus a whole pile of files. If so, that makes a lot of sense, because it would be much faster to package up and download.
If that is what is going on, I have some vb.net software that can link to the altair simh. It can link in two ways - either by saving a batch file that does a whole lot of things on startup and then starting the simh and running that batch file. Eg - load program x, compile program x, save program x, write program x back to the PC directory. So a compile of C or BASIC becomes a very fast single step operation. The other way to interface to the simh is to run the simh and attach it to a telnet terminal. And then from vb.net, hook into that telnet terminal with a winsock .dll This is done locally, but it probably could be done to another PC on the internet too. This is useful for moving groups of files from a PC directory into a 1 meg (or 8meg) .bin simh image. So you can copy all the .COM files with one keypress. Or all the Wordstar files. Or Basic files. It makes the whole process of creating disk images so much quicker when it is just a checkbox process.
The altair simh only lets you have 8 drives A to H. CP/M can only go to 8meg. But that means a SD card could hold drives A-Z with 8 meg in each =26*8. Ok, maybe a bit less because drive A would be the ram drive.
I'd better keep a close eye on this thread - all very exciting stuff!
Your RS232 board doesn't surprise me. Did you think about or are you using a prop for this, or just real gates?
My VB6 program is no where as sophisticated as yours. But it allows me to build the 64KB RAM/EPROM from files and I just added the CPM disk image (converting on the fly stripping the 137 byte sectors into 128 bytes) as well and combined it into a single 1MB file using the whole 512 byte sectors since it is being preloaded by spin/pasm into the SRAM. So it is as you said above. Heater just made me a 960KB drive with the files on it. As I said above, I created a blank file and a file with sectors filled with the sector number in each sector, each 8MB (ie. 32MB) and just copied them, renaming them DRIVE__A.CPM thru to H, and then copied to the microSD which uses FAT16. My board displays these files with code I wrote using the Femto SD Trainer code as the base. I also have an SD adapter and one of my laptops has an SD slot so I can read and write directly to the card
I think if can get the drives to be files within the FAT16 system, then it will be easy to copy any CPM files required to the SD via FAT and have a spin section that will read them and copy them onto one of the drive files. Then there will be no slow xmodem to worry about. Well, other than your wireless exploits. I am thinking of putting the z80 block move instruction into the driver and heaters ZiCog can just call the other cog directly to do this.
By renaming FAT files, it would be easy to swap hard drives, but I guess there wont be much requiring 8 x 8MB drives anyway.
While writing this I have been running a walking bit code (spin/pasm) through my SRAM to test its integrity - looks good.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Links to other interesting threads:
· Home of the MultiBladeProps: TriBladeProp, SixBladeProp, website (Multiple propeller pcbs)
· Single Board Computer:·3 Propeller ICs·and a·TriBladeProp board (ZiCog Z80 Emulator)
· Prop Tools under Development or Completed (Index)
· Emulators: Micros eg Altair, and Terminals eg VT100 (Index)
· Search the Propeller forums (via Google)
My cruising website is: ·www.bluemagic.biz·· MultiBladeProp is: www.bluemagic.biz/cluso.htm
You may read more about this on the following threads...
PropCMD···· http://forums.parallax.com/showthread.php?p=796034
ZiCog········ http://forums.parallax.com/showthread.php?p=788511
I also thought I would share my reasoning for the choice of microSD... see the attached photo. The top left is a USB connector with an 8GB microSD inserted (you can see the "8"). The top right is an adapter to take a microSD and convert to the SD format (you can see the slight recess at the bottom where·the microSD card to plugs into and is completely enclosed). The bottom is obviously the microSD card.
There is virtually no difference in price. For the time being, only use 2GB or less. The only issue is that SPI which is what we use to access the card is not mandatory for microSD. There have been reports that some microSD's do not support SPI although I have not found any.
The choice of miniUSB is·because it is used by most cameras and phones, so users are very likely to have a cable(s) already.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Links to other interesting threads:
· Home of the MultiBladeProps: TriBladeProp, SixBladeProp, website (Multiple propeller pcbs)
· Single Board Computer:·3 Propeller ICs·and a·TriBladeProp board (ZiCog Z80 Emulator)
· Prop Tools under Development or Completed (Index)
· Emulators: Micros eg Altair, and Terminals eg VT100 (Index)
· Search the Propeller forums (via Google)
My cruising website is: ·www.bluemagic.biz·· MultiBladeProp is: www.bluemagic.biz/cluso.htm
Post Edited (Cluso99) : 4/6/2009 4:23:00 AM GMT
This is not true[noparse]:)[/noparse] The SIMH AltairZ80 simulator suports 8 floppy drives of 1MB (A to H) and 4 hard drives of 8MB (I to L). The number of each is configurable in an include file when you rebuild the BIOS.
Adding hard drive simulation to PropAltair/Zicog has been on my todo list for a long while.
Now, SIMH has a config file which is read at start up which specifies which disk image file on the host file system (in our case FAT16) is used as which CP/M drive.
It has lines like:
attach dsk1 cpm2.dsk
It would be great to use a similar system on TriBlade, no copying of disk images around required.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
For me, the past is not over yet.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Links to other interesting threads:
· Home of the MultiBladeProps: TriBladeProp, SixBladeProp, website (Multiple propeller pcbs)
· Single Board Computer:·3 Propeller ICs·and a·TriBladeProp board (ZiCog Z80 Emulator)
· Prop Tools under Development or Completed (Index)
· Emulators: Micros eg Altair, and Terminals eg VT100 (Index)
· Search the Propeller forums (via Google)
My cruising website is: ·www.bluemagic.biz·· MultiBladeProp is: www.bluemagic.biz/cluso.htm
Thanks! for the very detailed update . It works in Google Doc's just fine.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Aka: CosmicBob
I have attached the zipped source files. The ZiCog Z80 code is not heaters latest (I am still using v005). There is an issue I need to fix by updating the TBP (TriBladeProp) driver. Currently I am resetting the cog to overcome this until I get time to fix (add a simple routine). All software to compile is in the demo zip file including the homespun compiler and the batch file to compile. There are 3 or 4 warnings about missing '#' in the jmp/ret instructions but this is in fact correct.
I have also attached the zipped disk files for the microSD card. It is a large file. Expand it first on your pc and then copy DRIVE__C.DSK to files·A through H. This is a blank CPM2.2 disk which can be used for 1MB drives A...H. Expand DRIVE_1MB.RAM. This file is used to load the CPM boot information into the RAM (sram) by the spin program. DRVCPM_2.DSK is the disk file containing all the CPM2 files, formatted as a 1MB drive. It is best to leave this untouched and just copy it over the DRIVE__A.DSK you have made above. This way anything you do will overwrite/add to the A CPM drive and you will still have an original. And CPM will boot to the A drive.
Once these files are ready, format your microSD card with FAT16 and preferably 32KB clusters. Then copy these files one by one. They will then be contiguous (using successive clusters) because we access them using physical addresses once the starting point is found under FAT16.
SURVEY is the CPM program that shows the disk drives and memory map. It will not display drives until you have done a DIR command.
As I haven't included the CPM3 disk file, edit the spin demo file to change the H drive to use DRIVE__H.DSK and recompile.
This version only uses Blade #2 and the PropPlug connected to your PC and PropTerminal.
I edit and save using PropTool (dont forget to save), then go to a dos box and run z.bat from the source folder (set the path etc - ensure there are no blanks in the path/file names). Check the compile produced an output binary file with no errors (other than the warnings mentioned above). Go back to PropTool and load the binary file to the prop ram. The switch to PropTerminal. You will have about 3 seconds to do this before the software starts on the prop. If you miss this, just type "y" to load the ram and continue.
If you have any questions just ask. I am a bit tied up for a couple of days (in case I am slow to respond).
Enjoy
PS. Later I will publish the entire code you require to boot from an SD card without the PC.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Links to other interesting threads:
· Home of the MultiBladeProps: TriBladeProp, SixBladeProp, website (Multiple propeller pcbs)
· Single Board Computer:·3 Propeller ICs·and a·TriBladeProp board (ZiCog Z80 Emulator)
· Prop Tools under Development or Completed (Index)
· Emulators: Micros eg Altair, and Terminals eg VT100 (Index)
· Search the Propeller forums (via Google)
My cruising website is: ·www.bluemagic.biz·· MultiBladeProp is: www.bluemagic.biz/cluso.htm
Post Edited (Cluso99) : 4/24/2009 4:53:38 AM GMT
Have you written a generic object for the SRAM configuration yet?
OBC
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
New to the Propeller?
Visit the: The Propeller Pages @ Warranty Void.
A megabyte of RAM for code and data, 2G file store, built in display, multiple I/Os, a pile of unused 32 bit CPUs. No I can't think of any other possible use for this[noparse]:)[/noparse]
I think this is driven by the fact that if you want to use 64K of RAM (or more) for running editors, compilers whatever on the Prop this emulation is the only possible way to do it at the moment.
Things could/will change dramatically when we get C compilers that can execute code from external memory or perhaps better for most Prop heads a Spin compiler/interpreter combo that works from external RAM !
Anyway, what about a 6809 emulation[noparse]:)[/noparse]
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
For me, the past is not over yet.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
JMH
The SD card driver is the Femto drivers with a published fix
I have said many times, emulators are NOT my only purpose of the TriBlade. It is just my focus. However, everything I have written is with general purpose use in mind, just like heaters ZiCog is deliberately NOT tied to the TriBlade.
We eagerly await the Prop spin/pasm or c compiler running on the prop.
Meanwhile we can run/compile other code under CPM, like Turbo Pascal, MBasic, Fortran, ADA, etc, etc, etc. And we can run SuperCalc (not tried yet), etc, etc, etc.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Links to other interesting threads:
· Home of the MultiBladeProps: TriBladeProp, SixBladeProp, website (Multiple propeller pcbs)
· Single Board Computer:·3 Propeller ICs·and a·TriBladeProp board (ZiCog Z80 Emulator)
· Prop Tools under Development or Completed (Index)
· Emulators: Micros eg Altair, and Terminals eg VT100 (Index)
· Search the Propeller forums (via Google)
My cruising website is: ·www.bluemagic.biz·· MultiBladeProp is: www.bluemagic.biz/cluso.htm
Does the text on the resistor network go under the small square on the outline?
For the micro SD card holder, are all four pads in back used as electrical connections?
For the pairs of capacitors in and near C16, C26, C33, what values do we use? Also, those are for overclocking with a 6 MHz crystal, right?
Does the switch go in any particular orientation?
Blade two, what are the U24A and U23A for as opposed to the versions without A?
What are the various holes for, notably the LK1, LK2, LK3, ...?
I suppose that the Blade 2 SRAM pins most definitely cannot touch the PCB (the pins you mention in your mod)?
Edit: another question: I don't have 240 ohm resistors or any 1K1 ohm resistors. Is it okay to substitute with something close (perhaps 250 and 1K respectively)?
Post Edited (SRLM) : 4/25/2009 4:52:36 AM GMT
1. The resistor network is soldered on the underside of the pcb and the usual "dot" indicating the common connection on the part goes to the square box end.
2. None of the pads at the back of the microSD connector are used in my design. They just provide stability for the connector in this application. Normally these back pads could be used for write protect and card inserted functions.
3. C16/17, C26/27 & C33/34 are for possible use in overclocking. No values are specified although I think 0.1uF and a 1uF (tantalum - so careful of polarity "+" closest to the vga connector) might be necessary from Sapheia's information. If required, they mount on the underside of the pcb. A 6MHz crystal hopefully should not require these.
4. There is no orientation for the switch. If the switch is in the left position·the power will be off, right it will be on.
5. U23A and U24A are for 2MB SRAMs which replace the U23 and U24 512KB SRAMs. They are EXPENSIVE and nothing uses them at present.
6. Links:
LK1 & LK2 allow Blade #1 & #2 to connect their P30/31 serial pins together. You cannot use the PropPlug to connect to these Blades when this is active. Beware that P30 connects to P30, and P31 connects to P31, so the terminal function on Blade #1 will need it's serial pins reversed in software,·otherwise there will be a bus conflict which may cause damage to the prop chips. Alternately use a small cable on these headers and cross them over.
LK3, LK4 & LK5 are designed to connect Blade #2 to Blade #3 together via the serial bus mentioned in LK1 & LK2. It can route the PropPlug to Blade #3 as well, including the reset line. See the schematic. Note these shunts can be all horizontal or all vertical for different pins
7. The modification to the SRAMs U23 and U24: These pins must not touch the pcb. If you are experienced, you may alternately drill the connection but·this is not wise. You may bend then cut the pin shorter on each SRAM to ensure it does not touch the pcb, but long enough to solder the wire to. Put the 2 wires under the ICs to hide most of the modification and keep the wires from getting caught. This is the professional way.
8. Just to reiterate... U12, U22, U32 & U25 must be soldered to the underside of the pcb as the pinout is reversed for this. They will not function if soldered to the top of the pcb !!!
9. U26 label should read FLASH, not EEPROM.
10. Q91 is inserted differently to the label. (see the errata A1)
11. For now, I suggest you do not fit U13 and U15 (SRAM and Latch for Blade #1) until such time as you have a requirement for it or someone produces a requirement for it. There are other possible uses for these pins if left unsoldered - such as allowing VGA AND TV concurrently with a cable, or more I/O.
12. I have not fitted (permanently) any of the EEPROMs as it is my intention (when time allows) to have Blades #1 and #3 code loaded by Blade #2 when in standalone (SBC = Single Board Computer) operation. The USB serial connection to the PC will go to Blade #3. All Blades will communicate between them via a common·high speed serial communications on a 1 or 2 wire bus (TBD).
13. I have not fitted the FLASH although I understand Mike Green is doing this.
14. You will see from the schematics, the LEDs are uncommitted, so you can wire them where you like. The pad will light the LED if it is logic high. I used a blue LED as LED2. I cannot recall which of the others were red and green. I only held wires to check the props were functioning.
Good luck and I await to hear you are running... If you have any questions, please ask.··
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Links to other interesting threads:
· Home of the MultiBladeProps: TriBladeProp, SixBladeProp, website (Multiple propeller pcbs)
· Single Board Computer:·3 Propeller ICs·and a·TriBladeProp board (ZiCog Z80 Emulator)
· Prop Tools under Development or Completed (Index)
· Emulators: Micros eg Altair, and Terminals eg VT100 (Index)
· Search the Propeller forums (via Google)
My cruising website is: ·www.bluemagic.biz·· MultiBladeProp is: www.bluemagic.biz/cluso.htm
One alternative on SRAM Pins is to cut conections around holes.
U23 have only conections on solder side.
U24 have on both sides ... On Component side it have conection to one line on PCB.
It is that I modified it ...
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Nothing is impossible, there are only different degrees of difficulty.
For every stupid question there is at least one intelligent answer.
Don't guess - ask instead.
If you don't ask you won't know.
If your gonna construct something, make it·as simple as·possible yet as versatile as posible.
Sapieha
1)I've been socketing all the chips, and I'm wondering about the U23/U24 business. If I get a socket to fit the extra four pins, how easily can I switch between the two versions of SRAM? My guess is that I'll have to do some resoldering...
2) So for the microSD, is it okay if I bridge the four connections on the U23 side?
3) Will the overclocking caps hurt me if I only have a 5Mhz crystal?
4)A question that I edited in above at the same time you were posting: "Edit: another question: I don't have 240 ohm resistors or any 1K1 ohm resistors. Is it okay to substitute with something close (perhaps 250 and 1K respectively)?"
Thanks for telling be about the tantalum capacitor polarity. Of the three on the board (right!?) I've only gotten to C91 and it's in backwards...
On question 3.
Extra capasitors have little efect on 5-6MHz crystal.
It give sxtra stablity on PLL system in Propeler chip most with frequensies 14-15MHZ x PLL8
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Nothing is impossible, there are only different degrees of difficulty.
For every stupid question there is at least one intelligent answer.
Don't guess - ask instead.
If you don't ask you won't know.
If your gonna construct something, make it·as simple as·possible yet as versatile as posible.
Sapieha
2. No problems if these pins are bridged as long as they are not shorting to the power plane underneath. Just meter the connection to verify this. (I presume you are not referring to the holes (vias) on the left side of the microSD connector near U26 - these cannot be shorted.)
3. No they will not cause any issues if installed. They just provide better decoupling and storage reservoir (in laymans terms). You can always solder these later by bending the legs slightly and soldering on the underside. Unless you want to significantly overclock, don't bother with them.
4. I am unsure. It will affect the colors not sure about the actual display.
5. C91 needs to be corrected. C11 (electrolytic) and C12 & C13 all require to be·correctly (polarity)·inserted.
6. Since you are asking about polarity, I presume you have fitted the diode D4 correctly - the small block is the white bar end·on the diode.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Links to other interesting threads:
· Home of the MultiBladeProps: TriBladeProp, SixBladeProp, website (Multiple propeller pcbs)
· Single Board Computer:·3 Propeller ICs·and a·TriBladeProp board (ZiCog Z80 Emulator)
· Prop Tools under Development or Completed (Index)
· Emulators: Micros eg Altair, and Terminals eg VT100 (Index)
· Search the Propeller forums (via Google)
My cruising website is: ·www.bluemagic.biz·· MultiBladeProp is: www.bluemagic.biz/cluso.htm
I originaly tried your ram test (with blue led on P24) which did not work, which i guess is due to the mods???. Next I tried the Test201, with this I get one screen full of number dumps. The numbers start wit 10,11,12 ...get up to 7f turn to 77s for some more rows and thenstart at ff count down to 80 and the 77s come back until i) complete gets a row of its own. Reading the code I think that it should test all the pages and the i) should be a number indicating that all 1024 pages have been vetted
Am I using the wrong test prog or have I got another puzzle waiting, whoopee.
So its a yippee, possibly.
I got one of my TriBladeProps assembled now and can boot CPM from your posts
I assume the files on the disk are the cbios that is being used? I want to look at using 128 byte sectors and packing 4 per SD sector to be more efficient and want to look at CPM 3 long term to get to 512MB hardrives
Have you enabled using console going to Blade 1 with VGA and keyboard ? I am interested in testing that out next.
This is exciting and fun
Dave
Yoda: I packed 4 x 128 byte sectors into the DRIVE_1M.xxx file. It has the CPM boot info (Z80 ram) in the first 64KB followed by the 960KB CPM2.2 Ram Disk (drive A). However, I decided not to pack the other drives as 4 x 128 bytes and just use the first 128 bytes as the data for each sector. I wasn't sure of the overheads used in the card itself to perform the write to the flash, and I was not concerned (at this point anyway) about available space on my 1GB and 2GB cards. You may notice, that while the files are 32MB (for an 8MB drive), only a 1MB drive is actually used. On the agenda is the 2 or 4 HD drives which can be 8MB each. Changing the BIOS will of course increase this, but later.
I did have the console working to Blade #1, but not satisfactory to me, and not at present. I went back to getting CPM3 running. As soon as I get time (shortly) I will get the bank switching for CPM3 running. Then I can go back to the Terminal in Blade #1. I think Mike is working on this anyway.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Links to other interesting threads:
· Home of the MultiBladeProps: TriBladeProp, SixBladeProp, website (Multiple propeller pcbs)
· Single Board Computer:·3 Propeller ICs·and a·TriBladeProp board (ZiCog Z80 Emulator)
· Prop Tools under Development or Completed (Index)
· Emulators: Micros eg Altair, and Terminals eg VT100 (Index)
· Search the Propeller forums (via Google)
My cruising website is: ·www.bluemagic.biz·· MultiBladeProp is: www.bluemagic.biz/cluso.htm