with the sd card placed so that the gold tabs are facing you at the top, the pins used are the "middle 7" of 9. The one that is displaced by the corner diagonal (9) and the last of the strait ones (8), which is half width, are not used.
1 is /CS
2 is DI data input
3 is VSS
4 is VDD
5 is SCLK
6 is VSS(2)
7 is DO data output
hope that this is what you needed.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Style and grace : Nil point
Thanks for the links. The pinouts are all ok, but it looks like I need five pullup resistors. I'm not 100% sure why the pullups are needed on some pins (eg DO), but not having one on pin 8 could be the problem.
@Cluso, I'm getting -1 back from that routine, not 0 or 1, so not sure what is going on there. But I can swap back and forth to the triblade board and debug that by comparing the two.
Drac: -1 means the statement is true. That is, a cog was started because
cog:=cogid+1
and
result := cog > 0 'i.e. if cog > 0 result := true (-1) else result := false (0)
The only catch in spin is that true := -1 and false := not true so false is <> -1
Does this make sense?
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔ Links to other interesting threads:
Nooooo... a dead prop?? I was hoping for a photo but I guess the photo might now be one of a charred mess?
I've just got home from work and fired up the board. I'm going to find this /CS bit of code, one way or another. I take it that Spin simply can't print debug messages in an easy way from within sub-objects?
Ok, if that is the case, time for some workarounds. I can at least print debug messages from the main subroutine, so I can narrow down where it calls and object and when it fails to return from that object. So, in the main object is this code:
PRI FindSDblock | i, n, r 'read blockno of file for count into buffer
r := sd.startSD(@ioControl) 'start the SD routines
UART.str(string("SD started"))
CheckError(not r)
UART.str(string("calling mount_explicit"))
r := sd.mount_explicit(spiDO,spiClk,spiDI,spiCS) 'mount the SD
uart.dec(r) 'print r
It prints "calling mount_explicit" but never prints an r value, so isn't returning from this object.
Unfortunately, mount_explicit is a complex object with all sorts of sub calls, so not easy to trace. So I thought I would start at the beginning and insert this bit of code in the fsrwfemto_rr001 object:
pub mount_explicit(DO, Clk, DI, CS) | startBlk, sectorspercluster, reserved, rootentries, sectors
{{
' Mount a volume. The address passed in is passed along to the block
' layer; see the currently used block layer for documentation. If the
' volume mounts, a 0 is returned, else abort is called.
}}
abort(99) ' insert this to abort the object with a value?
However, that still doesn't return back to the main object. What am I doing wrong here? I thought the abort command aborted a private or public method with a value? Which would be printed with the uart.dec(r) Or is something wrong with my syntax here?
Help, as always, is most appreciated. Partly this is a learning experience with Spin (which I'm finding much harder than Pasm). So everyone's patience is most appreciated.
Dr_A: If your called method exits with an abort then the caller has to catch it. So you need a slash in the call like so:
r := /sd.mount_explicit(spiDO,spiClk,spiDI,spiCS)
r := \sd.mount_explicit(spiDO,spiClk,spiDI,spiCS)
Basically method calls can be many levels deep. If one of the low level methods aborts then so does it's caller, and so does it's caller, and so does it's caller until we end up at the top level and the whole Prop stops.
So somewhere in the chain a "/""\" must be used to catch the failure and react accordingly.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
For me, the past is not over yet.
I think I have found the problem with my larger board. It was not accepting EEPROM programming and also doing strange thing with commands from the Propcomm bit. I had pit in the 5 pin Four 10K resistor strip reversed, so instead of 10K pullups I had a mind bending matix of pin to pin 20Ks (probably).
It was whilst trying to get around this I cocked up the PSUs a frazzled my third prop, RIP. I have found one, but at £10 (Farnell).
Moral of the story, Slow down and get it right.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Style and grace : Nil point
My prop DIPs cost >US$20 so dont be upset about Farnell's cost.
Drac: Sorry, just havent had time to look for you. You could just directly place a char into the tx buffer in hub and increment the pointer. However, I am pretty sure there is not much space in the pasm code.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔ Links to other interesting threads:
At work for another 8 hours then will have time to check this out. Heater's explanation makes a lot of sense. Though there are quite a number of abort statements and no \. Maybe they are legacies from earlier testing, but it could explain the whole thing hanging rather than coming back with a meaningful error message. Abort statements seem like 'exit sub' in vb.net, and I've tried to avoid those more recently as they seem to be, dare I say it, an example of poor programming. Better to nest things properly. But that is just a personal opinion.
Some good things might come out of this, as I've had some instances where the triblade hangs on bootup because I forgot to re-insert the sd card. It would be great to have some messages on the screen eg SD card not found, plus other diagnostic messages relating to the card if those are possible. So maybe some different code? (Which I think Cluso was working on anyway for other reasons).
This SD card code is turning out to be some of the hardest coding I've done in ages. Much harder than rewriting the CP/M Bios burning an eprom each time, for instance. And Cluso said it would be easy! Hey, Cluso, don't tell me anything else will be easy, ok *grin*
@Toby, I am very interested in the schematic of the SD card bit. I think it is simply 4 lines with pullups to the prop, and then a fifth pin with a pullup too. I'd just like to confirm that the tracks go to the actual pins on the sd card as I think they do. Then again, if you have a resistor array in the wrong way...
Drac: You should not be having this many problems...
Time to go back to basics. Get PropDos running on the TriBlade - you only need to change the address pins.
When this runs (should be simple) then try it on your board with different pins. If that fails use an SD adapter to use your current microSD and see if you have an SD card fault.
Otherwise, run the fsrw24 and test your card. Lots of diagnostics in this code. (use the search link in my signature for "Shes so pretty")
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔ Links to other interesting threads:
I have put only one 10K pullup on the /CE to the SD (propcomm) as when I was messing with PropDos et al before there didn't seem to be any change having all the lines, includung the /INT line, raised or not. I put the one on this version only because I was having problems.
I use the middle 7 only, 2 x GNDS, +3.3V,DI, DO, SCLK and /CE. The outside ones (8 and 9) dont even get holes in the PCB.
All the probs I had were due to a SIL pack giving me effectivly 20K star interconects on P28 _P31. As it was just a slightly changed version that already worked I blamed the new bit. The PropCOMM does winge about no SD and just halts so there may be come lines of code to be pinched there.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Style and grace : Nil point
I had a problem with fsrw23-1 whereby it would hang on trying to use a dead SD card with no meaningful error message. Turned out to be a missing "\" on methods in the serial_terminal.spin test harness like :
Many thanks for the helpful advice. Tried the sd card from the triblade but no luck (and put it back in the triblade just to confirm it hadn't been zapped and it is ok).
Fired up PropDOS. Changed the text. object from a TV to a uart object, so it sends back to a terminal. Commented out the color changes for the moment. Running via a terminal program.
Boots up fine, then gives an error message.
Propeller MiniDOS V1.6
By Jeff Ledger
No SD card present.
SD card expected on:
13, 14, 15, 9
System halted.
No pullups (but Toby, you said you didn't have any and it worked). Tried three different cards.
I'm still thinking a wiring error. Something like DO and DI swapped (but I have checked and rechecked). On the propeller, CS goes to physcial pin 14, DI goes to physical pin 20, CLK goes to physical pin 19 and DO goes to physical pin 18. On the SD card, pin 1=CS, 2=DI, 3=Gnd, 4=3V3, 5=Clk, 6=Gnd, 7=DO.
I've attached a photo.
If anyone can spot a mistake it would be very much appreciated!
Addit: I have had some success. I put in the pullup resistors. PropDos passes the tests. Now doing some more testing but it would appear I have one working SD card (the micro one from the triblade) and no less than 6 cards of various sizes and brands that do not work. At least I have something to work with now - the wiring is correct etc. These SD cards are difficult little beasties though!
Dr_A. That seems very unfortunate. I have only ever owned 3 micro SD cards and they all worked. Until one decided to write protect a bunch of blocks at the beginning. No idea how that happened but I can still read all blocks and write to all the others. Makes it no good for FAT though.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
For me, the past is not over yet.
2GB cards are the max that many equipment can recognise because it it the FAT16 limit. So they should be available for a long time to come, although much larger cards can be bought now. It is almost impossible to buy <1GB cards at retailers now.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔ Links to other interesting threads:
Time to delve back into the code. I've got it doing a DIR on propdos, though it doesn't seem to want to load actual files yet. Having double and triple checked pinouts (plus the fact that it can do a propdos DIR) I think says that the connections are right.
Interesting comment from Cluso re Sandisk. The only one that works for me is the Sandisk brand (1Gb). Yet all these other brands work fine in other devices such as a camera. I'm going to get some more of that brand just to simplify things. Has there been any discussion anywhere about different brands of disk? I'm thinking of timing issues etc. And also mildly concerned about the triblade failing to boot the sd card about one in ten times. I think something might be operating right on the limits here.
Addit: A thought. I found somewhere that while sd cards are specified for 45 or 80mA, they say one should allow up to 200mA. I've got a 100mA reg on the 3V supply - maybe this isn't quite enough for some brands?
The only ones that I have working also happen to be the same sort as each other. They are naned as "Integral" and www.integralmemory.com.
The lowest current that I have ever offered them is probly 500mA (reg) - 150mA for other things.
"Thankfully" the 9V was onto the small board and it didn't have the SRAM or the SD on it. (Next test will be attaching a kite, flying in a thunderstorm)
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Style and grace : Nil point
Propdos certainly is a nifty program. Much like CP/M in many ways!
I can do a DIR and I can TYPE out an entire text file in PropDOS. This really exercises the SD card in terms of current consumption and it is working fine, so I think that makes a power supply problem less likely.
I've got three Sandisk brand cards on the way.
I found another brand of SD card (Kingmax) and that works. Legend brand 512Mb does not work.
Next thing is to try hacking into the zicog code. Cluso has spent some time modifying the SD driver code so that it can share prop pins with other things. I'm looking at a design that simply has 4 dedicated pins, so the upside is that it will be simpler and use existing code. But the downside is that I guess I'm on my own here! I think I have all the bits working though; keyboard, vga screen, sd card in PropDOS and low level read and write to memory. Maybe this will be useful to someone porting Zicog over to other platforms?
Based on all the mods made so far (pullups etc), I sent off to get a new board made. (Actually 10 boards). I'm going for switching regs for both the 3V3 and 5V supplies.
Next step is to try to work out what has been changed from fsrwFemto to fsrwFemto_rr001.spin.
Ok, got that working. Just needed to drop back to the original code, then change 'startSD' to 'start', and change 'mount_explicit' to 'mount'. Got this far (hangs when gets to memory access as I haven't tied that in yet). I haven't changed a huge amount in Cluso's code so it should be possible to do this with ifdefs and run development of the various platforms in parallel (including Cluso's tiny ramblade which I think will really be exciting!). My code aims to drop in existing modules for keyboard, vga, sd card, and then just add some custom code to replace Cluso's triblade ram driver.
I am playing with one attempt to bolt Z80 to props. I tried to play safe with a tv version of PropDos and multiplexed Addr/Data lines. The Z80 and the Prop run on the same volts, varied by the PSU and the switch. The KBD hates anything less than 4 Volts (still thats a few less than 9 !!)
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Style and grace : Nil point
Ah. Very interesting. This is very sad I know, but I think I can guess some of your chips based on that photo of a board. Ok, big 40 pin on right is propeller. 40 pin on left is Z80. I see a 32 pin chip - 512k ram? Or maybe smaller. 3 Latches?
In a Z80 book I've got is a 'minimum' Z80 circuit in old-school electronics. A Z80, some ram, some eprom and a single latch for output and a single buffer for input. Well yours goes a little less than that. So is your plan to fill up the Z80 ram with data, then do a reset and set it going? Z80 clock comes from the propeller?
Nice board BTW. Old school single sided with links. And proper sized chips!
Its about my level ! I tend to re use all my stuff over and over again, most of my components are scrounged from scrapped kit in the first place. I call it recycling, everbody else calls it Scrooge in Steptoe's yard. I even experimented with electolosis to reclaim the feric chloride. The layout is done with FreePCB which I have become locked into, I must try EAGLE and I have access to DipTrace via work. In the meantime I can knock up things single sided with stripped down telephone wire.
The death of my third 40 pin Prop had made me think more about SM work but I have to pass wires between chip pins and with Toner tranfer that means 100 mil pitch. I shouls think about comertial PCB makers and the get those fantastic vias and thru holes.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Style and grace : Nil point
You both have some interesting board designs going on for ZiCog/Prop/CPM. Might I suggest you both start new threads for them.
In fact I'd like to be able to play with both of them myself so having their own homes would be great for following what progresses. Sadly I have very little time for my Prop Hobby at the moment.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
For me, the past is not over yet.
Comments
I have only just got home, and have left a load of stuff back in the office. The data I used to wire up the full sized SDs was from
http://pinouts.ru/Memory/sdcard_pinout.shtml
http://www.ucontroller.com/documentation/SDCardDoc.html
with the sd card placed so that the gold tabs are facing you at the top, the pins used are the "middle 7" of 9. The one that is displaced by the corner diagonal (9) and the last of the strait ones (8), which is half width, are not used.
1 is /CS
2 is DI data input
3 is VSS
4 is VDD
5 is SCLK
6 is VSS(2)
7 is DO data output
hope that this is what you needed.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Style and grace : Nil point
Thanks for the links. The pinouts are all ok, but it looks like I need five pullup resistors. I'm not 100% sure why the pullups are needed on some pins (eg DO), but not having one on pin 8 could be the problem.
@Cluso, I'm getting -1 back from that routine, not 0 or 1, so not sure what is going on there. But I can swap back and forth to the triblade board and debug that by comparing the two.
Cheers, James
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
www.smarthome.viviti.com/build
cog:=cogid+1
and
result := cog > 0 'i.e. if cog > 0 result := true (-1) else result := false (0)
The only catch in spin is that true := -1 and false := not true so false is <> -1
Does this make sense?
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Links to other interesting threads:
· Home of the MultiBladeProps: TriBlade,·RamBlade,·SixBlade, website
· Single Board Computer:·3 Propeller ICs·and a·TriBladeProp board (ZiCog Z80 Emulator)
· Prop Tools under Development or Completed (Index)
· Emulators: CPUs Z80 etc; Micros Altair etc;· Terminals·VT100 etc; (Index) ZiCog (Z80) , MoCog (6809)
· Search the Propeller forums·(uses advanced Google search)
My cruising website is: ·www.bluemagic.biz·· MultiBladeProp is: www.bluemagic.biz/cluso.htm
Do you think that there will a claim under the sale of goods act, after all I only overvolted it by a small factor of 3:1.
I wish I had a cat, something has got to be kicked !
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Style and grace : Nil point
At least you have spares.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Links to other interesting threads:
· Home of the MultiBladeProps: TriBlade,·RamBlade,·SixBlade, website
· Single Board Computer:·3 Propeller ICs·and a·TriBladeProp board (ZiCog Z80 Emulator)
· Prop Tools under Development or Completed (Index)
· Emulators: CPUs Z80 etc; Micros Altair etc;· Terminals·VT100 etc; (Index) ZiCog (Z80) , MoCog (6809)
· Search the Propeller forums·(uses advanced Google search)
My cruising website is: ·www.bluemagic.biz·· MultiBladeProp is: www.bluemagic.biz/cluso.htm
I've just got home from work and fired up the board. I'm going to find this /CS bit of code, one way or another. I take it that Spin simply can't print debug messages in an easy way from within sub-objects?
Ok, if that is the case, time for some workarounds. I can at least print debug messages from the main subroutine, so I can narrow down where it calls and object and when it fails to return from that object. So, in the main object is this code:
PRI FindSDblock | i, n, r 'read blockno of file for count into buffer r := sd.startSD(@ioControl) 'start the SD routines UART.str(string("SD started")) CheckError(not r) UART.str(string("calling mount_explicit")) r := sd.mount_explicit(spiDO,spiClk,spiDI,spiCS) 'mount the SD uart.dec(r) 'print rIt prints "calling mount_explicit" but never prints an r value, so isn't returning from this object.
Unfortunately, mount_explicit is a complex object with all sorts of sub calls, so not easy to trace. So I thought I would start at the beginning and insert this bit of code in the fsrwfemto_rr001 object:
pub mount_explicit(DO, Clk, DI, CS) | startBlk, sectorspercluster, reserved, rootentries, sectors {{ ' Mount a volume. The address passed in is passed along to the block ' layer; see the currently used block layer for documentation. If the ' volume mounts, a 0 is returned, else abort is called. }} abort(99) ' insert this to abort the object with a value?However, that still doesn't return back to the main object. What am I doing wrong here? I thought the abort command aborted a private or public method with a value? Which would be printed with the uart.dec(r) Or is something wrong with my syntax here?
Help, as always, is most appreciated. Partly this is a learning experience with Spin (which I'm finding much harder than Pasm). So everyone's patience is most appreciated.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
www.smarthome.viviti.com/build
Post Edited (Dr_Acula) : 11/5/2009 12:57:11 PM GMT
r := /sd.mount_explicit(spiDO,spiClk,spiDI,spiCS)
r := \sd.mount_explicit(spiDO,spiClk,spiDI,spiCS)
Basically method calls can be many levels deep. If one of the low level methods aborts then so does it's caller, and so does it's caller, and so does it's caller until we end up at the top level and the whole Prop stops.
So somewhere in the chain a "/""\" must be used to catch the failure and react accordingly.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
For me, the past is not over yet.
Post Edited (heater) : 11/5/2009 3:20:34 PM GMT
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
For me, the past is not over yet.
What details would you like from any Pics?
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Style and grace : Nil point
It was whilst trying to get around this I cocked up the PSUs a frazzled my third prop, RIP. I have found one, but at £10 (Farnell).
Moral of the story, Slow down and get it right.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Style and grace : Nil point
Drac: Sorry, just havent had time to look for you. You could just directly place a char into the tx buffer in hub and increment the pointer. However, I am pretty sure there is not much space in the pasm code.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Links to other interesting threads:
· Home of the MultiBladeProps: TriBlade,·RamBlade,·SixBlade, website
· Single Board Computer:·3 Propeller ICs·and a·TriBladeProp board (ZiCog Z80 Emulator)
· Prop Tools under Development or Completed (Index)
· Emulators: CPUs Z80 etc; Micros Altair etc;· Terminals·VT100 etc; (Index) ZiCog (Z80) , MoCog (6809)
· Search the Propeller forums·(uses advanced Google search)
My cruising website is: ·www.bluemagic.biz·· MultiBladeProp is: www.bluemagic.biz/cluso.htm
Some good things might come out of this, as I've had some instances where the triblade hangs on bootup because I forgot to re-insert the sd card. It would be great to have some messages on the screen eg SD card not found, plus other diagnostic messages relating to the card if those are possible. So maybe some different code? (Which I think Cluso was working on anyway for other reasons).
This SD card code is turning out to be some of the hardest coding I've done in ages. Much harder than rewriting the CP/M Bios burning an eprom each time, for instance. And Cluso said it would be easy! Hey, Cluso, don't tell me anything else will be easy, ok *grin*
@Toby, I am very interested in the schematic of the SD card bit. I think it is simply 4 lines with pullups to the prop, and then a fifth pin with a pullup too. I'd just like to confirm that the tracks go to the actual pins on the sd card as I think they do. Then again, if you have a resistor array in the wrong way...
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
www.smarthome.viviti.com/build
Post Edited (Dr_Acula) : 11/6/2009 4:11:12 AM GMT
Time to go back to basics. Get PropDos running on the TriBlade - you only need to change the address pins.
When this runs (should be simple) then try it on your board with different pins. If that fails use an SD adapter to use your current microSD and see if you have an SD card fault.
Otherwise, run the fsrw24 and test your card. Lots of diagnostics in this code. (use the search link in my signature for "Shes so pretty")
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Links to other interesting threads:
· Home of the MultiBladeProps: TriBlade,·RamBlade,·SixBlade, website
· Single Board Computer:·3 Propeller ICs·and a·TriBladeProp board (ZiCog Z80 Emulator)
· Prop Tools under Development or Completed (Index)
· Emulators: CPUs Z80 etc; Micros Altair etc;· Terminals·VT100 etc; (Index) ZiCog (Z80) , MoCog (6809)
· Search the Propeller forums·(uses advanced Google search)
My cruising website is: ·www.bluemagic.biz·· MultiBladeProp is: www.bluemagic.biz/cluso.htm
I have put only one 10K pullup on the /CE to the SD (propcomm) as when I was messing with PropDos et al before there didn't seem to be any change having all the lines, includung the /INT line, raised or not. I put the one on this version only because I was having problems.
I use the middle 7 only, 2 x GNDS, +3.3V,DI, DO, SCLK and /CE. The outside ones (8 and 9) dont even get holes in the PCB.
All the probs I had were due to a SIL pack giving me effectivly 20K star interconects on P28 _P31. As it was just a slightly changed version that already worked I blamed the new bit. The PropCOMM does winge about no SD and just halts so there may be come lines of code to be pinched there.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Style and grace : Nil point
term.dec( \sdfat.mount_explicit(term#sd_DO, term#sd_CLK, term#sd_DI, term#sd_CS) )
Discussed here http://forums.parallax.com/showthread.php?p=824166
I you sure your SD cards are good? (Silly question I know)
Sounds like TriBlade may well be missing some backslashes as well.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
For me, the past is not over yet.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Style and grace : Nil point
Fired up PropDOS. Changed the text. object from a TV to a uart object, so it sends back to a terminal. Commented out the color changes for the moment. Running via a terminal program.
Boots up fine, then gives an error message.
No pullups (but Toby, you said you didn't have any and it worked). Tried three different cards.
I'm still thinking a wiring error. Something like DO and DI swapped (but I have checked and rechecked). On the propeller, CS goes to physcial pin 14, DI goes to physical pin 20, CLK goes to physical pin 19 and DO goes to physical pin 18. On the SD card, pin 1=CS, 2=DI, 3=Gnd, 4=3V3, 5=Clk, 6=Gnd, 7=DO.
I've attached a photo.
If anyone can spot a mistake it would be very much appreciated!
Addit: I have had some success. I put in the pullup resistors. PropDos passes the tests. Now doing some more testing but it would appear I have one working SD card (the micro one from the triblade) and no less than 6 cards of various sizes and brands that do not work. At least I have something to work with now - the wiring is correct etc. These SD cards are difficult little beasties though!
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
www.smarthome.viviti.com/build
Post Edited (Dr_Acula) : 11/7/2009 5:38:59 AM GMT
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
For me, the past is not over yet.
As stated befor the outside two connections are not used ( champhered and half width ones)
From left to right on pic, the SD connectins are-
DO -> P9 (A1)
GND
SCLK ->P28
VSS -> +3.3V (or 9V if you're a pillock)
GND
DI -> P8 (A1)
/CS -> latch Q6
On the PropCOMM
P0 -> DO
P1 -> SCLK
P2 -> DI
P3 -> /CS
All the above is for full size cards.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Style and grace : Nil point
Post Edited (Toby Seckshund) : 11/7/2009 9:59:25 AM GMT
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Links to other interesting threads:
· Home of the MultiBladeProps: TriBlade,·RamBlade,·SixBlade, website
· Single Board Computer:·3 Propeller ICs·and a·TriBladeProp board (ZiCog Z80 Emulator)
· Prop Tools under Development or Completed (Index)
· Emulators: CPUs Z80 etc; Micros Altair etc;· Terminals·VT100 etc; (Index) ZiCog (Z80) , MoCog (6809)
· Search the Propeller forums·(uses advanced Google search)
My cruising website is: ·www.bluemagic.biz·· MultiBladeProp is: www.bluemagic.biz/cluso.htm
I an just concerned that the relentless rise in capacity will leave us without new cards.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Style and grace : Nil point
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Links to other interesting threads:
· Home of the MultiBladeProps: TriBlade,·RamBlade,·SixBlade, website
· Single Board Computer:·3 Propeller ICs·and a·TriBladeProp board (ZiCog Z80 Emulator)
· Prop Tools under Development or Completed (Index)
· Emulators: CPUs Z80 etc; Micros Altair etc;· Terminals·VT100 etc; (Index) ZiCog (Z80) , MoCog (6809)
· Search the Propeller forums·(uses advanced Google search)
My cruising website is: ·www.bluemagic.biz·· MultiBladeProp is: www.bluemagic.biz/cluso.htm
Interesting comment from Cluso re Sandisk. The only one that works for me is the Sandisk brand (1Gb). Yet all these other brands work fine in other devices such as a camera. I'm going to get some more of that brand just to simplify things. Has there been any discussion anywhere about different brands of disk? I'm thinking of timing issues etc. And also mildly concerned about the triblade failing to boot the sd card about one in ten times. I think something might be operating right on the limits here.
Addit: A thought. I found somewhere that while sd cards are specified for 45 or 80mA, they say one should allow up to 200mA. I've got a 100mA reg on the 3V supply - maybe this isn't quite enough for some brands?
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
www.smarthome.viviti.com/build
Post Edited (Dr_Acula) : 11/8/2009 2:27:03 AM GMT
The lowest current that I have ever offered them is probly 500mA (reg) - 150mA for other things.
"Thankfully" the 9V was onto the small board and it didn't have the SRAM or the SD on it. (Next test will be attaching a kite, flying in a thunderstorm)
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Style and grace : Nil point
I can do a DIR and I can TYPE out an entire text file in PropDOS. This really exercises the SD card in terms of current consumption and it is working fine, so I think that makes a power supply problem less likely.
I've got three Sandisk brand cards on the way.
I found another brand of SD card (Kingmax) and that works. Legend brand 512Mb does not work.
Next thing is to try hacking into the zicog code. Cluso has spent some time modifying the SD driver code so that it can share prop pins with other things. I'm looking at a design that simply has 4 dedicated pins, so the upside is that it will be simpler and use existing code. But the downside is that I guess I'm on my own here! I think I have all the bits working though; keyboard, vga screen, sd card in PropDOS and low level read and write to memory. Maybe this will be useful to someone porting Zicog over to other platforms?
Based on all the mods made so far (pullups etc), I sent off to get a new board made. (Actually 10 boards). I'm going for switching regs for both the 3V3 and 5V supplies.
Next step is to try to work out what has been changed from fsrwFemto to fsrwFemto_rr001.spin.
Ok, got that working. Just needed to drop back to the original code, then change 'startSD' to 'start', and change 'mount_explicit' to 'mount'. Got this far (hangs when gets to memory access as I haven't tied that in yet). I haven't changed a huge amount in Cluso's code so it should be possible to do this with ifdefs and run development of the various platforms in parallel (including Cluso's tiny ramblade which I think will really be exciting!). My code aims to drop in existing modules for keyboard, vga, sd card, and then just add some custom code to replace Cluso's triblade ram driver.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
www.smarthome.viviti.com/build
Post Edited (Dr_Acula) : 11/8/2009 12:00:44 PM GMT
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Style and grace : Nil point
In a Z80 book I've got is a 'minimum' Z80 circuit in old-school electronics. A Z80, some ram, some eprom and a single latch for output and a single buffer for input. Well yours goes a little less than that. So is your plan to fill up the Z80 ram with data, then do a reset and set it going? Z80 clock comes from the propeller?
Nice board BTW. Old school single sided with links. And proper sized chips!
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
www.smarthome.viviti.com/build
The death of my third 40 pin Prop had made me think more about SM work but I have to pass wires between chip pins and with Toner tranfer that means 100 mil pitch. I shouls think about comertial PCB makers and the get those fantastic vias and thru holes.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Style and grace : Nil point
You both have some interesting board designs going on for ZiCog/Prop/CPM. Might I suggest you both start new threads for them.
In fact I'd like to be able to play with both of them myself so having their own homes would be great for following what progresses. Sadly I have very little time for my Prop Hobby at the moment.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
For me, the past is not over yet.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Style and grace : Nil point