The Courier dropped off the prototypes for three new PCB designs today...

Bill Henning

After lunch, I will begin testing the three new PCB's the courier delivered about an hour ago:

(As usual, Sapieha did a beautiful job on the PCB layouts)

1) SD-uSD

A 10 pin module with both full size SD and micro SD support - at the same time!

I will start a thread for this board after I test it and verify the PCB is fully functional.

Fits the following platforms:

- Parallax Proto Board (USB and Serial)
- all Mikronauts platforms (Morpheus, Propteus, CPUModule, and others)
- breadboards (Demo Board, Propeller BOE)
- Propeller Platform (using ppBridge)
- C3 (using c3Bridge)
- QuickStart (using qsBridge)

Sapieha wanted to make this board - so it got made!

2) qsBridge

Adds support for using 10 pin modules and large prototyping boards with QuickStart - it will be very handy with miniMods...

I will start a thread for this board after I test it and verify the PCB is fully functional.

3) QuickLMM+

Stacks on top of QuickStart, and features:

- up to six SOIC8/TDFN8/DIP8 QSPI memory devices (in pairs to form an 8 bit bus)
- micro SD slot
- optionally provides two 10 pin headers for my 10 pin modules
- I2C expansion header
- HCOMM header for SerPlug or 485Plug
- hacker port allowing eight more qspi devices to be attached

I will start a thread for this board after I test it and verify the PCB is fully functional.

Hope you like them!

Microcontrolled

These look awesome! Nice job!

Where did you get them made?

Bill Henning

Thank you!

In Asia... (sorry wifey would kill me if I revealed the pcb house)

Microcontrolled wrote: »

These look awesome! Nice job!

Where did you get them made?

mindrobots

This is getting exciting with all the new boards to play with. I like the dip/smt mounting for the 8 pin packages - nice touch!

evanh

Slightly hijacked thread ... I'm in the market for a simple prototyping board that just has VGA, to make a large readout from, and preferably can be cleanly modified to take a 30 volt dc supply.

I had a look through the stickies and the Parallax shop but I didn't find any big list of boards at all. I was expecting a huge pile to wade through, I guess I'm not looking in the right places yet ... suggestions welcome.


Evan

Bill Henning

check your PM

idbruce

@Bill

Very nice looking boards. I hope you do well with them. I can't wait to see them finished. Please excuse the following plug

@evanh

I'm in the market for a simple prototyping board that just has VGA, to make a large readout from, and preferably can be cleanly modified to take a 30 volt dc supply.

I believe you should then show your support for this thread

http://forums.parallax.com/showthread.php?133037-Dear-Mr.-Parallax-High-Voltage-Propeller-Proto-Board

Or perhaps you can gain an idea or two to make the board you want. There are some tips to make higher voltage capable boards in that thread.

evanh

Thanks guys. Bill, good ideas, I'll have a better look.

LD, lol, first reply from Leon made me laugh. He immediately misunderstood what you meant by industrial. There is many many situations where a quick hack is the order of the day. Plenty of stuff is built in-house, or via local contractors, because it's so simple or it's bigger than one machine.

This is also why PLCs still rule the industrial world. They've always been an open design so the end users get to make their own edits and re-engineering.

evanh

Just started reading a little more of that thread ... wow, there was a lot of negativity and narrow sighted excuses there!

It almost seemed like they were scared you were going to force them all to pay extra on every board they purchased.

Bill Henning

idbruce:

Thank you!

evanh:

You are welcome!

Bill Henning

Updates:

SD-uSD:

built one - now I just have to test it

qsBridge:

built two - have found one error so far - P26 & P27 are in the wrong order on the EXP4 10 pin connector

QuickLMM+:

built one - need to finish my driver so I can test it...

softcon

When you get them all ironed out, I think I'd like to get a board #2, as I already have 2 qs boards, and will likely have more before the year is done.

Bill Henning

Thanks softcon, I will be posting a separate thread for qsBridge when I have the production run in hand.

So far it is looking pretty good, I have only found one error on the prototype (P26 & P27 are in the wrong order on the "EXP4" 10 pin connector); EXP1, EXP2 and EXP3 are verified; I still have to check the optional external power / 3v3 regulator.

softcon wrote: »

When you get them all ironed out, I think I'd like to get a board #2, as I already have 2 qs boards, and will likely have more before the year is done.

Bill Henning

SD-uSD is verified!

As soon as my bulk order of full size SD sockets arrive I will be making them available on my site.

I tried qsBridge with the SD-uSD adapter on EXP3 (P16-P23); it was kind of neat to see the SD card activity on the QuickStart's LED's.

Cluso99

Nice work Bill & Sapieha

Bill Henning

Thanks Cluso!

Bill Henning

I thought some of you may be interested in an update.

1) SDuSD

(click on the image for a larger image)

The attached photo shows one of my QuickStart's with a qsBridge sporting TWO SD-uSD dual memory card adapters.

Yep, this means that with some changes to fsrw, it would be possible to have 128GB (using four 32GB cards) on a prop!

I've known for a while that SD-uSD works, the problem has been the price of the 3M SD sockets - which are not only hard to find, but are expensive as well. I do have a small stock of them for the glue-adverse

Fortunately I've found a better SD socket that works, however for now it will need a bit of Crazy Glue (or hot glue) to keep the SD card tightly attached to the printed circuit board, as the original PCB layout did not have mounting pads where the new sockets expect them. I still have some queries out for better prices on the 3M connectors - I hope to get responses by friday.

I expect to release SD-uSD full kits on Monday, pricing to be announced at that time.

2) qsBridge

As you can see from the photo for SD-uSD I am using qsBridge in-house

I have found one error, P26 and P27 were accidentally reversed on the EXP4 10 pin connector. This is not really a big issue, as P28-P31 are dedicated to I2C and RX/TX, so it is very unlikely anyone would populate or use EXP4.

I will be making a revision 1.1 of qsBridge that corrects the error.

3) QuickLMM

I started working on the driver... then was distracted by other business. I have resumed working on it.

In anticipation of Propeller 2, I have purchased two logic analyzers - one running at up to 200Msps and one at 500Msps - which will help me with debugging all sorts of interesting future producs :-)

They should arrive soon.

So far, I've found two minor errors with the PCB, so there will be a rev 1.1 board soon.

(See Post#1 in this thread for descriptions of these boards)

Bill Henning

Sorry guys, due to the holidays and business matters, I am running behind in bringing out SD-uSD, qsBridge and QuickLMM.

I have my logic analyzers now, and should be able to get the 8 bit driver running RSN. Meanwhile, I found two minor errata with the QuickLMM PCB, and I expect to get a new revision of the PCB's around the end of February.

David Betz

Bill Henning wrote: »

Sorry guys, due to the holidays and business matters, I am running behind in bringing out SD-uSD, qsBridge and QuickLMM.

I have my logic analyzers now, and should be able to get the 8 bit driver running RSN. Meanwhile, I found two minor errata with the QuickLMM PCB, and I expect to get a new revision of the PCB's around the end of February.

Are you writing a PropGCC cache driver for your QuickLMM board? How is it going?

Bill Henning

Hi David,

Yep, I am working on a PropGCC cache driver for QuickLMM ... initially I am working on the ram driver (and making sure it works well), then I will add support for flash chips. I designed the board specifically for PropGCC & Catalina. I also intend to patch the SD driver to support the uSD slot on the board.

As you might suspect, Chip's DE0-Nano prop2 emulation is responsible for a lot of the delay

Right now my plan is to map driver addresses 0x00000000-0x7fffffff as ram, and 0x80000000-0xffffffff as flash, this allows me to use just two instructions to determine if the access is for ram or flash.

When I last worked on the drivers, I had weird problems in 8 bit mode, however the 4 bit mode ram test (not cache) drivers worked fine.

My best guess was that it was a timing issue, so I ordered a couple of hardware logic analyzers (200Msps and 500Msps) to be able to isolate the issue.

David Betz wrote: »

Are you writing a PropGCC cache driver for your QuickLMM board? How is it going?

David Betz

Bill Henning wrote: »

Hi David,

Yep, I am working on a PropGCC cache driver for QuickLMM ... initially I am working on the ram driver (and making sure it works well), then I will add support for flash chips. I designed the board specifically for PropGCC & Catalina. I also intend to patch the SD driver to support the uSD slot on the board.

As you might suspect, Chip's DE0-Nano prop2 emulation is responsible for a lot of the delay

Right now my plan is to map driver addresses 0x00000000-0x7fffffff as ram, and 0x80000000-0xffffffff as flash, this allows me to use just two instructions to determine if the access is for ram or flash.

When I last worked on the drivers, I had weird problems in 8 bit mode, however the 4 bit mode ram test (not cache) drivers worked fine.

My best guess was that it was a timing issue, so I ordered a couple of hardware logic analyzers (200Msps and 500Msps) to be able to isolate the issue.

I'm glad to hear you're working on a PropGCC cache driver for your QuickLMM board. Let me know if you need any help. You should be able to get the SD card working just by changing pin definitions in a .cfg file for your board. What kind of CS mechanism do you use?

Also, I don't think it will work to map RAM starting at 0x00000000. PropGCC assumes that hub memory starts at address 0x0.

Bill Henning

Thanks David, and will let you know... you have always been very helpful.

I am using a small variation on the PropCade chip select mechanism, so that change should be easy - and will incidentally verify that PropCade SD support will work.

Thanks for the heads-up re/ PropGCC usage, I'll have to check out some of the current cache drivers to see what the current usage is once I have the "raw" access routines passing all of my memory tests for 8 bit ram mode.

My plan is:

1) get the raw ram routines running (slow version) passing all my tests
2) wrap them up as a cache driver
3) optimize the raw routines like crazy for speed
4) add flash support

I will probably insert a 2.5 above, adding SPI ram support to the driver for PropCade as well.

David Betz wrote: »

I'm glad to hear you're working on a PropGCC cache driver for your QuickLMM board. Let me know if you need any help. You should be able to get the SD card working just by changing pin definitions in a .cfg file for your board. What kind of CS mechanism do you use?

Also, I don't think it will work to map RAM starting at 0x00000000. PropGCC assumes that hub memory starts at address 0x0.

Wossname

I really like that hybrid footprint for your DIP/SOIC component. That's a really slick solution and doesn't waste any real estate either. Cool idea :thumb:

David Betz

Bill Henning wrote: »

Thanks David, and will let you know... you have always been very helpful.

I am using a small variation on the PropCade chip select mechanism, so that change should be easy - and will incidentally verify that PropCade SD support will work.

Thanks for the heads-up re/ PropGCC usage, I'll have to check out some of the current cache drivers to see what the current usage is once I have the "raw" access routines passing all of my memory tests for 8 bit ram mode.

My plan is:

1) get the raw ram routines running (slow version) passing all my tests
2) wrap them up as a cache driver
3) optimize the raw routines like crazy for speed
4) add flash support

I will probably insert a 2.5 above, adding SPI ram support to the driver for PropCade as well.

Essentially, the XMM kernel calls the cache driver for any address above 0x20000000. You can map RAM vs. flash any way you want but all of the external addresses have to be above 0x20000000. At least that's the way it's setup now.

Edit: Sorry, the address is 0x20000000 not 0x40000000 like my original post indicated.

Bill Henning

Thank you - Sapieha did a fantastic job making that combined TDFN/SOIC/DIP 8 pin pattern.

Wossname wrote: »

I really like that hybrid footprint for your DIP/SOIC component. That's a really slick solution and doesn't waste any real estate either. Cool idea :thumb:

Bill Henning

[moved here from David's thread]

Hi Ray,

I expect to get fully functional QuickLMM+'s in stock in early to mid March - my PCB manufacturers are shut down for "Chinese Spring Festival" (Chinese New Year holidays) until February 18th, and it normally takes two weeks for me to get a production batch of boards. So far, I've found two minor issues - a trace that needs cutting on the prototype, and needing to resize the PCB a bit to make the second set of mounting holes line up.

The good news is that gives me plenty of time to get the PropGCC XMM driver running for the QSPI ram chips

QuickLMM+ stacks on top of the QuickStart, so you can access Vin on the QS header - and if you power the QuickStart from USB, Vin would be 5V; otherwise you can use an external 5V supply

Rsadeika wrote: »

So, when can we expect a memory add-on for the QS? Another thing that I would like to see is access to 5V, the modules that I work with require 5V, but no way to access that with any add-on boards that Parallax sells.

Ray