Pi Pico[W] Peripheral Expansion Box Side Port Device

[JasonACT]

When I got my eBay package earlier in the week, there were a few surprises, one of which was a piece of white plastic which I swear I've seen before but couldn't remember what it was for...  I nearly chucked it in the bin because it didn't seem to fit anything, until I did this tonight...

 

 

Another mystery is history.

[JasonACT]

6 hours ago, JasonACT said:

best case scenario of 192nS

Make that 384nS, I had forgotten to factor in the PSRAM is running at half the speed of the Pico, oops.

[+chue]

16 hours ago, chue said:

Will test 250 Mhz and 1 NOP next.

That worked too.  @JasonACT if you would like to see tests using other values, let me know.

[JasonACT]

1 hour ago, chue said:

That worked too.  @JasonACT if you would like to see tests using other values, let me know.

That's great news, thanks, it should make it simple for anyone making one of these.  If it doesn't work right away, just bump it up one NOP and 2 MHz. 

[RickyDean]

Here will be the production look of the modified (uson pads added) PPEB.

[wiloson]

On 4/26/2024 at 6:57 AM, JasonACT said:

New PPEB firmwares to test, mainly for @wiloson but also @chue and @RickyDean too.

 

These are the 2MB SAMS and 8MB SAMS versions (8MB only for use on the v3 board, or piggybacked mod) but both are running at 250MHz and one NOP.

 

You can now change the settings for these in autoload.cfg to increase them using a line saying "MHZ=252" up to 280, and "NOPS=2" up to 9.

 

I've also cleaned up the speech PWM to use 32KHz refreshes instead of 16KHz (I'm far too old to hear 16KHz, but there was some 16KHz noise in test-recordings I had made in Audacity, and they are gone now - your pets and/or grandkids will be grateful).

PPEB2.zip 761.5 kB · 3 downloads

Wow, thank you for this.   I found that 3 NOP worked best for me with it failing more at 2 and 4.    I had to increase the MHz to 255MHz before I was able to pass all the memory tests.     I will probably try upping the MHz to see when it starts to fail again and choose something in the middle and then reduce the NOPs to 2 or 1 to see if it will work.    

[JasonACT]

4 hours ago, wiloson said:

Wow, thank you for this.   I found that 3 NOP worked best for me with it failing more at 2 and 4.    I had to increase the MHz to 255MHz before I was able to pass all the memory tests.     I will probably try upping the MHz to see when it starts to fail again and choose something in the middle and then reduce the NOPs to 2 or 1 to see if it will work.  

FYI - It ignores odd numbers for the MHZ value, so if you specified 255 then it was actually running at 250.

 

On my NTSC TI I can specify 9 NOPs and 252MHz and pass all tests, but one of my PAL machine needed 258MHz for that to be stable.

 

Oh, and I don't think the Pico likes 274 & 278 - that's what my notes say, the I/O lines stop functioning for some reason - so the next build I release will ignore those two values.  Maybe I should be using the lowest working even number when something unsupported is specified, I might make that change too, instead of defaulting to 250.

 

I'm glad it's working for you.

[Gary from OPA]

I am still waiting for the extra bare board I grabbed from @RickyDean to arrive in the Great White North (called Canada), but in the meantime I met up with a local Canuck @wiloson and picked up one of his extra units he nicely built recently, so soon I will have two units to play with and test out, and will be trying out the one I grabbed today, sometime tomorrow and report back my findings on how it works.

 

I have two /4a consoles and one QI one, so a good set of systems to test them on.

 

 

[JasonACT]

On 4/29/2024 at 3:51 PM, JasonACT said:

Oh, and I don't think the Pico likes 274 & 278 - that's what my notes say, the I/O lines stop functioning for some reason - so the next build I release will ignore those two values.  Maybe I should be using the lowest working even number when something unsupported is specified, I might make that change too, instead of defaulting to 250.

OK, with any luck I've got this right now.  Also for the number of NOPS, I'll pick 9 if more are specified.

 

Find attached my experimental firmware(s) that allows you to plug in a USB memory stick to a USB-HUB, which overrides DSK3 by default (but the autoload.cfg can have a setting "USBD=" 1 through to 9 if you don't like DSK3).  You can't "change directory" on this temporary device, you are always in the root directory, but I think most of my DSR functions should still work (rename, mkdir etc.)

 

Makes for a nice way to get a file off the running TI onto a memory stick you can read on your PC.  Sadly, no disk-sector access to this device other than the direct file-sector access that's pretty easy to implement.  I've tested this with an SD card in a USB converter, and the SD card was one I had formatted properly like it needs to be for use in the usual Pico SD socket...  I think that's still a requirement.  You, of course, still need an SD card in the Pico SD slot with all the DSRs etc. to boot properly.

 

Let me know if you find any bugs  2MB and 8MB versions are in the archive...

PPEB2.zip

[+chue]

16 hours ago, JasonACT said:

Find attached my experimental firmware(s) that allows you to plug in a USB memory stick to a USB-HUB

I couldn't get this working on my Pico PEB.  SAMS burn-ins failed within the first 7 pages of the test.  I started at 250 mhz and 1 NOP, incrementing mhz and NOPs by 2 and 1 respectively.  I stopped at 266 mhz and 9 NOPs.  No luck at all running.

 

I went back to the prior firmware and all is well.

 

edit: testing was done on the silver and black TI only. I didn’t test on the other consoles.

[JasonACT]

1 hour ago, chue said:

I couldn't get this working on my Pico PEB.  SAMS burn-ins failed within the first 7 pages of the test.  I started at 250 mhz and 1 NOP, incrementing mhz and NOPs by 2 and 1 respectively.  I stopped at 266 mhz and 9 NOPs.  No luck at all running.

 

I went back to the prior firmware and all is well.

 

edit: testing was done on the silver and black TI only. I didn’t test on the other consoles.

I've reviewed my code changes between the two releases, nothing sticks out as affecting the timing sensitive CPU1 code.  All my changes were in the CPU0 area which shouldn't make a difference.  Is it possible there was a bad connection on the side-port while you were running the tests and removing/re-inserting the device to re-program the old firmware has rectified it?  I've had that before myself.

[+chue]

Just now I reflashed the USB firmware and ran the test again and same result.  Fails on the first 7 pages of the test.  Note I always flash with the Pico PEB disconnected from the TI, which requires removing and reinserting the Pico.

 

So I tried on my beige (non-QI) TI and same result there, failure.

 

I pulled out a different Pico PEB and loaded the USB firmware on that.  Fails on both TIs.

 

I changed to the prior version of the firmware on both Pico PEBs and the SAMS tests pass on both TIs.

 

If you would like to see another test, let know.  For now the prior firmware is great for me, so I will use that (outside of testing of course).

 

Edit: all tests here were done at 250 mhz, 1 NOP

[JasonACT]

@chue Here's an archive with 4 files named B..E - which has my progressive changes leading up to the USB stick firmware - I'd call that one 'F' that's failing, but if you can tell me which one fails first for you, it might help me track down which of my edits is responsible.  I hope it's 'B' because I can do something about that one.

PPEB2_Tests.zip

[+chue]

Indeed, it fails on "B".  Testing was done on a single Pico on the silver and black TI.  I tried 250 mhz + 1 NOP, 266 mhz + 9 NOPs.

 

[JasonACT]

13 hours ago, chue said:

Indeed, it fails on "B".

OK, that was the one and only edit in the CPU1 code, it's been reverted in these firmwares.

PPEB2.zip

[+chue]

Looks good!  SAMS burn-in passes on your latest firmware.  The memory test ran 3 times @ 250 mhz 1 NOP.  Tested on one Pico and the black and silver TI.

[JasonACT]

I've enabled exFAT formatted drives in this version of the firmware (you can still use FAT32 too).

 

This allows larger drives to be plugged in, but use a powered USB hub if/when appropriate!

PPEB2.zip

[Mudmin]

Man. I'm excited about this.  I'm having a bit of trouble following the thread through.  Are the gerbers in Speech_gerberv2 the latest ones for this board?

[+chue]

There's a V3, which allows 8 MB SAMS memory.  It was "unverified" at the time of the post.  I don't know if anyone has built one yet.

 

See here for more information: 

 

[+chue]

14 hours ago, JasonACT said:

I've enabled exFAT formatted drives in this version of the firmware (you can still use FAT32 too).

 

This allows larger drives to be plugged in, but use a powered USB hub if/when appropriate!

SAMS burn-in passes on this, so I will be using it going forward.  Tested on one pico and one TI.

[RickyDean]

27 minutes ago, chue said:

There's a V3, which allows 8 MB SAMS memory.  It was "unverified" at the time of the post.  I don't know if anyone has built one yet.

 

See here for more information: 

 

I'm waiting for my boards to come in, they are currently in Savannah, 80 miles from me, but DHL has a hold for some reason. I believe it's weather related though.

Edited Friday at 07:16 PM by RickyDean
spelling

[Mudmin]

1 hour ago, chue said:

There's a V3, which allows 8 MB SAMS memory.  It was "unverified" at the time of the post.  I don't know if anyone has built one yet.

 

See here for more information: 

Thank you. That's the file I had.  The gerbers were just in a zip file that said something about speech, so I wanted to make absolutely sure. I appreciate you sharing your hard work.

 

T

[JasonACT]

Now with a built-in Digi-Port (don't use this firmware with a TI RS232 card - It'll clash).

PPEB2.zip

[Artoj]

5 hours ago, JasonACT said:

Now with a built-in Digi-Port (don't use this firmware with a TI RS232 card - It'll clash).

Great work Jason, fast to act when you got the knack. your card will plug into my MPEB extension board. A definite must have card, regards Arto.  

[+chue]

16 hours ago, JasonACT said:

Now with a built-in Digi-Port (don't use this firmware with a TI RS232 card - It'll clash).

SAMS burn-in passed.
 

Just one question, what is a Digi-Port? My google-fu didn’t give good results.