Pokeymax v4 bring up thread

[_The Doctor__]

if there is a bonding point on both a thinner coating is used for the larger area. It serves as a plane, and a sink. Stability.

[foft]

I think you can just tin the pads with leaded then solder the bga right on. Or even just solder it as is. That is what I did on my last one. Just be careful not to damage the solder mask at all. A little flux is a good idea, not too much or it moves the solder balls!

 

I just bought a stencil for my next ones, going to give that way a try. I have an infra-red reflow oven.

 

I also designed/ordered a little toslink and 3.5mm socket breakout board.

[woj]

Right, so I am now facing the possibility that my cheap Chinese USB Blaster clone is too cheap 😕 Working with the freshly soldered Ultimate Cart, after many hurdles with permissions and such I got the Quartus software to say "Hardware not attached". The board is powered, I tried different connecting sequences, and check with the multi meter for shorts and that the JTAG pins of the FPGA are connected to the header, and that the FPGA is getting 3.3V. So I am running out of options here. Unfortunately, I do not have either another confirmed USB blaster to check the board, or another JTAG board with correct connector to check the cable. Both U1MB and SIDE3 have the much smaller JTAG headers, and truth be told, I am afraid to connect them anyhow. Of course, the soldering could have gone wrong, but any examination so far suggests it did not. I did find posts of people having similar problems with those knock off cables. I need to check if the lab at my work have the real one, but I will not get access until next Monday at least even if it is there... This is a sort of necessary thing to have a go at PokeyMax...

[_The Doctor__]

You have to use the older drivers for the blaster, the newer one operates at the wrong rates.

Edited May 8 by _The Doctor__

[madness77]

41 minutes ago, woj said:

Right, so I am now facing the possibility that my cheap Chinese USB Blaster clone is too cheap 😕 [...]

Been there, done that: 

 

I threw the Chinese blaster in the trash and bought a Polish one from KAMAMI, all problems with chip detection/programming disappeared.

https://kamami.pl/programatory/561847-zl32prg-kamami-usb-blaster-pro-programator-usb-dla-ukladow-pld-firmy-altera-zgodny-z-usb-blaster.html

Edited May 8 by madness77

[ijor]

44 minutes ago, woj said:

Right, so I am now facing the possibility that my cheap Chinese USB Blaster clone is too cheap

I think it is a very bad idea to use those cheapest USB Blaster clones. There are decent clones by "known" chinese brands that cost just a couple of extra dollars. But in your case, with the amount of work you invested, IMHO, you should get an original USB Blaster. Note that the original Blaster is sold both by Altera/Intel and Terasic. The Terasic one is much cheaper. And, of course, there is the Usb Blaster II that it's much more expensive but it is not really needed unless you work with huge, and ridiculously expensive, FPGAs that otherwise take too long to program.

 

Edit: Note that the clones are really Usb Blaster I clones (despite that they claim they are Usb Blaster II clones).

Edited May 8 by ijor

[_The Doctor__]

My cursor is in the picture, ignore it

this is the driver you need that works with damn near all of the Altera blasters be they chinese knockoff or not.

[woj]

20 minutes ago, _The Doctor__ said:

My cursor is in the picture, ignore it

this is the driver you need that works with damn near all of the Altera blasters be they chinese knockoff or not.

That's a good hint, and it might come down to trying this, the problem is I am on Linux 🤔

[foft]

Do you see it on lsusb?

Also do you have the udev rules?

Edited May 8 by foft

[woj]

4 minutes ago, foft said:

Do you see it on lsusb?

Also do you have the udev rules?

Yes and yes, the second one was a problem, as I could only get things going the furthest as root, but I fixed that with udev rules for the regular user. I also now tried connection everything through a USB hub, not only because one website suggested it, but also because it once solved a serious instability issue with a totally different USB device for me some years back. Still no luck. And it can of course still be poor soldering job.

[foft]

So to be clear Quartus finds the USB blaster but the blaster can't find the FPGA?

 

Can you scan the jtag chain and the fpga is found, just you can't flash? Or you can't scan the chain?

Edited May 8 by foft

[madness77]

12 minutes ago, woj said:

Yes and yes, the second one was a problem, as I could only get things going the furthest as root, but I fixed that with udev rules for the regular user. I also now tried connection everything through a USB hub, not only because one website suggested it, but also because it once solved a serious instability issue with a totally different USB device for me some years back. Still no luck. And it can of course still be poor soldering job.

Are you able to detect the connected FPGA? Like:

 

$ ./jtagconfig -d
1) USB-Blaster(Altera) [1-1.3]
  031820DD   10M08SA(.|ES)/10M08SC (IR=10)

  Captured DR after reset = (031820DD) [32]
  Captured IR after reset = (151) [10]
  Captured Bypass after reset = (0) [1]
  Captured Bypass chain = (0) [1]
  JTAG clock speed 6 MHz

 

Go see my post in the thread about Ultimate, see what errors the given commands give, maybe we can work something out.

Edited May 8 by madness77

[woj]

12 minutes ago, madness77 said:

Are you able to detect the connected FPGA? Like:

No, chip is not detected, it says hardware not connected. I will probably continue with this tomorrow, I have a long list of things to try out.

[woj]

18 minutes ago, foft said:

So to be clear Quartus finds the USB blaster but the blaster can't find the FPGA?

Correct.

18 minutes ago, foft said:

Can you scan the jtag chain and the fpga is found, just you can't flash? Or you can't scan the chain?

I can't scan the chain it seems.

[foft]

Not sure what error you are getting, I don't recognise it.

 

If I type jtagconfig without the blaster connected to USB I get this:

>jtagconfig
No JTAG hardware available

 

If I type jtagconfig with the blaster but no FPGA I get this (or connected but the board not powered):

>jtagconfig
1) USB-Blaster [1-5.3]
  Unable to read device chain - JTAG chain broken

 

If I type it with an FPGA I get something like this:

>jtagconfig
1) USB-Blaster [1-5.3]
  031820DD   10M08SA(.|ES)/10M08SC

 

So, I never get 'Hardware not connected'. In Quartus I go to Tools->Programmer, then Hardware setup and select the USB blaster. Then I can choose the blaster.

Then I can scan the chain. It tells me if the chain is broken (e.g. now connected or power off):

 

So I'm a bit confused by the error you are getting. 'hardware not connected'. I've not got that in any case here.

 

Finally this is the device I see on USB:

>lsusb | grep Altera
Bus 001 Device 055: ID 09fb:6001 Altera Blaster

 

Edited May 9 by foft

[woj]

Nothing attached to my computer:

 

woj@impreza:~/intelFPGA_lite/23.1std/qprogrammer/bin $ ./jtagconfig 
No JTAG hardware available

Plug in just the USB-blaster:

 

woj@impreza:~/intelFPGA_lite/23.1std/qprogrammer/bin $ lsusb | grep Altera
Bus 003 Device 035: ID 09fb:6001 Altera Blaster
woj@impreza:~/intelFPGA_lite/23.1std/qprogrammer/bin $ ./jtagconfig 
1) USB-Blaster variant [3-3]
  Unable to read device chain - Hardware not attached

 

Connect the blaster to a powered-up board:

 

woj@impreza:~/intelFPGA_lite/23.1std/qprogrammer/bin $ ./jtagconfig 
1) USB-Blaster [3-3]
  Unable to read device chain - Hardware not attached

 

In the last two calls to jtagconfig it spends considerable time looking for things. And the same things happens if I run everything as root, with the small difference (this is really weird), that running jtagconfig as user does not leave the jtagd process running, but does when I run it as root...? And yes, in the GUI I can see the device as I do in jtagconfig, still I just see it, it does not react.

 

I will try an older Quartus for starters. I also have a junk spare Windows laptop, last night I searched for this 2.4.16.0 Windows driver, to no avail so far (without paying to some of those driver database pages :/), but I did get a good another hint that it is in fact one that is needed to make things work (a FPGA programming course description from a reputable university).

 

Edited May 9 by woj

[foft]

Ok that does look like a broken chain.

 

BTW jtagd crashes and needs killing often for me.

 

I’m using quartus 20.1, but I guess the latest is also fine.

[woj]

I could go as far back as Quartus 17.0, this does not change anything.

[foft]

Could check on your scope that you see TCK and TDO/I toggle when you scan it. To see if its a board issue or a blaster hardware issue.

[woj]

TCK goes from high to low once on the first attempt of connection after freshly connecting the blaster to USB, then it stays low for all subsequent connection attempts. TDI and TDO stay high at all times, no sign of life there. Both with the board connected or not. Do I conclude correct that this is indeed the cable that is not functioning?

 

EDIT: I should also add, I tried connecting to the blaster with OpenOCD hoping the config file I made guessing from the information out there is correct, this did not even recognize the interface. I did not play too long with this, this is even more black magic to me. I also installed the Quartus programmer on the Windows laptop with the default drivers that it comes with, I made Windows happy with the USB blaster device, but the programmer does not even see it / list it.

 

All in all, seems like a lost case, I need to get a working cable.

Edited May 9 by woj

[foft]

I guess you could:

i) Double check the 10 pin cable is good

ii) Open the blaster and check for any stupid soldering issues etc

 

P.S. TDI should be going low. e.g. if IDCODE is sent:

https://www.intel.com/content/www/us/en/docs/programmable/683207/current/supported-jtag-instructions.html

https://medium.com/@aliaksandr.kavalchuk/diving-into-jtag-protocol-part-1-overview-fbdc428d3a16

Edited May 9 by foft

[woj]

Both things crossed my mind, so far I had time to flip the cable seeing that on some genuine products they have ends marked. But seeing high signals on the pins suggests at least those have connection. And yeah, since I received one UAV with the header soldered in front side back, and since my NOS 130XE had (a) one resistor loose, (b) the main power cap reversed, I have very low trust in production quality 🤔

[woj]

The cable is sound, the board seems sound too (measured it here and there), however, now looking at the Internet, this the cheapest of the cheap it seems, with really low chances of working, I will need a different one. 

[woj]

Had an incidental chance to pass by the office doing something else (it's holiday today), and look what I have found in my colleague-on-long-term-absence's office next door, just lying on the desk like that! Needless to say, the programming worked on the first try, now that all my device permissions and all were all set right. Now just need to get back to my Atari to see if the cart works. Nevertheless, the device I bought is junk!

[woj]

Can't edit no more - F... Yeah! I am PokeyMax 4 DIY ready! On Saturday I will make another Ultimate Cart using a slightly different soldering technique, and make sure my result is repeatable. Also need to clean my desk, I made a huge mess when fighting this