exxos's Atari, Amiga & retro forum

I haven't had any more time yet to play with this switching setup, although I do notice that the mux wiring that I built is identical to the mux wiring in c't's configuration. The mux IC is very slightly different, but I would imagine inconsequentially.

However, the c't switcher has this at the top:


Does anyone know what is the function of the connected GAL on the PAK board?

A lesson was learned tonight: Do not accidentally connect BR to +5V when your CPLD is a 3.3V device. :lol:

So, I toasted the Xilinx chip on the TF536. Luckily I bought one spare when buying the chips for the TF cards I'm going to build up myself, so I pulled off the dead one and stuck on a new one, no problems there.

But I think I maybe don't have any hardware that can program it. I've got a TL866II PLUS, a PICKit 3, and a USB Blaster. Can any of these program the Xilinx CPLD, or do I need to acquire yet another programmer?

I thought those chips were 5V tolerant ?

exxos wrote: 03 Aug 2020 20:18 I thought those chips were 5V tolerant ?

I saw something on the datasheet about being 5V tolerant for programming. Don't know if it is tolerant on the whole bus. (I was wondering about that, though, since obviously it's sitting on a 5V bus, so it seemed strange that putting 5V on BR would be a problem.)

In any case, I did break it. I was using the Stacy internal DMA bus for grounding BR, and mistakenly put it on a +5V pin, which is right next to GND, and the TF quit. I noticed after that that the regulator was running really hot and producing only around 1V on its output, and 3V3 and GND were shorted. I checked the regulator and the accompanying caps and they were fine, so CPLD was the next thing to test. The short went away when I pulled off the old CPLD, and with the new CPLD on, the regulator is back to producing 3.3V.

@derkom you can program them with a raspberry pi but not tried that myself. There's a guide on here somewhere I think?

P.s. hopefully the ram has survived too.

PhilC wrote: 03 Aug 2020 20:36 @derkom you can program them with a raspberry pi but not tried that myself. There's a guide on here somewhere I think?

I see a couple of those projects (OpenOCD, xc3sprog), but it looks like I'll still need a cable I don't own.

P.s. hopefully the ram has survived too.

Indeed, although I do have the chips for my other boards if need be.

derkom wrote: 03 Aug 2020 08:36 Does anyone know what is the function of the connected GAL on the PAK board?

It's described in the c't article that was posted: it drives /AS, /LDS, /UDS and /VMA on the PAK. The additional gate makes these output go high-Z.

czietz wrote: 03 Aug 2020 20:57 It's described in the c't article that was posted: it drives /AS, /LDS, /UDS and /VMA on the PAK. The additional gate makes these output go high-Z.

Ah, thanks. I missed that in the article.

I see that I can get a proper Xilinx programmer for not too much, but of course that won't help me tonight. :D Probably a good idea to have one anyway, for firmware updates, etc.

I'm running right up to my summer holiday, so I'm afraid all of this is about to go on hold for a couple of weeks anyway, and then I'll get ahold of the Xilinx programmer and pick up where I left off.

Okay this turned out to be a lot easier than I thought, once I was directed to the right information. I've got the firmware on the new CPLD and the TF once again boots. I can't easily check the RAM at the moment, but will figure that part out eventually.