Falcon refuses to boot without diag cart

Problems with your machine in general.
mikro
Posts: 474
Joined: Mon Aug 28, 2017 11:22 pm
Location: Kosice, Slovakia
Contact:

Falcon refuses to boot without diag cart

Post by mikro »

Hey guys,

I could you some of your vast experience here :)

Back in Oct 2018 I bought a capacitor kit for my Falcon. It was lying and moving around with me for the past 3.5 years until I finally decided to put it in action (I've also wanted to finish one huge point in my TODO list, CosmosEx internal installation&guide which requires me to take out the PCB).

So fine, I went ahead and recapped the beast. I had to wait for a couple of weeks for the last remaining caps to arrive because the package turned out to be incomplete. So I could verify the result only now, when soldered the last pieces. I powered the Falcon up and saw the Atari logo, yay. Switched it off and next time, it wouldn't even initialise video (i.e. produce any kind of sync signal, not even the infamous "out of range"). Better said, it would like in 1 out of 20 retries (pressing reset didn't have any effect on it). Tried with diag cart, same. Not in a happy mood I rechecked the board the next day and to my relief, I found out that one cap, C181, was soldered the wrong way (yes, what a shame!) The cap was all right (it was actually rated for 100V, so no worries here), verified with multimeter, still 1uF. Swapped it, soldered back, powered the machine up and ... success! It would boot.

Switched on, off, tried the diag cart... everything worked. Then I connected my IDE DOM and ... it wouldn't start anymore (same symptoms -- not even video sync). Pretty pissed I inserted the diag cart again and ... it works. And it's like that to this moment. I insert the cart - everything (video, audio, keyboard, all internal tests, RAM, IDE, ...) works. I remove it -- no boot at all.

So far I have tried:
  • verified all values and orientations of the caps
  • cleaned the PCB on both sides as much as I could to remove any random metallic parts
  • pushed all ICs, cleaned all sockets
  • resoldered all the caps (I even replaced the opposite one with a brand new, just to be sure)
  • stripped it completely naked except (even without FPU) except RAM and NVRAM
At one or two occasions I had a short stroke of luck and it would boot even without diag cart. Switch off, swich on - no boot. So obviously something is not working properly, most likely some broken/short-circuit connection but how on earth should I diagnose it?

Looking into TOS 4.04 boot in Hatari, there is not that much happening before the video init kicks in -- except the ROM it accesses the memory controller and PSG. ROM obviously works (otherwise the cart wouldn't boot) so I don't understand why diag cart doesn't show any errors for PSG/memory then?

I have one thing to verify -- that "short strokes of luck" maybe weren't totally random. My feeling was that it always happened after a longer period of time, i.e. like when a certain cap would discharge. (this would be also supported by my finding of the opposite cap -- the Falcon was lying on the desk overnight, i.e. I would switch it on after a longer pause). This could maybe (?) imply that over that 3.5 years some of the caps went bad but then again, why that bloody cart makes it all work like as if nothing is wrong?
User avatar
stephen_usher
Posts: 5580
Joined: Mon Nov 13, 2017 7:19 pm
Location: Oxford, UK.
Contact:

Re: Falcon refuses to boot without diag cart

Post by stephen_usher »

I'd check the pull-ups etc. If loading the address and data buses changes things maybe the levels are marginal and they float to a different (higher) voltage, above the TTL switch level, if the ROM is plugged in.
Intro retro computers since before they were retro...
ZX81->Spectrum->Memotech MTX->Sinclair QL->520STM->BBC Micro->TT030->PCs & Sun Workstations.
Added code to the MiNT kernel (still there the last time I checked) + put together MiNTOS.
Collection now with added Macs, Amigas, Suns and Acorns.
User avatar
rubber_jonnie
Site Admin
Site Admin
Posts: 10472
Joined: Thu Aug 17, 2017 7:40 pm
Location: Essex
Contact:

Re: Falcon refuses to boot without diag cart

Post by rubber_jonnie »

I'm not 100% certain about the Falcon, but on the ST TOS detects the diag cart very early on in the boot up process, so it may be the same for the Falcon. If so then I'd suspect a rom problem, possibly address lines.

Have you used the diag cart to test your ROM at all? It may flag something up.

Also, if you can connect your serial port at boot time with the diag cart, it may show something that you can't see normally during boot. It may not, but it's worth a shot.

Worthwhile checking all the address/data lines have continuity, because if you have handled the board a lot, it is entirely possible something has failed mechanically, like a fractured track or via.

Again, not 100% as I don't have a Falcon, but check any socketed ICs and PLCC sockets in particular (There may only be the ROM in a socket TBH), and clean chip legs, use contact cleaner on the sockets and make sure that any PLCC sockets have the contacts properly in contact with the ICs in them.

@stephen_usher's suggestions are definitely worth following too.

I recently repaired an STE for a forum member and one the physical damage was made good, it still wouldn't boot and it was the CPU socket not making contact properly. Likewise when I did get it booting, having the Blitter enabled made it crash and the Blitter socket had a similar contact problem.
Collector of many retro things!
800XL and 65XE both with Ultimate1MB,VBXL/XE & PokeyMax, SIDE3, SDrive Max, 2x 1010 cassette, 2x 1050 one with Happy mod, 3x 2600 Jr, 7800 and Lynx II
Approx 20 STs, including a 520 STM, 520 STFMs, 3x Mega ST, MSTE & 2x 32 Mhz boosted STEs
Plus the rest, totalling around 50 machines including a QL, 3x BBC Model B, Electron, Spectrums, ZX81 etc...
mikro
Posts: 474
Joined: Mon Aug 28, 2017 11:22 pm
Location: Kosice, Slovakia
Contact:

Re: Falcon refuses to boot without diag cart

Post by mikro »

rubber_jonnie wrote: Fri Mar 25, 2022 2:52 pm I'm not 100% certain about the Falcon, but on the ST TOS detects the diag cart very early on in the boot up process, so it may be the same for the Falcon.
Yes, it is the same in Falcon. However that detection is done on a software level, i.e. the ROM boots, it reads a long word from the cart area and if a magic value is found, it jumps ever there. So ROM, at least as far as this check goes, must work.

Thanks for the suggestions, I'm gonna dig deeper then... :)
Steve
Posts: 2570
Joined: Fri Sep 15, 2017 11:49 am

Re: Falcon refuses to boot without diag cart

Post by Steve »

@mikro Very odd!

First, make sure you have C24 the right way around: https://www.exxosforum.co.uk/forum/viewt ... =17&t=1698
As C24 is incorrectly installed from factory, it's important to install new capacitors in the correct way otherwise (unlike the wangcap) they'll bulge & leak (this happened to me, which is how it was discovered)

I would also ask: Is the machine stock, does it have any other modifications that may have been knocked/dislodged during the re-cap, a clock patch? nvram? etc.

I would then air-blow the board to maybe dislodge any solder 'bits' that may have accidentally got loose. Maybe pour some IPA over IC's and go over lightly with a carbon brush.

Maybe check the resistance of the capacitors to make sure one hasn't shorted, perhaps one in the batch could have been faulty...
User avatar
rubber_jonnie
Site Admin
Site Admin
Posts: 10472
Joined: Thu Aug 17, 2017 7:40 pm
Location: Essex
Contact:

Re: Falcon refuses to boot without diag cart

Post by rubber_jonnie »

mikro wrote: Fri Mar 25, 2022 3:02 pm
rubber_jonnie wrote: Fri Mar 25, 2022 2:52 pm I'm not 100% certain about the Falcon, but on the ST TOS detects the diag cart very early on in the boot up process, so it may be the same for the Falcon.
Yes, it is the same in Falcon. However that detection is done on a software level, i.e. the ROM boots, it reads a long word from the cart area and if a magic value is found, it jumps ever there. So ROM, at least as far as this check goes, must work.

Thanks for the suggestions, I'm gonna dig deeper then... :)
That would be my assumption too, that the ROM is at least getting part way started. Good luck!
Collector of many retro things!
800XL and 65XE both with Ultimate1MB,VBXL/XE & PokeyMax, SIDE3, SDrive Max, 2x 1010 cassette, 2x 1050 one with Happy mod, 3x 2600 Jr, 7800 and Lynx II
Approx 20 STs, including a 520 STM, 520 STFMs, 3x Mega ST, MSTE & 2x 32 Mhz boosted STEs
Plus the rest, totalling around 50 machines including a QL, 3x BBC Model B, Electron, Spectrums, ZX81 etc...
User avatar
DoG
Posts: 1125
Joined: Sat Apr 07, 2018 12:26 pm

Re: Falcon refuses to boot without diag cart

Post by DoG »

...and it is not stuck in reset?
User avatar
Badwolf
Posts: 2231
Joined: Tue Nov 19, 2019 12:09 pm

Re: Falcon refuses to boot without diag cart

Post by Badwolf »

It's not the invalid NVRAM problem is it? Not sure how that manifests, but the diag cart probably won't care.

Software for the reset is available here: https://www.exxosforum.co.uk/forum/viewt ... ?f=25&t=30

BW
DFB1 Open source 50MHz 030 and TT-RAM accelerator for the Falcon
DSTB1 Open source 16Mhz 68k and AltRAM accelerator for the ST
Smalliermouse ST-optimised USB mouse adapter based on SmallyMouse2
FrontBench The Frontier: Elite 2 intro as a benchmark
mikro
Posts: 474
Joined: Mon Aug 28, 2017 11:22 pm
Location: Kosice, Slovakia
Contact:

Re: Falcon refuses to boot without diag cart

Post by mikro »

Hooray, I can report success although I can't really take big credit for that.

First I was trying to think hard about its symptoms: if I ruled out any mechanical damage (I was quite careful by desoldering), the whole issue just didn't make sense. @Badwolf is correct that the diag cart doesn't care about NVRAM but at the same time, it also resets its content (and I did see the effect of it during one of the rare boots) so this could have been safely ruled out as well.

Reset stuck as per @DoG's suggestion also wasn't possible (the cart can't ignore reset signal + there's a chance the normal boot would get as far as video init even with a repeating reset signal).

C24 didn't have any effect on the issue, most likely because it is not used when connected to VGA. However I had one false hooray here, more on this later. Checked also clockpatch (really just the basic one here), nope, everything as it should be.

I couldn't detect any continuity errors nor see anything wrong so I had two options: painfully go IC by IC, most likely with someone's help with LA or take @Steve's suggestion and really make sure that every millimeter of the PCB has been cleaned. And since I have been experimenting with this new method... what can I lose, right? (ok, actually quite a lot, I could have ended up with a Falcon with more than one issue to solve :D). My thinking was that if there's some unreliable, hard-to-detect problem, that wash will either fix it or break it for good. And luckily for me, it did freakin' fixed it!!!

I verified it from numerous angles: first, it boots every time now. After a minute, 10 minutes, one hour, always. Second, it boots in CT60 mode (that didn't happen at all before). And third, it boots from IDE, I could run DSP demos, FPU demos, memory / disk io hungry demos... all fine. And as a bonus, my Falcon now looks like taken from production line. So... wow.

And the promised false alarm: I hadn't noticed anything strange with the wrongly oriented C24 so far. But after this huge success I wanted to switch to RGB, just to be sure. And what do I see? Not syncing RGB input on my LCD. Before that I would get a "signal error" message but I would see the picture and all. And now just slight noise on a black screen. Hmm... so I go ahead, switch the C24 and ... it didn't help. ;-) I take another cable and connect it to my CRT... terrible, terrible picture. So I am like OMG, it's true! And I blew the C24 in the process! Painstakingly I find another 470uF capacitor, replace it and ... still no change. Wow. Then I remember that I have yet another RGB cable -- I use that one and ... finally, a nice image! Put the previous C24 back, still ok. And in that moment I realise that thanks to the NVRAM reset, Falcon boots at 60 Hz in RGB. Changed it back to 50, connected to the LCD again and of course it worked. ;-) Apparently, 60 Hz/15 kHz is too much for it while 50 Hz is still tolerable. Moral story of this is: always double check stuff, ideally with different equipment. But I kept the C24 as per Exxos/Steve's suggestion, surely I don't want to hear an explosion in a few days.

So what can I say... next time I'll be paying far more attention to my desoldering and environment, to avoid any small particles and blobs floating around.

Thanks everyone for brainstorming, it is definitely helping for not giving up! :)
Steve
Posts: 2570
Joined: Fri Sep 15, 2017 11:49 am

Re: Falcon refuses to boot without diag cart

Post by Steve »

Yay! Glad it's sorted!! :)
Post Reply

Return to “HARDWARE ISSUES”