Problem with QL

Nagging hardware related question? Post here!
Post Reply
beanz
ROM Dongle
Posts: 3
Joined: Tue Sep 08, 2015 1:52 am

Problem with QL

Post by beanz »

Hi,

My QL recently stopped working properly and I'm looking for some help on troubleshooting it.

When turning on the memory test appears to pass ok but then it goes to black screen. I installed a Minerva rom figuring it was a memory problem and that the minerva rom would identify that. Well it still passes the memory test without any reported problems and displays the screen with Minerva in the bottom right hand corner after the memory test but then hangs there. If I hit the reset button maybe once in every 20 times the "press F1/F2" box on the lower left will appear but then freeze up and/or both the microdrive LEDs will light as part of that freezing.

I've had a memory problem before and fixed that after the Minerva identified which IC was the problem but again it seems to pass the memory test fine, I also tried it with a trump card and the memory test took longer because of the additional memory but still seemed to pass it ok defaulting to the "minerva" black screen (see photo attachment) but then hanging there.

Any clues from this which component(s) might be causing the problem? Thanks guys.

Image


martyn_hill
Aurora
Posts: 909
Joined: Sat Oct 25, 2014 9:53 am

Re: Problem with QL

Post by martyn_hill »

Hi Beanz

Not sure if this is expected or not with your version of Minerva, but I notice that the Minerva logo itself is all white in your picture, whereas my various Minerva versions always show the logo in red.

I also note that you have one of Vitaliy's excellent multi-ROM's installed (with the latest external ROM selector switch.)

Whilst you mention that it was problematic even before Minerva (and thus before fitting the multi-ROM), my guess is a bad ROM socket/connection, or else some rogue IC (bus-transceiver or address multiplexer) hogging the CPU bus intermittently (but after the RAM test - which suggests after some heat has built-up.)

A dodgy ZX8301 might exhibit such issues too, but I bet you've already switched that IC before.

During the memory test, does it look like the 'right' colours are presented in the tweed pattern - see the avatar for Cristian elsewhere on this forum, as a comparison if you're not sure what the tweed-pattern should look like...)

Good luck! It's always 'fun' fault-finding a QL...


beanz
ROM Dongle
Posts: 3
Joined: Tue Sep 08, 2015 1:52 am

Re: Problem with QL

Post by beanz »

martyn_hill wrote:Hi Beanz

Not sure if this is expected or not with your version of Minerva, but I notice that the Minerva logo itself is all white in your picture, whereas my various Minerva versions always show the logo in red.

I also note that you have one of Vitaliy's excellent multi-ROM's installed (with the latest external ROM selector switch.)

Whilst you mention that it was problematic even before Minerva (and thus before fitting the multi-ROM), my guess is a bad ROM socket/connection, or else some rogue IC (bus-transceiver or address multiplexer) hogging the CPU bus intermittently (but after the RAM test - which suggests after some heat has built-up.)

A dodgy ZX8301 might exhibit such issues too, but I bet you've already switched that IC before.

During the memory test, does it look like the 'right' colours are presented in the tweed pattern - see the avatar for Cristian elsewhere on this forum, as a comparison if you're not sure what the tweed-pattern should look like...)

Good luck! It's always 'fun' fault-finding a QL...
Hi, thanks for the reply.

I'm using a cable that only gives monochrome output hence the color of the Minerva being "gray"...it's just for testing etc I guess I could hook it up to color to check for the correct colors.

Could an empty rom socket still cause a problem if there was a short on it? I've tried the multirom in both sockets and get exactly the same results, again with the original roms it's just black screen straight after the tweed memory test.

You mentioned bus-transceiver or address multiplexer..which ICs would these be...I'm pretty adept at soldering, not so much at trouble shooting, so I'm not afraid to change out ICs.

Thanks.


martyn_hill
Aurora
Posts: 909
Joined: Sat Oct 25, 2014 9:53 am

Re: Problem with QL

Post by martyn_hill »

Hi Beanz

Ahh - well that would explain the lack of colour in the logo! Forget that particular part.

Before launching in to further ideas, you might want to search this forum for some very similar requests for help (I'd just be repeating much of what has been written here over time.)

The part that is unusual here is the time gap between normal start-up and failure. Hence my thinking that some edge of tolerance component is warming-up during the start-up process, only to fail slightly later than we typically observe problems (during the mem test is usually where we first observe a fault.)

You'll want the schematic for your version of the QL motherboard (Iss5 or Iss6/7) to start with - which are both available from the World of QL (kidding - I mean Dilwyn's amazing repository of everything QL...)

A basic or common-sense (I mean no insult!) approach takes you quite far - in lieu of an intimate knowledge of QL HW when troubleshooting - I mean to say, don't presume you "don't know" just because its a QL - much successful QL hardware troubleshooting can be carried-out with just a reasonable electronics thinking-hat on, plus a multi-meter if you've got one.

Once you've got the right schematic from Dilwyn's site (find the PDF versions - they're as accurate as you're going to find), identify the following ICs:

DRAM (you know already): ICs 1..16 (4164)
Address multiplexers: IC 19 and IC 20 (LS257)
Bus transceiver: IC 21 (LS245)

The multiplexers and transceiver are soldered-on, and replacing is not a bad idea - adding DIP sockets in the process.

Now, one other idea that might point elsewhere in troubleshooting is that the other custom IC (ZX8302) and the secondary uP (8049) are not particularly significant in the earlier stages of the booting process, IIRC, so it could be that, once the OS really starts to move, any faults with those components might affect a fully functioning system.

E.g. persistent 'Interrupts' being fired, bringing the OS to its knees trying to respond - both those ICs are connected to the interrupt lines. Just one line of enquiry.

You'd want to replace the 8049 if you haven't already in any case (fondly called the 'Intelligent Peripheral Controller' or IPC - IC24) with a Hermes from RWAP - its a plug-in replacement, as is the ZX8302 (referred to as the Peripheral Chip - PC - IC23.) Again RWAP can supply these quite readily/cheaply.

...and did you already check/swap-out the main custom IC - the ZX8301 (IC 22)? Even the main 68008 CPU could be at edge of tolerance.

To avoid doubt, you can and _must_ leave one empty ROM socket whenever you use any of the available EPROM replacements (Minerva, or Vitaliy's Flash multi-ROM.)

I also assume that you have already removed any other expansion cards which could be playing silly-b*ggers with the CPU bus if faulty.

One last thought for now - it was working before (albeit, following earlier memory issues that were thought to be fixed.) What has transpired since it last worked - what oddities showed themselves in the last days before it finally failed to boot-up again, but you thought were irrelevant at the time?

Try to be patient and systematic in what you're doing - you want the beast working again as soon as poss of course, but getting intimate with the QL HW can pay dividends in the long run.

Of course, it might prove easier just to save-up for a replacement motherboard or even full QL that can be found from time to time on SMR or eBay...

Good luck!


User avatar
1024MAK
Super Gold Card
Posts: 592
Joined: Sun Dec 11, 2011 1:16 am
Location: Looking forward to summer in Somerset, UK...

Re: Problem with QL

Post by 1024MAK »

IC's 19, 20 and 21 are not likely to be faulty if the machine is managing to complete it's memory test. And the heat rise in the relatively short time it takes to do a memory test, will not increase the internal temperature of these 74LSXXX series chips by very much.

If a heat sensitive chip is thought to be the cause, try some freezer spray.

If you have another working QL, you can try swapping the socketed chips between them, to see if the fault follows the chip...

Mark


:!: Standby alert :!:
“There are four lights!”
Step up to red alert. Sir, are you absolutely sure? It does mean changing the bulb :!:
Looking forward to summer in Somerset later in the year :)

QL, Falcon, Atari 520STFM, Atari 1040STE, more PC's than I care to count and an assortment of 8 bit micros (Sinclair and Acorn)(nearly forgot the Psion's)
martyn_hill
Aurora
Posts: 909
Joined: Sat Oct 25, 2014 9:53 am

Re: Problem with QL

Post by martyn_hill »

Mark's right of course, though I've found the larger ICs can heat pretty rapidly if there's a fault in them or elsewhere.

The three ICs that typically get quite warm under normal circumstances are the CPU, the ZX8301 and the PAL chip (MC1377 - IC28). Most times I have problems of the sort you describe, the ZX8301 gets REALLY hot within seconds.

One of the other 'classic' issues that I'm sure you've already looked in to (and given the symptom, unlikely in this case but trivial to check) is the 'bent-pin' syndrome at the edge-connector. These can be surprisingly tricky to spot - try a piece of card/paper and insert between the two rows, looking for either a bent pin touching another, or a missing/snapped pin.

A missing pin itself wouldn't cause a problem without an expansion card inserted, but the question then becomes - 'where did it go?' as it could now be resting across other pins, possibly forced in to the recess of the connector.

Once the lid is off the QL, its just as well to eliminate all the easy candidates - even if unlikely given the specific symptoms. Trying to get super-clever too early can sometimes miss the simpler causes...


User avatar
tofro
Font of All Knowledge
Posts: 2685
Joined: Sun Feb 13, 2011 10:53 pm
Location: SW Germany

Re: Problem with QL

Post by tofro »

Once Minerva comes up to clear the screen and show it's logo on a steady screen, the basic functions of ROM, RAM, video and CPU seem to be verified. 8301 seems to be all-right. Apparently, the computer is in an endless loop waiting for some hardware that doesn't properly answer: Keyboard, Microdrives would be my next candidates.

I would check the following in this order:
  1. Check the ROM is properly inserted and replace with a known good one.
  2. Verify non-essential + and -12Volts are present
  3. Check with a known working 8302
  4. Check with a known working 8049
  5. Check that IPC (8049<->8302) comms is working, look for bent pins in that area
Stuck interrupt is another thing that comes to mind when looking at the symptoms, which could be caused by a bent pin on the expansion port.


ʎɐqǝ ɯoɹɟ ǝq oʇ ƃuᴉoƃ ʇou sᴉ pɹɐoqʎǝʞ ʇxǝu ʎɯ 'ɹɐǝp ɥO
beanz
ROM Dongle
Posts: 3
Joined: Tue Sep 08, 2015 1:52 am

Re: Problem with QL

Post by beanz »

Lots of good information! thanks guys. I'll try out the suggestions.


User avatar
tofro
Font of All Knowledge
Posts: 2685
Joined: Sun Feb 13, 2011 10:53 pm
Location: SW Germany

Re: Problem with QL

Post by tofro »

martyn_hill wrote: Trying to get super-clever too early can sometimes miss the simpler causes...
Martyn,

very much liked your last comment. Often a reason to run in the wrong direction for quite a distance and get lost between the trees...

Tobias


ʎɐqǝ ɯoɹɟ ǝq oʇ ƃuᴉoƃ ʇou sᴉ pɹɐoqʎǝʞ ʇxǝu ʎɯ 'ɹɐǝp ɥO
martyn_hill
Aurora
Posts: 909
Joined: Sat Oct 25, 2014 9:53 am

Re: Problem with QL

Post by martyn_hill »

Hi again Beanz

Having just completed a complex repair on one of my own Iss5 QL boards (see separate post from Cristian) - albeit, by undoing some of my own soldering errors - one observation that I document in the other post might prove useful for your investigation here. I won't re-post in full here, but note the 'passes memory test, then hangs after showing Minerva logo' - not 100% identical to your observation, but suspiciously similar. In my case, it was the ROMOE line being held permanently active (due to a bodged soldering job of mine.)

Good luck!


Post Reply