Help fixing a dead QL

Nagging hardware related question? Post here!
asapreta
ROM Dongle
Posts: 4
Joined: Thu Apr 18, 2019 2:27 pm

Help fixing a dead QL

Postby asapreta » Fri Apr 19, 2019 4:16 pm

Hi all!!

A new owner of a QL here from Brazil but it is dead atm.

Before being registered I found a lot of information here that lead me to do some improvements on this dead QL.
But now I need your assistance to try to fix it.
I will try to give as many information as I can.

It is an issue 5 board. It arrived giving me these screens while booting:

Image

Image

Reading some posts I found it could be RAM or a dead 74LS157 or the 74LS245. Removed the two LS157 and using an IC tester found one blown. Replaced them using sockets then I got a white screen.

Digging more info, found white screen is related to bad RAM or the 74LS157. But as I have already replaced them, I decided to check the RAMs. So I found Minerva ROM as able to check them.

Downloaded it and installed following this picture below as I didn't found any other information (or misunderstood them) about configuring the board jumpers to use ROM/EEPROM.

Image

This lead me to some more questions, like: does it work correctly this way or I need to install following this schematic:

Image

Why I am asking this? Later I am going to show how strange is the behavior of this QL.

When I booted the MInerva 1.98 for the first time I got this screen:

Image

For me it was clear I still have problems with the LS157 and RAMs.
But now the "stranger things" started to happen.

I only can get a "normal" image with the error from Minerva if I use the DEAD LS157...
Something like this:

Image

But why if the LS157 is broken?

If I use two good LS157 I can´t read the screen.
One guy from the QL group on FB send me a Basic program that I could find the RAM IC by entering the info from the Minerva boot. I used it on a emulator and was given as that IC1, IC2 and IC3 were bad each time I entered the info (I replaced IC2 then IC3 was bad then IC1). Now It still wont boot properly and says IC1 is bad (I have about 30 4164 new ICs here and all of them won't let me pass this error)

This lead me to replace 74LS245. No change.

I removed the Minerva ROM and reinstalled the original one. And messing with the LS157's, I used two 74HC157. I know it is not the correct but give a try. When I combined the HC157 with, and only with it, the original good LS157 from the QL, i got this screen:

Image

And now I am really lost because, If I try to boot Minerva with this "setup" (HC157 and LS157) I can't read the error code screen.

Would you guys give me any advice? Is it a bad ULA? Bad ROM? I don't want to mess the board more than it needs.

Thanks in advance for any help.

Cheers.


User avatar
Dave
SandySuperQDave
Posts: 2412
Joined: Sat Jan 22, 2011 6:52 am
Location: Austin, TX
Contact:

Re: Help fixing a dead QL

Postby Dave » Fri Apr 19, 2019 4:55 pm

The pattern in the screen shot is ROM copied into RAM.

There is no failure pattern of RAM that would cause that image. Solid black or white bars indicate 8 ICs all returning a 0 or 1 bit together. If one RAM IC was asserting or de-asserting on address lines, it would have to be either A14 or A12 or both to make a pattern that changes in that way.

The 8301 only cares about A16 and A17, so it is likely ok.

The only place all those items come together is the CPU and the crappy CPU socket on your UK Issue 5 QL...

So yes, confusing.


User avatar
Pr0f
Gold Card
Posts: 369
Joined: Thu Oct 12, 2017 9:54 am

Re: Help fixing a dead QL

Postby Pr0f » Fri Apr 19, 2019 5:47 pm

ROM addresses seem ok - otherwise it wouldn't get as far as the F1/F2 prompt - but clearly a failure somewhere in RAM still - solder blobs, shorted tracks - lifted tracks - worth checking for all these

I have a semi-dead QL that has a shorted track under the RAMS - it's a bugger to lift and replace them - and oh so easy for stray solder to get in and ruin your day.

It has similar symptoms to your QL. Switch everything off and if you have a multimeter - check all the data lines to the ram chips. Pins 2 and 14 should be shorted together, and not to any other adjacent pins on the same chip - also for the top bank of 6 chips and the 2 farthest chips from the processor, none of the pins 2/14 should be shorted to any of the other, and the same for the lower first 8 chips next to the processor (closest to the edge of the board - the longest row of RAM chips) One upper chip and one lower chip will share common data line.

The address lines may also be shorted and cause an issue like you have (this is my problem), these lines are 5,7,6,12,11,10,13 and 9 (for lines A0-A7) respectively. Check no adjacent or opposite pins are shorted

For example - consider A2 - this is next to A0 and A1 (so pins 5 and 7 either side of 6), and opposite pins are 10,11 and 12 - you need to test for shorts to the tracks that pass through between pins too!

It's a painful slow process I'm afraid.

Datasheet:

http://pdf1.alldatasheet.com/datasheet- ... 4164B.html


asapreta
ROM Dongle
Posts: 4
Joined: Thu Apr 18, 2019 2:27 pm

Re: Help fixing a dead QL

Postby asapreta » Fri Apr 19, 2019 6:36 pm

Pr0f wrote:ROM addresses seem ok - otherwise it wouldn't get as far as the F1/F2 prompt - but clearly a failure somewhere in RAM still - solder blobs, shorted tracks - lifted tracks - worth checking for all these

I have a semi-dead QL that has a shorted track under the RAMS - it's a bugger to lift and replace them - and oh so easy for stray solder to get in and ruin your day.

It has similar symptoms to your QL. Switch everything off and if you have a multimeter - check all the data lines to the ram chips. Pins 2 and 14 should be shorted together, and not to any other adjacent pins on the same chip - also for the top bank of 6 chips and the 2 farthest chips from the processor, none of the pins 2/14 should be shorted to any of the other, and the same for the lower first 8 chips next to the processor (closest to the edge of the board - the longest row of RAM chips) One upper chip and one lower chip will share common data line.


I have checked if they are connected but not for shorts yet.


The address lines may also be shorted and cause an issue like you have (this is my problem), these lines are 5,7,6,12,11,10,13 and 9 (for lines A0-A7) respectively. Check no adjacent or opposite pins are shorted

For example - consider A2 - this is next to A0 and A1 (so pins 5 and 7 either side of 6), and opposite pins are 10,11 and 12 - you need to test for shorts to the tracks that pass through between pins too!

It's a painful slow process I'm afraid.

Datasheet:

http://pdf1.alldatasheet.com/datasheet- ... 4164B.html


Will double check for shorts and the correct connection to address lines.

But for me the most intriguing aspect is the "selection" of the LS157 it is choosing to "work".
For me it should go, as far as it can, with both Minerva or original ROM, but probably I will need to do the Minerva "correct" installation.

Thanks for you time and answer.


User avatar
Pr0f
Gold Card
Posts: 369
Joined: Thu Oct 12, 2017 9:54 am

Re: Help fixing a dead QL

Postby Pr0f » Fri Apr 19, 2019 8:14 pm

Changing the 157 for the dead one may indeed indicate a shorted address line - as these multiplex the address lines, it may be the dead one prevents the issue loading devices upstream...


Martin_Head
Gold Card
Posts: 421
Joined: Tue Dec 17, 2013 1:17 pm

Re: Help fixing a dead QL

Postby Martin_Head » Mon Apr 22, 2019 10:52 am

It sounds like you have the facilities to unsolder IC's without damaging them. You also say you have lots of spare RAM chips.

So why not unsolder all the RAM chips, check the PCB, fit IC sockets, then replace all the RAM chips at once. If this fixes it, then test the old chips one at a time to see which ones are faulty.



Who is online

Users browsing this forum: No registered users and 5 guests