MTX 512 (debugging thread)

User avatar
Dave
Posts: 810
Joined: 11 Aug 2012 18:16
Contact:

Re: MTX 512 (debugging thread)

Post by Dave »

It might be worth making a quick check of the voltage on the sense lines that I described in the second link. That might or might not show anything, but it is a quick test to do while you are looking at the VRAMs

regards
Dave
GozdniJezek
Posts: 27
Joined: 07 Jun 2019 20:49

Re: MTX 512 (debugging thread)

Post by GozdniJezek »

Will also do that, thank you.
Martin A
Posts: 464
Joined: 09 Nov 2013 21:03

Re: MTX 512 (debugging thread)

Post by Martin A »

It look to me like the pattern has a repeat of 64, which would normally suggest the A6 address line might be involved.

Of course Texas labelled the VDP backwards, so maybe check AD1 for shorts and such ?

AD1 goes to the VRAM at 7G so that could be a suspect.
GozdniJezek
Posts: 27
Joined: 07 Jun 2019 20:49

Re: MTX 512 (debugging thread)

Post by GozdniJezek »

Good idea will also check for that. Thank you.
User avatar
1024MAK
Posts: 625
Joined: 24 Dec 2012 03:01
Location: Looking forward to summer, in Somerset, UK

Re: MTX 512 (debugging thread)

Post by 1024MAK »

The key-switches used suffer from similar problems to the keyboard switches as used on Acorn Electrons, BBC Micros and BBC Masters. The contacts inside become dirty, or a thin layer of oxide forms on the contacts. This results in either, the contacts not making electrical contact, or making intermittent contact.

As Dave says, the first thing to do, is to operate (fully depress each key, then let go) fifty to one hundred times each.
If this does not work, some can be brought back to life by using contact cleaner spray. But be careful, just test on one at first, as some makes/brands can affect some types of plastic.

If that does not help, the removal/replacement is the next course of action.

Mark
GozdniJezek
Posts: 27
Joined: 07 Jun 2019 20:49

Re: MTX 512 (debugging thread)

Post by GozdniJezek »

Hello, it's me again :lol:
So, first a recap:
I have socketed all of the video RAM and I've replaced all of the 4116 with a known good one. the picture after it was like:
IMG_20190713_123944.jpg
IMG_20190713_123944.jpg (270.51 KiB) Viewed 2505 times
Than i checked the voltages again, and I've figured out, that sometime during the replacement and resocketing I've lost the -5V :o
I've checked the voltage regulators at 1H, the left most gives 12V, the middle one gives 5V and the right most gives -5V, so my guess is that it must be some passive element, or a connection (hopefully unlikely) between the video RAM and the voltage regulator (still need to check which one it might be).
1.) The question now is, is the video RAM the only one requiring the -5V line?
2.) I know the missing -5V will damage the 4116, so i swapped in the old 4116 and I'm turning on the machine for a short period of time, what else do you suggest?
3.)What damage could I've done to other components, with missing -5V.
4.)Since I'll be meddling with the power supply, I was also thinking about recapping, does anyone sell capacitors kits, or do i have to buy it separately and do you think that is a good idea?
5.)The -5V goes through the fuse and over the large capacitor that is on top of voltage regulators, am I correct?
I can also hear a crackling noise, hopefully just from a bad RAM. Any advice would be appreciated and as always thank you for your answers. :)
Martin A
Posts: 464
Joined: 09 Nov 2013 21:03

Re: MTX 512 (debugging thread)

Post by Martin A »

I'm no expert on the power supply, but circuit diagram http://primrosebank.net/computers/mtx/r ... _power.htm shows that :

The TIP2955 is part of the +5 supply, not the -5.

The -5v is derived from the -v supply (-12 ???) using the resistor R60, Zener diode ZD3 and capacitor C49.

I would guess of those 3, it's the Zener that the more likely reason for the missing 5v. (at least until Mark sees this and can give you a more qualified opinion :D )
User avatar
1024MAK
Posts: 625
Joined: 24 Dec 2012 03:01
Location: Looking forward to summer, in Somerset, UK

Re: MTX 512 (debugging thread)

Post by 1024MAK »

The left most voltage regulator is a 7812 and provides the +12V rail.
The middle voltage regulator is a 7805 and with a TIP2955 helps provides the +5V rail.
The right most device is not a voltage regulator, but a power transistor. This is a TIP2955. It boosts the current to increase the available current to the +5V supply and works in conjunction with the 7805 voltage regulator.

The -5V supply is provided by resistor R60 (1kΩ) and Zener diode ZD3 (5.1V / 5V1).

The -5V supply is derived from the unregulated -V supply.

Resistor R60 limits the current and Zener diode ZD3 clamps the voltage at about -5.1V, ZD3 acts as a simple shunt stabiliser.

When they fail, zener diodes normally go short circuit. What do you have on either side of resistor R60 (1kΩ) {don't get confused with R62, the board markings are not very helpful}. Or you can test on D16 or D18 anode leads.

If the voltage on D16 or D18 anode leads is heathy (about -12V) and R60 measures 1kΩ (test with the power off), either ZD3 (5V1) has failed "partly" short circuit, or one (or more) of your 4116 video DRAM chips has failed and is/are dragging down the voltage. Or there is a short circuit between the -5V tracks and 0V/GND (or a circuit that connects to 0V/GND).

See also page 30 (PDF page 33) of the Service Manual.

See these pages from Dave's site: MTX Power Supply Details, Memotech MTX Power Problem

The only important thing the the -5V supply is used for is the 4116 DRAM chips. So it is very unlikely that anything else has been damaged.

Mark
GozdniJezek
Posts: 27
Joined: 07 Jun 2019 20:49

Re: MTX 512 (debugging thread)

Post by GozdniJezek »

Thank you, I'll try this out. I'll get back to you on the results :)
GozdniJezek
Posts: 27
Joined: 07 Jun 2019 20:49

Re: MTX 512 (debugging thread)

Post by GozdniJezek »

Hello,
it's been almost a year of silence, but other things got in between but now it's time to get back to this wonderful project.
To recap what has been done to the machine in the meantime. I recapped it, swapped all of the video RAM and socked it, changed the regulators and a diode (beside C49 and ZD2, beside regulator), because my helping hand shorted the power transistor and magic smoke appeared. In between for a couple of minutes all worked perfectly, however now I'm having issues with -5V line. I've removed all of the socketed video ram, but after a couple of seconds after power on, the voltage (on just the -5V line, other voltages are fine) goes to 0V. So my question is how to go on debugging this? I'm thinking it must be the system ram or does anything else in the machine use the -5V? How would you go on debugging? At this stage I also have an oscilloscope available if it will be any help at this stage.
TL;DR: I'm having issue with collapsing -5V line, any help appreciated. Thank you for your answers.
Post Reply