Resurrecting an MTX500.

About original Memotech hardware.
stephen_usher
Posts: 199
Joined: 27 Nov 2016 19:58

Re: Resurrecting an MTX500.

Post by stephen_usher » 18 May 2019 17:56

With respect to 7A the inputs, they are:

Address lines A5+6+7 outputting to LS04 6B and input to LS08 7E (which controls the GI input on the two LS138s, 7C and 7D)

A4+/IORQ+/MI(via LS04 6B) outputting to the other input to LS08 7E

Output of the first input via LS04 6B+A2+A4 outputting into LS00 5B along with A3 which drives the /PE8T11 line.

LS08 7E has just been replaced this morning and that's had no effect.
LS04 6B was the previous chip to be replaced, which had no effect.
LS00 5B has also been replaced and this didn't change the issue.

I've swapped the CPU with two other working ones. The older CPUs do seem to be marginally better with respect to this fault.
I've tried swapping Z80CTC and that made no difference.

I could try replacing the two LS138s again.

stephen_usher
Posts: 199
Joined: 27 Nov 2016 19:58

Re: Resurrecting an MTX500.

Post by stephen_usher » 18 May 2019 18:28

What I have noticed while 'scoping is that the address bus seems to be only driven to ~3.4V. There are a very few times it goes above this. The data bus, however, is a lot cleaner and far nearer to 5V. This is consistent for all the data lines and all the address lines.

This could be an issue if there's any CMOS logic accessing the address bus as it's below the minimum voltage for a "high".

Martin A
Posts: 414
Joined: 09 Nov 2013 21:03

Re: Resurrecting an MTX500.

Post by Martin A » 18 May 2019 20:54

I only mentioned the TIP as that was a problem with one of my MTXs. I spent 2 weeks chasing a "timing problem" of random crashing and inability to run programs for any length of time, which turned out to be nothing to do with the timing. Anyway, if the big 10R resistor R62 next to rom B isn't hot then it's not that.

I scoped the address bus of the MTX with the Z80A in it and the output seems to be fairly consistent at 4v. The screenshot happens to be A7 but the other address lines were similar.

The data bus seemed to jump between 4v and 5v highs, which is sort of expected as the roms are CMOS replacements while the CPU is NMOS so the level would depend on which one is driving. (I gave up trying to get a shot with both 4 and 5v highs so no picture!)
A7 at ROM B.jpg
4v signal 8" from the CPU
A7 at ROM B.jpg (32.92 KiB) Viewed 498 times

stephen_usher
Posts: 199
Joined: 27 Nov 2016 19:58

Re: Resurrecting an MTX500.

Post by stephen_usher » 18 May 2019 21:47

OK, I think I may change that next, but that'll have to wait a couple of months now as I'm packing things away (and I don't have the part anyway).

User avatar
1024MAK
Posts: 563
Joined: 24 Dec 2012 03:01
Location: Looking forward to summer, in Somerset, UK

Re: Resurrecting an MTX500.

Post by 1024MAK » 18 May 2019 23:56

CMOS devices designed to work with 5V TTL systems should be okay with logic one/high levels of 2V and greater. So all the Zilog Z80 series CMOS devices, 74HCTxxx chips and all compatible EPROM, EEPROM and SRAM chips. However with modern EEPROM and SRAM chips, it’s best to double check with the data sheet, as some devices these days do not have TTL compatible inputs.

CMOS Z80 CPUs are known to work in ZX81s and ZX Spectrums without problems.

There are a couple of caveats: Reset and Clock inputs to MPU/CPU or similar often do not conform to TTL signal levels.

The other things to keep in mind are: does the bus or signal line have a pull-up resistor(s)? And with the data bus, other devices other than the CPU drive it.

Mark

stephen_usher
Posts: 199
Joined: 27 Nov 2016 19:58

Re: Resurrecting an MTX500.

Post by stephen_usher » 19 May 2019 09:59

Which version of the TIP2955 should I look for?

There seem to be two main types, one with a positive collector voltage and the other with a negative one.

User avatar
Dave
Posts: 748
Joined: 11 Aug 2012 18:16
Contact:

Re: Resurrecting an MTX500.

Post by Dave » 19 May 2019 13:39

There is only one 2955, some datasheets are common to the 2955 and 3055, they are PNP and NPN. Might that be what you are looking at?

stephen_usher
Posts: 199
Joined: 27 Nov 2016 19:58

Re: Resurrecting an MTX500.

Post by stephen_usher » 19 May 2019 14:17

Dave wrote:
19 May 2019 13:39
There is only one 2955, some datasheets are common to the 2955 and 3055, they are PNP and NPN. Might that be what you are looking at?
Nope. Farnell (and RS) have two versions, one with +70V and the other with -60V, both manufactured by ST Electronics. I'm wary as in my other MTX I put a 2955 in which I'd got from Maplin (type wasn't specified) and it blew the regulator.

User avatar
1024MAK
Posts: 563
Joined: 24 Dec 2012 03:01
Location: Looking forward to summer, in Somerset, UK

Re: Resurrecting an MTX500.

Post by 1024MAK » 19 May 2019 15:08

Sometimes the people copying / writing the descriptions and specifications on the suppliers web site / catalogues don’t fully understand the datasheet. And sometimes the datasheets are misleading.

The voltages and currents of PNP transistors are opposite compared to NPN transistors, so are normally prefixed with a negative symbol.

On one of the datasheets it says this:
For PNP type voltage and current values are negative
At Farnell, these TIP2955 types are suitable:

https://uk.farnell.com/search?st=TIP2955

https://uk.farnell.com/on-semiconductor ... st=TIP2955

https://uk.farnell.com/stmicroelectroni ... st=TIP2955

https://uk.farnell.com/multicomp/tip295 ... st=TIP2955

At RS:-
https://uk.rs-online.com/web/c/semicond ... 22:true%7D

This type is suitable:-
https://uk.rs-online.com/web/p/bjt-bipo ... s/4859705/

Do carefully check that you get what you ordered, as sometimes the wrong components end up in the wrong storage bin, or the order picker makes a mistake and grabs the wrong item. One major supplier has even put the wrong chip in an anti-static bag, sealed it, put the sticker with the (wrong) part number on it and sent it to a customer. So don’t just check the label, check the part number / ID on the device itself.

But as I said earlier, if the 5V supply rail voltage is stable with little noise and no ripple visible on a ‘scope trace (using AC coupling), changing the TIP2955 will not sort out the problem.

Mark

stephen_usher
Posts: 199
Joined: 27 Nov 2016 19:58

Re: Resurrecting an MTX500.

Post by stephen_usher » 19 May 2019 15:48

Thanks
1024MAK wrote:
19 May 2019 15:08
Do carefully check that you get what you ordered, as sometimes the wrong components end up in the wrong storage bin, or the order picker makes a mistake and grabs the wrong item. One major supplier has even put the wrong chip in an anti-static bag, sealed it, put the sticker with the (wrong) part number on it and sent it to a customer. So don’t just check the label, check the part number / ID on the device itself.
I had that just over a week ago with Farnell. I ordered some CD4013B chips and that's what the anti-static bag label said, but they were 74HC75 chips.

Well, given that the resistor is/was getting hot it's probably time to change the TIP anyway.

Post Reply