Same Problem on 2 Timex Boards

[TMAOne]

It's late and I may be tired but this is bugging me,...

I have two boards, both Timex 1000's, which won't boot. I don't get any of the more interesting pyrotechnic puke, just the faint vertical striping that I find is the usual background to an otherwise healthy screen. Perhaps the pattern is a little more pronounced than usual; I'm not certain. At any rate, both boards exibit the same symptom--the K prompt is nowhere to be seen.

I only have enough chips left from other misadventures to populate one of these two boards at a time. (The other board has been a "donor"--I've tossed a bad ULA from one ZX81, a CPU from another. One ROM lost a leg chip-swapping but could be salvaged if needed. The donor's single chip 2K RAM is still there.)

Of the 5 chips (including extra RAM), each one has been individually tested in a working Timex, successfully. So all the ICs seem to be good. Okay, then it's the mother(s), right? Or chip seating. Seating has been done many times as I try to isolate and diagnose,.. hence the ROM getting a purple heart for a lost leg.

I'm second-guessing myself because the observations seem twitchy. I find it very suspiscious that I have 2 boards exhibiting the same issue!

I have read that a "no prompt" condition is often caused by the RAM not working. Well, I have two of them that do work when swapped into a good computer. I can envision a bad CPU doing this, but the Z80 seems happy when switched with another working one too.

Okay so I replaced both electrolytic capacitors in both machines. No difference. Voltages appear good everywhere I look. I haven't run a continuity check from each and every chip leg to the solder side of the board, but will try that next. (The soldering on both boards is bright and neat.)

Am I missing something really obvious?

Ian

[gozzo]

Not much else can go wrong! Modulator failure is extremely unlikely(assuming you are using the RF out and not a composite modification? but give that channel select switch a squirt of cleaner, then wiggle it a bit!),so forget that for now! check both transistors, the ceramic resonator thingy , check for any broken tracks and shorts between tracks,make sure the regulator is giving 5v out, as for anything else, I would have though failure of the resistors,diodes,etc. was unlikely...but possible...????!!!!

[Proteus]

Hi TMAOne, could you post a screenshot of the image you get? Do you use composite mod or the modulator?

[RetroTechie]

Ceramic resonator could be the problem, or one of the transistors. No clock signal = no image. Do you have a logic probe or oscilloscope, to check clock signal on Z80 pin 6 ?

Another possibility is disabled ROM or RAM memory due to a problem with /RAMCS or /ROMCS signal (cracked circuit trace / short circuit), resulting in blank pattern data. Last: can you blind-type

LOAD "" [NEWLINE]

and see if loading patterns appear? (you never know, right?)

[TMAOne]

Proteus, I am using the one-transistor composite mod from ZXTEAM's site. I have never succeeded in getting a VHF signal to this particular little TV set I use for the ZX81/Timex's. Only the UHF modulator on my old original kit ZX81 seems to be able to drive it for some reason. Although the TV was the princely sum of $20 at a going out of business sale, it has a composite video input, so works perfectly with the mod on all my other boards.

Photo below. The black banding is a camera effect, the vertical stripes are what I see.

RetroTechie, at your suggestion I plugged a keyboard back in (oh so carefully), and after blind typing for a LOAD "", I DID get the typical drunken tartan effect! I assume that means the clock is ticking.

You guys are awesome.

There are no shorts or broken traces that I can see. Both these Timex's came from garage sale type sources and I don't think they'd ever been opened until I did it. They look well manufactured, for what that's worth, which is nothing.

I do have an old oscilloscope but it never worked properly, and as a result I don't know how to use it well anyway. It's an ancient tube jobber and likely not fast enough for this. No logic probe either. eBay has cheap ones for $11. Would one of those be of any value?

Ian

P.S. Just replaced both transistors with 2N3904. No joy.

I'm going to rig up a video circuit for the "other" dead board, because I find it hard to believe that a discrete component failed on TWO boards, over the possibility that I'm just doing something stupid late at night and don't see it.

[...] Yeah, okay, swapped the chips back to the "parts donor" and the pattern is NOT the same. It's unstable video noise now on Board #2. So I don't have such an unlikely coincidence. I guess that makes me feel good enough to retire to bed,...

[sirmorris]

I seem to remember having problems using a 3904 in a video mod - I just don't think it was fast enough. You'd be better sticking wit the specified part. I know that doesn't solve the problem though

C

[Paul]

I seem to remember having problems using a 3904 in a video mod

I use 2N3904 for my video mod inside the modulator case see here without any problems.
FYI
Paul

[sirmorris]

I stand corrected

[TMAOne]

Sorry, I was unclear.

The video modifiation uses an S9013 transistor, which seems to work well. I have done nothing inside the stock video modulator. It is still connected, just unused.

I replaced the two motherboard transistors, originally MPS2713, with 2N3904. That has made no apparent difference, but I have the original transistors and could put them back if need be.

Ian

[PokeMon]

I seem to remember having problems using a 3904 in a video mod

I use 2N3904 for my video mod inside the modulator case see here without any problems.
FYI
Paul

Depend on what you do.
This simple emitter follower does not need any amplification (gain about 1) and so it should follow a 3 MHz video signal without problems.