Page 1 of 3

Re: Issue 4A with no colour!

Posted: Mon May 06, 2019 10:43 pm
by 1024MAK
Yeah, but keep in mind that with any in-circuit testing, the results may be affected by the other components on the board.

Mark

Re: Issue 4A with no colour!

Posted: Mon May 06, 2019 11:11 pm
by Muttley Black
That's true. But TR1 should give you this measurements ("open", both on collector and emitter with neg at base) in-circuit testing. I also test it on my spectrum's after your post.

Re: Issue 4A with no colour!

Posted: Tue May 07, 2019 12:29 am
by Liveinabin
Muttley Black wrote: Mon May 06, 2019 9:40 pm TR1 looks faulty to me. Cause with neg at base, you should get "open", both on collector and emitter if i am not mistaken. If you are sure that you measure TR1 as TR1 and TR2 as TR2 then replace TR1. I say that because your TR1 measurements fit more to TR2 transistor. Anyway, to be 100% sure replace both TR1 & TR2 with new BC549C transistors and tell as if the problem solved. The other transistors looks ok.

Also post your voltages just to have a picture about them. With neg probe to ground measure voltages on pins 1,8,9 and 16 from the IC6 and post them here.
I’ll check my measurements again and I’ll change both components (it’ll be a good chance to have a proper look at the tracks as make sure I haven’t shorted/damaged anything). I’ll report back tomorrow night with findings and more readings.
Thanks so much for helping me out here. Really appreciate it.

Re: Issue 4A with no colour!

Posted: Tue May 07, 2019 12:44 am
by 1024MAK
If you are getting a good monochrome picture. By that I mean good contrast and brightness, then there can’t be much wrong with TR1 and TR2, as the monochrome (‘luminance’) signal passes through both of these.

Also TR1 has nothing to do with the colour part of the signal. The encoded colour signal comes from pin 13 of IC14 (LM1889) via C65 (100nF) to TR2 base, where it is mixed with the ‘luminance’ signal.

An oscilloscope would be a great help with fault finding this problem.

Mark

Re: Issue 4A with no colour!

Posted: Tue May 07, 2019 7:34 am
by Liveinabin
The mono image is perfectly fine (well, for a Spectrum).

Re: Issue 4A with no colour!

Posted: Tue May 07, 2019 9:17 am
by 1024MAK
So, some questions.

First, have you used a magnifying glass to carefully visually examine the areas of the video section where you did any work. Both on the top side of the board and on the underside of the board. Do this under a good bright light, or better still, in good daylight. You are looking for any of the following: cracks or breaks in the PCB tracks/pads, dry solder joints (cracked, distorted, discoloured, dull), solder splashes (including thin solder threads, balls of solder stuck in flux), components fitted the wrong way round, components where a leg has been fitted in the wrong hole.

Tell me about the display/monitor/TV that you are using. Did this Spectrum work okay before with it? Have you tried this Spectrum on another display/monitor/TV? Does this display/monitor/TV work okay with other 8 bit computers?

I understand that you have a digital multimeter. Tell me about your digital multimeter. Does it have a frequency range? Does it have a 200mV AC range?

Do you have any other test gear?

Do you have another ZX Spectrum? If yes, which model?

Mark

Re: Issue 4A with no colour!

Posted: Tue May 07, 2019 1:17 pm
by Muttley Black
All this "transistors" story happens because [mention]Ast A. Moore[/mention] ask him to test the transistors and friend didn't know the way to do it. So i drive him to JoulesperCoulomb video that shows the way. After that, he post the measurements and we found out a faulty transistor. For sure the first suspect for missing color is the LM1889 but we cant ignore the fact that is a faulty transistor on the board.

Anyway. I've been here, on this forum not much time, but for a reason. i have read ALL threads on hardware category and what i want to say is that both [mention]1024MAK[/mention] and [mention]Ast A. Moore[/mention] can be easily named "ZX Resurrectors". And it is not only that. They are both very well technical educated members of this forum and they are always happy to help. So you are in good hands my friend. :D

Re: Issue 4A with no colour!

Posted: Tue May 07, 2019 1:31 pm
by Ast A. Moore
Aww, thanks, Muttley, but Mark is by far much more knowledgeable in this area than I am.

Re: Issue 4A with no colour!

Posted: Tue May 07, 2019 7:59 pm
by Liveinabin
Thank you all. I’ll try and get some time tonight/poss tomorrow to go over the Spectrum carefully.

Re: Issue 4A with no colour!

Posted: Tue May 07, 2019 9:11 pm
by Liveinabin
1024MAK wrote: Tue May 07, 2019 9:17 am So, some questions.

First, have you used a magnifying glass to carefully visually examine the areas of the video section where you did any work. Both on the top side of the board and on the underside of the board. Do this under a good bright light, or better still, in good daylight. You are looking for any of the following: cracks or breaks in the PCB tracks/pads, dry solder joints (cracked, distorted, discoloured, dull), solder splashes (including thin solder threads, balls of solder stuck in flux), components fitted the wrong way round, components where a leg has been fitted in the wrong hole.

Tell me about the display/monitor/TV that you are using. Did this Spectrum work okay before with it? Have you tried this Spectrum on another display/monitor/TV? Does this display/monitor/TV work okay with other 8 bit computers?

I understand that you have a digital multimeter. Tell me about your digital multimeter. Does it have a frequency range? Does it have a 200mV AC range?

Do you have any other test gear?

Do you have another ZX Spectrum? If yes, which model?

Mark
I’m busy with the magnifying glass right now. Aside from a tiny bit of messiness around TR1 and TR2 caused when desoldering them, I can’t see anything else..yet. I will desolder those transistors (carefully) to get a better look at the board underneath.

My monitor is an LG 24” LCD. I use it as it runs everything; the RF input likes ZX81s and Spectrums (and even Fairchild channel F and old pong machines) and the composite input is happy enough with everything else. It’s been fine with my other Spectrum’s (both now composite modded) and it liked this busted one back when it was RF.

I attached a picture of my multimeter.

As for other test equipment. Sadly no. I understand an oscilloscope would be a benefit but aren’t they astronomically expensive for a hobbyist? If there’s a halfway affordable model, I’ll happily pick one up.

As for other Spectrums. I have 2 more. An issue one 16K and a Samsung 3B 48K



Image

Re: Issue 4A with no colour!

Posted: Wed May 08, 2019 10:23 pm
by Muttley Black
1024MAK wrote: Tue May 07, 2019 12:44 am An oscilloscope would be a great help with fault finding this problem.
Mark
I bit off topic but, can you suggest a value for money oscilloscope? Just for spectrum repairs. What is the minimum requirements for? Thank you.

Re: Issue 4A with no colour!

Posted: Wed May 08, 2019 11:47 pm
by Ast A. Moore
Muttley Black wrote: Wed May 08, 2019 10:23 pm What is the minimum requirements for?
For an analog scope, I’d say at least 20–40 MHz of bandwidth. If you’re looking at a digital scope, you’ll need something like 50–100 MHz of bandwidth with 200 MS/s and enough memory depth.

A two-channel scope is often beneficial, but even a single-channel scope will do the job nicely.

An aftermarket analog scope can be obtained dirt cheap and is a pleasure to use. If it works, that is. You might end up with something you’ll spend more time repairing than working with.

Cheap digital scopes have gotten better, but decent ones can still be quite pricey and take a little getting used to.

Re: Issue 4A with no colour!

Posted: Thu May 09, 2019 9:06 am
by Liveinabin
So given that I'm currently bereft of any other test equipment; am I right in assuming the component we really needed to test was the LM1889 video encoder chip (as I'm figuring that it is responsible for the colour, mostly)?
Reason I ask is, I'm about to order some bits anyway (keyboard membrane, voltage regulator..) and a new LM1889 is only £3. If we're even vaguely suspicious about the chip, I'll happily pay £3 to find out :)

Re: Issue 4A with no colour!

Posted: Thu May 09, 2019 9:13 am
by Ast A. Moore
Liveinabin wrote: Thu May 09, 2019 9:06 am If we're even vaguely suspicious about the chip, I'll happily pay £3 to find out :)
In that case, go for it, definitely.

Re: Issue 4A with no colour!

Posted: Thu May 09, 2019 10:52 am
by 1024MAK
The cheapest will be a used analogue ‘scope. I recommend one with a minimum bandwidth of 25MHz. Or higher. Ideally the analogue bandwidth should be at least ten times greater than the frequency of the signal you want to measure/see. Especially with digital waveforms that are supposed to have fast rise times, fall times and are supposed to be so called ‘square waves’ (but are actually pulses).

The best value digital ‘scopes are made by Rigol and their basic (cheaper!) models are popular amongst the home ‘retro’ computer community. The greater the sampling rate and the greater the memory, the better the resolution of the displayed signal waveform.

Note that if you do want to buy a new Rigol, buy it from an authorised dealer, as otherwise the warranty will not be valid.

Dual/two channel ‘scopes, with a separate trigger input are good. Four channel can be very useful (but cost a lot more). And if you are feeling flush with money, you can get ‘scopes that include multiple digital inputs like a logic analyser. These are known as MSO - mixed signal oscilloscopes.

As well as stand alone digital ‘scopes, there are USB PC ‘scopes. I use one from pico. These have the high frequency circuitry in a box that you connect to a PC. Then you run an application on the PC, and get a display on the PC of the signal along with a control panel. Most waveforms that I post online are from my pico. Note that despite using a PC, these are also not cheap.

In answer to your question about buying spares, yes, if you can afford it, it’s always a good idea. It’s not often that the LM1889 chip fails, they are normally reliable. I suspect one of the peripheral components or connections.

If you are buying spares, buy a new crystal for X2 (4.43361875MHz).
See the service manual for more about the sub-carrier (colour) circuitry (here).

Mark

Re: Issue 4A with no colour!

Posted: Thu May 09, 2019 11:54 am
by Liveinabin
1024MAK wrote: Thu May 09, 2019 10:52 am In answer to your question about buying spares, yes, if you can afford it, it’s always a good idea. It’s not often that the LM1889 chip fails, they are normally reliable. I suspect one of the peripheral components or connections.

If you are buying spares, buy a new crystal for X2 (4.43361875MHz).
See the service manual for more about the sub-carrier (colour) circuitry (here).

Mark
Oh yes, I forgot about the crystal. Right, all bits ordered including a matching crystal for X2 and a new LM1889 (although, as has been said, it's probably fine). About the crystal, is there a 'right' way to orient it? Do these things have a polarity?

Re: Issue 4A with no colour!

Posted: Thu May 09, 2019 12:21 pm
by Ast A. Moore
Liveinabin wrote: Thu May 09, 2019 11:54 am Do these things have a polarity?
Nope, they don’t. Although, their physical orientation in space will have a small effect on their operation causing deviations form their standard frequency. (Just a technicality; no need no worry about it.)

Re: Issue 4A with no colour!

Posted: Thu May 16, 2019 8:32 am
by Liveinabin
Oooookay! Update time.

Got a new LM1889. Fitted that. No difference (didn't think there would be but worth a pop, although desoldering that was a pain)
New crystal in X2. Still mono.
Reflowed and, in some cases resoldered and replaced some of the new capacitors around the video bit. Nope. Still greyscale.
Desoldered the BC549Cs at TR1 and TR2 (mainly to have a good look at the traces around them) and fitted new ones.
Added a 100uF transistor into the composite mod (which does indeed brighten things up)
Went over the whole board with a big magnifying glass, cleaning up as I went.

Still black and white! Gah!

I do have one more thought. I swapped TR1 and TR2 for BC549Cs at the start of this escapade (sadly ditching the old transistors) and I've tried this Spectrum on a couple of TVs: My reliable LG is greyscale and a little (also LCD) Samsung shows, it turns out, no picture at all!
Tried my other working, comp modded Spectrum on both displays (a Samsung 3B with the original TR1 TR2 transistors) and it displays on both: Well on the LG and poorly on the Samsung TV (but then the Samsung TV IS crap).

Might it just be the BC549Cs not being compatible with my TV? I read that the 'improvement' they give may not work with flat TVs. Might it be worth me fitting a couple of ZTX313s instead?
That's really all I can think of to try now.

Re: Issue 4A with no colour!

Posted: Thu May 16, 2019 6:43 pm
by 1024MAK
The “monochrome” parts of the video picture require a larger bandwidth (meaning higher frequency) than the colour carrier, so as I said above, I’m not convinced. But you are free to try.

Mark

Re: Issue 4A with no colour!

Posted: Thu May 16, 2019 7:58 pm
by Muttley Black
You have already checked if 12 volts present on LM1889 (pin14), yes?

Re: Issue 4A with no colour!

Posted: Thu May 16, 2019 9:28 pm
by Liveinabin
Muttley Black wrote: Thu May 16, 2019 7:58 pm You have already checked if 12 volts present on LM1889 (pin14), yes?
I’m getting 11.56v
That’s with ground probe on the plate near the voltage regulator and the positive probe on pin 14, if that’s correct.

Re: Issue 4A with no colour!

Posted: Fri May 17, 2019 9:01 am
by 1024MAK
LM1889 (IC14).

Test between pins 14 and 5.
Then between pins 15 and 5.
Then between pins 16 and 5.
Then between pins 12 and 5.

Mark

Re: Issue 4A with no colour!

Posted: Sat May 18, 2019 1:29 pm
by Liveinabin
Hallo
11.58v on all of those pins.

Re: Issue 4A with no colour!

Posted: Sat May 18, 2019 2:38 pm
by Liveinabin
Ah OK, further news :)

Noticed that both of my Spectrums have the exact same revision of ULA so.... carefully popped both out and swapped them.
ULA from my Samsung gave no picture at all on this spectrum. ULA from this spectrum gave garbled mess on the Samsung (random block characters and colours). Checked the seating of the chips a couple of times, same thing. Swapped em back. Luckily my Samsung 3b is still fine but there's no video at all from this unit.
I dunno, reckless to try it perhaps.

SO.. I've sadly now hit my limit in the time and money I can really put into this Spectrum, and I'm sure I'm now only making things worse (even if I fixed it, still needs a new kb membrane).
Thanks SO much for your help, Mark and Muttley Black. I really appreciate it. If either of you would like a free busted Speccy to use for bits (or repair/burn for fun etc), post free, just PM me your address and I'll send it over.

I'm not done with Spectrums yet - just this one ;)

Re: Issue 4A with no colour!

Posted: Sat May 18, 2019 3:00 pm
by 1024MAK
Sorry to hear that.

Watch out for the incoming PM...

Mark