Spectrum Computing

That’s the same problem as experienced by plenty of 16k/48k/plus machine owners. One, or both of the transistors used in the DC/DC converter circuit have been killed. Or less likely, “the coil” has failed.

In a +2 grey, the relevant transistors are: TR8 (ZTX213) and TR9 (ZTX650). TR9 is almost certainly dead, TR8 may or may not be faulty.

Replace TR9 with a ZTX651 or a ZTX653. Replace TR8 with a ZTX750 (matching lead-out) or a good quality NPN small signal transistor (may have a different lead-out).

After removing TR9 test for shorts on the +12V and on the -12V supply rails before soldering in a new transistor.

Mark

Thanks I have just ordered some from Phil Ruston so should have them by the end of the week hopefully or Monday and then I can have a go and try and fix it.

Hi,
Just as an addendum to this I found that D18? or the 5V1 whichever that one is was broken and I mean physically cracked into two halves. I haven't got the 12V back online yet, still awaiting replacement TR9 and TR8. If that fails I have a working spare coil I can try.
So once they are in and I have tested it I'll update this thread.

Replaced TR8 and TR9 today, now have 12V on the Edge connector and on pin 11 of the TEAC. Cannot get it tuned in on the modulator but the RGB socket works. Just need to sort the noise on the audio that +2 Greys were prone to.
In all other aspects it runs fine so have saved it from the scrapyard

Morpheus wrote: ↑Sat Aug 03, 2019 5:36 pm Just need to sort the noise on the audio that +2 Greys were prone to.

From my thread on WoS:

Yep already done!

Noise is hum from the Cassette unit - disconnect it and the noise/hum vanishes. I probably won't use it much but need to fit an external Ear socket so might have to live with it. I do have external AY interfaces to get stereo AY minus the Beeper if I really want to listen to the music/fx. I am contemplating replacing the PAL with a GAL 16V8 but I would need to clip the PAL out as I wouldn't be able to desolder it without damaging the PCB given my experience with other parts on the PCB it's very flaky.

See this thread (Correction of +2 hardware errors) over on World of Spectrum

Mark

Morpheus wrote: ↑Sun Aug 04, 2019 8:22 am I am contemplating replacing the PAL with a GAL 16V8

I’d advise against that. This is a mod that will alter the behavior of the +2. We already have plenty of emulators that have that same “mod.” You’ll also lose the “precious” ULA snow effect. I’d leave it be.

Morpheus wrote: ↑Sun Aug 04, 2019 8:22 amneed to fit an external Ear socket

Be aware that you’ll also need to replicate the original 128K circuitry that cuts of negative swings of the audio signal fed into the ULA’s EAR pin. While it might be okay as is, it is rumored that the ULA might be damaged if the signal goes below 0V. You basically need a diode to take care of that. A capacitor in series will also remove the DC offset. Check the relevant part of the 128K schematic for more info.

Thanks, I’ll leave the PAL as is based on your advice. I wasn’t keen on clipping it out as it was a one-way choice. I was looking at the mod that connects the ear input of a new mono socket to pin 5 to the LA6324. Then I can use my TZXDuino I was also going to try extra grounding on the cassette unit where the hum originates from.

If you add the extra circuitry, you could replace the 3.5 mm jack with a single TRS one and wire it the same way it’s wired on the +3. If you don’t want to bother with all that, you can just buy a generic cassette adaptor. They’re cheap and work well.

Ast A. Moore wrote: ↑Sun Aug 04, 2019 2:17 pm If you add the extra circuitry, you could replace the 3.5 mm jack with a single TRS one and wire it the same way it’s wired on the +3. If you don’t want to bother with all that, you can just buy a generic cassette adaptor. They’re cheap and work well.

If I remember you made a video of this using something called a Schmitt trigger?

Morpheus wrote: ↑Mon Aug 05, 2019 4:52 am If I remember you made a video of this using something called a Schmitt trigger?

Nope. Definitely wasn’t me. Perhaps you’re referring to this video from Joules per Coulomb:

https://youtu.be/qUIv-A_DOc0?t=1014

Yes, sorry I think you posted it a while back and I thought you had made it. My bad

Morpheus wrote: ↑Tue Aug 06, 2019 9:33 am Yes, sorry I think you posted it a while back and I thought you had made it. My bad

No worries. I’m actually flattered.

Sorry for reviving an old thread, but I have this exact same problem with a +3. Very dark, grey output from the RGB port.

Looking at the motherboard I can’t find TR8 or TR9 at all, so I’m assuming the +3 is different or I’m being stupid in some way!

Anyone got any suggestions for a +3?

Not sure if it related but the output from the RF unit is completely black. The RGB scart compatibility modification has been done by the previous owner.

Thank you.

Treguard wrote: ↑Fri Jan 01, 2021 8:35 pm Sorry for reviving an old thread, but I have this exact same problem with a +3. Very dark, grey output from the RGB port.

Your TV expects composite video, instead of RGB. The +3 doesn’t have a composite video signal, and what you’re seeing is the composite sync signal fed into the composite video input. If you can manually force your TV to accept RGB video via SCART, do so. If not, you’ll need to make a small modification to your +3 (replace R44 with a smaller-value resistor, such as 330–470Ω) for the TV to accept RGB video automatically.

Here are a few places to peruse: one, two, three.

A +2 grey is uses a completely different PCB with very different circuitry. So the details about TR8 and TR9 are not relevant to your board.

Are you are trying to connect via a SCART lead? If yes, do you understand that a SCART connector can carry a choice of various different video signals? TVs normally default to composite video, but the +3 does not (as it left the factory) output a composite video signal on it’s monitor/RGB port. It only produces RGB video signals.

Hence I tend to agree with Ast A. Moore.

Mark

I bought it on ebay, and it came with a RGB - Scart cable. I also bought a Scart to HDMI convertor that was advertised as RGB and +3 compatible. That also doesn't seem to be able to lock onto a signal.



Looking at the circuit, board and based on the listing, I think the modification you suggest has already been done (photo attached) but I'm colourblind and can't read the resistor coding so I don't know if it's correct.

Does this circuit board look right to you?

If RGB isn't going to work, has anyone got any idea what might cause the RF to produce a completely black screen?

I might send it back, but the rest of the machine is immaculate, so I'd like to fix it if I can.

Perhaps I just need an older telly. This is the only one I have with SCART at all. All the rest are HDMI only.

Thanks.

Treguard wrote: ↑Sat Jan 02, 2021 1:06 pm Looking at the circuit, board and based on the listing, I think the modification you suggest has already been done (photo attached) but I'm colourblind and can't read the resistor coding so I don't know if it's correct.

Hmm. It’s an odd choice to use a high-precision resistor for R44, but it looks like it’s a 610 Ohm one with a ±1% tolerance. It’s still too high, though. I’d replace it with a 330 Ohm one.

It's 680 shirley.
That is theoretically enough, though on the high side so less margin if the 12 volt rail sags a bit. I used 560 in my +3 for that reason.
330 is lower than I've seen anyone recommend before. It seems like increasing current for no benefit to me.

Guesser wrote: ↑Sat Jan 02, 2021 4:19 pm It's 680 shirley.

D’oh, you’re right. It’s a gray band.

Guesser wrote: ↑Sat Jan 02, 2021 4:19 pmThat is theoretically enough, though on the high side so less margin if the 12 volt rail sags a bit. I used 560 in my +3 for that reason.
330 is lower than I've seen anyone recommend before. It seems like increasing current for no benefit to me.

Well, the spec calls for 1–3V. As it is, we have 12/(1,000/75), which is 0.9V. 330 Ohm will give us 2.7V, which is, I agree, on the higher end, but still within spec. 560 gives you 1.6V. I think I stuck 470 in my +2A originally. Can’t remember off the top of my head, because in my case it’s moot; I select the RGB input on my TV manually. (Haven’t tested if the TV can switch inputs automatically, actually.)

That's not how you calculate a potential divider

Heh, no it’s not, but the ballpark values are close enough.

Okay,

12×75/(1000+75)=0.83V
12×75/(560+75)=1.4V
12×75/(330+75)=2.22V

Okay, time to ask the slightly obvious questions...

Is this TV known to work with other devices that produce a SCART RGB signal? E.g. DVD players?

If the computer has been modified, it’s likely that that cable is correct, but it is still worthwhile checking. Either visually by checking the wiring after removing the connector covers. Or electrically by using a multimeter.

Does the disk drive spin a disk?

If no, it’s possible that the PSU is not producing a +12V output. This would affect the +12V supply to the monitor/RGB port. It would also affect the RF/UHF TV output.

Mark

All this talk of resistors.

Have I got a surprise for you lot... hopefully soon...