Inves Spectrum+ - Investigating the Green Tint (Part 2)

[@bradkamrath]

Super impressed with your never-give up attitude, Noel! It is so satisfying to see you finally get to the root cause of the issue and there are a LOT of lessons you are teaching electronics buffs old and young alike!

[@NoelsRetroLab]

Thank you so much! Glad you enjoyed it.

[@tony359]

What I like about your videos is that you're not editing out the dead ends. Everything is documented, including mistakes. That makes your videos very educational and enjoyable! Thanks!

[@NoelsRetroLab]

Thank you so much! I'm glad you're enjoying even the dead ends. I also think they're quiet interesting.

[@galier2]

Who needs crime novels when we have such compelling detective work! Awesome.

[@gartmorn]

Like you say at the end of the video you've come out of the process with a working computer and also a lot more knowledge and understanding of how it works which is a real bonus! A big part of fixing things is the learning process!

[@moterov4]

I was thinking to tell you in the previous video, that I was very suspicious of the RF Module, but sorry, I didn't want to seem "listillo" without proofs ;-). The coils have a ferrite core that over the years lose magnetization and it is necessary to recalibrate them in height so that they give the original value in Henries (if possible) or replace them. You can try to retune the chroma signal with the oscilloscope either with these or with new coils. But it's a difficult process (I remember in FM tuner). Your videos are great and I always learn new things. Congratulations.

[@NoelsRetroLab]

Haha, yeah I was also suspicious of it, but managed to convince myself that that wasn't the problem. Oh well, live and learn. Glad you're enjoying the videos though!

[@BertGrink]

I too learn new things each time! :D

[@andygozzo72]

those ferrite cores cant 'lose' magnetism as they have no 'standing'/quiescent magnetism to start with, but i suppose possible they may absorb moisture and their magnetic permeability 'could' change in some way ...

[@moterov4]

@@andygozzo72 OK, but do not you think that the ferrite core acquires some magnetism in the first instant that intensity passes through the coil, in the same way that we magnetize a screwdriver?

[@andygozzo72]

@@moterov4 it 'concentrates' magnetism, increasing inductance of any coil its within, it has none of its own , in these cases, it doesnt remain magnetised when the coil current goes away..... ferrite magnets such as used on speakers are a different thing

[@delmonti]

brilliant bit of video journalism there. Really enjoyed the process.

[@NoelsRetroLab]

Thanks so much!

[@rdoetjes]

That’s so cool! I didn’t know there were Speccy clones. Well I hardly know anything about speccy’s that’s why these sort of channels are so cool and educational.

[@alonecoder600]

There were more clones than the original Spectrums :) The total number (including xUSSR, Spanish, Portuguese, Brazilian, Argentinian, Polish, German, Czech, Hungarian, Romanian, Italian, Indian clones) is somewhere between 13.5 million and 22 million.

[@sarahwhitman1754]

This was epic, your perseverance is impressive Noel!

[@NoelsRetroLab]

Thank you! That's the fun thing about these old machines: I feel that you can completely understand them and figure out what's going wrong with them. As soon as you get into more modern computers... they get too complicated! 😃

[@pparadigm]

As always, very good. You are very inspiring Noel. Thank you.

[@paoloselvini3449]

Simply amazing! What kind of technical background/skills do you have Noel? You're a true PRO! You do not find such competence very often and so well described. WOW, totally impressed!

[@NoelsRetroLab]

Thank you! I studied Computer Systems Engineering, but that was only partly useful for this. Most of my knowledge is self taught with low-level programming, and then a few years ago I went deep into retro electronics by repairing my own computer.

[@paoloselvini3449]

@@NoelsRetroLab I studied Computer Engineering too but I am waaay far from your level of expertise! I was thinking that it could be interesting to make a RetroLab video where you speak a bit about how it all began, how did the lab start to grow, which were the first machines you looked at and fixed, where/how did you deepen/improve your skills in "retro-electronics"... stuff like that. Of course you would skip any personal information, but in a way I think it would be very informative for guys (like me) eager to become a bit knowledgeable and skilled at doing stuff like you do and show us so well. Think about it, no rush ;-)

[@Xoferif]

I really enjoy the methodical way you work through problems! Much more educational than my actual education! =)

[@NoelsRetroLab]

Thank you! That's what I'm shooting for, so I'm really glad to hear that comment :-)

[@AdmiralHorror]

These videos are awesome! Please keep making them, and wishing you a million subs in the near future.

[@NoelsRetroLab]

Thanks! Once my current project is over, I should be back to making them more often.

[@IgnacioMonge]

I love your videos. Diagnosing step by step like a detective. Amazing Lopis!

[@scottlarson1548]

This is why I don't miss analog video.

[@supermarketsweeps]

i was gutted when you said you was going to put it on the shelf for another day , well chuffed you fixed it !

[@NoelsRetroLab]

I thought I had gone as far as I could! Fortunately I was able to go one more step 😃

[@kurbeco]

Really enjoyed learning about this as you were fault finding. Thank you 👍

[@NoelsRetroLab]

You're welcome! Glad you enjoyed it! 😀

[@konturgestalter]

This video is incredible. i literally sit here and take notes in every video

[@NoelsRetroLab]

So glad to hear that!

[@Dejotaerre]

Best channel about Speccy ever!

[@NoelsRetroLab]

Thank you! Glad you're enjoying it! 😃

[@chainq68k]

There's so much to learn from this video. How to methodically diagnose parts of a complex circuit, and why always start with triple-checking your initial assumptions. :) Too bad you didn't try to fix or at least analyze the workings of that RF modulator, to understand where the green tint really came from, even if of course just jumping to the composite mod is a totally practical solution in 2020. Still, very enjoyable fix video! Looking forward to part 3.

[@NoelsRetroLab]

Thank you! Glad you liked it! Yes, I have to admit I don't know much about the inner workings of RF modulators. And besides, they're all crammed and hard to work in :-) So yeah, composite is plenty good for now.

[@gertsy2000]

"Never give up, never surrender!", So basically you had to up the signal to get a non-tinted picture. As a simple Composite Mod 'break-out' had the same problem?, Or was it because you didn't break the composite line to the modulator in the first test? In any case well done Noel! You are a positive example to us all. You deserve more than 2k subscribers. Great troubleshooting in a consistent and well produced package. I look forward to every video you make.

[@NoelsRetroLab]

The transistor was just to boost the signal and not get is so dark. The culprit was the RF modulator that somehow was adding that tint during the modulation process. Just bypassing it and using the composite signal at its input was not green, but it was dark, which tricked me last video thinking it was still greenish :-/

[@up2tech]

As Sherlock Holmes once said: «Once you eliminate the impossible, whatever remains, no matter how improbable, must be the truth.»

[@onkelnb]

Great Video! I hope for more background information about the Inves Spectrum+. Very interesting ZX Spectrum clone!

[@NoelsRetroLab]

Thanks! Yes, that will be in part 3 finally, now that it seems to be working.

[@onkelnb]

@@NoelsRetroLab Great! I wonder if Investronica also continued building an "Inves Toastrack" after Sinclair was sold to Amstrad. As they did most of the development, it seems to me like the easiest option. Isn't it somehow strage that they put lots of efford into copying the quite outdated Spectrum+?

[@NoelsRetroLab]

@@onkelnb They never did as far as I know, so it would be a super rare find it someone found one. I would imagine it would have been pretty easy for them to do though.

[@frankowalker4662]

Great diagnostics. And thank you for suggesting TeraTerm, it's perfect.

[@NoelsRetroLab]

Thanks! Glad that worked. TeraTerm felt like a breath of fresh air after trying to use HyperTerminal! :-)

[@home-w8y]

Capacitor came off without any resistance ...haha...I see what you did there!! ;)

[@damianvila]

Finally! Yay! Was waiting for this one. Great video Noel!

[@NoelsRetroLab]

Thanks! :-)

[@BertGrink]

The block diagram at 0:48 took me by surprise, because the Sinclair 48K models have YUV color encoding, or at least the early models did.

[@NoelsRetroLab]

Right! That's one of the huge differences internally with this clone. In part 3 I'll get into some of the actual differences and how they affect things running on it.

[@BertGrink]

@@NoelsRetroLab Kewl man :D I'm looking forward to the next episode already.

[@ElectronGordo]

Bravo! I wouldn't figure it out, that was the modulator and also requires a more powerful signal the composite video... so transistor... amazing! thanks

[@MegaMukluk]

This might be stupid, but in the last episode you talked about a 7812, the to-92, and your supply voltage was only 11.6V. That should be a problem, since the 7812 needs 14.6V minimum for a stable voltage out. What is the 7812 used for? It would seem stupid to supply more than 15V in since the MC1377P seemed to have a maximum of 14V in, and a typical 12V in.... Is the 7812 output connected to the MC1377P supply voltage? If the board has 12V in, the 7812 is useless. I know that you fixed it already, but I am curious as to the usage of the 7812 voltage regulator.

[@uwezimmermann5427]

someone else might have already commented this, but the capacitor in the QL is not mica but a polystyrene foil capacitor and the one in the Inves is a regular ceramic one. As long as you have a carbon film resistor as the other part in the circuit, a high-quality capacitor would not make any sense here...

[@HoboVibingToMusic]

Grats on 2k subs! :D

[@eebaker699]

Great video. I like your troubling shooting techniques. Can be transferred to other systems. 😁

[@NoelsRetroLab]

Thanks! Yes, that's the idea. I know not many people are going to be repairing Inves Spectrums, but a lot of this is general knowledge and troubleshooting which can be applied to other systems as well. Glad you enjoyed it! 👍

[@fawad.rashid]

Wow i am being off topic here however i love your t-shirt, isnt that the character from Knight Lore ??. I must have spent endless hours playing that game on my CPC6128!

[@NoelsRetroLab]

Yes! Me too! Loved that game! :-)

[@KeanM]

New subscriber via EEVblog forums. Really good troubleshooting work. Always important to recheck your assumptions when things don't make sense.

[@NoelsRetroLab]

Thanks! Glad you enjoyed it!

[@macstar8317]

Great work. Really interesting vids.

[@NoelsRetroLab]

Thank you very much!

[@retrocomputeruser]

Noel, that seems so strange because once the colour circuit gets to the composite level, all RGB colours are going through one signal. That is so weird how the modulator was the culprit. I can fully understand why this was the last resort on finding the fault.

[@NoelsRetroLab]

Right. I don't know how exactly the RF modulator works, but it must do something with the color subcarrier to throw it slightly off.

[@andygozzo72]

@@NoelsRetroLab the rf output circuit of them is 'push pull' so its possible one half may have gone out of balance to the other, so distorting the waveform a bit ...

[@jesusarias4320]

The 7805 heatsink is a bad place to put the ground clamp of the scope probe. Try instead to clamp the ground to some decoupling capacitor near your signal in order to reduce the noise

[@TaberBucknell]

Congratulations on finding the solution! That was quite the adventure. Question ... So why when you made your own RGB TV connection with the RGB connections, breadboard, and cable did the green tint still appear? Wasn’t that upstream of the faulty RF modulator?

[@NoelsRetroLab]

Thanks! Glad you enjoyed it. With the RGB out we had no green tint at all (it was just a tad dark). Which is to be expected because those RGB signals were purely digital. So we knew it had to be somewhere after that.

[@pev_]

So the RF modulator somehow puts an asymmetric load on the composite signal that makes it green... that is weird.

[@SimonJackson13]

A low pass filter loading effect on the subcarrier signals? Output impedance of composite?

[@LambertZero]

Somewhere in the middle of that ordeal I said "just take the RGB TTL singals, feed them into rgb2hdmi, and enjoy a pixel-perfect picture" 😀

[@OmgSaySomething]

After figuring out the composite signal level was low, I wonder why did not just feed the amplified signal to the modulators input?

[@sa3270]

I kept thinking about that song Kermit the Frog used to sing, "It's Not Easy Being Green", as I watched this.

[@thomasandrews9355]

I was screaming at the screen for a while saying "CHECK THE RF MODULATOR" haha, I've spent HOURS on issues due to things like shorts that I caused...Learned to check for those things early on.

[@NoelsRetroLab]

Haha, I know! I would have checked it earlier had I not discarded it last week as not being faulty. Oh well :-/

[@thomasandrews9355]

@@NoelsRetroLab trust me, spent about 15 hours learning to check things. i spent 3 months fixing my first amiga 2000. i've fixed over 15 in the last 4 months as part of my new repair business lol.

[@JorgeCarvalho_web_dev]

Hi Sir, you rule!!! Are you an hobbyist or do you have some good electronic background basis like electronic engineering? Love you videos :) This channel will very soon compete with some popular ones like gadgetuk and others. Keep the good job, I feel that I have learned something today.

[@NoelsRetroLab]

Thank you! It's just a hobby, I'm a game developer professionally. I did study Computer Systems Engineering in college, but I never really *got* electronics until I started getting my hands dirty with these computers years ago. So I really believe anybody can do it with a bit of patience and learning as much as they can from videos and books.

[@KolliRail]

This is an extremely good fault finding video. Top!

[@NoelsRetroLab]

Thanks! Glad you liked it!

[@konturgestalter]

can you tell me by any chance where i do get these clamps you attached to the jumper wires to connect to the chip when you tested the video output?

[@NoelsRetroLab]

Those actually came with the oscilloscope probe. They fit right on top of the probe and snap in place. Super handy when I need to keep them in place and test other things.

[@Akiasergal94]

congrats on finding it

[@leojrgensen2819]

Thanks for another educating video 👍

[@ThePillenwerfer]

That 'mica' capacitor you fitted looked more like polystyrene to me.

[@NoelsRetroLab]

I'm not familiar enough between the two to know the difference. Do they have similar characteristics? Meaning, is it possible that the encoder suggestes a mica capacitor but the used a polustyrene one because it's equally accurate?

[@ThePillenwerfer]

@@NoelsRetroLab I'm no expert but as far as I know both mica and polystyrene are regarded as being more accurate and stable than other types. I've found this: www.electronics-notes.com/articles/electronic_components/capacitors/polystyrene.php which says "polystyrene capacitors can often be used in place of silver mica or ceramic disc capacitors."

[@NoelsRetroLab]

@@ThePillenwerfer That makes sense! I didn't know the difference because I don't often encounter them in these computers. Thanks for pointing that out!

[@thegreatjonzini]

Quite the journey ;) well done!

[@NoelsRetroLab]

Thanks!

[@penniesshillings]

Very useful! Love your videos!

[@zxrenew5642]

Superb Sir!

[@NoelsRetroLab]

Thanks!

[@rastersoft]

What do you think about trying to implement the floating bus? This computer, like the Plus3/Plus2A, doesn't have the "floating bus effect", so some games doesn't work because they wait forever trying to detect when the ULA starts to read the screen. But there is a mod for the Plus3 that fixes this (spectrumforeveryone.com/technical/performing-the-in-ff-mod-on-a-spectrum-2a-3/ ) and it should be quite easy to implement it in the Inves too, just by adding a resistor between the ULA and the Z80 data buses (I don't know it you know this great description of the inners of the Inves Spectrum plus. where it explains details like how can it have no memory contention at all: www.zxprojects.com/inves/ )

[@NoelsRetroLab]

That's interesting. I didn't think about actually implementing the floating bus here, but you're right, it's totally doable. I'm writing it down for a future project, or maybe next time I revisit the Inves. Thanks!

[@laurentvoisin8129]

So talented 😮 !

[@kriswillems5661]

Very impressive. Are you an electronics engineer?

[@NoelsRetroLab]

Thanks! No, I'm a game developer. I did study Computer Systems Engineering in college, but I didn't learn "real" electronics until I started tinkering with computers on my own years ago.

[@kriswillems5661]

@@NoelsRetroLab I can see you really understand this stuff. I am an electronics engineer. Thumbs up!

[@andygozzo72]

those modulators are 'usually' reliable...but.... all my 3 acorn electrons have had their modulators changed at some point, they're date coded after the electron was discontinued so cant be factory originals...maybe there was a bad batch of them???? and , some orics have trouble with low amplitude rf out, .... hmmm... i have a spectrum plus 3 with video 'issues' but changed the modulator with one i got from a lot of 10 from a surplus parts seller, still didnt cure it! maybe those ones were also 'suspect'??!

[@NoelsRetroLab]

Weird! I've had the occasional RF modulator not work (although I don't know if it's because someone tried doing something with it), but never a partial failure like this. So they've been pretty reliable for me.

[@andygozzo72]

@@NoelsRetroLab yep, never yet come across a definite 'dud' one, but those electron replacements are a bit suspicious... thing is , not easy to test for 'partial' failure...i think i have the internal schematic somewhere, its input is actually a 'pull down' emitter circuit of the two push pull rf output amps., not a base input circuit, so to drive it you pull a current down, rather than drive a voltage into it, you usually have external resistor(s) on its input to ground so you can voltage drive it..

[@andygozzo72]

it can be difficult to repair them as component positioning can be critical in some parts of them , due to the very high frequency

[@OthmanAlikhan]

Thanks for the video =)

[@NoelsRetroLab]

My pleasure!

[@SeanChYT]

Awesome!

[@jeffm2787]

Good video, but LOL, little ripples on the clock and you have that long ass ground lead on the scope probe. Edit: OK, you have some other signals that are better, but still you have ringing on them from the ground lead. Not a big deal.

[@thomasandrews9355]

Those are MLCC caps, they break very easily. Not surprised it broke haha.

[@NoelsRetroLab]

Yeah, I'm still wondering how it was even working. It was just suddenly in my hand without any force. I suspect it would have broken next time the computer would have been jolted, so maybe better that way.

[@andygozzo72]

@@NoelsRetroLab yep, they break very easily, i tried to remove some from unwanted boards to keep as spares, not many survived...

[@donaldklopper]

Ouch indeed

[@NoelsRetroLab]

😜😃

[@jbinary82]

Impressive!!

[@javierrivas6604]

Increiblebleble!!!! Ole tus güeb.... :-)

[@fernandoquinto7716]

I don't have good memories about this model.

[@juanbrualla9465]

WOWWWWWWWWWW :)

[@romandomashenko8950]

Это был качественный секс! ;)

[@joshpayne4015]

First? Love your videos, Noel.

[@NoelsRetroLab]

Thanks!

[@TheShadow0515]

Of course there's no resistance. It's a capacitor.