MX12-Luma Mod schematic advice

[PAUL_TEMPLE]

Hi,

I’m working on re-documenting the MX-12 Luma Mod. (Which I will upload when complete)

It was done in 2006 by Brian Murray however he was killed in a bike crash before he could verify his documentation. MGCJerry did add some notes to the documentation, correcting a few small errors. I managed to scrape it off the waybackmachine recently.

I’ve re drawn the schematic for the extra multiplexer chip which should be installed.

Everything makes sense except one part I am confused about, this is unclear in the documentation also.

I am concerned that if directly connected to, the existing multiplexer chip IC4 (PIN 12) (cmos4053) could be overloaded. I am not sure if the pin would need isolation from the existing circuit. It is fed by a transistor with a very low AC voltage. It is not a question of it receiving 2 x +5v or shorting it to ground.

I would think that if IC 4 (PIN 12) did need isolation, this would be mentioned in the original documentation, (as the chrominance section is isolated and documented) however I would enjoy confirmation that it’s fine to direct connect, because this step is super vague.

Part of this step uses a ceramic cap which bridges two pins on the new multiplexer. I was planning to use something in the pf range for this, maybe 470 / 220, though I’m not certain. I don’t think it’s a critical value or it would (probably!) be mentioned.

This concerns the luma, Y part of the mod, the pins connecting to (IC4 PIN 12) are luma signals, either from source1 or source2, or the original. The truth table for the 4052 output selection is shown in the schematic.

If anyone has any advice on these two points, I would be happy to hear your insight!

Here are the pics of the sections in question - I can send on pdfs if a deeper insight is needed.

Ta friends,

(black text is B.Murray, red text is MGC.Jerry)

[revjon]

So, I kind of don’t think it’s a 4052? The schematics you share that look like they’re from a service manual line up better with a 4053. The 4052 has two 4:1 multiplexer/demultiplexer units, and the 4053 has 3 2:1 multi/demultiplexers, the pins of which match the pins on the schematic from the service manual.

That said, if the question is whether bridging one side of the multiplexer/demultiplexer to the other with a capacitor will cause it to overload, I very much doubt it. Since the chip doesn’t do any amplification, just switching, I think it would be glitchy (which I assume is what is desired), but I don’t think you run any risk of messing up the chip unless you exceed the input voltage (Vcc + 0.5V per the datasheet), which does not seem likely based on your schematic.

[PAUL_TEMPLE]

Hi! Thanks for your insight. Sorry if it was not clear:

The 4053 you are correct is in the machine already - IC4 and IC3. Yes, these are screenshots I have made from the service manual, not the mod docs. They show IC 3 4 and 5.

There is an extra chip which is installed into the machine as part of the mod, which is a 4052. This is what I have drawn the schematic of connection in kicad.

So the extra 4052 connects to the 4043s. It is just switching indeed, it’s not amplifying.

My question is - Will connecting the output of the 4052 LUMA signal, to the pin 12 of IC4, potentially cause an overload either in the IC4 or downstream?

I ask because, pin 12 of IC4 is already fed by the base of Q30, the white colour signal.

Thanks again! I think I will just connect it up as it is stated.

[revjon]

Ahhh, ok. Sorry, the similarities between the two chips had me confused, I should have read more closely. I still think that it’s not going to overload anything since there’s no amplification involved.

[PAUL_TEMPLE]

I re-read my post and it’s pretty rambling! No worries : )

I understand that there’s no amplification in the chip itself,

however, could it be considered a form of amplification: to apply two different signals to the same pin at the same time, where there was only a single signal previously… ?

[revjon]

the caveat is that i am not an electrical engineer, just a hobbyist with lots of experience with CMOS chips and video synthesis…
I don’t think that having the two signals on the same pin has an additive effect. The two signals just get jumbled together, but the peak-to-peak voltage doesn’t get higher than whatever voltage the system is running at. Given that these are two CMOS chips, I’m assuming everything is 0 to +5v in this part of the circuit.
To test this hypothesis, just now I went down in the basement and played around with my CMOS-based video synth. I took two different oscillators from two different chips and hooked them up to each other and then measured the resulting signal using my oscilloscope, and it was a noisy mix of the two frequencies, but the amplitude wasn’t any higher (just a bit of fuzz on the top and bottom caused by the unsophisticated mixing). So I feel pretty confident it will be fine in your case.

[PAUL_TEMPLE]

Thank you so much! I am also not an EE. I taught myself. Getting better all the time, thanks in part to exchanges like this!

I believe you are correct in your hypothesis, though it’s great you tested it! Thanks!
It is all 5v logic yes.

I’m really curious about your CMOS v.synth - do you have a project page for it, I’d love to find out more. I take it you DIYd it?

I designed a nice eurorack comparator module recently, which can bridge audio and video freqs. I have some pcb/panel sets for it.

In the past I created an audio CMOS computer with three gates, inverters, attenuverters etc, from datasheet/cmos principles. I’m sure I could use it with my videosynth though it’s a bit wonky. I think to redesign it in eurorack as well at some point.

Thanks again for this conversation, I have enjoyed it.

I’ll let you know when I complete this modification!

[revjon]

yes, I did DIY it, I’m working on a project page for it now and plan to post about it here as soon as it’s published!

[PAUL_TEMPLE]

Look forward to that! I tried out the Luma mod yestereen, it worked kind of, you could see it was doing it partially -

Then I realised that I’d misread the silkscreen and connected to R79 and not R71!
(that thing is densely populated and the silkscreen is an artwork)

I think it will work fine when I connect to the right part of the circuit!

[PAUL_TEMPLE]

Yes the keyer mod is working much better now. Although the mixer needs some calibration/attention. I think the chroma phase is out - one input is very washed out looking. This is affecting the keyer mod.

If I put the same input on both inputs, I can see a phasing effect when you wipe between channels.

I’m hoping it might be a trimpot solution, but maybe it’s a failing capacitor.