CEM3381 Dual VCA

[nosaj]

Hi Juan,

What did you have that could substitute for the PA381?
Thanks,
nosaj

[J M Fahey]

Sorry missed a couple posts, will answer in sequence even if later ones change the situation, so we leave no gaps in troubleshooting.

Ok jumpered CN10 Pins 3and5 which are In1 and In2, Applied 1khz sine. Applied +5vDC to R53 (TP3) Scoped IC11-1 and varied the DC input(which makes the sine go up and down like a Volume pot)

OK, thatīs what I expected to see, the VCA IS functioning as expected, at least that half.
As you see, it does not matter what happens "inside it", we have a "black box" called "VCA" and it does the function itīs supposed to.
Now to the second half which in principle looks damaged:

jumping over to IC11-7 no signal but varying the DC voltage The scope line stays flat but jumps to different points on the scope while staying flat.

Ouch, looks like we have a damaged PA381 ... which of course is unavailable.

As a last ditch test, replace IC11 TL072.
Usually against shotgunning but itīs available, cheap, you certainly have a few in stock and it takes less than 5 minutes.
Not much hope but since itīs so easy and quick

Repeat above tests and if no good news, we go to Plan B: a crude but workable VCA which can reasonably do the deed.

EDIT:
now answering your second post:

Anythoughts on what I'm seeing? Here's my thoughts on this IC7 pins are not working correctly. I tested IC11 by inputting 1k tone at jp1 bside and get signal at ic11pin7. Anything else I can try ?

Ok, IC11 seems to be working, so .... we abandon the dead PA381 and since no replacements are available, we continue with Plan B.
Let me edit the schematic so I include the Mod.
stay tuned.

[nosaj]

Sorry missed a couple posts, will answer in sequence even if later ones change the situation, so we leave no gaps in troubleshooting.

OK, thatīs what I expected to see, the VCA IS functioning as expected, at least that half.
As you see, it does not matter what happens "inside it", we have a "black box" called "VCA" and it does the function itīs supposed to.
Now to the second half which in principle looks damaged:

Ouch, looks like we have a damaged PA381 ... which of course is unavailable.

As a last ditch test, replace IC11 TL072.
Usually against shotgunning but itīs available, cheap, you certainly have a few in stock and it takes less than 5 minutes.
Not much hope but since itīs so easy and quick

Repeat above tests and if no good news, we go to Plan B: a crude but workable VCA which can reasonably do the deed.

EDIT:
now answering your second post:

Ok, IC11 seems to be working, so .... we abandon the dead PA381 and since no replacements are available, we continue with Plan B.
Let me edit the schematic so I include the Mod.
stay tuned.

Thanks so much Juan.

[J M Fahey]

Ok, hereīs the FET VCA:



We ditch the dead PA381, we use the FET as a voltage controlled attenuator .
Attenuator is a series resistor (in principle we leave the original 39K) and a FET to ground, working as a variable resistor.
Its value will go from around 100 ohm (which is not perfect Zero but in any case *a lot* of attenuation, should be more than enough) to infinite, so no attenuation, when we vary Gate voltage from 0V to +5V.
NOTE: since available control voltage is positive, we need a P channel Fet, not the more common N channel type.
Most Fets dissappeared from the market, but J176 is still available.

We pull the dead PA381 but use its PCB holes as reference and we can mount a couple "new component" legs on them; extra component legs should be tack soldered to needed pads, if already occupied.
I guess you will have to add the two J176 from below, the copper side, should not be a problem.
If too close to chassis, then you will have to solder it from above, do whatīs most comfortable to you.

This will vary volume as expected, maybe not as smoothly, but acceptable.

If level is too high or too low, gain can be adjusted; with my values it "should" be about unity gain with +5V, which I guess was the original gain, but if not, itīs adjustable, up or down, varying a couple resistors.

Ok, get the P Fets and try it.
Good luck.

[nosaj]

Ok, hereīs the FET VCA:

[ATTACH=CONFIG]49131[/ATTACH]

We ditch the dead PA381, we use the FET as a voltage controlled attenuator .
Attenuator is a series resistor (in principle we leave the original 39K) and a FET to ground, working as a variable resistor.
Its value will go from around 100 ohm (which is not perfect Zero but in any case *a lot* of attenuation, should be more than enough) to infinite, so no attenuation, when we vary Gate voltage from 0V to +5V.
NOTE: since available control voltage is positive, we need a P channel Fet, not the more common N channel type.
Most Fets dissappeared from the market, but J176 is still available.

We pull the dead PA381 but use its PCB holes as reference and we can mount a couple "new component" legs on them; extra component legs should be tack soldered to needed pads, if already occupied.
I guess you will have to add the two J176 from below, the copper side, should not be a problem.
If too close to chassis, then you will have to solder it from above, do whatīs most comfortable to you.

This will vary volume as expected, maybe not as smoothly, but acceptable.

If level is too high or too low, gain can be adjusted; with my values it "should" be about unity gain with +5V, which I guess was the original gain, but if not, itīs adjustable, up or down, varying a couple resistors.

Ok, get the P Fets and try it.
Good luck.

2 pchannel jfets and 2 100k resistors for r74 and r76 looks like the only changes on your schematic that I'm seeing.

Will check parts I have shortly. Fooey..all I have are N channel Jfets, just ordered some off the bay Vishay j176 50 for $10.


Thanks Juan

[Enzo]

I checked my inventory sheet, only Curtis chips I stock are:
CEM3340, 3360, 3372, and 3394.

But the thought occurs to me, didn't you say you ONLY have the amp module and will be using it alone - not with the rest of the keyboard?

If so, then why have the remote voltage control over volume in the first place? Replace the VAC with a pot.

[nosaj]

I checked my inventory sheet, only Curtis chips I stock are:
CEM3340, 3360, 3372, and 3394.

But the thought occurs to me, didn't you say you ONLY have the amp module and will be using it alone - not with the rest of the keyboard?

If so, then why have the remote voltage control over volume in the first place? Replace the VAC with a pot.

No as soon as it's working I want to put it back in the piano. I only just have the power amp right now.

Jason

[nosaj]

Ok, hereīs the FET VCA:

[ATTACH=CONFIG]49131[/ATTACH]

We ditch the dead PA381, we use the FET as a voltage controlled attenuator .
Attenuator is a series resistor (in principle we leave the original 39K) and a FET to ground, working as a variable resistor.
Its value will go from around 100 ohm (which is not perfect Zero but in any case *a lot* of attenuation, should be more than enough) to infinite, so no attenuation, when we vary Gate voltage from 0V to +5V.
NOTE: since available control voltage is positive, we need a P channel Fet, not the more common N channel type.
Most Fets dissappeared from the market, but J176 is still available.

We pull the dead PA381 but use its PCB holes as reference and we can mount a couple "new component" legs on them; extra component legs should be tack soldered to needed pads, if already occupied.
I guess you will have to add the two J176 from below, the copper side, should not be a problem.
If too close to chassis, then you will have to solder it from above, do whatīs most comfortable to you.

This will vary volume as expected, maybe not as smoothly, but acceptable.

If level is too high or too low, gain can be adjusted; with my values it "should" be about unity gain with +5V, which I guess was the original gain, but if not, itīs adjustable, up or down, varying a couple resistors.

Ok, get the P Fets and try it.
Good luck.

Why would I need to put jfets on opposite side of board. i was thinking about using a 8pin DIP socket to just plug the jfets into. That should be fine I would think, no?

nosaj

[nickb]

Kinda missed this thread. I'm 95% sure I have some PA381's if you still want one...

[Enzo]

There y'go^^^

Why would I need to put jfets on opposite side of board. i was thinking about using a 8pin DIP socket to just plug the jfets into. That should be fine I would think, no?

It is the circuit that matters, put the parts where you like.

[J M Fahey]

Why would I need to put jfets on opposite side of board. i was thinking about using a 8pin DIP socket to just plug the jfets into. That should be fine I would think, no?

nosaj

It was not an "order" but a "suggestion", accompanied by "do whatīs most comfortable to you."
That said:
1) I donīt trust IC sockets and even less here, they wonīt firmly grab the thin Fet legs.
2) in any case some Fet connections are NOT to IC holes/pads anyway.
3) in holes 3 and 5 youīll need to insert BOTH a resistor leg and a Fet leg .... hard to solve with an IC socket.
4) top Fet Source goes to R77 ground pad (on bottom) or scratched and tinned top leg ... your choice.
5) bottom Fet Source goes to R75 ground pad (on bottom) or scratched and tinned top leg ... your choice.
6) thereīs more than one way to skin a cat ... I suggest a couple just in case
7) just read nickbīs post ... if an original is available, by all means use it

[nosaj]

It was not an "order" but a "suggestion", accompanied by "do whatīs most comfortable to you."
That said:
1) I donīt trust IC sockets and even less here, they wonīt firmly grab the thin Fet legs.
2) in any case some Fet connections are NOT to IC holes/pads anyway.
3) in holes 3 and 5 youīll need to insert BOTH a resistor leg and a Fet leg .... hard to solve with an IC socket.
4) top Fet Source goes to R77 ground pad (on bottom) or scratched and tinned top leg ... your choice.
5) bottom Fet Source goes to R75 ground pad (on bottom) or scratched and tinned top leg ... your choice.
6) thereīs more than one way to skin a cat ... I suggest a couple just in case
7) just read nickbīs post ... if an original is available, by all means use it

i just didn't understand quite why 3 explained it for me.

Just waiting for nickb response, jfets are on the way.Can't hurt to have parts to instantly try something instead of waiting in the future.

Thanks all.

nosaj

[nosaj]

Kinda missed this thread. I'm 95% sure I have some PA381's if you still want one...

That's a definite PM sent with my info.
Thanks,
Jason

[nickb]

That's a definite PM sent with my info.
Thanks,
Jason

..and responded to. I'll have to venture into the dustiest & darkest places of my Old Parts Emporium.

[J M Fahey]

i just didn't understand quite why 3 explained it for me.

Just waiting for nickb response, jfets are on the way.Can't hurt to have parts to instantly try something instead of waiting in the future.

Thanks all.

nosaj

P-Fets are an ignored and very useful tool, great advantage being they can be used as switches or attenuators being contolled by **positive** voltages, which are more easily available (specially on pedals ) .

Look at Phasers; in order to use more common (specially way back then) N Fets as variable phase shifters, they are *forced* to "elevate" Sources so Gates can be made "negative" (actually "less positive" ) for control.

P Fets provide elegant simple solutions, they can be straight driven by, say, Logic voltages, output of microprocessors, etc. , which can provide *any* voltage you imagine ..... between 0 and +5V that is

I always have at least a dozen handy, saved my bacon many times.