Need help debugging big box micro synthesizers

[Kevin Mitchell]

Hey gang. I've started a similar thread on this earlier in the year and had to bench things as life is unpredictable. I have a couple of micro synthesizers, a guitar one (V2) from the early 80s and a bass one (v3) from the 90s.

The 80s unit I had purchased from an ebay pawn shop is not functional - despite the ad noting otherwise. The bass version was likely stomped out of frustration from not working as half the sliders had to be replaced but after the repair it's still not functional. Sadly my friend who had given me the unit to fix up had passed away recently. So I feel I have some unsettled business getting the unit functioning again.



Here's a schematic with some notes on the corrections



Let's start with the first one. I've followed the signal through the unit and verified that all of the voices and filters are working. The signal dies out as it's entering the last OTA, the VCA.
I've circled the spot where the signal dies. It's strongly present before the 47k resistor at the + input.


So I had two thoughts. The first thing I had done was replace the EH1048 with a socket and tried a few CA3094 chips with no changes. Thinking it's a fault in the circuit that opens the VCA (the Attack Delay) I had explored and found that the OTA is getting voltages -10v to about +6.5v with the speed working in respect with the attack delay slider. Point 6 on the schematic is indeed high on silent as it should be. So I think the sweep is working fine?

I cannot for the life of me get the VCA to open up as it should and I'm just about at the end of my rope with this one. It's very frustrating. Like running a marathon only to eat pavement as you're approaching the finish line.

Any ideas? 
-KM

[Kevin Mitchell]

A little unrelated rant...

Just to show authenticity I'm pretty sure the 80s unit was in the first V2 production run as the latest date stamps are from August of 1980 on the sliders and the ICs were stamped between 1979 and early 1980 with EH house marks (they relabeled many of their ICs back in the 70s & 80s).


Felt it was worth noting as I see "vintage 70s & 80s" listings all over the place that are BS. Easiest way to tell is that the external transformer "export" models didn't exist until the 90s reissue started up. Something I hope folks are weary of when browsing the market as the majority of the guys trying to sell these don't care for honesty but rather make up a ridiculous price on false claims - thus hurting the honest market (see reverb & ebay). They take a 90s unit that is beat up to all hell, call it "vintage" and jack up the price. It's sickening really.

-KM

[Scruffie]

Checked for continuity between pins 2 & 3? If R97 had for some reason gone short, your signal would be tied to ground.

[Kevin Mitchell]

Checked for continuity between pins 2 & 3? If R97 had for some reason gone short, your signal would be tied to ground.

I've done a resistance test on R97 and got just above 1K. So it can't be shorted there.

I'll post voltages from A14 and a scope shot of the sweep (pin 5) shortly.

-KM

[Kevin Mitchell]

Okay. New discovery. I was measuring the VCA sweep on the wrong side of the 47k resistor (R98). After the resistor (on pin 5) it's stuck at about -9V with no shorts detected. I pull the IC and it sweeps correctly at that spot. So something is going on that is keeping pin 5 low - and it's not anything connected to pin 5. I've tried quite a few CA3094 variants with no luck so it doesn't seem to be an internal short.

Here's a shot of of the R98/Q4 junction. On attack it goes to about -9V and raises to about +6.5V. The speed is controlled by the Attack Decay slider. This scope show's it at max (slowest attack).


So when the IC is installed pin 5 is stuck at -9V even though R98 is carrying the controlled voltage. Just like the guitar signal - it goes *poof* once it hits the chip.
Continuity tests shows no shorts to the negative rail.

Idle voltages for A14 are...
1-  -363mv
2- 0 (GND)
3- -135mv
4- -9.6V (V-)
5- -8.9V
6- 0
7- 7.7V (V+)
8- 0

The mystery only deepens 
-KM

[Scruffie]

Pin 5 is controlled by current, not voltage.

You could try tacking a suitable sized resistor from +V to pin 5 (check the datasheet for its max current, I can't remember if it's 1 or 2mA) and see if you can just keep it open, that would narrow it down to a problem with the OTA or a problem with the Attack section.

I do notice one thing in your voltages, isn't the supply meant to be +10V/-9V or something (going from memory here) 7.7 might be low on the positive side.

[Kevin Mitchell]

Usually it's powered by it's 24VAC transformer. Since it's out of the enclosure it's being powered by the EH-24VDC power supply so it's a little lower than the datasheet implies. +8 and -10 appears to be a common measurement for folks. Regardless of the power supply it's still behaving he same.

I'll try to open the VCA with a resistor. But it doesn't explain the OTA sucking the signal down to nothing. I've tried so many with the same results.

Even though the filter seems to work I wonder if there's something wrong with the trigger circuit causing this mess.

-KM

[Scruffie]

Usually it's powered by it's 24VAC transformer. Since it's out of the enclosure it's being powered by the EH-24VDC power supply so it's a little lower than the datasheet implies. +8 and -10 appears to be a common measurement for folks. Regardless of the power supply it's still behaving he same.

I'll try to open the VCA with a resistor. But it doesn't explain the OTA sucking the signal down to nothing. I've tried so many with the same results.

Even though the filter seems to work I wonder if there's something wrong with the trigger circuit causing this mess.

-KM

Ah, okay, as I said, I was going from memory so ignore the voltage question

No it wont explain the issue but it narrows it down, if the OTA opens, you know the chip isn't the issue, if it doesn't you know it is.

Those old EHX boards do get the occasional hair line trace fracture, the caps dry up, I even had one with a shorted mylar cap which I wasn't expecting, that one took a couple of hours to track down. If there's no rational cause, you know where to look so it's trial and error.

While you're there, you should probably provide voltages for the 3 transistors in the attack delay network.

[PRR]

> the signal dies. It's strongly present before the 47k resistor at the + input.

100% dead? Or just 98% dead? The 47k:1k divider is a heavy cut. And you expect milliVolts at the pin because it is a naked transistor and any more would overload.

[anotherjim]

Is C29 good? I know you have signal across it, but so would a piece of wire! Maybe a DC offset current would force the OTA inputs into either saturation or cut off so no signal gets through.
R104 also has to be good? OTA current output needs something to dump current into in order to develop a voltage signal output.

[Kevin Mitchell]

100% dead? Or just 98% dead? The 47k:1k divider is a heavy cut. And you expect milliVolts at the pin because it is a naked transistor and any more would overload.

I'd say about 98%. It's very, very low but it's still there.

Is C29 good? I know you have signal across it, but so would a piece of wire! Maybe a DC offset current would force the OTA inputs into either saturation or cut off so no signal gets through.

I'm pretty sure the capacitor is OK. I had to remove it to do a capacitance test and it seems okay.
I don't think R104 (the 1k pulldown at the output) is problematic as it wouldn't cause pin 5 to stick as it is - same thing with C29 I think. But it wouldn't hurt to check.

I'm open to anything at this point. I'm going to take voltages for the trigger section. Maybe someone can tell me if something looks off - even though the filter seems to sweep accordingly. I haven't figured out how functional the trigger is. Though that's hard to do when you have NO OUTPUT! 

-KM

[Kevin Mitchell]

Those old EHX boards do get the occasional hair line trace fracture, the caps dry up, I even had one with a shorted mylar cap which I wasn't expecting, that one took a couple of hours to track down. If there's no rational cause, you know where to look so it's trial and error.

While you're there, you should probably provide voltages for the 3 transistors in the attack delay network.

Thanks, Scruffie.
I'll check these out tonight. I believe I had replaced one with a 2n3904 earlier this year. If I didn't then someone before me did  Was likely a shot in the dark effort. I should probably record voltages across the board anyways. I'll need them to sift through the 90s unit that likely needs much more work (the reason why I have two of these to begin with).

-KM

[Kevin Mitchell]

It took a while but I've gathered voltages and sorted out the designations.

Code Select Expand

MICRO SYNTHESIZER idle voltages - no input, sliders up, EH24VDC power

A1 & Q7 are measured with main power ground, before the diode bridge
Everything else is measured with sleeve ground

Power while off - 35V
on - 28V

V+ is 7.73v
V- is -9.6v

A1-RC4558 (80s house marked RC4558N, 90s JRC 4558D)
1/O: 22.8v
2/N: 10.33v
3/P: 10.33v
4/-: 740.5mv
5/P: 10.33v
6/N: 0v
7/O: 0v
8/+: 27.26v

A2-RC4558
1/O: 1.83mv
2/N: 12.4mv
3/P: 9.4mv
4/-: -9.6v
5/P: 190.8mv
6/N: 194.2mv
7/O: 194.2mv
8/+: 7.73v

A3-RC4558
1/O: -56.3mv
2/N: -.3mv
3/P: 0
4/-: -9.6v
5/P: 0
6/N: 2.2mv
7/O: 17mv
8/+: 7.73v

A4-RC4558
1/O: 2.34v
2/N: .2mv
3/P: 0
4/-: -9.6v
5/P: -51.5mv
6/N: -34.7mv
7/O: -53.6mv
8/+: 7.73v

A5-RC4558
1/O: -8.24v
2/N: -11.4mv
3/P: -56mv
4/-: -9.6v
5/P: -56.2mv
6/N: -54.6mv
7/O: 4.5v
8/+: 7.73v

A6-RC4558
1/O: -2.3mv
2/N: .5mv
3/P: 0
4/-: -9.6v
5/P: -57.5mv
6/N: -53mv
7/O: -536.5mv
8/+: 7.73v

A7-RC4558
1/O: -8.2v
2/N: 0
3/P: -8mv
4/-: -9.6v
5/P: 0
6/N: 10.8mv
7/O: -8.14v
8/+: 7.73v

A8-CA3094/EH1040 (80s house marked EH1048, 90s Intersil CA3094AE)
1/O: 2.3mv
2/N: 0
3/P: 0
4/-: -9.6v
5/F: -9.16v
6/E: 105mv
7/+: 7.73v
8/C: 41mv

A9-RC4558
1/O: -1.81v
2/N: -1.81v
3/P: -1.82v
4/-: -9.6v
5/P: 2.18v
6/N: 2v
7/O: 7.28v
8/+: 7.73v

A10-CA3094/EH1040
1/O: 1.3v
2/N: 7.2mv
3/P: 5mv
4/-: -9.6v
5/F: -8.93v
6/E: 82mv
7/+: 7.73v
8/C: 7.73v

A11-CA3094/EH1040
1/O: 1.67v
2/N: 0
3/P: 4.4mv
4/-: -9.6v
5/F: -8.93v
6/E: 445mv
7/+: 7.73v
8/C: 7.73v

A12-CA3094/EH1040
1/O: -1.2mv
2/N: -140.7mv
3/P: 0
4/-: -9.6v
5/F: -8.95v
6/E: -7.3v
7/+: 7.73v
8/C: 7.73v

A13-CA3094/EH1040
1/O: -5.72v
2/N: -1.84v
3/P: -1.87v
4/-: -9.6v
5/F: -9.12v
6/E: -6.9v
7/+: 7.73v
8/C: -1.86v

A14-CA3094/EH1040
1/O: -378.6mv
2/N: 0
3/P: -133mv
4/-: -9.6v
5/F: -8.96v
6/E: 53mv
7/+: 7.73v
8/C: 0

A15-RC4558
1/O: -7.1v
2/N: 199.6mv
3/P: 194.8mv
4/-: -9.6v
5/P: 195mv
6/N: 206.2mv
7/O: 12.4mv
8/+: 7.73v

A16-RC4558
1/O: 355.8mv
2/N: 197.5mv
3/P: 194.2mv
4/-: -9.6v
5/P: 193.8mv
6/N: 195mv
7/O: -780mv - travels + or - 100mv
8/+: 7.73v

A17-CD4013 (80s RCA CD4013BE, 90s TC4013BP)
1/: -4.76v
2/: -4.78v
3/: -6.96v
4/: -9.6v
5/: -4.79v
6/: -9.6v
7/: -9.6v
8/: -3.79v
9/: -9.6v
10/: -7.42v
11/: -9.6v
12/: -7v
13/: -2.7v
14/: 0

A18-CA3094/EH1040
1/O: -1v
2/N: -8.3mv
3/P: 2.9mv
4/-: -9.6v
5/F: -9.23v
6/E: 0
7/+: 7.73v
8/C: 0

Q1-2N5087
1/E: -9.13v
2/B: -530mv
3/C: -55.6mv

Q2-2N4302 (80s 2N4340, 90s BF245A)
1/D: 220mv
2/S: 7.73v
3/G: -998mv

Q3-2N5087
1/E: -1.81v
2/B: -2.46v
3/C: -1.84v

Q4-2N5088
1/E: 6.57v
2/B: 7.13v
3/C: 7.73v

Q5-2N5088
1/E: 7.28v
2/B: 7.72v
3/C: 7.73v

Q6-2N5088
1/E: -9.6v
2/B: -9.53v
3/C: 7.72v

Q7(Q5 IN SCHEMATIC)-6N1110 (80s 6N1111, 90s B826)
1/B: 26.7v
2/C: 18v
3/E: 27.3v



Does anything look odd for A14 or Q4-6?

-KM

[Kevin Mitchell]

I've taken a snap of bot sides of R98 - E of Q4(blue) and pin 5 of A14(yellow). Trigger is at max, everything else it at 0. Interference noise is high when the blue channel is high, low when the blue channel is low. Only when the blue channel is traversing high to low a buzz comes through the output. That's the only time something audible shows at the output. It's normally low on silent and raises high with the rate dependent on the attack delay control.



I've replaced A14 with ones that I know are working quite a few times. Just like before this issue is becoming a rabbit hole. I think I know where it's failing, but I don't know as to why.

-KM

[Kevin Mitchell]

Still stuck.
Considering voodoo magic.

-KM

[StephenGiles]

I would suggest you replace Q5 & Q6 to eliminate any problems there.

[Kevin Mitchell]

I would suggest you replace Q5 & Q6 to eliminate any problems there.

I figured they weren't the issue but at this point I'll have all of the ICs socketed before finding the culprit.

Be back in a bit - replacing Q4-Q6 with fresh 2N5088 transistors.

-KM

[Kevin Mitchell]

For a moment there was distorted signal coming through the output but I noticed an a/b board wire had board which has been an ongoing problem handling the thing. I fixed it and we're back at phase 1 but now it's easier to trigger the VCA. Though it's garble at the moment - there's more action than before.

Testing for broken wires and if that checks out I'll replace A5(RC4558). Perhaps the IC before the transistors is the source of the problem.
EDIT: Nope. So I'm back at the original problem though it may be a sightly improved.

I wonder if I burnt out a replacement transistor (Q4 or Q5). Nothing obvious to report though.

-KM

[Eb7+9]

Considering voodoo magic.

curious ... do you have a way of applying exact +9v/-10v from an external bench supply ?!

[StephenGiles]

Next step - lift R98 (47k) Q4 side, then connect to wiper of a 100k linear pot. Connect outside lugs of the pot to 0v and -9v. The VCA should now work. If it doesn't , something is definitely amiss around CA3094 A14