Vintage Synth Repair

----- Original Message -----

From: Adrian Corston

To: vintagesynthrepair@yahoogroups.com

Cc: Hugh Vartanian ; Peter Mörck

Sent: Saturday, July 27, 2013 5:33 PM

Subject: Re: [vintagesynthrepair] CS-70m power supply issue, -15V rail is dead

I should correct a minor mistake I just made, sorry:

I measured +5V on the -5V test point, but -26.7V on Tr9's emitter. The -5V test point is the other side of the 4.7ohm resistor from Tr9's emitter.

If anyone would like me to post a link to the PCB layouts for this circuit then please let me know.

Cheers,
A.


On 28/07/13 12:51 AM, Adrian Corston wrote:

I hate to reply to myself, but...

I did some more work on this tonight, replacing a bunch of components. Since my last email I have replaced:
- All ICs
- Tr9
- All electrolytic caps except for the big three (4700uF x 2, 10000uF x 1)
- All zeners
- All 1S1555 (with 1N4148)
- All 2SC1815 except for one - Tr 6, because I ran out of new 2SC1815
- All 2SA1015

I have previously replaced Tr20 as mentioned below.

Both faulty behaviours remains as described below - in summary:
* +5V on the Tr9 emitter (instead of the expected -5V)

^^^^^^^^^^^^^^^^^ correction: this should have read "+5V on the -5V test point". ; Sorry about that.

* On power-up Tr20 emitter starts at approx. -12V but rises to +0.6V within 10-15 seconds then remains there (instead of the expected -15V)

This is despite a perfectly reasonable -26V on the collectors of both Tr9 and Tr20.

I re-measured all legs of the five main power transistors:

Tr20 B=+340mV, C=-26.7V, E=+0.85V
Tr15 B=+7.3V, C=+26.7V, E=+17.6V
Tr9 B=-26.7V, C=-26.7V, E=-26.7V
Tr5 B=+26.7V, C=+26.7V, E=+26.1V
Tr1 B=+3.9V, C=13.7V, E=+3.4V

Peter asked for the voltage at Tr23's base; it's +13.5V. Given this comes from the op-amp in IC2 and IC3 pin 2 is +1.9mV and IC3 pin 3 is +9.2V this is what I'd expect (i.e. the positive power rail of the IC3 minus a bit).

I also tested all PCB traces and did not find any breaks (obviously I couldn't easily test for shorts though).

The schematic can be found here.

Unless anyone has any better suggestions, my next plan is to start replacing resistors and ceramic capacitors. After that I'm out of ideas :-(

Thank you everyone for your suggestions so far!

Cheers,
A.


On 22/07/13 11:47 PM, Adrian Corston wrote:

On 22/07/13 4:39 AM, Peter Mörck wrote:

I would expect it to be around -15.7V or so when the circuit is operating properly. Exactly how much is it?

How much is base voltage of Tr21, Tr22 and Tr23?

Check if D7 is good.

On 22/07/13 6:31 AM, Hugh Vartanian wrote:

Hi,

A little about the operation of this circuit when operating normally. Hopefully there is a hint in there that will get to the bottom of this! Note, no distinction is made between what is obvious and what is less so in the description. Also, the ideas of ;'more' or 'less' voltage or current are a little funny when talking about positive and negative values. Generally the conversation uses absolute values or magnitudes when using such language.:

1. Tr20 is the main pass transistor, across which is the difference between the unregulated voltage at the collector and the regulated side at the emitter. When it's base is pulled negative, current flows through from the emitter turning on the transistor and pulling the (amplified) current through the collector, driving the emitter to a negative voltage. If there is an open circuit in this transistor, there will be no output. (see below)

2. Tr21 functions as another amplifier (a 'darlington' connected transistor pair with Tr 20) to pull base current out of Tr20, with the low currents being pulled by the rest of the circuit. So, pull Tr21 base negative, amplify that to pull current through Tr20 base, and give us the juice for the -15V supply.

3. If one takes Tr22 and Tr23 out of the circuit (disconnect their collectors), then Tr21 turns on with the current through the 1K resistor, which turns on Tr20. In this situation the emitter of Tr 20 will be at the unregulated supply voltage (minus a few tenths of a volt.). (IF the rest of the synth is disconnected from the -15V supply, then one could do this test to check out Tr20 and Tr21, but I don't think we need to try this just yet and -26V cannot get to the rest of the synth circuitry safely!.)

4. We want regulation of the output voltage, so we must have a way of draining off the current from the 1K resistor and not sending it into the pass transistor(s). Tr22 or Tr 23 do this.

5. First off, Tr22 is there to keep the -15V supply off until the +15V supply is turned on. If negative supply is on and the positive supply is off then Tr22 is turned on through the 47K resistor on its base. This pulls the current from the 1K resistor to ground, keeping it from turning on Tr21. When the positive supply comes up, then Tr22 is shut off by the slight positive voltage on the D7/27K resistor junction. ;If the voltage at the D7/27K/47K/Tr22Base junction is not positive several tenths of a volt when power is up, then something is the matter in the upper circuit, probably an open 27K resistor (if the 1K above it was open, then the positive output wouldn't be working.).

6. Next is the regulation function. The regulator operates to keep the -15V output a mirror of the +15V output. Pin 3 of IC3 is connected to the junction of 2x 10K resistors, one to +15V and one to -15V. IC 3 will drive through Tr23/Tr21/Tr20 until the junction at the 10K resistors is almost exactly 0V, halfway between + and -15V. If that voltage is positive, meaning +15 is bigger than -(-15V), then we need 'more' -15V. The positive voltage on pin 3 is amplified by the op-amp, making pin 1 more positive, turning off Tr23 and letting more current from the 1K go into the Tr21/Tr20 transistors, giving us more -15V. If the voltage on pin 3 is too negative, i.e. -(-15V) is greater than +15V, then the output of IC3 goes negative, which turns on Tr23 more and shuts off Tr21/Tr20 more, giving us less drive and a 'lower' -15V. By this action, the -15V is kept at a mirror of +15V.

7. It looks like the .22 ohm resistor ought to be part of a current limiting circuit, but I don't see that represented in the schematic. Measured as not open, so it is fine.

8. If something was dragging down -15V, Tr20 would be very hot. Somehow we are just not getting juice through Tr20.

9. The +.6V measured on the -15V rail is probably leak through from the + supply through the circuitry in the absence of any -15V drive. If there was -0.6V on the -15V rail, I'd say the upper 10K resistor is open.

10. Interesting voltages to know, once the transistor thing is straightened out, are pin 3 of IC3 (should be nearly 0V), pin 2 of IC3 (also 0V), pin 1 of IC 3 (somewhere slightly negative, maybe a volt or so), Tr23 collector (-16 or 17V, about 1.5V less than -(+15V)). Also the E/B/C voltages on Tr20 and Tr21. Measure these and report them.

11. Interesting that the 1K resistor is only hot when the bad transistor is installed. I would expect it to be hot if there is no -15V output. Otherwise there is an open circuit in here somewhere. Possible that Tr20 is bad???? An open circuit in the collector would do this. If you pull Tr20, there should be a diode measured with the meter from the Emitter to the Base and from the Collector to the Base. If either of these is open circuit (or shorted), in both directions with the meter on 'ohms' or preferably the 'diode' scale, then this transistor is bad. No hot resistor, no output. Same measurements on Tr21, same outcome (cold 1K and no output). Measure these transistors between EB, CB, and might as well CE with the meter on the diode scale, meter probes in both directions, and publish the results.

12. BTW, small Vce sat differences don't matter when substituting transistors for function is this circuit (even most circuits unless they are RF). Most important is that the right transistor type is used ( for bipolar transistors PNP or NPN (FETs more complicated)), the PINOUT is confirmed for the E/B/C connections, voltage margin is OK (60 is fine for this circuit), and the current capability is OK (only an issue here for the Tr20 pass transistor. Everything else is small signal and almost any substitution will work fine to test this circuit. The transistor substitution listed above (2sb590/2sa1534a) should be fine. Check between the pins with the meter as listed above, both transistors, with the diode range, between each pair of pins. Maybe the substitute was good and the resistor got hot because both Tr20 and Tr21 are bad(!).

Looking forward to voltages and transistor readings!

Hugh

Thank you both for your suggestions.

Unfortunately, one thing I did tonight was to accidentally short out pins on Tr9 to the case of Tr5. I was being so careful, so of course the very first time I forgot to check them before applying power they shorted out *sigh*. So after checking and replacing F2 and it now seems to be mostly working as before, with the exception that the -5V rail is now reading almost exactly +5.00V (!?), and the base and emitter of Tr9 which appear to be shorted to the unregulated supply voltage (-26V) on the collector. I will get a replacement for Tr9; I am pretty sure it's toast. Should I replace other components in that area (maybe the 220uF electrolytic or the 10K and 20K resistors and IC2?) I feel pretty stupid, and now have the five off-board power transistors all sealed up in insulated snap-lock bags to avoid it happening again :-(

As far as the -15V circuit is concerned, I took the following readings. I also tried to work out where the fault may be coming from and did some extra voltage measurements along that path, so hopefully this will help with identifying the root cause.

The 1K resistor across the collector and base of Tr21 gets very hot even with the original Tr21 (2SB560) so that was a red herring, sorry. I've put the original transistor back in now.

All component value measurements below were made with the component under test removed from the circuit.

Tr20 base = +0.180V
Tr21 base = -0.280V
Tr22 base (D7/27K junction) = -0.783V [This looks wrong! Focus on that circuit then]

27K measures 27.0K
D7 tests OK with a 0.626V voltage drop - I replaced it anyway, with a new, similarly tested 1N4148
47K (from Tr22base to the unregulated negative supply) measures 46.9K
1K (from Tr16 base to unregulated positive supply) measures 1.00K

[So the components of the voltage divider look OK, something else must be driving the D7/27K junction negative. Is it Tr19/Tr17?]

27K/1K junction (Tr16 base) = +1.326V
Tr17 base = +0.136V
Tr19 base = +0.700V

[That comes all the across from Tr13, Tr14]

Tr13 collector = +0.693V
Tr13 base = +0.054V
Tr14 base = +0.690V
ZD3 should be a 6.8V zener (marked on the PCB but not in the service manual - I didn't test it though)
ZD3 cathode = +7.48V

Hopefully that gives some more insight - should I check more or is that enough to identify the faulty part(s)?

I feel like I can now almost understand what's happening, but I'm not quite there yet. I think Tr14/Tr13/Tr19/Tr17 might be there to stop both the +15V and -15V rails from operating unless the +10V is OK, or something along those lines.

Due to the rapid heating of the 1K resistor I was hesistant to leave power on on for longer than a few minutes at a time, and I noticed that some of the values above were not constant and changed for about the first minute or so after power was applied. For example Tr20 base started at around +0.190V and kept dropping about 2mV/sec. Also the -15V rail reads about -12V for one refresh of my digital multimeter when power is first applied, then it starts increasing fairly rapidly until it reaches the +0.6V mark where it then remains, again taking a little less than a minute to do so.

Hugh, your breakdown of aspects of the circuit's function and the process of debugging what's going wrong is fantastic and I found it incredibly helpful. Thank you so much! I really appreciate it.

Peter, sorry I forgot to measure the base of Tr23. I'll make sure I remember to do it tomorrow night.

Cheers,
A.

On Sat, Jul 20, 2013 at 3:09 PM, Peter Mörck <peter@...> wrote:

>From my limited knowledge I react on the voltage 0.6V because it sounds like a PN-junction voltage drop.
>So I would guess that something, coming from 0V/ground, via a reverse PN-junction, reaches the emitter of Tr20.
>For example, its own base-emitter junction, if the base of Tr20 is 0V. Is it? *no, base is > 0V*

I would expect it to be around -15.7V or so when the circuit is operating properly. Exactly how much is it?

How much is base voltage of Tr21, Tr22 and Tr23?

Check if D7 is good.

;

BR,

Peter

----- Original Message -----

From: Adrian Corston

To: vintagesynthrepair@yahoogroups.com

Sent: Saturday, July 20, 2013 5:31 PM

Subject: Re: [vintagesynthrepair] CS-70m power supply issue, -15V rail is dead

On 20/07/13 5:50 PM, Adrian Corston wrote:

I'll be having another look at the unit tonight and will try out the excellent suggestions received.

Here we go... my responses are in red in the quoted text below.

Change ANY electrolytic capacitors in that -15 circuit. *done*

Diode "check" any diodes ( make sure they conduct one way and NOT the other) *I don't see any in the -15V circuit*

If there are any tantalum capacitors ( usually teardrop shape) replace them too *not relevant - there are none*

The -15' and +15' values are probably right as they're derived from a zener diode + a PN drop, so 15-0.7 and -15+0.7, roughly.
Sounds about right. *OK*

From my limited knowledge I react on the voltage 0.6V because it sounds like a PN-junction voltage drop.
So I would guess that something, coming from 0V/ground, via a reverse PN-junction, reaches the emitter of Tr20.
For example, its own base-emitter junction, if the base of Tr20 is 0V. Is it? *no, base is > 0V*

Also switch it off and check for a short to ground from Tr20 base. *no short found*

If IC3 is socketed you could try replacing that. Who knows what will happen if that's fried... *IC3 replaced (and socketed for next time)*

Also check for shorts between Collector-Emitter on Tr22 and Tr23. *no shorts found*

I'd also check the electrolytic on the emitter of TR20 isn't pulling anything to ground *electrolytic replaced*

Also check for any shorted tantalum caps throughout the synth... *not relevant, rest of synth is disconnected*

5. No voltage is put out for -15V. Other voltage positions operate correctly.
Probably cause: Tr21 - Tr20 - 0.22ohms.1P *tr20 replaced, tr21 replacement pending, 0.22ohm resistor measures OK*

None of the above things that I tried had any effect; the -15V rail is still sitting at around 0.6V.

I have no 2SB560 (or well-documented substitute) handy for tr21, but I did find a 2SA1534A (in another old power supply) which has a similar spec. The 2SA1534A is only 60V rather than 100V, but I assume that won't matter in a -15V circuit (I only measured 26V for Vce in the circuit with the 2SB560 in place). Pinouts are the same for the two parts. However according to my wonderfully-useful transistor comparison book it is not "lo-sat" (low collector-emitter saturation voltage) whereas the 2SB560 is. According to data sheets:

2SB560 Vce(sat) = -0.3V typ, -0.8V max
2SA1534A Vce(sat) = -0.2V typ, -0.4V max

Datasheets:
2SB560: http://www.futurlec.com/Transistors/2SB560.shtml
2SA1534A: http://www.alldatasheet.com/datasheet-pdf/pdf/12706/PANASONIC/2SA1534.html

This looked near enough to me, so I tried substituting it. Unfortunately the 1K resistor between tr20/tr21 collectors and tr22/23 collectors immediately started getting too hot to touch, so I removed power straight away. I will try again next week after I find a real 2SB560.

Thanks again,
A.