Maestro PS 1 - you no go slow

[kaycee]

Maestro PS 1 clone on a replica board, like the one at the top of this page and the schematic beneath.

http://www.diystompboxes.com/smfforum/index.php?topic=102281.0

Had it going, but now it won't phase on the slow setting. Medium and Fast still operate as normal. Obvious culprit would be a cold joint or loss of continuity, but I've been over the sections around the switching and the LFO opamp and still no joy. I've been over the rest now, re flowing joints and testing continuity, changed out the LFO opamp. Flat out of ideas..

I have -12v, +12v and Gnd connections where I'd expect throughout. Adjusted the bias and offset and have amplitude at full, these all work as expected too.

Any ideas where to look, specifically in relation to the slow?

[Govmnt_Lacky]

Check resistor and cap values. Might be a 100K where a 1M is supposed to go. 

[kaycee]

I've already replaced all the resistors in that top left corner where the switching lives in order to check for wrong values and cold joints, even the off set trim.

[jonnyeye]

Check the voltage at Q9's emitter with the speed set to slow - it should be small but positive (about 0.2-0.3V).  If it's too low the oscillator will fail to oscillate.

(If you want to dig deeper, the oscillator is voltage-controlled; range from about 0.2V up to 12V, lower voltages being slower oscillation.  Q9 buffers the control voltage generated by the switches.  The 10k/220uF cap on Q9's base sets the ramp speed. I played around with this oscillator in SPICE a while back because some day I will build one of these and I wanted the speeds to be controllable... haven't got to it yet...)

[kaycee]

Thanks Jonny, that's something to have a look at over what I've done already, might get me somewhere, cheers

[kaycee]

Ok, Jonny, your suspicions confirmed. Readings for Q9 on slow. Supply voltages +12.2/-11.3

C -11.3
B 0.36
E 0.19

It's nearly phasing.

On medium the voltages on the emitter goes up to 1.46 and on fast it's 3.6

I've tried several othe PNP transistors in there to no avail, doesn't drive the voltage up on the emitter. Odd that it worked in the first place.

So, please, how do I drive up the emitter voltage to make the LFO run in slow mode? And if I do this will I need to adjust the other resistors in the switch path to maintain the original medium and fast phase settings?

[kaycee]

Bumperoo Boogaloo

[jdub]

I assume, of course, that adjusting the upper trimpots has no effect?

[jonnyeye]

Sorry, summer got in the way of me responding...

The simple fix is to reduce the value of the 1M resistor to Q9's base. Try 820k or 680k.  There will be an imperceptible change in the medium/fast speeds, no need to worry there.  But my sim has the fast speed having an emitter voltage of about 9V compared to your 3.6V, (the 1.46V medium is basically spot on) so there's perhaps something else going on here... (Also, I'm hoping that the base of Q9 is -0.36V, otherwise the stage shouldn't be working...)  Perhaps measurement of all the values is in order - did you maybe use 47k for Q9's emitter resistor?

[jdub]

Adjusted the bias and offset and have amplitude at full

Ooops, missed that statement, please disregard my previous post...+1 to what jonnyeye wrote 

[kaycee]

Thanks, I appreciate people taking an interest.

So, I checked the emitter R, it's a 4k7 as should be. The 1M also checked out.

I removed the 1M and wired in a couple of flying leads to a 1M pot and adjusted until I got slow phase (yay!). Took the pot out and I'm reading 120k to have it going at a slow rate, much higher than that and it's getting quite quick, much less and it stops again. So that's good progress.

Supply readings as before, I'm running a 9volt PSU into a charge pump, through a 12v regulator, then a bipolar chip. +12.1 & -11.3

My voltage readings are now.

Slow

E 0.95
B 0.37
C -11.3 (same at all settings)

Medium
E 2.05
B 1.51

Fast
E 3.5
B 8.6
I've tried different opamps incase its a variable there, the same with HA17458, TL062 and LM1458.

If nothing stands out to chase further then I'll settle for it as is and look out for another board for another run at it some time. A friend went to some effort to etch this one for me, so I'd really like it to be as good as I can get it for sentimental reasons.

[jdub]

Glad to hear you're getting somewhere, KC.  Just out of curiosity, are you taking your PSU output voltage readings while under load- that is, from the board? The reason I ask is that I never could get mine to work with various 9V/charge pump setups, never had enough output current, got voltage sag and the charge pump chips would get very hot.  I had to stick with an 18v AC supply into a bipolar supply board...just thinking out loud, wondering if problem could be related to current supply.

[kaycee]

Hi John, yes that idea crossed my mind as well. I'm taking the readings from the rails on the board under load. I haven't noticed the charge chips getting hot, maybe I've only been checking the board ones, I'll check next time I run it.

I was wondering about sag, tried running it on 12v going in, gets about 18 after the doubler, goes back down to the stated voltages after the regulator. And, yes I understand voltage isn't current and what you mean by sag.

A little thought in the back of my mind is the difference between the positive and negative supplies? I'm regulating the charge pumped positive before the bipolar chip, maybe I should be using regs on both the supplies after the bipolar chip.

Next time I run it, I'll put 12v into a bipolar and run it without the pump just in the name of science. Thanks for chipping in, I am taking note of your comments too

[jdub]

I'm actually running mine at ~+/-15v, otherwise I would've chimed in with voltages already.  I'll take a look anyway for comparison purposes. What chips are you using for charge pump and bipolar- LT1054s? What exact trannies are you using for Q9 and Q10? Also, can you post pics?  Dropping that 1M to 120k seems pretty excessive, so something must be going on. 

[kaycee]

I'm using 7660S in both instances, I ran it for 20 mins and neither was even warm. My power supply on this is one of those Maplin Eagle ones, 2000ma, 24VA, I suppose I should measure the current draw, duh!

For Q9 I have 2N5807, Q10 MPF102, the rest are 2N5485's matched. It also ran on BF245's but the bias adjustment was smaller.

I tried mine on 18v and it distorted, it does run at 9 too, but probably lacks headroom. I test using a signal tester most of the time, once things are working right I try with a guitar signal.

I agree, the difference between 1M and 120k is too much for something not to be wrong, which is why I'm still chasing it. I'm away over the weekend so I'll maybe post pics next week. I do like the ramping feature when changing speeds, that's fun!