help me understanding what's going on please

[OG]

As mentionned in previous posts last week, i'm trying to debug an Electric mistress standard which looks just like this :
http://img.photobucket.com/albums/v317/StephenGiles/EHelectricmistress_UK.jpg
(thank you Stephen Giles)
Wires were missing, so I have rewired it, replaced the trimpots, the 741 and a few electrolitic caps, and it "nearly" works.
The problem is this : no sound even in bypass mode until I check voltage (or resistance ) and touch the base of the 2n4354 with the mutimeter probe. It seems doing this switches the transistor on ( from 0 v I get +-655 millivolts at the base) then bang sound is there again, in bypass or (great) flange mode.
So what 's going on?
Thanks.
OG

[gez]

Are you measuring the same reference voltage at pin 2 of IC5 as indicated in the schematic (6.8V)?

[OG]

Yes ,
the voltages seem correct as indicated on the schem : 6,5 volts , 13 volts at the collector (not 12 but correct according the EH tech note) with a 18 volts supply.
Thanks

[gez]

What is the voltage at pin6 of the 471?  If it's high, then it could be that you're grounding the base of the trannie with your probe when testing it, just enough to turn it on?

Edit:  whoops, you squeezed in a reply just before I did (ignore above).

[gez]

If you're measuring 13V (that would be about right by my calculations) at the collector then I don't understand why it isn't working.

[Paul Marossy]

Just a wild guess, but maybe the transistor(s) is bad?

[niftydog]

dodgy solder joint? rework the transistors leads.

Measure it's bias resistors out of circuit to be sure they are within tolerance. It could be that placing your DMMs resistance accross the circuit is altering it just enough to work.

[OG]

1-Sorry but I don't understand what you mean exactly. i'm rather new to all this even if i'm trying to learn more and more of the theory behind this. ( a hell of a job).
2- it doesn't seem to be a connection problem , i did re solder the leads.
3- If i remove the transistor and destroy it by accident, there is still the question of the equivalent pnp tranny.
well there is some usefull information courtesy of RG now on another post, i have to check, but ...
Maybe Mark Hammer could be helpfull on this topic, i would really like to understand.
Mark, are you with us?

[niftydog]

Perhaps it would be an idea to get a replacement or equivalent transistor and try it in circuit. The original may be "half way" dead already.

Try this; turn off the circuit, disconnect the battery or power source, unplug all the leads and remove the transistor from the circuit. Now, measure the resistor that connects between pin 6 of the op amp and the transistor base. Also, measure the 100k resistor that connects to the emitter. Also, the 22k and 30k below the op amp as well as checking over the 100k potentiometer.

are any of the caps you changed associated with this part of the circuit?

[OG]

Thanks for the suggestions. I 'll try to find an equivalent transistor
( thank you R G for the very instructive post about equivalence) and will take the opportunity to check the resistors, and maybe replace them with metalfilm.
I have already changed the 2 trimpots, and the only cap i changed in this part of the circuit is the 33uF electro. So I will try to replace the two others, who knows?
I'll tell you.

[OG]

Ok, this little thing really begins to get on my nerves...
basically, i have changed everything around and including the 741 (excepted the zener but the voltage is good so...), all the electro caps on the print and the two trimpots. I'm reluctant to change the 4558 because i don't have a spare and a tl072 will change the sound somehow.
I tried to replace the 2n4354 with a 2n3906 and i get exactly the same thing : no sound ( nearly no voltage at the collector!) till i touch the base with the lead of the multimeter then pure mistress flange...
I checked the different voltages and here is what i get, in two columns,
no sound / circuit working :
741 :
1) 0 / 196 mv
2) 6.54 v / 6.54 v
3) 223 mv / 6.53 v
4) 0 / 0
5) 0 / 19 mv
6) 16.97 v / 11 v
7) 17.17 v / 16.26 v
8") 0 / 0
4558 :
1) 4 to 2 mv / 3.87 v
2) 4 to 2 mv / 3.86 v
3) 117 mv / 3.42 v
4) 0 / 0
5) 117 mv / 3.42 v
6) 2 mv / 3.86 v
7) 2 mv / 3.86 v
8") 104 mv / 12.86 v
2n4354 :
base : impossible to check of course / 15.95 v
collector : 103 mv / 12.86 v
emitter : 16.96 v / 16.55 v
base-emitter : 20 mv / 663mv
emitter-collector : 16.96 v / 2.91 v

This is really too much for my basic knowledge (and head) so i hope someone from the community will be able to help me. Damn, it seems so close to work
Thanks in advance.
OG

[niftydog]

does the voltage at pin 6 of the 741 vary when you turn adjust VR1?

Have you checked and measured the resistors I mentioned? Especially, check the 47k resistor attached to the base.

Is the trimpot of the correct value and have you tried adjusting it?

When the circuit is not working, the transistor is not being switched on... obviously. But the problem is that that one transistor is affecting all of the voltages you listed as it is effectively the power switch for the whole circuit.

Here's my guess on what's happening. You attach power. 6.8VDC (6.54 v ) appears at pin 2 of the 741 thanks to the zener. There's effectively zero volts (223 mv ) at pin 3 at this time. Hence, the output (pin 6) will be a very low voltage, initially very close to 0VDC. This turns on the transistor and power flows into the rest of the circuit. Now, the voltage at pin 3 is affected by the resistor network below the op amp. At this point, the voltage at pin 3 must become greater than the voltage at pin 2, thus taking the output (pin6) to a high voltage (16.97 v) thus turning OFF the transistor.

This may appear as not fully on but not fully off either!

Definately check all those resistors out of circuit, especially that 47k!

You are very close my friend, don't give up yet!

[OG]

Thank you very very much for your support.
I did check the resistors and changed them actually, i'll try to see what happens when moving the trim ( from what i remember there is a very narrow region where you have sound and no distortion, and i also remember comments by Mark Hammer telling that in general bias is critical with flangers)
I noted what i had at the wiper of vr1 : i have 115 mv when there is no sound and, 3.74 volts when the circuit is working (and base-collector voltage is 3.14 v ).
yesterday, i tried with another 4013, no change, with a 2n3906 in place of the 2n5087, no difference.
then i checked again difference places to find a bad contact of some sort, and something even more strange happened, there were a lot of points on the print turning the circuit on, mainly those connected to +12v (or to each other of course): on the sad it was pins 3/5/7/8/10/11/14, on the 4013 pins 1/2/4/9/12/13/14, the collector on the 2n5087 and the 4354, pin 3 on the 741 and even the wipper of one of the trim... then i tried again to disconnect the battery, and it wasn't happening this way anymore but only when touching the base of the 4354 !
So..., i'll try again to check the voltages and i still have the possibility to call an exorcist.
Who put a spell on my mistress i ask you?

[niftydog]

I noted what i had at the wiper of vr1 : i have 115 mv when there is no sound and, 3.74 volts when the circuit is working (and base-collector voltage is 3.14 v ).

again, that's all simply because the circuit is effectively switched off when there's no sound and switched on when there is sound.

What you need is for it to switch itself on and stay switched on.. without the need for your multimeter probe to be hanging accross it!

Do some careful tweaking of that VR1 and get back to the forum!

[OG]

I took time to check  the resistors again, and they are ok.
When i turn the trim pot, there is no notable difference in voltage at pin 6 of the 741( now replaced by a tl071), the only difference is the voltage change from ov to 6.50v at the wiper when you turn the trim full down(=to ground) or full up. Moving the trim, there is a very small region where you have clean flanged sound , then with more distortion, then sound with no flanging with a more or less distorted signal. but there is no place where i can turn the transistor on without touching the base with the lead of the multitester.
could it be a problem with the lm339 ? there is no socket, so i didn't try to change it...
Honestly i'm a little bit lost.
thanks.
OG

[niftydog]

Moving the trim, there is a very small region where you have clean flanged sound , then with more distortion, then sound with no flanging with a more or less distorted signal. but there is no place where i can turn the transistor on without touching the base with the lead of the multitester.

ah ok... so you have the multimeter connected to the base WHILST turning the pot to get the good sound? Just clarifying...

All I could suggest at this point is this;

Try "replacing" the multimeter with a large value resistor and see if it has the same effect. That is; wire a large value resistor (say 1M to start with) from the base to ground.

This might tell us if we're dealing with the possibility that your multimeter is outputing a voltage or if it's simply loading the circuit with it's impedance.

If this works, I would suggest you leave the resistor in place and start using the effect!

[OG]

Well my friend, hats off, it works!
I tried to put a 1Meg resitor from base to ground, and the circuit is now alimented without touching the 4354 base with the lead of the ohmmeter.
In fact i noticed that it even switched on with a simple wire...
I also tried with a 10 Meg, but it didn't work anymore.
So, i still don't understand what really goes on doing this, and if you could explain it to me , it would be the icing on the cake!
But..., there is a small change in sound, so i was thinking about other possibilities.
Tell me if i'm wrong : 1meg in // with the existing 47k to the base is equivalent to 44.89K ohm. Will i have the same result if i only replace the 47k resistor with a same value trimpot (omitting the 1meg resistor) and tweak it till i have the same 44.89K value, or maybe even closer to the fixed value, assuming this would less change the sound.
Or if i understand a little, with the previous experiments in mind,  the problem is bad biasing with this 47k and is related to dc only. If i put a bypassing cap across the 1meg resistor to ground, will this give the same result considering dc, and at the same time will i get rid on the effect on the -ac- signal.
This is all very new to me, and even if i'm still trying to learn more and more theory maybe i got it all wrong...
Anyway Nifty, thank you very much for your invaluable help so far.
You have a friend in Belgium.
OG

[OG]

sorry deleted cause posted twice

[niftydog]

i still don't understand what really goes on doing this, and if you could explain it to me , it would be the icing on the cake!

Well, my guess is that the op amp was not sinking enough current to allow the transistor to turn on properly. The extra path from base to ground allowed just smidgen more current to flow, hence turning on the transistor.

1meg in // with the existing 47k to the base is equivalent to 44.89K ohm.

They're actually not in parallel though. The 1M is from base to ground whereas the 47k is from base to the 741 output.

there is a small change in sound, so i was thinking about other possibilities.

Yes, the 1M was purely a starting point. What you want to do now is measure the voltage you're getting at the V+ point with the 1M in place. Ideally you want it close to the "proper" spec, whatever that is. It may require some tweaking of that resistor value.

Perhaps it'd be easier to replace the 1M resistor with a 1M pot and tweak it until you're getting a good voltage reading (or a good sound.) Actually, I'm thinking that you may even want a higer value pot than that. The higher the better in terms of the AC component of the signal at that point.

In fact i noticed that it even switched on with a simple wire...

Yes, it would turn on that transistor, but any audio in that path will be shunted to ground.

If i put a bypassing cap across the 1meg resistor to ground, will this give the same result considering dc, and at the same time will i get rid on the effect on the -ac- signal.

Close, but in effect you'll be getting rid of the AC signal... entirely! This would have much the same effect as a wire to ground. In a broad sense, caps are AC short circuits and DC open circuits.

Check the V+ voltage again, tweak the resistor size and I'll think you'll be fine.

[niftydog]

ok, thinking about it some more, I really think you need to get that zener voltage up to 6.8V. I mean, they've gone to the effort to note that voltage on the schematic, it must be fairly crucial to the operation. That zener should have a tolerance of 5%, meaning that the one you have is quite low but within spec. Still, I think it'll make a difference.

The 741 is effectively a regulator. The 11V at pin 6 that you measured when the circuit is working is, IMO, an averaged value of a very quickly oscillating voltage.

(The op amp must be able to sink the current, that was a stupid speculation on my part.)

no sound / circuit working :
741 :
1) 0 / 196 mv
2) 6.54 v / 6.54 v
3) 223 mv / 6.53 v
4) 0 / 0
5) 0 / 19 mv
6) 16.97 v / 11 v
7) 17.17 v / 16.26 v
8") 0 / 0

2n4354 :
base : impossible to check of course / 15.95 v
collector : 103 mv / 12.86 v
emitter : 16.96 v / 16.55 v
base-emitter : 20 mv / 663mv
emitter-collector : 16.96 v / 2.91 v

Just noticed that pin 7 and the emitter seem to have been measured incorrectly. They should be the same voltage as they are connected to the same point.

Also, with pin 2 at 6.54 and pin 3 at 0.2V, pin 6 should be close to zero, but you have it as 16.97V.

16.97V happens to be the same as the emitter voltage... so this leads me to believe that something is awry with these figures.

(all the above while circuit is not working)