Yamaha PM700 console upgrade/mod

[keefaz]

You can experiment some mods on one channel to lower the current, I would remove TR6 and TR7 then link opa604 output (pin 6) to the bipolar cap (BP 47/16) with a jumper

 

[spaceludwig]

You can experiment some mods on one channel to lower the current, I would remove TR6 and TR7 then link opa604 output (pin 6) to the bipolar cap (BP 47/16) with a jumper

Would you know what function those 2 transistors are performing?

 

[abechap024]

Maybe try increasing the feedback loop capacitor from (27pf, 47pf?, 100pf??) to a higher value.

EDIT: I'm not sure what opamp your talking about, but those compensations do look pretty lean for some modern fast opamps.

The only cap i am seeing in a feedback loop of ic 2 is the 22pF cap in parallel with the 33k. I don't know how changing that value to a higher cap would change anything...

Increasing that cap value lowers the bandwidth the opamp is amplifying, sometimes those caps play a very important role in stability.

Would you know what function those 2 transistors are performing?

You can think of those transistors as  a souped up output stage of the opamp, the have more current drive (maybe thats where the extra current is going) and a lower impedence, example they could dirve an output transformer, while most opamps can't

 

[spaceludwig]

Increasing that cap value lowers the bandwidth the opamp is amplifying, sometimes those caps play a very important role in stability.

Increase by what, multiples of 2? like 22, 47, 100, etc...?

You can think of those transistors as  a souped up output stage of the opamp, the have more current drive (maybe thats where the extra current is going) and a lower impedence, example they could dirve an output transformer, while most opamps can't

OK, but why put them there when you have the individual signals going to the summing amplifier which is there, presumably, to perform that function? Is it because the signal is getting split 3 ways (out + monitor 1 & 2)? Actually, monitor 1 seems to have it's own opamp.

Also, I don't understand why the current draw is lower with the stock opamp vs a newer one.  I have no compunction about removing those transistors but it annoys the hell out of me not to understand why the newer IC's pull more juice.  Hahahaha, I think I'm starting to go crazy with this mod....

edit: Monitor 2 has it's own opamp. Monitor 1 is sharing the signal going to the summing amp.

 

[abechap024]

Well, as been stated before, oscillations will cause an opamp (and its output stage) to draw more current, and even a ne5534 can be a fidgety chip sometimes without the proper compensation.

a safe bet would be increasing the feedback cap to 47pf to see if that cures it....if you don't have one laying around you can maybe "borrow" a random capacitor (pretty much any smaller value will work for testing) and see if that helps.

I feel your pain brother!!! Oscillations....ugh!

But we don't know for sure yet if that is the complete problem. we shall see...

 

[spaceludwig]

Abe,

Can I test it "freehand", i.e. without soldering and de-soldering, like holding it with pliers and touching the 33k resistor leads? Some of the traces have lifted and I don't want to damage them further.

BTW, I really appreciate your, and everyone else's, help with this 

 

[abechap024]

the only problem with that is, the pliers or fingers or whatever you are holding the capacitor in place with, may influence the capacitance of the capacitor and/or add stray capacitance to the circuit...

 

[spaceludwig]

Just to be sure: When I use the original opamp (TA7136P) in place of IC 2 the current draw of the circuit is around 40mA. If I use a newer opamp (opa604, 5534, etc) with all pins EXCEPT output (pin 6) connected it draws the same amount. However, the moment I wire the output it jumps over 100mA for a total of 130+ mA. This is the result of oscillation, correct?

 

[keefaz]

Just to be sure: When I use the original opamp (TA7136P) in place of IC 2 the current draw of the circuit is around 40mA. If I use a newer opamp (opa604, 5534, etc) with all pins EXCEPT output (pin 6) connected it draws the same amount. However, the moment I wire the output it jumps over 100mA for a total of 130+ mA. This is the result of oscillation, correct?

(edit) Sorry I made a confusion with opamps pins

I don't know, I think I would experiment with unsoldering the transistors at output of IC2 and connect the opamp output to the bipolar cap

 

[abechap024]

Just to be sure: When I use the original opamp (TA7136P) in place of IC 2 the current draw of the circuit is around 40mA. If I use a newer opamp (opa604, 5534, etc) with all pins EXCEPT output (pin 6) connected it draws the same amount. However, the moment I wire the output it jumps over 100mA for a total of 130+ mA. This is the result of oscillation, correct?

its impossible to say for sure without being there....just trying to give you ideas

 

[MikeClev]

Just to be sure: When I use the original opamp (TA7136P) in place of IC 2 the current draw of the circuit is around 40mA. If I use a newer opamp (opa604, 5534, etc) with all pins EXCEPT output (pin 6) connected it draws the same amount. However, the moment I wire the output it jumps over 100mA for a total of 130+ mA. This is the result of oscillation, correct?

I'll go out on a limb and say yes here. Also as Keefaz says, yank those transistors out and jumper between the opamp output and the bipolar cap. Those transistors are inside the feedback loop of the opamp, so will contribute to the oscillation. Those transistors and surrounding circuit were designed around the properties of the original TA7136P opamp, so switching it out obviously destabilises things.

 

[spaceludwig]

... I don't know, I think I would experiment with unsoldering the transistors at output of IC2 and connect the opamp output to the bipolar cap

That worked. Total current draw for input channel circuit with IC 3 pulled out (not planning on using the monitor channels) is a smidgen under 30mA.  This solves my power supply problem too as it is now relieved of producing an extra 1.2A. Merci beaucoup!

I will do a listening test tomorrow then, assuming all is good, change all the input channels as per the above and post updates. Thank you all for your help, it is very much appreciated.  I will sleep well tonight.

 

[spaceludwig]

I finally solved the "oscillation" problem. If I'm not mistaken, it wasn't oscillation at all.

I don't know if my analysis is correct but after staring at the schematic I noticed the 100k resistor going to pin 5 and 1k going to the base of TR6. If I'm not mistaken this forms a voltage divider, correct? Since pin 5 was not hooked up on my adapters I guess the full 15 volts was going to the transistor which I imagine caused the huge current draw. Anyhow, I hooked up pin 5 and sure enough current draw fell to less than half, around 65mA.

So the question for me now is do I hook up pin 5 and retain the integrity of the channel's sound (Do those transistors contribute something to the sonic character of the signal?) or do I remove them and get the signal straight from the opamp? In the end I guess it boils down to whether or not those transistors are having an effect on the quality of the output signal.

 

[abechap024]

Glad your making headway. That is always a good feeling!


Transistor buffer outputs are generally considered a good thing. They give a low impedance drive that lets the opamp just work on doing what it do, then the transistors worry about pushing the signal around the board.

I have transistor buffers on my board and I look at it as a definite asset. they add some girth to the signal (I like to think anyway)

On the flipside, any extra component you push your  signal through is going to add some distortion to the signal (for better or worse) and noise, in the case with transistors. You might find that the opamps you upgraded to can do just fine and don't need the buffer transistors.

But I think console designers use those designs for a reason. I would leave them in.


You could do some  listening tests...

;D

 

[spaceludwig]

You could do some  listening tests...
;D

That's what I'll do now. I suspect the difference on 2 channels is tiny, but the cumulative effect on 12 more readily noticeable.

And yes, a great feeling of satisfaction when you resolve a problem that consumes your time, thought and energy. However, I find it equally important and satisfying to understand what the problem was to begin with.

I have a few mods I would still like to make but I want to bring this part to a close before I move on. Thanks again to everyone contributing to this thread.

 

[spaceludwig]

So, I've spent the last few weeks getting the gremlins out of some channels, soldering patch-bay (which I HATE , surprised I have any sensation left in my fingers), trying to calibrate the inputs - which are not precisely matched - with 1Khz signal. Everything now appears to be in order.

Since the console has 2 outputs - one balanced, one not - I decided to build  a buffer so I could use my headphones to monitor the output. It's basically a cmoy in a case I made of some transluscent plastic crap I bought at the dollar store with legs fashioned from nuts, nylon bolts, rubber legs & glue from home depot. Pics below:

The sound is clearer than the stock console headphone amp and I like the fact I am hearing the mixbus directly after the output transformer.

I would now like to proceed with my next mod which is to the EQ on the input channels. Basically, I'd like to add some options to the shelves and the mid. I've started another thread and will update this one as soon as I've decided how I'm going to go about it and confirmed that it actually works.

Cheers