I can't get this amp working. The schematic has 5 different ground symbols, 2 are connected by 47 Ohms. The speaker lies between the two center taps. U4 is driven by the smaller power supply. The signal and its NFB go into U4, (pin 5 and 6) and U4 has no Local NFB. The Peavey Co. doesn't sell a service manual and are otherwise no help. I either get no output from U4 or it slams rail-to-rail. On LTSpice it doesn't work, but that might have something to do with the relationship between the center taps which is undefined. Interesting design, eh ?

 

Floating power supply design. Quite common in power amps of that era as all the small signal stuff can be done with an opamp, instead of high voltage small signal transistors.
If you want to simulate it, give the op-amp ±12v supplies referenced to signal ground.
The output is the centre tap of the transformer. Just model that power supply as a ±50V (or whatever it is, schematic isn’t clear) reference to output, and what would normally be the output (emitters of power transistors) connects to signal ground.

What should happen is when the input signal goes positive, he op-amp output goes negative turning on the PNP transistors, which drags the negative of the ±50V supply towards ground, making the centre tap of the ±50V supply go positive, which delivers a positive voltage to the speaker.

if the output of the op-amp (can’t read its number) deviates from 0V it should turn one set of power transistors on.

 

Certainly that circuit schematic is unconventional, big time!
A bit of a challenge to follow.
My guess is that if that opamp IC is not failed, that there is a problem in the feedback loop, either a leaky capacitor or an open resistor. So checking voltages is in order. AND, any poor connection in the common circuit can cause your observed problem.
ALSO, I suggest checking that large stack of diodes that set the operting bias for all of the following stages.
Consider that many times, when others avoid some design choice, there is a good reason for avoiding it. One diode failing open or short will disrupt the whole circuit. OR even a resistor faling out of tolerance.
I would not call it a good design.

 

Floating power supply design. Quite common in power amps of that era as all the small signal stuff can be done with an opamp, instead of high voltage small signal transistors.
If you want to simulate it, give the op-amp ±12v supplies referenced to signal ground.
The output is the centre tap of the transformer. Just model that power supply as a ±50V (or whatever it is, schematic isn’t clear) reference to output, and what would normally be the output (emitters of power transistors) connects to signal ground.

What should happen is when the input signal goes positive, he op-amp output goes negative turning on the PNP transistors, which drags the negative of the ±50V supply towards ground, making the centre tap of the ±50V supply go positive, which delivers a positive voltage to the speaker.

if the output of the op-amp (can’t read its number) deviates from 0V it should turn one set of power transistors on.

Thank you. The Op Amp is RC4558D (typical). Supply is +/- 47V. My question remains: Without a Local NFB on the Op Amp, the output would slam rail to rail. Somehow the NFB going into Pin 5 is supposed to prevents this. As the schematic shows, the power amp outputs to the center tap of the 20V secondary and to the speaker negative, while the speaker positive goes to the 35V center tap with part of the voltage going to pin 5. All components between Speaker positive and Pin 5 test good and still I can't get it to work. That leaves only 8 ohms between the two center taps. Maybe my problem with LTSpice is I don't see how to accommodate more than one ground. LTSpice doesn't see the 35V center tap as a ground and my book on LTSpice doesn't cover this. Do you know of any books that I could read? Thanks again.

 

Certainly that circuit schematic is unconventional, big time!
A bit of a challenge to follow.
My guess is that if that opamp IC is not failed, that there is a problem in the feedback loop, either a leaky capacitor or an open resistor. So checking voltages is in order. AND, any poor connection in the common circuit can cause your observed problem.

All components check well. I wonder what the range of voltages going to Pin 5 should be. Not to mention the relationship between the two center taps. I know that they can't be connected, but is the 8 ohm speaker enough ? I had to replace the transformer that showed signs of cooking. Thank you.

 

That circuit is the most unconventional audio amplifier circuit that I have ever seen!! If it is typical of the PEAVY brand, I would refuse any requests to service such an amplifier.
My complaint is that by altering the common points and power lines it has been made so very much different that even the simulator gets confused.

 

That circuit is the most unconventional audio amplifier circuit that I have ever seen!! If it is typical of the PEAVY brand, I would refuse any requests to service such an amplifier.
My complaint is that by altering the common points and power lines it has been made so very much different that even the simulator gets confused.

It's nowhere near as weird as you think.


As you can see, with the emitters fixed at 0V, the bases only need to swing ±2V (if they are Darlingtons) or ±6V if they are MOSFETs. (They are ususally Darlingtons with 2SC5200/2SA1943 as the output stage). That about of voltage swing can be done with an op-amp, meaning that the high voltage transistors in the Voltage Amplifer stage are no longer required.

 

I'll add two further (and rather more subtle) points.
1) As the top transistor turns on, the output goes more negative, so the phase of the feedback has to be swapped. It goes to the non-inverting input of the op-amp.
2) As the output stage is now common-emitter and not common-collector, it has Miller capacitance. And there's no longer such a handy place to connect dominant-pole compensation, so stability might be a problem. As a result (and also due to using a mediocre op-amp) this type of amplifier tends to have mediocre high-frequency response.

 

The peavey circuit puts the whole high current power supply circuit dancing up and down with the speaker output! I call thata poor choice design. REALLY!!
PLUS, it demands more effort to keep hum out of the output, because the whole power transformer has the output voltage everywhere. AND it demands a second power transformer secondary.
ONE MORE thing is that it will certainly confuse any service tech who is not in on the difference. So it gives them more service business= more profit. The amplifiers I built and sold have a "forever" warranty , except on the tubes and output transistors.

 

The trade off is extra transformer winding (which probably doesn't cost much if ordering hundreds at a time) vs. lack of high voltage voltage amplifier stage components. The hum situation is no different to the normal arrangement, and the transformer secondary having the signal superimposed only means that the output has a small capacitive load to drive.
If the repair tech does nothing more than swap output transistors, then he would be oblivious to the difference. If he needs to connect the 'scope then he might get confused!
It's not a circuit I like, because it doesn't perform well, but it is actually rather common, used by Crown and C-Audio and the myriads of copies.
Its reliability has a bad reputation because there are an awful lot of fake 2SC5200/2SA1943 transistors out there which have found their way into some of the copies.

[Edit] In a mixer amp on any amp with any pre-processing such as balanced line inputs there will already be a low-voltage winding on the transformer for the small signal circuitry, however, for a stereo amp this circuit requires TWO high voltage windings.

 

'It's weird, therefore it's bad' is not a sound (!) argument. There must be thousands of these things still operating around the world. Yes, Ill see your personal incredulity and raise you one ad populum.

One of my cousins found a KB-60 in a barn where it has been abandoned for 20 years, and gave it to me to resurrect. The speaker cone was still intact, and - get this - all of the ICs were in sockets. !!! Recapped everything, upped all of the electrolytics in the signal path by 4x, and swapped out the opamps with NE5532's. This dropped the noise floor by over 6 dB, and fattened up the bottom end. The amp used a not-very-common Switchcraft switching jack, so I called them and got a sympathetic person to send me some "design samples". My brother now uses it as his bass amp.

The rebuild was a trip down memory lane. I remember way-back-when, working out why the amp output was hard-grounded. A memorable analog moment. In all of its various forms, this circuit is a rite of passage for anyone doing component-level maintenance of band equipment.

ak

 

It's nowhere near as weird as you think.

As you can see, with the emitters fixed at 0V, the bases only need to swing ±2V (if they are Darlingtons) or ±6V if they are MOSFETs. (They are ususally Darlingtons with 2SC5200/2SA1943 as the output stage). That about of voltage swing can be done with an op-amp, meaning that the high voltage transistors in the Voltage Amplifer stage are no longer required.

Excellent graphics and summary.

ak

 

That circuit is the most unconventional audio amplifier circuit that I have ever seen!!

Really? Check out Carver and McIntosh gear from the 70's.

ak

 

Interesting design, eh ?

Please post a larger / higher-resolution image, and/or a link.

Thanks.

ak