OK, I purchased and installed the output protection relay board and after firing it up I had some hum on the right channel. 
. So in the process of trying to track down the source, I managed to touch something that bricked it. (DC at output) Well, not really, because I painstakingly went through all of the components and circuits trying to find a bad component, and everything looked fine. Grrr. Then I decided that I needed to figure out how the thing is supposed to work so that I could track the voltages through the driver board. (driver board had to be the culprit) I took a photo of the trace layout, flipped it and overlayed with the component layout photo. I brought those into Visio and added components and traces. You know how much a PITA it is to reverse engineer even a fairly simple circuit? Well this circuit is not all that simple... Anyway I then got a copy of LTSpice and entered the circuit from the service manual (there are two versions in the manual), then compared the actual circuit from the unit to the spice drawing. Not the same. A third version! They are all fairly similar but there are some differences mainly at the signal input and somewhat different resistor values in various places. I wanted to try to get the DC bias analysis from the spice model to figure out why I had DC at the output. I never really got that to work, partly I think because the amp uses a transient at startup to bootstrap the driver into an "on" state. That is what I think, anyway.
Result. I gave up on the spice model but I learned a lot about the circuit in the process. Carver inverts the signal to the right channel at the input. There is an op-amp at the input to each channel (LF356) and in the manual he drives the left side into the non-inverting input and the right side into the inverting input. On my amp, it is different. He drives both on the non-inverting input. He then added another op amp with unity gain to the right channel to invert the input phase. This turned out to be the problem as that op amp must have gotten fried and was putting out about 2V DC with zero volts input. What I did for now was just to hook the right input directly to the non-inverted input op-amp and bypass the extra op-amp. That seemed to set everything straight.
My question then is... Why did Carver invert the right channel phase? I mean he makes a big deal of it in the output connectors and in the service manual, but is it really necessary? Is this a trick to even out the power draw in the amp reducing distortion?
Oh, and I now have driver board layouts and circuit schematics. I still have no idea how the protection circuits are supposed to work, and I don't really understand the signal flow from input to output of the driver board.
Anyway, the beast works, and I'll probably order an RC4136N quad op-amp if they still make them. It's funny that he had used this extra quad op amp on the right side but only used one section and shorted the other three.
Sterling
. So in the process of trying to track down the source, I managed to touch something that bricked it. (DC at output) Well, not really, because I painstakingly went through all of the components and circuits trying to find a bad component, and everything looked fine. Grrr. Then I decided that I needed to figure out how the thing is supposed to work so that I could track the voltages through the driver board. (driver board had to be the culprit) I took a photo of the trace layout, flipped it and overlayed with the component layout photo. I brought those into Visio and added components and traces. You know how much a PITA it is to reverse engineer even a fairly simple circuit? Well this circuit is not all that simple... Anyway I then got a copy of LTSpice and entered the circuit from the service manual (there are two versions in the manual), then compared the actual circuit from the unit to the spice drawing. Not the same. A third version! They are all fairly similar but there are some differences mainly at the signal input and somewhat different resistor values in various places. I wanted to try to get the DC bias analysis from the spice model to figure out why I had DC at the output. I never really got that to work, partly I think because the amp uses a transient at startup to bootstrap the driver into an "on" state. That is what I think, anyway. Result. I gave up on the spice model but I learned a lot about the circuit in the process. Carver inverts the signal to the right channel at the input. There is an op-amp at the input to each channel (LF356) and in the manual he drives the left side into the non-inverting input and the right side into the inverting input. On my amp, it is different. He drives both on the non-inverting input. He then added another op amp with unity gain to the right channel to invert the input phase. This turned out to be the problem as that op amp must have gotten fried and was putting out about 2V DC with zero volts input. What I did for now was just to hook the right input directly to the non-inverted input op-amp and bypass the extra op-amp. That seemed to set everything straight.
My question then is... Why did Carver invert the right channel phase? I mean he makes a big deal of it in the output connectors and in the service manual, but is it really necessary? Is this a trick to even out the power draw in the amp reducing distortion?
Oh, and I now have driver board layouts and circuit schematics. I still have no idea how the protection circuits are supposed to work, and I don't really understand the signal flow from input to output of the driver board.
Anyway, the beast works, and I'll probably order an RC4136N quad op-amp if they still make them. It's funny that he had used this extra quad op amp on the right side but only used one section and shorted the other three.
Sterling
To to try to answer your question on why the one inverted channel it is done to give the ability to make a bridged mono output from both channels. The two channels are in fact out of phase. The wiring on the stock binding posts reversed the pos and neg of one channel so it would deliver in phase out put. By bridging the two outputs the mono channel is created by using the two hot legs of each channel to the speakers. When in bridged mode the amp sees a load equal to 1/2 the connected speaker load. so if you have 8 ohm speaker the amp sees a 4 ohm load. This also more than doubles the output power and cuts the damaging factor in half. This bridged setup is useful to drive a sub or in a multi amp system.
inverting the op amp on one channel is done by having the dual op setup on that one channel. Installing a Whiteoak pl14-20 board you loose this ability, how offen does one need to bridge an amp other than Perry and Me that is.
If if you added Don's protection board as I did then you must check the path of each pos an neg input to the board from the backplane. This should reverse the binding posts from stock on one channel. Check if the binding posts are now out out of phase. With the DCP board in is imparitive that the ground and hot go through the board correctly. If you have done this correctly the binding posts will now be out of phase and this can be easily corrected by revering the positive and negative on one speaker cable. Bob jockeyed the wiring in the original to make it simple to get back in phase at the outputs.
The last thing is the protection circuit is kind of smoke and mirrors combined with some function. This circuit is a derivative of the Pl32 board used in the d-500 phase linear. It reduces the input signal to the opamp on both channels to a point where the variance dectected is reduced back to normal range though lowering output
with the added DCP board you have increased the safety margin to the point where I think the on board cuircuit is redundant and can be eliminated if you choose.
Here are photos of the DCP board setup on a stock c-500 output. The board I built for my c500 is built inverted compaired to the phase 400. The c500 is shown in the top half of the two assembly pictured and the phase 400 in the lower half. This was necessary due to the outputs and inputs location being inverted on the chassis between the c500 compared to the phase 400
Attachments
Last edited:
