Troubleshooting audio inverter circuit

Thread Starter

Dean Rantala

Joined Sep 27, 2018
Here is what I am working with, I have tried to label everything as best as possible:


So... This is the inverter board for my Harman Kardon 75+. The thing works great, but it is the power-up that is giving me issues.

First, I will explain a little backstory. I had to re-work the entire power supply (for the 30VDC). It was originally fed from a ~42VDC transformer/rectifier pair. From there, it was fed into a R/C network, some zeners and a couple power transistors to provide +30 and +12 volts (30 for pre-amp and inverter circuits and 12 volts for the tuner board). The board seemed to create a (give or take) 30 minute ramp up from 0 -->> 30vdc. The original board was shot, so I re-built one using a couple LM317's and a simple linear ramp circuit (about 18-ish seconds). Long story short: the 30 volt supply to the above circuit slowly ramps to the full 30 volts.

Now, the purpose of the above board is due to this being a quad or stereo amp (depending on a switch on the back). In quad mode, all 4 amps are essentially directly fed the direct signal. However, in stereo mode - the amps are bridged 2+2. So the signal to one of each of the pairs must be inverted. That is essentially what the above circuit does. I run this ONLY in stereo.

The issue is how the inverter board comes up. As power is ramping up, there is a strong positive and negative DC offset at the speakers. The offset starts positive for half the ramp, followed by a negative offset. Obviously, it is really hard on the speakers - I have smoked a couple cheap bench speakers (only for testing stuff - so no big loss.. yet). After the +/- offsets, the amp runs great.

I have determined (correct me if wrong here) that the offset must have something to do with how C753 and C754 (as an example for the left channel) has to charge when power is applied. I *think* the transistor (for each channel of the inverter board) is essentially a dual class a amplifier that couples one amp to the collector, the other to the emitter side.. thus "inverting" each signal. At least that is how I understand it. I think the time needed to charge across the two output capacitors is essentially what is causing this +/- offset... correct?

Any insights as to what the issue could be? Or this is simply how the circuit is essentially supposed to be? Do I need a longer ramp-up time to minimize the offset? Perhaps it would be easier to build some sort of clamp circuit? Or better.. maybe a custom-built inverter board with a couple NE5532 op amps would simply perform better?

I have already replaced all the capacitors AND both transistors.. and I doubt there are bad resistors on BOTH channels causing an equal offset.

- Dean


Joined Sep 9, 2010
I can't answer your questions but I have one for you: Why are you calling this your inverter? Don't you mean amplifier?


Joined Apr 5, 2008

I would rather call the circuits a phase splitter.
The rear signals are in anti-phase with the front signals.


Thread Starter

Dean Rantala

Joined Sep 27, 2018
Well... Thats what it is called in the schematic/manual.. "inverter board". As mentioned... It basically inverts (splits, reverses, whatever) the phase 180 degrees to one amp.

This is required when the stereo is in stereo mode becauce the 4 amps are bridged 2+2. So each channel is using two amps (bridged).

An update....

I have bypassed the entire pre amp and inverter circuits with a simple ne5532 and 100k pot... Damn... Much better! No DC offset and sounds better than new. On par with my adcom gear imho.

So.... Im working on gutting entire pre amp stage and replacing with couple custom made pcb's and some 5532s. The quad feature will permanetely be disabled in the proccess.