THAT1512 / Transformer 'blind test' mic preamp

[jamesk]

@DaveDC I reversed all the diodes as they were all using the same symbol. Or are you saying that D6 is wrong on the schematic?

 

[DaveDC]

No, D6 is okay as long as it is really the correct polarity, with cathode connected to the +48V output

 

[jamesk]

@DaveDC Just double checked and yes, cathode is connected to the +48V output. The 0.17V between the output and adjust pins is very strange, and I don't understand where R6 is sourcing all it current from. It should only have 2mA through it, not 30.

 

[DaveDC]

Doesn't the anode of D4 want to be GND?

 

[jamesk]

I just desoldered D5, D6, R5 and R6 and i still have 85.8V on the output and 85.6V on the adjust pin.

 

[jamesk]

@DaveDC It's connected to the unregulated +ve rail so up the available voltage for the +48V regulator, which it seems to be doing. The regulator doesn't seem to be regulating at all, though.

 

[DaveDC]

Did you destroy the regulator?

 

[jamesk]

@DaveDC I did wonder that. I'm trying to think how having those diodes all in back to front could have fried that regulator.

 

[DaveDC]

Backwards power diodes can do a lot of damage.

BTW, what is your transformer turns ratio?

 

[jamesk]

@DaveDC I removed the regulator entirely, along with R5, D6, D5 and R6, so that the only thing connected to the +48V output was C12 and still measured +55V on the output, which had me completely baffled. I then removed C12 and measured 0V. I then reconnected the other components, including the regulator, but with C12 removed, and still measured 85V on the output.

Perhaps the regulator is fried and is passing the full voltage to both the output and adjust pins. That 85V could then have destroyed C12, which is only rated at 63V. I still don't understand how I'm measuring 55V with the regulator disconnected, though.

The transformer has dual 18V secondaries.

 

[jamesk]

@DaveDC Okay, I've removed the regulator again and it turns out there's a dead short between the input and adjust pins, so the full unregulated voltage is ending up at the adjust pin. It looks like the regulator is fried, I think! Still not sure about that 55V reading with it removed, though.

 

[DaveDC]

With no load, the input to the +48 regulator should be 3X the input to the +17 regulator. Is that what you are measuring?

 

[jamesk]

@DaveDC The +55V was measured with the regulator removed entirely, so the only component connected to the output pin was C12. Removing that gave 0V.

 

[DaveDC]

I asked you about the voltages at the inputs to the regulators.

 

[jamesk]

@DaveDC Sorry, I thought I hadn't been clear about where I was measuring 55V. I've just checked again and yes, I'm getting the correct voltages at the inputs to the regulators. around 30V at the input to the 17V regulator and around 90V at the input to the 48V regulator.

 

[DaveDC]

Wow those voltages are pretty high. You will be wasting a ton of power in your regulators if that measurement is real.

Are you sure you have 230 VAC mains voltage and 18 VAC windings? You may have HF hash on your mains and your voltmeter is giving you a peak voltage instead of average. Try loading your meter with a few k ohms shunt and remeasure. And check the waveforms with your scope.

 

[jamesk]

@DaveDC I've just scoped the various voltages. I'm getting 30V peak on the transformer secondaries. My mains voltage is 240V so by my calculations, I should be getting 26.8V peak, not 30V.

The transformer output is clipped like this.

 

[DaveDC]

This is no-load, not even the regulators? I’d expect some flat topping when the diodes turn on under load and the source resistance of the transformer drops the voltage. But no-load should not be like this. You might abandon your PCB for now and dead-bug just the transformer, rectifiers, and filter caps until you can explain what is happening.

 

[jamesk]

@DaveDC That's with the fuses removed so nothing connected to the secondary, not even the rectifier.

 

[DaveDC]

If you have a spare bare board, assemble it piece by piece starting with the transformer, and measure it carefully as you add each new piece. If you see something you don’t understand, don’t proceed until you have an explanation. Use the unloaded transformer to view your mains waveform. Use an AC voltmeter to measure both the primary and secondary voltages to confirm the transformer turns ratio. Then add some diode (correcting for polarity) and check the waveforms. You may need to load your scope slightly to swamp the effects of stray capacitance and leakage. Once the diode waveforms agree with theory, add the capacitors one by one, again confirming the waveforms and voltages.