Serious Clipping one channel RIAA preamp.

[brucetassin]

I built the Oatley K303 RIAA preamp some time ago. It works very well BUT
both channel have some clipping, the left more so. I tried a few MM cartridges
with the same results. I'm thinking that R1 (47k) might not be ideal on the left channel, but works better on the right channel. I added a screen shot of the waveforms. Any suggestions would be very helpful.


Thanks

 

[vinylkid58]

If anyone needs, I can upload the schematic.

Probably a good idea. You'll get more help that way.

jeff

 

[scott wurcer]

I built the Oatley K303 RIAA preamp some time ago. It works very well BUT
both channel have some clipping, the left more so.

What are the levels here, running anywhere near clipping during normal play on a 32V rail seems suspicious. This would indicate no dynamic headroom at all.

 

[brucetassin]

I added the schematic.
The power supply is +47 vdc single rail.
I use Audacity for capture, I'm not sure what the scales on the screenshot mean.

 

[rayma]

The power supply is +47 vdc single rail.

If only one polarity of the output signal is clipping, the DC bias point(s) could be off.
Are the drain DC voltages around half the supply voltage?

Try a simpler input waveform, like a triangle before the inverse RIAA, to see the
problem more clearly.

 

[scott wurcer]

The schematic has 32V supply on the first stage and 18V on the drain for 1.4mA current in the first FET. With 100 Ohms in the source you should see ~140mV. You should check the bias levels to make sure that the FET's are not fakes. With the values on the schematic you should get way over line level before clipping.

 

[brucetassin]

Time for the volt meter.
Thanks for your help.

 

[davidsrsb]

Just measure the four jfet drain voltages and post here.
While you have Audacity, use analyze - plot spectrum. The high gain on the left channel might be due to a dry joint in the passive RIAA and this will show in the frequency spectrum

 

[brucetassin]

So here are the voltages and the graph. I don't know how to choose which channel the graph shows.

 

[scott wurcer]

So here are the voltages and the graph. I don't know how to choose which channel the graph shows.

The second stage on the left channel is 2X low on the designed current draw. The numbers given would require the highest Idss range (BL) on the 2SK246.

 

[davidsrsb]

The input seems to be designed for GR selection 2SK170 fets and seems about right
The output is designed for GR 2SK246 fets and it looks like you have a "Y" in CH1.
Still there is plenty of voltage swing available.

Running 50V electrolytics at 50V or more is not a good idea

For the spectrum, I meant analyse L and R separately

 

[PRR]

Looks like one channel has a lot more gain than the other.