Solid State Pultec-Odd Frequency Response

[craigmorris74]

I've built a stereo Pultec using HS-66 input and output transformers and an API 2520 makeup amp.

The frequency response on both channels looks odd, with a low and high frequency bump at all times, even bypassed.

I've attached the measured response and the schematics I used for the build.

Any help with this would be greatly appreciated.

Craig

 

[radardoug]

The problem is these days that there are all these computer programs that masquerade as test gear, and are then operating by the uneducated. Pull your Pultec out of circuit and do a loopback test on your test gear. You need to test and verify your test gear. That app you are using, NO SCALES! Its crap. The low frequency spike looks very much an artifact.

 

[craigmorris74]

The problem is these days that there are all these computer programs that masquerade as test gear, and are then operating by the uneducated. Pull your Pultec out of circuit and do a loopback test on your test gear. You need to test and verify your test gear. That app you are using, NO SCALES! Its crap. The low frequency spike looks very much an artifact.

The measurements were taken with REW, a pretty standard program for measurement. My setup is verified. A DA/AD loopback is flat. My other gear tests as expected. I chopped the scale off accidentally, sorry about that. The lines are in 1dB increments.

 

[radardoug]

So that bass response definitely looks like an artifact of your setup. Its too resonant. The top end looks like transformer responses. Is your bypass a hard bypass removing all circuitry, or does it just stop the boost/cut?

 

[pucho812]

I had a pultec years ago where the client claimed it sounded like 5K.
Well sure enough it did a 5K bump but this was because after 50 years of duty, a solder joint went wonky on the low frequency pot. once fixed all was right as rain. confirm connections.

 

[craigmorris74]

So that bass response definitely looks like an artifact of your setup. Its too resonant. The top end looks like transformer responses. Is your bypass a hard bypass removing all circuitry, or does it just stop the boost/cut?

I had a pultec years ago where the client claimed it sounded like 5K.
Well sure enough it did a 5K bump but this was because after 50 years of duty, a solder joint went wonky on the low frequency pot. once fixed all was right as rain. confirm connections.

Thanks for the replies.

Bypass only removes the EQ circuitry.

I probably need to bypass the EQ and just test the amp. The low frequency pot would be a good place to start looking.

 

[craigmorris74]

I've ruled out the makeup amp. It has a pretty flat response.

The high frequency lift is caused by an error somewhere in the filter. I've attached the response graph.

I'm trying to trace the circuit and find the mistake, but haven't found it yet. Would there be anything in particular that would cause this?

 

[radardoug]

So now you are down to 1/2 dB difference. Not really worth worrying about.