All discrete preamp (line driver)

This is based on a 1970s design of a line preamp using all discrete transistors. This is my first prototype, so forgive the messy looking board, the mismatched transistors, and the trimmer resistor hanging by wires.

I just used transistors that I have in my parts box, and used some TO-3 power transistors for that "Neve" vintage look... even though a TO-92 will work just fine. Also replaced some resistor and capacitor values to suit my taste.

The output trafo came from buttachunk and according to him it came from an MCI416 console. It sounds great.

I still need to do the mic preamp portion which will go in front of this unit... This will be powered by 48Volts. (Don't want to deal with separate voltages for phantom and board, so it's just one 48Volt) . In my testing, I powered it with 40Volts (that's as high as my bench PS can go).



It sounds sweet, effortless reproducing bass and low frequencies, and to my ears seems there is a slight scoop in the mid frequencies. It sounds nice in my opinion. And headroom... there is lots of headroom in this baby.

And some oscilloscope shots...

27Volts peak-to-peak @ 1Khz


27Vpp @ 10Khz


22Vpp @ 100Khz


And some square waves

33Vpp at 1Khz


33Vpp @ 10Khz

The rise time is alright, but the falling edge is kinda slow....

22Vpp @ 100Khz


Okay, that last shot there at 100Khz doesn't look anywhere near a square wave.

Are these peak-to-peak values enough, or too much?

Anyways, I plan on building a mic pre with input trafo and phantom power to go in front of this unit for a complete preamp stage.

Buttachunk, it's the "booster" amp circuit on the Ya-ma-ha PM-1000 mixer channel (after the EQ section).

I found out why the trailing edge of my square waves aren't exactly steep... the rise time of the transistor I selected was 35ns, and its fall time was 50ns. That accounts for the inclined slope which gets worse at higher frequencies.

I don't know about that making a difference - I would think it has more to do with the circuit configuration, it looks class-A to me - they usually have very unsymmetrical slew rates. 50 ns is something like 20 MHz, and I suspect the transformer won't pass much at that frequency anyways.

> it looks class-A to me... they usually have very unsymmetrical slew rates

Yes, it is a Class A. Thanks for that info. I didn't know that.

Can we have an idea of the schematic ?

Best,

O :grin:

> that last shot there at 100Khz doesn't look anywhere near a square wave.

Can you hear the difference between the bent 100KHz square/ramp-wave and the original 100KHz square-wave?

If you want a Video Amplifier, build a video amplifier. If you want an audio amplifier, don't test it as a video amplifier. Look for unexpected behavior when driven hard above the top of the band, sure. But some ramping and rounding and ripples are not problems.

> Are these peak-to-peak values enough, or too much?

If you can pass 6KHz Sine up to almost the same level as mid-band clipping, you probably are in good shape. Full power without gross warpage up to 20KHz may be better, or may be no different on actual sounds that you would listen to.