If You Don't Have a SIT, Fake It...

[vgeorge]

Hi,

I put on a breadboard your first schematic as you can see on the pictures, to try it, as I need a pre with a little bit of gain.
I made it as per the attached schematic, with the following changes: R2 = 47K, R4=180K // 5pico silver mica (I also had some overshoot on my 10khz square, but now looks fine as you can see, also wanted to reduce gain a bit so I used an 180k), for R7 I used a trimmer in order bias the circuit (in my case around 510ohm), and for current source I used an IXYS IXTP01N100D set at 35mA.
For voltage regulation, a Salas BIB set at 30V and 210mA.
I used a 10k pot at the input.
Measurements on the scope were fine, with a bandwidth of 330kHz.
However, when I listen to it I got a quite high level of noise for my liking. Keep in mind that I use horn speakers which are more than 100dB sensitive. On less efficient speakers, this maybe tolerable, but not in my case.
Gain was still high for needs. As I understand, gain is determined by the ratio of R4/R2. Is it possible to reduce their values, without having to add a bigger cap at the input?
Despite that, I quite liked the sound initially, especially the bass was very good.

Any help on all these will be appreciated.

[wrenchone]

If you used a pot to set the bias level and kept in in, this might account for the noise you mention. Even a nice Noble 50k volume pot will hiss like a nest of snakes with some DC across it (I know from sad experience). If the pot is still in place, try replacing it with a fixed resistor equal to the pot resistance at its current setting (just use a meter to measure) and you may be very pleasantly surprised. Since this is, after all, a line-level amplifier with low gain, I can't think of anything else that could account for audible levels of noise, even with sensitive speakers.

BTW, kudos to you for actually taking the initiative to build this circuit and listen to it. You actually beat me to listening tests. I'm still playing around with my distortion test setup so I can get realistic and consistent readings.

[vgeorge]

Yes, I still have the pot. I will replace tonight and report.
What do you think about the gain? Is it possible to reduce the ratio R4/R2, without having to add a bigger cap at the input?

[wrenchone]

What are you trying to do? The circuit is presently set up for a gain of 5 . If you want much less gain than that, why not use a follower? It's easier and less problematic.

[vgeorge]

Only that I would like a little bit less gain (3 to 4). I may try it as a follower also. Thanks.

[Salas]

Try this g=~3 with lower impedance in the inverting amp's Rf/Ri for lower hiss maybe?

[vgeorge]

Will try it.
Also, with the 117 I use at the input, I replaced the trim with a 330ohm resistor, instead of 680, in order to correct bias the circuit.
I only wonder what the influence of the electrolytic at the input to the sound. Will see.

[Salas]

Q1 170 was just what I had in the lib. You bias with your IDSS, and new R2,R3 ratio plays a role (R5,R4 in wrenchone's schema). So your 330 may get revised or not. See to have VR4/R4=<IDSS Q1 and ~16V before C2 to GND.

[wrenchone]

Salas, the whole point of this exercise was to be able to push up the impedance of the feedback network and yet have some drive available for the mosfet gate. The way you show it, you really need that electrolytic (or a great honkin' film cap), plus a buffer to drive the buffer.

[wrenchone]

One variation I'm thinking of trying is to replace the bottom resistor of the output voltage centering/gain network (R5, per the schematic in Post 41) with an active current sink. A low IDSS, low leakage electrometer fet like the PN4117 would be just the ticket as an active current sink for the few 10s of microamps that are needed to properly set the bias. Just a thought...

I'm currently working on a board that incorporates all the cascode enhancements shown in previous posts, plus an onboard shunt regulator. I'm curious to see how it will perform in my distortion test setup.It would be interesting as well to have it drive a simple buffer and a pair of headphones for a subjective evaluation. I may adopt this approach so I can avoid ripping up my living room setup every time I have a bright idea...