Hi Wrench,
here is my take on "turning a (MOS)FET into a SIT/triode". Fair Zin, low Zout, BW(-3dB)=1MHz, rail-to-rail swing (almost), high PSRR, Id about 15mA, sounds wonderfull - what more could a man want ?
Built, tested, measured (except for THD) - here is the schematics:
Juma, as always another jewel from you. Thanks.
BS170 and MMBF170 (SMD) are easily available. Is this schematic to be used as a preamp or the input stage for a follower output stage?
I'd recommend BS170 because of higher Pd_max.
Both, but if you wish to use it to drive a small power follower some more gain is needed (make R7=10k).
Juma, in the Preamp for F5 thread you state that this circuit's output will be 8 Volts. So with a Follower stage biased at 1 Amp, this will be a good 8 Watt SIT amp. Ofcourse, a 4 ohm load will require a bias of 2 Amps for the same output power. Then this will no longer be a small power follower. What is your idea of "a small power follower"?
Hi Wrench,
here is my take on "turning a (MOS)FET into a SIT/triode". Fair Zin, low Zout, BW(-3dB)=1MHz, rail-to-rail swing (almost), high PSRR, Id about 15mA, sounds wonderfull - what more could a man want ?
Built, tested, measured (except for THD) - here is the schematics:
... a volume pot?
Where is a good place for a volume pot
Or is it useful to use an additional input buffer like B1 and use the volume pot there?
thanks & cheers
Thorsten
It depends on whose circuit you choose, I guess. I deliberately kept impedances fairly high in my implementation so that I could park a pot at the input - indeed, I have a few Noble 10k "black brick" pots that would be perfect for the application. Using a buffer to drive a buffer makes no sense to me.
In order to address the noise issue, I went out and bought a Y cable that interfaces a pair of RCA jacks to a stereo 1/4" jack. With this, I can use a pair of sensitive and fairly strident Sony headphones (MDR-V6) to listen for noise with the inputs of my current preamp setup shorted. I'm curious about what I'll find. Right now, one side of my preamp is faster than the other. This may be due to an accident I had in the early phases of bringing up the board, in which case matching another pair of input jfets may make them identical. We'll see...
In order to address the noise issue, I went out and bought a Y cable that interfaces a pair of RCA jacks to a stereo 1/4" jack. With this, I can use a pair of sensitive and fairly strident Sony headphones (MDR-V6) to listen for noise with the inputs of my current preamp setup shorted. I'm curious about what I'll find. Right now, one side of my preamp is faster than the other. This may be due to an accident I had in the early phases of bringing up the board, in which case matching another pair of input jfets may make them identical. We'll see...
I put it at input (10k log).
Considering the type of the feedback topology it's obvious that using a pot this way will influence the impedance of the signal source (by adding a part of pot's resistance in series with it) and the gain of the preamp (slightly). Fortunately those influences are small enough not to affect any of important characteristics (measurement or sound-wise).
Pass-band is most affected at the point of half value of the pot but even then the band stays wider than 100kHz.
So, all in all, no damage done...
Simulations indicate that the 10k pot doesn't do much of anything to the distortion characteristics of the circuit - I used a value appropriate to a 10k linear pot at half-mast. Obviously, a higher impedance pot will disturb the preamp circuit more, but load the input less. 10k is the lowest value I would consider inflicting on my system components, though it would be an interesting exercise to dig into some of them and install decent output buffers - so many projects, so little time... This one here started off as an idle whim - that's how it goes.
The effects of a pot are something I can check with my analyzer after I home in on a setup that gives consistent results.The results I've gotten so far are already pretty good for a circuit with so few parts, even though they don't match the simulations (which of course, are noiseless, with a perfect input generator, and have no hum pickup and power supply interactions). I'll be reasonably satisfied with the circuit when I get 1KHz and 10kHz readings that match at an input excitation that gives me 0.5V to 1V output. This will clip any amp I have in my system at present, as I usually size the gain for max output at 1V input.
A circuit less finicky about input impedance would be the "liniac inspired" circuit I posted here a while back. It is also a preamp with gain, but is non-inverting. It's currently playing in my living room. When I come up with something satisfying to replace it for a while, it'll take its turn in the analyzer "hot seat".
I have access to an Audio Precision System Two at work, but I'd need to take some weekend time to go there and wrestle with both the setup (not the best by any means) and the user interface.Time spent here at home perfecting a test enclosure/setup will pay off there, both in time spent hair-tearing, and in ultimate results. I'm also building a Pete Millett sound card interface, but I don't expect that effort to bear fruit any time soon.
The effects of a pot are something I can check with my analyzer after I home in on a setup that gives consistent results.The results I've gotten so far are already pretty good for a circuit with so few parts, even though they don't match the simulations (which of course, are noiseless, with a perfect input generator, and have no hum pickup and power supply interactions). I'll be reasonably satisfied with the circuit when I get 1KHz and 10kHz readings that match at an input excitation that gives me 0.5V to 1V output. This will clip any amp I have in my system at present, as I usually size the gain for max output at 1V input.
A circuit less finicky about input impedance would be the "liniac inspired" circuit I posted here a while back. It is also a preamp with gain, but is non-inverting. It's currently playing in my living room. When I come up with something satisfying to replace it for a while, it'll take its turn in the analyzer "hot seat".
I have access to an Audio Precision System Two at work, but I'd need to take some weekend time to go there and wrestle with both the setup (not the best by any means) and the user interface.Time spent here at home perfecting a test enclosure/setup will pay off there, both in time spent hair-tearing, and in ultimate results. I'm also building a Pete Millett sound card interface, but I don't expect that effort to bear fruit any time soon.
Last edited:
I did the headphone experiment, and the phones were dead quiet - no difference between one side connected to the UUT, and the other connected to nothing. If I touch the ground side of the input connector, I get a faint trace of line frequency hum (very, very faint). If I touch the input directly, it's knitting needle through the ear time. I think my setup is pretty quiet. All the resistors of consequence are metal film. The input connectors were left open. If I shorted those (or used a 5k resistor), I suspect the beast would be even more resistant to casual touch around the input connectors.
Hi Wrench,
here is my take on "turning a (MOS)FET into a SIT/triode". Fair Zin, low Zout, BW(-3dB)=1MHz, rail-to-rail swing (almost), high PSRR, Id about 15mA, sounds wonderfull - what more could a man want ?
Built, tested, measured (except for THD) - here is the schematics:
I like this one
How would you evaluate it between all those fine circuits you and others have made in the past? I am thinking power follower driving duty...
Just to name a few:
-BoZ
-JBoZ
-B1 (sound character)
-Triode with CCS load
-Other relevant
Na, building will be done nomatter what within a few weeks
I'll be starting another thread in the Pass Labs section called L'Fake, a response to the L'Amp featured there that uses the scarce Yamaha SITs. The mosfet/jfet combination I showed at the beginning of this thread is perfect for making a fake SIT with a largish mosfet like the IRFP140/240 or the IRFP044/044N, among others. I also have in mind some largish TO-247 Fairchild "house numbered" devices I got for a relative pittance from Electronic Goldmine. The nice thing is that this setup will work with just about any mosfet that's large enough to dissipate the required power. They just about all turn on substantially at 4V or so, so setting up the bias current and output centering is relatively straightforward.
Anyway, I've done sims with both resistive and current source loading - the current source load sims better in terms of THD. For someone that wants to mess around with a simple SE Class A amp using readily available parts, this could be the ticket. All sims I've done so far are dominated by 2nd harmonic, with the 3rd down 10-20 db, and the rest much lower.
Anyway, I've done sims with both resistive and current source loading - the current source load sims better in terms of THD. For someone that wants to mess around with a simple SE Class A amp using readily available parts, this could be the ticket. All sims I've done so far are dominated by 2nd harmonic, with the 3rd down 10-20 db, and the rest much lower.
The Goldmine part turns out to be an HUF75329G, with much less gate capacitance than the other two devices I mentioned previously (~1000pF vs 1500 or more). It's rated for 55 V, so will be perfect for a Pass-ish single-ended amp. Goldmine is sold out of these devices, unfortunately, though those curious enough might be able to find the same thing elsewhere, or something similar or better in Fairchild's current lineup of devices.
I wouldn't obesess about it though, as just about any mosfet capable of dissipating the required power and voltage would work.
I wouldn't obesess about it though, as just about any mosfet capable of dissipating the required power and voltage would work.
To drive a small PF (5-10W, 24V PS, see post #65) I'd use a Mu follower variant:
Attachments
If this fakes a SIT or a triode, I guess my ZEN4 is already a fake SIT amplifier?
And when I get around to upgrade it with the Semisouth FET, will it still be?
Hi Juma,
I'm just collecting parts to build your nice pre amp for my F3.
For tweaking my Naim gear I have a nice PSU which provides +/-15 V powered by two 15-0-15 V centertapped r core transformers. Regulation is alternatively ALW Superregs with AD797 or Teddy Pardo Regs.
While the Naim pre is also a SE amp, the symmetrical PS is a nice step forward.
I guess it would also be nice to drive your pre amp with a symmetrical PS. What has to be changed in your circuit and which regulation would you prefer?
thanks in advance
Thorsten
I don't see the advantage in using the symmetrical PSU. Circuit would lose its elegance, and coupling cap(s) would still be present.
It can be redesigned but it wouldn't be better...
The circuit has high PSRR and I didn't notice the improvement by changing the PS from LM317 to shunt reg...
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.