Hello.
I searched the forum and did not found any measurement for low noise high voltage mosfet or jfet.
I want to use them as a buffers (and also 1 as gainstage) in a valve amplifier circuit, with a supply of 350Vdc.
Do you have any recommendation ?
It would be better if you told me how to choose for low noise from reading the datasheet.
Thank you.
I searched the forum and did not found any measurement for low noise high voltage mosfet or jfet.
I want to use them as a buffers (and also 1 as gainstage) in a valve amplifier circuit, with a supply of 350Vdc.
Do you have any recommendation ?
It would be better if you told me how to choose for low noise from reading the datasheet.
Thank you.
Noise testing at the operating point would be best, but in general look for high transconductance.
A buffer application would not be critical.
A buffer application would not be critical.
Some people use a "cascode circuit" to synthesize a high voltage device using two individual semiconductors. Maybe you might want to study that concept and see whether it accomplishes your goals.
I see lots of great looking search hits when I google for cascode tutorial
I see lots of great looking search hits when I google for cascode tutorial
Thank you rayma. gm = id / vgs. Should I look for High current mosfets. Sorry I am really new to this.
Thank you mark that is very intesting, because if I understand correctly the gain is the same, but it allow me to use lower voltage/lower noise fets. That is really interesting.
Thank you mark that is very intesting, because if I understand correctly the gain is the same, but it allow me to use lower voltage/lower noise fets. That is really interesting.
No I don't really have a circuit it's a "workbook" question.
I need them to lower the impedance to use a smaller potentiometer.
But I understand now it's not (gm = id / vgs) it's gm = delta id / delta vgs.
So it's the "change" in voltage and current not the "static" id and vgs.
I need to understand how to interpret the datasheet.
I need them to lower the impedance to use a smaller potentiometer.
But I understand now it's not (gm = id / vgs) it's gm = delta id / delta vgs.
So it's the "change" in voltage and current not the "static" id and vgs.
I need to understand how to interpret the datasheet.
Last edited:
Transconductance is the slope (tangent) at the operating point. This can be estimated by
drawing a tangent line to the curve and calculating its slope.
If you intend to use a 10k volume control as a load, most any suitable fet will be ok.
Will the fet be within a feedback loop?
drawing a tangent line to the curve and calculating its slope.
If you intend to use a 10k volume control as a load, most any suitable fet will be ok.
Will the fet be within a feedback loop?
Last edited:
Thank you. Not in a feedback loop. And I will just use a simple source follower.
I know most of them will work, i am just trying to optimize every bit of noise I can avoid. Sorry.
So it's just like tubes you have to draw on load lines, at the mA the mosfet will operate. Ketje showed me how to do it for tubes. I will have a lot of work to do !
I will try to look again on the forum to see if there are some mosfet that are popular for low noise. If any of you have some favorite please tell me.
And thank you again guys for for your answers and for being always kind and helpful.
I know most of them will work, i am just trying to optimize every bit of noise I can avoid. Sorry.
So it's just like tubes you have to draw on load lines, at the mA the mosfet will operate. Ketje showed me how to do it for tubes. I will have a lot of work to do !
I will try to look again on the forum to see if there are some mosfet that are popular for low noise. If any of you have some favorite please tell me.
And thank you again guys for for your answers and for being always kind and helpful.
I doubt high voltage mosfets are noisier than valves.
For the gain stage I would use a cascode, then the small signal part only sees a low voltage and Miller effect is avoided
Using a mosfet in high voltage linear circuits needs care, often the device SOA is not given and you need to limit Vgs while clipping, switch on/off etc
For the gain stage I would use a cascode, then the small signal part only sees a low voltage and Miller effect is avoided
Using a mosfet in high voltage linear circuits needs care, often the device SOA is not given and you need to limit Vgs while clipping, switch on/off etc
Do you have 300 Volt signals?
Usually an audio potentiometer works with signals of less than a few Volts at its input. And the buffer driving it can use 10V or 20V power.
And the hiss of a JFET may go way up at higher voltage!
And there are few places in audio that the noise impedance can't be handled by a well-scaled BJT instead of a JFET. (MOSFETs are prone to 1/f rumble so rarely a first choice in low-noise work.)
And if this potentiometer *reduces* signal, then the hiss is dominated by the stage *after* the pot. Not the buffer before the pot.
Having 350V when you need 30V is a minor problem. There are many ways to reduce DC voltage, and you can also design-in filtering.
Thank you everyone.
Yes it's this preamp. I will replace the 1M potentiometer in this circuit by a 5K one if possible.
But I did not post it because I wanted to know the theory also. I did not want to only have the suggestion and basta. I was affraid you guys will probably say, you don't need a low noise part everything will work fine.
That is why, sorry.
Yes it's this preamp. I will replace the 1M potentiometer in this circuit by a 5K one if possible.
But I did not post it because I wanted to know the theory also. I did not want to only have the suggestion and basta. I was affraid you guys will probably say, you don't need a low noise part everything will work fine.
That is why, sorry.
How can I do this without adding too much noise (or lowering the first gainstage also)
Frederico-
If the Altec preamp is "noisy", you are using it wrong.
However, just because something is vintage does not mean it is good.