Hello all
I have just finished a very special Hybrid 50 W power amp using a valve as V amplifier and some MOSFETs as Current amplifiers.
In between those two stages there is a driver.
What would be the 'best in sound MOsfet' ? N channel and 500 V.
The result as it is is really wonderful but the Mosfet I use could be not the best (IRF 740). This power amp is to give out excellent quality sound, not just noise...
What is your opinion and advice ?
I have just finished a very special Hybrid 50 W power amp using a valve as V amplifier and some MOSFETs as Current amplifiers.
In between those two stages there is a driver.
What would be the 'best in sound MOsfet' ? N channel and 500 V.
The result as it is is really wonderful but the Mosfet I use could be not the best (IRF 740). This power amp is to give out excellent quality sound, not just noise...
What is your opinion and advice ?
Try IRFP450, SiHFP450 from Vishay.
These are rated to 500V, 14A cont and 56A surge and 190 watts.
I would suggest a quasi output stage using nmos.
Of course, the IRF710 series is I think 3A but as you say 500V.
Hugh
These are rated to 500V, 14A cont and 56A surge and 190 watts.
I would suggest a quasi output stage using nmos.
Of course, the IRF710 series is I think 3A but as you say 500V.
Hugh
Thanks Hugh
Still, I read and understand that the higher current rating and Voltage rating, the bigger the chip and the bigger the Gate capacitance which is the nuisance.
Therefore I was looking for some low poser high voltage Mosfet, since it will just need to conduct about 20 mA average.
A huge >15 Amps may be too large, I m afraid...
Still, I read and understand that the higher current rating and Voltage rating, the bigger the chip and the bigger the Gate capacitance which is the nuisance.
Therefore I was looking for some low poser high voltage Mosfet, since it will just need to conduct about 20 mA average.
A huge >15 Amps may be too large, I m afraid...
Yes I don't understand either why do people always go with 100x current over rated MOSFET - you end up driving more nonlinear device capacitance to no good effect - can even end up with "second breakdown" like behavior from local hot spotting that would be avoided by using more appropriately sized, higher on resistance parts better suited to linear operation
IXYS makes 100 mA 1 kV N-depletion MOSFET, Zetex, Supertex make 500 V small signal complements
IXYS makes 100 mA 1 kV N-depletion MOSFET, Zetex, Supertex make 500 V small signal complements
Last edited:
Hi Guys
Without a schematic of the circuit it is essentially impossible to answer the question. Everything so far is a guess.
At the very least, a description of what the mosfet is doing would help.
Is it just a buffer into the output tube grid?
Is it a gain stage with signal taken from the anode?
For either case, unless it is driving the speaker output directly or through a transformer, the power rating may not need to be very high. There are qualifications that need to be considered.
P-channel mosfet ratings can be taken at face value for linear use where n-channel mosfet ratings must be de-rated. Most n-channels are designed for switching and use the smallest die that will do the job. Linear is more wasteful so needs better heat dissipation and management. The usual route is to use a larger die for linear, so you just go to a larger device.
On the other hand, as jcx points out, you want to keep gate capacitance small so better to find a compromise - or rather, the most appropriate choice - for the application. There are many low gate charge devices available now with high voltage, current and power ratings. If your app requires significant power to be handled by the mosfet, then you should go to case styles larger than TO-220.
Remember that even as a buffer operating at high voltage, say because you wanted direct coupling between the tube gain stage and the mosfet, there is waste heat to consider. The idle current through the mosfet will heat it. A TO-220 in free air has a thermal resistance of 65C/W and it will be very easy to have the mofet running very hot - too hot without some heat sinking. The larger cases are 45C/W so get less hot for a given idle current.
Have fun
Kevin O'Connor
Without a schematic of the circuit it is essentially impossible to answer the question. Everything so far is a guess.
At the very least, a description of what the mosfet is doing would help.
Is it just a buffer into the output tube grid?
Is it a gain stage with signal taken from the anode?
For either case, unless it is driving the speaker output directly or through a transformer, the power rating may not need to be very high. There are qualifications that need to be considered.
P-channel mosfet ratings can be taken at face value for linear use where n-channel mosfet ratings must be de-rated. Most n-channels are designed for switching and use the smallest die that will do the job. Linear is more wasteful so needs better heat dissipation and management. The usual route is to use a larger die for linear, so you just go to a larger device.
On the other hand, as jcx points out, you want to keep gate capacitance small so better to find a compromise - or rather, the most appropriate choice - for the application. There are many low gate charge devices available now with high voltage, current and power ratings. If your app requires significant power to be handled by the mosfet, then you should go to case styles larger than TO-220.
Remember that even as a buffer operating at high voltage, say because you wanted direct coupling between the tube gain stage and the mosfet, there is waste heat to consider. The idle current through the mosfet will heat it. A TO-220 in free air has a thermal resistance of 65C/W and it will be very easy to have the mofet running very hot - too hot without some heat sinking. The larger cases are 45C/W so get less hot for a given idle current.
Have fun
Kevin O'Connor
Hi all
Struth, I did write that the Current amp was made with Mosfets, no valves.
A triode valve is used as a V amp, and as I wrote the Mosfet I am looking for is to be used as a buffer. No gain. It is a 'Mosfet Mu Follower' design, which feeds a home designed mosfet amp buffer.
I don't want to start a discussion about the whole design, it works wonderfully and shows sqaure waves up to better than 100 khz with no distortion.
I just want to upgrade my IRF buffer stage between V and I amps with a better To.
The current which flows through the valve is 20 mA and the the Mosfet buffer does not need more than one amp current potential.
I did try some low power high volt Mosfet but the square signal gets worse so until now the IRF 740 is the best I ve tried.
Struth, I did write that the Current amp was made with Mosfets, no valves.
A triode valve is used as a V amp, and as I wrote the Mosfet I am looking for is to be used as a buffer. No gain. It is a 'Mosfet Mu Follower' design, which feeds a home designed mosfet amp buffer.
I don't want to start a discussion about the whole design, it works wonderfully and shows sqaure waves up to better than 100 khz with no distortion.
I just want to upgrade my IRF buffer stage between V and I amps with a better To.
The current which flows through the valve is 20 mA and the the Mosfet buffer does not need more than one amp current potential.
I did try some low power high volt Mosfet but the square signal gets worse so until now the IRF 740 is the best I ve tried.
Attachments
Sorry,
Didn't fully understand what you wanted; I have aphasia these days. I will keep out of this future.
Hugh
Didn't fully understand what you wanted; I have aphasia these days. I will keep out of this future.
Hugh
Hi Guys
Your mosfet is only conducting the current of the tube, which you say is 10mA. If the mosfet drops 100V then it dissipates 1W and might be a bit hot to the touch.
The IRF710 has pretty low gate capacitance as TO-220s go. It's gate charge is also low as a result.
You say the square wave performance is good and other aspects are good. You won't really "improve" the sound by changing this element in this circuit.
What do you deem an improvement to be sonically? Mu-followrs, SRPPs et al are not known for their warm tube tone - quite the contrary actually. The purpose of such designs is to extract very linear low-THD performance from the tubes, and this may not be exactly what you want, or at least it may not be what turns you on musically.
Larger mosfets will have higher gate capacitance and may slow the response. You can mimic that aspect simply by tacking in a cap across G and S. The same larger mosfet will have higher transconductance which might reduce output impedance and/or improve linearity.
Have fun
Kevin O'Connor
Your mosfet is only conducting the current of the tube, which you say is 10mA. If the mosfet drops 100V then it dissipates 1W and might be a bit hot to the touch.
The IRF710 has pretty low gate capacitance as TO-220s go. It's gate charge is also low as a result.
You say the square wave performance is good and other aspects are good. You won't really "improve" the sound by changing this element in this circuit.
What do you deem an improvement to be sonically? Mu-followrs, SRPPs et al are not known for their warm tube tone - quite the contrary actually. The purpose of such designs is to extract very linear low-THD performance from the tubes, and this may not be exactly what you want, or at least it may not be what turns you on musically.
Larger mosfets will have higher gate capacitance and may slow the response. You can mimic that aspect simply by tacking in a cap across G and S. The same larger mosfet will have higher transconductance which might reduce output impedance and/or improve linearity.
Have fun
Kevin O'Connor
Hi Struth
In fact you are just right ...
I was just imagining the IRF was not the best in class, there sometimes is a hearing difference between two devices that have the right and almost same capacitance parameters...
No, I don't want to modify the sound, I have been used to valve sound and have worked quite a bit with them, even making a very special double power triode amp in A class.
You are also right the IRF heats a little bit, I have put it on a heatsink.
If some here are interested I could find a way to post the schematics of that hybrid.
I had to design mine because I had made 4 of the Pass A Class Zen amps and found them rather disappointing and complex. Lots of heating for no reason. In my opinion rather poor result after lots of trouble.
I went the simplest way: the best for V amplification: one only valve in a Mu follower configuration. in goes about 2 V, out comes 40 Volts easily. With this schematic you can adjust separately all the valve parameters and that is a very good thing. It makes it just easy to have the best results.
Then, a buffer (to help the V amp output impedance to match the Mosfet grid one) on the Current amp which uses the double N and P Mosfets an the same chip. Found them cheap and avaiable at Farnells. They are perfectly tuned and need no extra gimmicks. Specially designed for Hi FI stuff.
The sound is really great.
In fact you are just right ...
I was just imagining the IRF was not the best in class, there sometimes is a hearing difference between two devices that have the right and almost same capacitance parameters...
No, I don't want to modify the sound, I have been used to valve sound and have worked quite a bit with them, even making a very special double power triode amp in A class.
You are also right the IRF heats a little bit, I have put it on a heatsink.
If some here are interested I could find a way to post the schematics of that hybrid.
I had to design mine because I had made 4 of the Pass A Class Zen amps and found them rather disappointing and complex. Lots of heating for no reason. In my opinion rather poor result after lots of trouble.
I went the simplest way: the best for V amplification: one only valve in a Mu follower configuration. in goes about 2 V, out comes 40 Volts easily. With this schematic you can adjust separately all the valve parameters and that is a very good thing. It makes it just easy to have the best results.
Then, a buffer (to help the V amp output impedance to match the Mosfet grid one) on the Current amp which uses the double N and P Mosfets an the same chip. Found them cheap and avaiable at Farnells. They are perfectly tuned and need no extra gimmicks. Specially designed for Hi FI stuff.
The sound is really great.
Hi Guys
Billy, it sounds like you have the performance you need. Time to go to another project.
Have fun
Kevin O'Connor
Billy, it sounds like you have the performance you need. Time to go to another project.
Have fun
Kevin O'Connor
- Status
- Not open for further replies.