Astute observation, making the part very interesting
Thanks! Based on my first very crude tests, I really like it. It is a very nice behaving device.
I have been testing this and some other similar devices today, and this only increased my appreciation for these devices, and especially the IRFPS3810.
With this simple setup I could trace the current based on the Vds for a fixed Vgs.
I could also see the large differences in each device sensitivity to thermal runaway.
The IRF3710 was almost impossible to get any sensible reading from, since it was very thermally unstable.
Every device with a poor maximum safe operating area (SOA) had a very fast and violent reaction to heat. This is what I expected.
The IRFP150 and the IRFPS3810 was very thermally stable and both has a large published SOA.
The old "classical" mosfet IRFP150 did not show any marked growth of current as a function of the Vds - as expected.
The IRLB3813 had a very sharp increase in current with increasing Vds over 8 volts.
The IRLP3034 was extreme in almost every way. It reacted violently to both incresing voltage and heat. Once over 5 volts and 1 ampere, the current increased very fast with both heat and Vds. It seems to have very steep curves, almost vertical over 15 volts and 3 ampere. It was very difficult to differentiate between the inherent electrical "triody" character and its disposition for thermal runaway.
The IRFPS3810 was the clear winner in almost every aspect in this simple test. Very little reaction to heat. Very smooth and nice growth in current based on the voltage between drain and source. Both with an external source-resistor and without one.
Real world hands on experience with these devices is a great and fun learning process.
I want to encourage everyone to buy a few IRFPS3810 and test in different single ended designs. They seem to have a great potential.
Regards,
Johannes
All you have to do now is drive that high capacitance.
I would prefer to work with your SITS, so if you have a bag of them laying around that you don't need I will gladly except them... Until that day I will have to do with more easily attainable devices, even though they have much more gate capacitance.
I don't see that gate capacitance as much of a problem. A good low output impedance preamp or a p-channel Mosfet as a source-follower driving the gate should be enough to reach the AM-band, even if I parallel many of them for more power and more subjective bass-"oohmpf!".
When listening to my first rats-nest test amp I drive the IRFPS3810 from the headphone output of my Samsung phone via a 100 ohm gate-resistor. I love the character of the IRFPS3810, and I can not hear any problems associated with the gate capacitances that I would not first attribute to the crappy headphone output.
Regards,
Johannes
IRFPS3810 Common Source Amp
Quick test at lunch break, thanks to my mighty junky scrap pcb platform.
R drain = 20 Ohm
R Source = 1 Ohm
R load = 8 Ohm via 20,000 uF
Vgs = 6.36 V
V supply = 30 V
I ds = 1 A
THD+N = 2.9% @ 1 Watt
Need your input to get to a verdict for this.
Quick test at lunch break, thanks to my mighty junky scrap pcb platform.
R drain = 20 Ohm
R Source = 1 Ohm
R load = 8 Ohm via 20,000 uF
Vgs = 6.36 V
V supply = 30 V
I ds = 1 A
THD+N = 2.9% @ 1 Watt
Need your input to get to a verdict for this.
Attachments
IRFPS3810 common drain amp
Common drain circuit with 62mH iron choke source loaded.
MoFo inductor options for use above 1kHz only
Very close results to the IRFP240
V supply = 30 V
I ds = 3 A
THD+N = 0.04
Your verdict?
Common drain circuit with 62mH iron choke source loaded.
MoFo inductor options for use above 1kHz only
Very close results to the IRFP240
V supply = 30 V
I ds = 3 A
THD+N = 0.04
Your verdict?
Attachments
Last edited:
common drain is not to my mind a very interesting application for a triode since in that config, it has full feedback already anyway.
Common source is probably where the fun is but a (my guess) roughly ~200R drain impedance means that you may be better off putting a transformer load on the drain and see what you get. See Mike Rotarchers experiments with the Vfets for some guidance.
Some other things you can try though while its set up as is on thee bench - try bypassing the source resistor with a big electrolytic cap and remeasure the distortion at the same op level.
Also what is the gain of the common source circuit you show ?
Common source is probably where the fun is but a (my guess) roughly ~200R drain impedance means that you may be better off putting a transformer load on the drain and see what you get. See Mike Rotarchers experiments with the Vfets for some guidance.
Some other things you can try though while its set up as is on thee bench - try bypassing the source resistor with a big electrolytic cap and remeasure the distortion at the same op level.
Also what is the gain of the common source circuit you show ?
The IRFPS3810 needs 6,8 volts between gate and source for 2 ampere current at 16 volts between drain and source?!?!
It is one strange device.....
Its Vgs of +6.8V may put it within "spitting range" [per Pass] of [-Vgs] for the available 2SJ28; to make a DEFiSIT.
Your verdict?
Thanks for testing these devices and reporting your findings here.
These "triody" devices might not always measure better then normal devices. Its all about finding the sweetspot in their transfer curves. 1,0 ampere is probably much too lean. Test with a lower voltage and more current.
I think you will find better measured performance with the IRFPS3810 in common source applications.
I tested the gate to source voltage rating of the IRFPS3810, and it did not like a few kilovolts static discharge.... I have to change the device in my test setup, and I am a bit busy at the moment.
Cheers,
Johannes
R drain = 20 Ohm
R Source = 1 Ohm
R load = 8 Ohm via 20,000 uF
Vgs = 6.36 V
V supply = 30 V
I ds = 1 A
THD+N = 2.9% @ 1 Watt
Need your input to get to a verdict for this.
Maybe you should play with the values a bit - there is only 9v dc on the
part.
I/V curves
@ Mr Pass, Mr Rothacher, Johannes, Kasey197
Great, many thanks to all of you.
The SPI file from IRFPS9810 gives me lot of trouble related to IV simulation, cannot display any movement to the Vds and Ids input. I use a dirty trick by termination both Drain and Source with 1 Ohm resistor, i think it is an unacceptable result
I simulated it in comparison with an SIT THF-51S's SPI (modified X parameter from Mr Rothacher's 2SK180).
Listen to your advices, I simulated them in a CS configuration with Inductor loaded to Drain output.
Cheers and gratitude from Pascal
@ Mr Pass, Mr Rothacher, Johannes, Kasey197
Great, many thanks to all of you.
The SPI file from IRFPS9810 gives me lot of trouble related to IV simulation, cannot display any movement to the Vds and Ids input. I use a dirty trick by termination both Drain and Source with 1 Ohm resistor, i think it is an unacceptable result
I simulated it in comparison with an SIT THF-51S's SPI (modified X parameter from Mr Rothacher's 2SK180).
Listen to your advices, I simulated them in a CS configuration with Inductor loaded to Drain output.
Cheers and gratitude from Pascal
Attachments
Last edited:
Final test in real life with irfps3810, irfp240 and thf51s in CS amp.
The IRFS3810 CSSE amp
Thanks for posting the mesurements MossFed.
Low numbers THD+N of SIT impress me in a positive way.
Think I need to experiment with my Sit-ish SIT's one of this days.
Interesting
Attachments
Thanks for posting the mesurements MossFed.
Low numbers THD+N of SIT impress me in a positive way.
Think I need to experiment with my Sit-ish SIT's one of this days.
Interesting
Wow, expensive presse papier!
Last Nail
Well, I managed to destroy one intentionally for getting IRFPS3810 DC SOA, max 50V/3A on my heatsink.
After tracking it in best THD+N within DC SOA (Max Ids = 2A)
What I found :
- Audio performance is lightly inferior to the IRFP240.
- THD+N value is out of SIT’s class.
Would appreciated your reflection on those Sitish mosfet.
Regards from Pascal
Well, I managed to destroy one intentionally for getting IRFPS3810 DC SOA, max 50V/3A on my heatsink.
After tracking it in best THD+N within DC SOA (Max Ids = 2A)
What I found :
- Audio performance is lightly inferior to the IRFP240.
- THD+N value is out of SIT’s class.
Would appreciated your reflection on those Sitish mosfet.
Regards from Pascal
Attachments
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
- Home
- Amplifiers
- Pass Labs
- A very nice looking SIT-ish Mosfet