I am planning to build the MOSFET amp attached, but I would like to replace the VAS transistors IRF610/9610 with better types, preferably MOSFETs. I have tried to find low cap, high wattage types without any luck. I also intend to run this amp without global feedback. Any suggestions?
Attachments
I forget where this circuit is from, but are you sure it is good? It looks like the designer wanted a symmetrical circuit while the mosfet is less complementary than bipolar?
And what do you mean with "without any luck"?. How about 2SK213-216/2SJ76-79? These are lateral mosfets not vertical like the IRF.
I have found in many amps using the IRF610/9610 that the mosfet is the limiting factor. Luckily the MOSFET used in the JLH 80W is a lateral one (tho the IRF also fits there)...
About IRF510 (as per Lineup suggestion), I'm aware that there are many manufacturers but from my experiment/observation/search, the IRF610 is the best (soundwise), not IRF510. For higher power IRF530 is my favorite.
Thank you for your replies. I forgot to say I have a 2x18V 500 VA transformer and two quite large heatsinks, so I am planning to run it in Class A, about 25 W into 8 ohms. This makes it possible to not have feedback from the output, but from the VAS instead. the design is my own, the symmetrical input is maybe not necessary...
The circuit topology is imo "tailored" towards class-B. For class-A, we already "have" Pass Labs' F4 which is also about 25W into 8 ohm, using similar components as yours. Of course, it can be more fun to have own design regardless of quality. But isn't it better to follow the proven track? What I mean is, if you ask for expert's help here, they may suggest you to change your amp into something that is completely different.
If THD is not the main concern, (about 0.02%), a symetrical differential + vas
using a pair of low power bjts to drive a few lateral fets pairs is the way to go.
Since the psu has relatively low voltage, you can increase the
quiescent current to something like 120mA/device and get
class A at higher level than with bipolar outputs using the same
current, thanks to the lfets caracteristics..
Indeed, it will blow out any F4 or other Fsomething like amps...
using a pair of low power bjts to drive a few lateral fets pairs is the way to go.
Since the psu has relatively low voltage, you can increase the
quiescent current to something like 120mA/device and get
class A at higher level than with bipolar outputs using the same
current, thanks to the lfets caracteristics..
Indeed, it will blow out any F4 or other Fsomething like amps...
The problem is 2SK213-216/2SJ76-79 are hard to find.
Dgikey has 2SK214 and 2SJ77
FETs - Single | Digi-Key
Dgikey has 2SK214 and 2SJ77
FETs - Single | Digi-Key
Last edited:
So maybe 2sk214/2SJ77 as vas transistors...
and 3 pairs of 2SJ162/2SK1058 as power outputs will do it....
I like your spirit!!!
20W is not too useful for me. There are so many good amps within this power rating that I cannot decide which one is better than the other.
But if it is subjectively better than F4 and is using laterals, I'm looking forward to building it
Federmann, the noise density vs frequency of your amp show
an overshoot at half a megahertz, which tells that your amp
has a gain overshoot at the said frequency.
The low pass input filter cancel this resonnance seen from
the input of the amp, but this will not change tha fact that
internaly, there s a large phase dip at this frequency, leading
to mediocre stability..
an overshoot at half a megahertz, which tells that your amp
has a gain overshoot at the said frequency.
The low pass input filter cancel this resonnance seen from
the input of the amp, but this will not change tha fact that
internaly, there s a large phase dip at this frequency, leading
to mediocre stability..
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.