Hybrid ZEN Amplifier + LM317 = Efficient and simple

The LM317 could offer some advantage over the more well-known modulated current source option implemented with a mosfet and a small bipolar transistor. I believe that the most obvious advantage could be in a better thermal control of the static polarization parameters, which could even allow a direct coupling of the output, if other conditions are ensured simultaneously.

Diego
 
I never looked into what output decoupling cap to use.

But for 8 ohm load, a 10000uF will give you -3dB at 2Hz, so minimum really.
Mouser still has some Nichicon LS, though already End of Life, like all audio components.
LKS1J103MESC

And then maybe a 100uF film cap in parallel for 200Hz upwards ?
Cheapest I can find is at DK :
B32778Z4107K000

Them together will set you back ~25€ per channel.
Probably more than all other components combined, if you do your own matching.
Not counting power supply, of course.


Patrick
 
Last edited:
This neat little amp looks like it would also work well with the IRFP048, a close descendant of the IRF044 as used in the original Zen amps.
I'm using the IRFP048 to test some new PCBs for the F6. It has characteristics close to my preferred, but nearly unobtanium FQH44N10.
@TungstenAudio
Are the FQH44N10 worth buying instead of the IRFP048 ? If so, maybe we could do a GB.
Rochester has 9000 in stock but the minimum purchase is 187 mosfet…all of these for $250. I also found an American seller on eBay in case they are worth buying.
Disclaimer: Never heard of Rochester before.
BTW, the DIYaudio store currently sells Matched IRFP048 (look for F-6 Transistor kit)

Eric
 
Picture of the ones from eBay, is the font similar to a real one ? Thanks

IMG_0746.png
 
All these high transconductance MOSFETs, including the IRFP044 mentioned before, have high Ciss.
That means to drive them in common source to a decent bandwidth requires low value resistors in the feedback network.
See Diego's post #45.

Unless of course you cascode, like in the Zen Variation 9.
Then you can go back to more common power-amp Zin of ~10k.

Penalty ? Needs higher voltage.
No free lunch.


Patrick
 
Last edited:
  • Like
Reactions: 1 user
The FQH44N10 has characteristics that work very well in several applications. I have used it where the amp is running on higher voltage rails, and at higher bias current. This part has been found to sound better in the ACA vs the IRFP240 or IRFP140 by several builders. It slays in my F6, giving it a sonic character nearly identical to my re-capped Threshold CAS-2.
Of the higher transconductance Mosfets, its gate capacitance, Ciss, and total gate charge are remarkably low, better than the IRFP140. I adjust the amp circuit by using somewhat smaller gate stopper resistors.

I bought a tube of 30pcs from the ebay seller with that picture. They appear to be genuine NOS from Fairchild. My previous parts were from OnSemi, so I can’t directly compare the marking, but the F logo looks correct. I will be using the Fairchild parts in my new F6 Diamond PCBs, and will have a direct sonic comparison then.
 
  • Like
Reactions: 1 user