Simplistic NJFET RIAA

[Salas]

Originally posted by guglielmope
but I where I can find info about matching fets or buy matched ones?
Thanks a lot for your help and interest
Grazie
Guglielmo

Idss matching

[RCruz]

Originally posted by salas

The standard 28V LED shunt that Rcruz made, he reported it dead silent. That is all I can personally know.

The shunt is really silent and ultra fast.

Just did the 330pF mod and now everything is getting into the right place.

Now I will play with the cart loading.

Amazing !

Ricardo

[guglielmope]

Hello Riccardo,
There isn' t any330p on the schematic, where I can find this update?
Do you had any problem with pcb?they was ok?
Thanks a lot
Guglielmo

[iko]

The only difference I noticed between the two versions of the regulator is that the power mosfets of the non-buffered one get very hot, even with a significant heat sink. The buffered version you mentioned above has cold mosfets; the darlington gets a little warm (I measured a 27 mA collector-emitor current). On the oscilloscope both show a nice regulated trace, and sound wise I could not notice any difference. The reason I mentioned it to salas was partly theoretical, furthermore, I like tweaking everything, no offense intended.

[Salas]

Rather odd to have cold Mosfets if the CCS is running same current in both versions. I would expect the shunt Mosfet to be somewhat colder since the driver and the idss buffer would consume about 37mA above load consumption VS the standard one, lightening its constant burden. But I don't see how the CCS Mosfet running at same current and same Vin - Vout can show such a thermal difference. Plus if they share the same heatshink they would eventually come to about the same centigrade.

[RCruz]

Originally posted by guglielmope
Hello Riccardo,
There isn' t any330p on the schematic, where I can find this update?
Do you had any problem with pcb?they was ok?

Hi Guglielmo

The original 46dB schematic indicates 16nF for C3 on the riaa circuit.

As I was using 15.72nF, I added 330pF so I got 16.05nF.

The result is very noticeable... I lost the frontal presentation of sax and trumpet, voices became less boxy and piano has a natural presentation. the 330p silver micas are burning nicely and the trebble is regaining it´s fluidity.

The pcb is very good indeed... everything worked at the first time !

Now I will replace some resistors and work on a propper case with good connectors.

Ricardo

[nicoch58]

Ricardo

[RCruz]

Originally posted by nicoch46
Ricardo

Thank you nicoch46

[iko]

Originally posted by salas
Rather odd to have cold Mosfets if the CCS is running same current in both versions. I would expect the shunt Mosfet to be somewhat colder since the driver and the idss buffer would consume about 37mA above load consumption VS the standard one, lightening its constant burden. But I don't see how the CCS Mosfet running at same current and same Vin - Vout can show such a thermal difference. Plus if they share the same heatshink they would eventually come to about the same centigrade.

Yes, it puzzle me too. I did some digging yesterday; my suspicion was that the bjt did all the work in the second stage of regulation, the shunt stage, and thus the second mosfet did no more work. I measured the current through the first mosfet, at about 97 mA, and that seems sane. The current through the second mosfet was less than 1 mA. Logically, next I removed the entire buffer+second mosfet. Right now I am running the attached configuration with great results I might say. The darlington can be replaced with other medium power transistor, I use it because it's what I had around in the junk box. I can say this with certainty, because I did compare yesterday with a lm317, and a lm338t. The shunt regulator is much better, both versions yours, salas, and the one that I slightly modified.

salas, so have you formed an opinion about the ruskie ft-3s?

[iko]

However, I would like say that I am just exploring some of these variations, and to recommend to everyone to run the salas shunt regulator.