F5 with 2SK2013/2SJ313

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Juma,
Can't figure it out. With source grounded, Vgs is at 0 and Jfet is fully on, passing full Idss current. This current drops a voltage across Rs which causes the Vgs to change. With this change the current drops somewhat due to the change until we reach equilibrium. With source tied to the negative rail, the Jfet should still be full on, but what detmines Vdrop across Rs. Can't be the same, because with Jfets with Idss of 10mA, we would be dropping 3.3V. That would turn the Jfet off.
 
Juma,
....With source tied to the negative rail, the Jfet should still be full on, but what detmines Vdrop across Rs. Can't be the same, because with Jfets with Idss of 10mA, we would be dropping 3.3V. That would turn the Jfet off.
All wrong. I can't find the way to explain it more clearly than I already did - please build one channel and measure it, then read my previous explanations once more and it will dawn to you eventually... ;)
 
Hi everyone, just finished my headamp(preamp), the one from post 42. Very happy with the sound, and a BIG improvement over my Asus soundcard. Bias at 400mA and heatsink around 45 degree celsius. Powered with two transformers, CRCRC and 7815/7915 regulators. I guess i need new headphones though.
Thank you Juma!
 

Attachments

  • IMAG0354.jpg
    IMAG0354.jpg
    840.6 KB · Views: 1,056
  • IMAG0355.jpg
    IMAG0355.jpg
    940.3 KB · Views: 1,000
  • IMAG0362.jpg
    IMAG0362.jpg
    965.7 KB · Views: 954
Cathode/Source/Emitter degeneration is often considered as a form of negative feedback and yes, it's used to control the gain and the working regime of the active device.
There's a lot of stuff written about it here in forum as well as in the general electronics literature - search button and google will provide you with abundant info about it...


Nice job holminator, glad you like it :up:
 
regulators

Hi everyone, just finished my headamp(preamp), the one from post 42. Very happy with the sound, and a BIG improvement over my Asus soundcard. Bias at 400mA and heatsink around 45 degree celsius. Powered with two transformers, CRCRC and 7815/7915 regulators. I guess i need new headphones though.
Thank you Juma!

Congratulations holminator.Very nice job.I think using a better regulator instead of 7815/7915 would be better.For example lm317/337 types.Thanks
 
Hi crt,
it's a fast amp and your pics show slight but well controlled overshoot - nothing to worry about. If you want to eliminate that overshoot some answers are needed so we can get the whole picture:
1. The scope picture of the input signal at 10kHz and 100kHz recorded under same conditions as output signal pics which you attached
2. what's the value of gate resistors at output MOSFETs' gates ?
3. what's the value of resistors in the feedback loop ?
4. do you have bypass caps on amp PCBs ?
5. details about amp's power supply ?
6. pictures of your amp's insides - wiring etc.
 
Hi crt,
it's a fast amp and your pics show slight but well controlled overshoot - nothing to worry about.

Well, if you say that :cheers:

If you want to eliminate that overshoot some answers are needed so we can get the whole picture:
1. The scope picture of the input signal at 10kHz and 100kHz recorded under same conditions as output signal pics which you attached
2. what's the value of gate resistors at output MOSFETs' gates ?
3. what's the value of resistors in the feedback loop ?
4. do you have bypass caps on amp PCBs ?
5. details about amp's power supply ?
6. pictures of your amp's insides - wiring etc.

1. Yes. 1V square wave
2. Bit bigger than schema. i use 243R
3. 150R / 3 in parallel with 10R
4. None
5 & 6 Already in my posting at page 20
 
Crt
I found pics in posts #192 & 193. It looks OK. You used a lot of small caps in PSU - it gets you rid of one sort of problem but it puts you in another sort of predicament. Never mind, it would not be practical to deal with it now.

I asked you for a picture of input signal. I don't know if your input signal has some overshoot. Check it, and if it does, it's OK. If it doesn't, there are three easy things to do (pick what's easier for you):
1. Disconnect one of three 150R power resistors in every feedback loop so that in your stereo amp instead of 4 x (3 x 150R) you get 4 x (2 x 150R). That will raise the gain and slightly change the frequency and phase response of the amp. If you choose to do this you'll have to check the bias/DC offset again.
2. Connect one 220pF cap in parallel to every 150R triplet (you'll need four 220pF caps). Those caps can be bigger (330 or even 470 pf) since the feedback loop has very small impedance. They can also be smaller - you'll have to experiment to find the lowest value that will eliminate the overshoot. This change affects the frequency and phase response of the amp, too.
3. Increase the value of gate resistors on the output MOSFETs but that's the less effective method.

Now you have something to play with, you can try all three methods and decide what you like the best - all of them will influence the sound, some for the better, some for the worse - it depends on your build's characteristics.

Have fun :D
 
5&6 - you need to write exact # of post

pages are browser dependent

:eek:


Crt
I found pics in posts #192 & 193. It looks OK. You used a lot of small caps in PSU - it gets you rid of one sort of problem but it puts you in another sort of predicament. Never mind, it would not be practical to deal with it now.

I asked you for a picture of input signal. I don't know if your input signal has some overshoot. Check it, and if it does, it's OK. If it doesn't, there are three easy things to do (pick what's easier for you):
1. Disconnect one of three 150R power resistors in every feedback loop so that in your stereo amp instead of 4 x (3 x 150R) you get 4 x (2 x 150R). That will raise the gain and slightly change the frequency and phase response of the amp. If you choose to do this you'll have to check the bias/DC offset again.
2. Connect one 220pF cap in parallel to every 150R triplet (you'll need four 220pF caps). Those caps can be bigger (330 or even 470 pf) since the feedback loop has very small impedance. They can also be smaller - you'll have to experiment to find the lowest value that will eliminate the overshoot. This change affects the frequency and phase response of the amp, too.
3. Increase the value of gate resistors on the output MOSFETs but that's the less effective method.

Now you have something to play with, you can try all three methods and decide what you like the best - all of them will influence the sound, some for the better, some for the worse - it depends on your build's characteristics.

Have fun :D

Ok will do experiment tomorrow :)
Which one you prefer to do first ? number 1 or 2

here output from signal generator connect directly to scope
 

Attachments

  • scope 10k.png
    scope 10k.png
    5.7 KB · Views: 489
  • scope 100k.png
    scope 100k.png
    5.9 KB · Views: 484
Last edited:
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.