fet boy

[pstcho]

thanks a lot for your replies

i replaced the 100r with 1W already ( nice haeter for the room hahahah )

ok ill order different FETs and make tests to see if it improves something, they are on sockets, so it will be easy

any advice on what other FEts to try ? now i have 2N5457 on both PCBs, i tried mahy different ones but nothing changed about the heat

also do you have any advice on what voltages to target around the transistors ?

 

[Admiral-dk]

The fact that R9 & R16 gets HOT shows that there is a serious flaw in the original Schematic + that also indicates a reduced Headroom on the output (not that you need 20Vpp.) .... You shouldn't measure more than aprox. 1V over them ...!!!

 

[RuudNL]

Well, the original schematic says: +24 V DC @ 40 mA.
The first FET stage is hardly drawing any current. (Rd= 15 K.ohm)
So most of the current is flowing throught the output transistors.
This means a dissipation of 12 x 0.04 = 480 mW in the 'top' NPN transistor.
40 mA though the 100 ohm resistor would mean a voltage of 4 Volt over it.
Dissipation in the 100 ohm resistor would in this case be 4 x 0.04 = 160 mW.

 

[Admiral-dk]

I have shown some exampels here - only of the Sink part of the PreAmp (the bottom NPN's).

The first is by far the worst - having the Trimmer connected directly between the +24V and the Base-Emitter to GND.
Turn the Trimmer to a low resistance and you have blown BOTH the Trimmer and the bottom NPN + very likely a lot of other components !!!

The next is a working solution, #3 has the Base noise decoupled. #4 & 5 are proper Constant Current Sources (Sinks).
And the bottom two examples are probably how you would make two Good Sinks cheap and reliable.

I should mention that these are calculated for the 40mA mentioned by RuudNL - they can pull a Headphone directly ...!
Personally I would aim for a lower Current -> Improved reliability.

Per

Edit : I have realized that I forgot to mention / show - replace the bottom NPN with a N-JFET - Source replace Emitter, Gate to GND and Drain replace Collector - the 100R will work - but not neccesary be to Optimal Value.

 

[RuudNL]

Shouldn't the 1.82K source resistor be adjustable, in order to bias the FET in such a way that it has about 12.6 V at the drain?

 

[Admiral-dk]

I must admit that I only commented on the really bad part of the design - but I will agree that either an adjustment of the current in the FET or a DC Feedback Loop to lock it to reasonable Drain voltage vould be a good thing.

 

[pstcho]

ok thats a lot of good info, ill take a loot during my next day off and check if i can make something out of those two boards

youre awesome guys, thanks a lot

 

[sketch-n-etch]

Hi all,

I'm trying to incorporate Admiral-dk's suggestion for a constant current source for this circuit (per the attachment provided), but I'm not sure I've done it correctly. Can someone who has a better sense weigh in on what I've got?

Also, I'm confused by the additional comment:

Edit : I have realized that I forgot to mention / show - replace the bottom NPN with a N-JFET - Source replace Emitter, Gate to GND and Drain replace Collector - the 100R will work - but not neccesary be to Optimal Value.

 

[Admiral-dk]

Your Circuit will work as is !!
Collector Current through Q3 ~ 0.6-0.7Vdc / 100R (R5) = 6-7mA > ~ 3.6Vp in 600R.

My comment works as follows :
Remove Q4 & R4. Replace Q3 with another 2N5457 (since you're using one now) - connect the Source to R5, Drain to Q2's Emitter, and Gate to GND.
If you keep R5 as 100R - you will have another Constant Current Source ~ 8mA. (give or take) ....
This is the simplest Current Source - and can even be bought as a two pin device (strangely enough sometimes called a Constant-Current Diode).

Hope this clarifies things.

Per

 

[sketch-n-etch]

Thanks so much for the reply / confirmation on that circuit.

So regarding the comment, that's just an alternative for a different constant current source?

 

[Admiral-dk]

So regarding the comment, that's just an alternative for a different constant current source?

Yes ! - one that might be cheaper in Production (fewer parts) ....

Per