diyAudio

diyAudio (http://www.diyaudio.com/forums/)
-   Solid State (http://www.diyaudio.com/forums/solid-state/)
-   -   help!ab class mosfet amp (http://www.diyaudio.com/forums/solid-state/61521-help-ab-class-mosfet-amp.html)

mbonus 27th July 2005 06:26 PM

help!ab class mosfet amp
 
2 Attachment(s)
hi!
I decided to build one power amp (with 2sk135/2sj50)
which works in AB class with power of 60w in 8ohms.
In croatia this mosfets are pretty expensive so I'm
thinking about replacing this mosfets with irfp240/irfp9240.
I heard that these hexfets need Vbe multiplier.
I'm beginner in electronics so I'll be grateful if
somebody can tell me can i just replace this mosfets or
help me what i need to change in my scheme so i can put
irf devices.
here is the scheme:

Craig405 27th July 2005 06:44 PM

I am not sure but 2sk1058/2sj62 might do it, someone said they had used these in an amplifier that used the mosfets you mentioned and they were ok.
Check the data sheets first though.

Craig

tlf9999 27th July 2005 06:47 PM

It is OK to use d-mosfets but with a Vbe or Vgs multiplier. It tends to be overly compensated with a Vbe multiplier and I usually use a small TO220 MOSFET as Vgs multiplier. You just need to be careful with the Vgs multiplier to make sure that you don't have a thermal runaway.

Steven 27th July 2005 07:36 PM

I think there is a fault in the circuit. I expect a connection from R14 somewhere to D7.

Steven

mbonus 27th July 2005 07:56 PM

2 Attachment(s)
you are right Steven.
my fault...
i corrected it now.

AndrewT 27th July 2005 08:03 PM

Hi,
What does D3 do? it seems it could be omitted.
V FETs must use a thermal compensator.
Drive to Q8 missing.
R1 is too big, try 680r and change C2 to 1nF to maintain rc time constant.
C7 & C8 could use a bigger decoupling cap in parallel, try 100uF+47nF.

amplifierguru 28th July 2005 12:49 AM

Hi Andrew,

D3 is needed to increase the drop from the 40V supply to bias the next stage AND to provide Tcomp of that following stage. Otherwise Vacross the 1K5 resistors is only 0.63V, not enough!
D3 raises this to about 1.25V enough with one Vbe Tcomp to provide a stable 0.6V for the 47R tail resistor of the diff'l Vas so defining the operating current as 6.4mA per leg.

Hi mbonus,

To answer the thread question, you will need a higher bias, variable up to 8V for the IRFP240/9240 pair and include source resistors of maybe 0R47. These MOSFETs have a +ve tempco of around 3 times that of a BJT but off a much higher Vgs so % tempco is actually about half the compensation needed to track for a BJT. That's why a std Vbe multiplier overcompensates.

You can provide this in a few ways -
1. A standard Vbe multiplier with degrading R in the emitter leg.
2. A Vgs multiplier.
3. Standard Vbe multiplier for half plus variable resistor for other half bias.
4. string of diodes in thermal contact for 1/2 + variable resistor for other half bias chain.

Hope that helps

Cheers,
Greg

mastertech 28th July 2005 01:25 AM

AndrewT

"Hi,
What does D3 do? it seems it could be omitted."

yes it could ,then youd need to adjust the colector resistors


cheers

mastertech 28th July 2005 01:27 AM

by the way,very nice design
job well done

mbonus 28th July 2005 06:23 AM

thanks andrew!
it will be much more easier to me
if somebody can draw in my scheme where
and how to connect this Vgs multiplier :)
so i can put irf-s.
(i tried to find 2sk1058/2sj62 but 2sj62
is also hard to find for me.)
thx!


All times are GMT. The time now is 11:54 AM.


vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2014 DragonByte Technologies Ltd.
Copyright 1999-2014 diyAudio


Content Relevant URLs by vBSEO 3.3.2