diyAudio

diyAudio (http://www.diyaudio.com/forums/)
-   Solid State (http://www.diyaudio.com/forums/solid-state/)
-   -   Thermal Stability of Lateral FET Driven Bipolar Follower. (http://www.diyaudio.com/forums/solid-state/228483-thermal-stability-lateral-fet-driven-bipolar-follower.html)

Andrew Eckhardt 23rd January 2013 03:59 PM

Thermal Stability of Lateral FET Driven Bipolar Follower.
 
1 Attachment(s)
Does anyone have an off-hand idea how far off from stable this circuit would be? The actual application will probably use more like ten pairs of bipolar outputs for a 2.66 ohm nominal load with 65 volt rails. With only 15mA emitter current (basically class B), I wonder if it would run away if it got hot..

Mooly 23rd January 2013 04:23 PM

My instincts say not very stable because you would need around what ? 2 volts or less across the 47 ohm which is 40 milliamps or less. The FET's wouldn't hold that steady enough to be used for biasing... I don't think so anyway. The bias would be prone to wander quite a bit.

No proof, just a feeling :)

Bensen 23rd January 2013 05:12 PM

1 Attachment(s)
I have the same opinion as Mooly.
I would change the R100 to a vbe multiplier.

See for example the Classé amps, these are proven designs.

Greetz

Andrew Eckhardt 23rd January 2013 05:17 PM

Great help.Thanks guys..

east electronics 23rd January 2013 06:49 PM

you may also wana take a look at Zeck amplifier schematics ...Long time ago but they have been doing something similar i think

AKSA 23rd January 2013 11:15 PM

The bias voltage for maybe 100mA quiescent might find there is almost 2.4 volts bias between the lat fets gates. You have the lat fets biased above 27mA, but this is a bit low and this would need changing the 47R to 12R to deliver around 100mA. At this higher lat fet bias, you are around zero tempco, but negative tempco at 400mA and not very efficient. The compensation would be quite severe at the output quiescent, and they would probably might into thermal overload if you only use a resistor to set the bias voltage.

Yes, a Vbe multiplier is a good idea, using a single bipolar like a BD139. Collector to base is 1K5 and base to emitter around 330+200R trimpot in series for adjustment.

Hugh

Cheers,

Hugh

Andrew Eckhardt 24th January 2013 02:03 AM

Thanks Hugh. That about dials it in. I can see that fiddling around without the Vbe multiplier is not going to go..

effebi 24th January 2013 09:34 AM

Quote:

Originally Posted by Andrew Eckhardt (Post 3338942)
Does anyone have an off-hand idea how far off from stable this circuit would be? The actual application will probably use more like ten pairs of bipolar outputs for a 2.66 ohm nominal load with 65 volt rails. With only 15mA emitter current (basically class B), I wonder if it would run away if it got hot..

Just a guess , but isn't .2 ohm a little small for ensure proper Ic balancing of 10 pairs? I would go for 0.47 at least. I have a similar design (10 pairs To3) on 2 Ohm that works happily with 1 Ohm emitter resistors.

effebi 24th January 2013 10:09 AM

In any case, if you are going with a long heat sink and forced ventilation place your bias spreader on the output side of the air flow. That side is going to be hotter and where your transistors will most probably fail if ever.


All times are GMT. The time now is 11:51 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