thermal compensation

djdamix · 28th March 2005, 06:27 PM

Still having problem with my thermal compensation : I use IRFP240 & 9240 and the VBE multiplier is on the heatsink (MJE340) however the compensation is too high I got distortion when the temperature rise because the quiescent current decreases too much (=0mA!)

I've also try the compensation use by A.Holton in the AV800 : IRF610 and 100Ohm but it seems to be too low the quiescent current rise until 600mA or more

I've read S.Douglas's book has it described some way to change the tempco by adding voltage source and else but it seems quite a hard solution ...

At this point I plan to use a diode in serie with the vbe multiplier (MJE340) and to put only the diode on the heatsink to have small tempco.

Any help or comment is welcome !

David

K-amps · 28th March 2005, 06:59 PM

Here's a suboptimal solution but works... add multiple layers on insulation on the bias device and heatsink to be tracked...

ChocoHolic · 28th March 2005, 08:55 PM

Hi David !
You can reduce the steepness of your temperature compensation
quite easy.
If you put a series resistor to the emitter of the MJE340 in your
Vbe multiplier.
The value of this resistor depends on the current which is running through MJE340 in your design.
I am now assuming about 10mA steady state current in the VAS:
A good starting point may be a 4.7 Ohm resistor in series to the emitter and then slowly increasing values if the compensation is still
to massive. Please note that you may have to readjust the idle current depending on the additional resistor.

If there is already resistor at the emitter of your
MJE-340- Vbe-multiplier, then you simply could increase this value.

BTW:
Which resistor values do have in series with the
source of your output MOSFETS?

Good luck
Markus

djdamix · 28th March 2005, 09:15 PM

K-amps => It could be a solution, I'll try but I think i'll increase the delay before the current adjusts ...

Markus, Thank you I'll try ! Last time your advices about DC offset were helpfull !! Now my amp is working I just have this quiescent current problem to solve and it will be ok.

The resistor in serie with the mosfet are 0.22 ohm

Thanks

David

ChocoHolic · 28th March 2005, 10:59 PM

Thanks for the flowers....
Let's hope that I understood your current fight correctly
and that my hint is helpful this time, too.

I think 0.22Ohms at the source of your output Fets
should be OK.

Bye
Markus

anatech · 29th March 2005, 01:37 AM

Hi djdamix,
Markus was referring to a series resistance in the emitter of your MJE340. This is exactly what they did in the Counterpoint amps.

When I changed to a bipolar type output the sound became much cleaner, less distorted. Never mind the bias was now easy to control.

-Chris

djdamix · 29th March 2005, 06:03 PM

I'm working on the thermal compensation & I was wandering what was the recommended quiescent current for AB mosfet amp. It seems that I need about 400mA to make the distortion disapear at 20Khz at about full power (according to my scope)

Does it seems ok ? Can somebody confirm ?

Thanks.

djdamix · 29th March 2005, 06:10 PM

=>Chris, what is the Counterpoint amps you mentionned ? Is there any schematic somewhere ? I'm interested !

David

ChocoHolic · 29th March 2005, 08:50 PM

Hi David !
I am also curious about the counterpoint amps.
My idea was simply to reduce the impact of the
temperature on the collector current of the Vbe multiplier BJT.
Emitter resistor are the most traditional way to reduce the temperature dependency of a transistor current....

I think your observation of distorsions at high frequencies is
reasonable. I guess your amplifier is the one which you have posted here:
improving my VAS
The open loop gain of this design is not very high and may additionally decrease already below 20 kHz. Your closed loop gain is high.
About 100. So there is not much feedback. You will have to work with
quite some idle current. 400mA do not make me wonder.
Probably you will find output powers which show even higher distorsion than at full load. Did you also try about 10V-20V peak output at 4 Ohms load .... ?

Good luck
Markus

Allexx · 29th March 2005, 08:57 PM

In my amplifier I use CFP output with BJT driving MOSFETs with drain connected to load. Vbe multiplier transistor of the same type as driver is installed on the same small heatsink of a driver transistor (does not matter which one). This way compensation of quiescent current is amazingly good. Maybe small common heatsink for drivers with Vbe mounted there is the best solution. Do not forget that output is CFP.

28th March 2005, 06:27 PM	#1
djdamix diyAudio Member Join Date: Jun 2004 Location: Lyon	thermal compensation Still having problem with my thermal compensation : I use IRFP240 & 9240 and the VBE multiplier is on the heatsink (MJE340) however the compensation is too high I got distortion when the temperature rise because the quiescent current decreases too much (=0mA!) I've also try the compensation use by A.Holton in the AV800 : IRF610 and 100Ohm but it seems to be too low the quiescent current rise until 600mA or more I've read S.Douglas's book has it described some way to change the tempco by adding voltage source and else but it seems quite a hard solution ... At this point I plan to use a diode in serie with the vbe multiplier (MJE340) and to put only the diode on the heatsink to have small tempco. Any help or comment is welcome ! David

Thread Tools	Search this Thread
Show Printable Version Email this Page	Search this Thread: Advanced Search

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Biasing/thermal compensation of Thermal Trak transistors	Bob Cordell	Solid State	156	11th December 2010 09:38 PM
Thermal Compensation Bias	hawkfeather	Solid State	4	23rd January 2006 09:57 PM
Thermal compensation	lumanauw	Pass Labs	1	23rd April 2005 08:32 PM
Thermal compensation	DarkOne	Pass Labs	5	30th December 2004 09:55 AM

New To Site?	Need Help?
Register to Participate Search Privacy Statement Contact Us	Frequently Asked Questions Did you forget your password? Mark Forums Read


	Please consider donating to help us continue to serve you. Ads on/off / Custom Title / More PMs / More album space / Advanced printing & mass image saving

28th March 2005, 06:59 PM	#2
K-amps diyAudio Member Join Date: Nov 2003 Location: Indiana	Here's a suboptimal solution but works... add multiple layers on insulation on the bias device and heatsink to be tracked...

28th March 2005, 08:55 PM	#3
ChocoHolic diyAudio Member Join Date: Dec 2003 Location: Munich	Hi David ! You can reduce the steepness of your temperature compensation quite easy. If you put a series resistor to the emitter of the MJE340 in your Vbe multiplier. The value of this resistor depends on the current which is running through MJE340 in your design. I am now assuming about 10mA steady state current in the VAS: A good starting point may be a 4.7 Ohm resistor in series to the emitter and then slowly increasing values if the compensation is still to massive. Please note that you may have to readjust the idle current depending on the additional resistor. If there is already resistor at the emitter of your MJE-340- Vbe-multiplier, then you simply could increase this value. BTW: Which resistor values do have in series with the source of your output MOSFETS? Good luck Markus

28th March 2005, 09:15 PM	#4
djdamix diyAudio Member Join Date: Jun 2004 Location: Lyon	K-amps => It could be a solution, I'll try but I think i'll increase the delay before the current adjusts ... Markus, Thank you I'll try ! Last time your advices about DC offset were helpfull !! Now my amp is working I just have this quiescent current problem to solve and it will be ok. The resistor in serie with the mosfet are 0.22 ohm Thanks David

28th March 2005, 10:59 PM	#5
ChocoHolic diyAudio Member Join Date: Dec 2003 Location: Munich	Thanks for the flowers.... Let's hope that I understood your current fight correctly and that my hint is helpful this time, too. I think 0.22Ohms at the source of your output Fets should be OK. Bye Markus

29th March 2005, 01:37 AM	#6
anatech diyAudio Moderator Join Date: Jun 2004 Location: Georgetown, On	Hi djdamix, Markus was referring to a series resistance in the emitter of your MJE340. This is exactly what they did in the Counterpoint amps. When I changed to a bipolar type output the sound became much cleaner, less distorted. Never mind the bias was now easy to control. -Chris

29th March 2005, 06:03 PM	#7
djdamix diyAudio Member Join Date: Jun 2004 Location: Lyon	I'm working on the thermal compensation & I was wandering what was the recommended quiescent current for AB mosfet amp. It seems that I need about 400mA to make the distortion disapear at 20Khz at about full power (according to my scope) Does it seems ok ? Can somebody confirm ? Thanks.

29th March 2005, 06:10 PM	#8
djdamix diyAudio Member Join Date: Jun 2004 Location: Lyon	=>Chris, what is the Counterpoint amps you mentionned ? Is there any schematic somewhere ? I'm interested ! David

29th March 2005, 08:50 PM	#9
ChocoHolic diyAudio Member Join Date: Dec 2003 Location: Munich	Hi David ! I am also curious about the counterpoint amps. My idea was simply to reduce the impact of the temperature on the collector current of the Vbe multiplier BJT. Emitter resistor are the most traditional way to reduce the temperature dependency of a transistor current.... I think your observation of distorsions at high frequencies is reasonable. I guess your amplifier is the one which you have posted here: improving my VAS The open loop gain of this design is not very high and may additionally decrease already below 20 kHz. Your closed loop gain is high. About 100. So there is not much feedback. You will have to work with quite some idle current. 400mA do not make me wonder. Probably you will find output powers which show even higher distorsion than at full load. Did you also try about 10V-20V peak output at 4 Ohms load .... ? Good luck Markus

29th March 2005, 08:57 PM	#10
Allexx diyAudio Member Join Date: Jul 2004 Location: Ottawa	In my amplifier I use CFP output with BJT driving MOSFETs with drain connected to load. Vbe multiplier transistor of the same type as driver is installed on the same small heatsink of a driver transistor (does not matter which one). This way compensation of quiescent current is amazingly good. Maybe small common heatsink for drivers with Vbe mounted there is the best solution. Do not forget that output is CFP.

Currently Active Users Viewing This Thread: 1 (0 members and 1 guests)