simple amplifier with mosfets

Stormo · 12th March 2004, 01:24 PM

Hi all, will this little baby work in real life ? -

Frank Berry · 12th March 2004, 01:38 PM

I may be wrong but ... no.
It looks like C3 is providing an a.c. short across the upper transistor.
Lose C3.

Charles Hansen · 12th March 2004, 02:43 PM

Also, it will be very difficult to achieve the correct biasing. One solution would be to use a Vbe (or Vgs) multiplier, mounted to the heatsink to allow thermal tracking.

Good luck!

traderbam · 12th March 2004, 03:28 PM

The biasing difficulty is this.
With no input signal you, ideally, want the source voltage to be half the power rail, or 43V. The voltage across R2 will be about 0.6V. So the voltage across R3 will be 42.4V and the current through R3 will be 42.4/860k = 49uA. This current must be the sum of the current through R2 and the base current of Q3. R2's current is about 0.6V/56k = 11uA, so Q3 base current must be 38uA. So Q3's base current is dominating the current through R3. Now, Q3's collector current needs to be about 43/10k = 4.3mA. Knowing Q3's base and collector currents we can estimate it's beta to be 4.3m/38u = 113.

So the source voltage will sit at half the supply voltage only if the beta of Q3 is 113. The trouble is that the beta characteristic varies from one device to the next, often 50% or more, and within a single device it varies with Ic and temperature and Vce. So the odds of beta being 113 are very slim. If the beta was 50% out so would be your source voltage. You need a bias arrangement that is relatively insensitive to beta.

Nelson Pass · 12th March 2004, 10:00 PM

Ditto on the above. Also:

I would consider splitting R1 into two values and attaching the
+ terminal of the 100 uF cap between them for bootstrapping,
which is what might have originally been intended.

Stormo · 13th March 2004, 02:27 PM

Thanks alot for your comments. I made some changes regarding the bias, and the bootstrap.

It's an easy little amp - will it run ?

peranders · 13th March 2004, 03:09 PM

The three diodes aren't enough to bias the mosfets. You will need 4-6 volts (compared to 2.1 V) and some temperature compensation.

Charles Hansen · 13th March 2004, 03:32 PM

Per Anders is right. Use a Vbe multiplier as I suggested above, and mount it on the heatsinks. To address the bias stability of the input transistor as Traderbam noted, add an emitter resistor on the input transistor.

millwood · 13th March 2004, 03:52 PM

alternatively, you can use a pot in place of the three diodes to get this thing going. It wouldn't be able provide any thermal stability, which is critical in your design as there are no source resistors on Q3/Q4.

the voltage drop on R4/R7 is about 40v (86/2-3v for the Vgs drop on Q3). and you need about 6v drop on between the gates of Q3/Q4 so you need a pot about 6/40*(3300+3300) = 990ohm so a 1.5k pot should do.

I would strongly suggest a pair of source resistors.

Michael Rothacher · 13th March 2004, 04:45 PM

Have a look at SuperFlea:

http://www.diyaudio.com/forums/showt...8376#post68376

Just look at one half. Of course the ultimate in simplicity for a complementary design is the ZV5

12th March 2004, 01:38 PM	#2
Frank Berry diyAudio Member Join Date: Sep 2003 Location: Midland, Michigan	I may be wrong but ... no. It looks like C3 is providing an a.c. short across the upper transistor. Lose C3. __________________ Frank

13th March 2004, 03:09 PM	#7
peranders Electrons are yellow and more is better! diyAudio Member Join Date: Apr 2002 Location: Göteborg, Sweden Blog Entries: 4	The three diodes aren't enough to bias the mosfets. You will need 4-6 volts (compared to 2.1 V) and some temperature compensation. __________________ /Per-Anders (my first name) or P-A as my friends call me

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
help about power amplifier mosfets	milad-k800	Car Audio	1	23rd August 2008 09:26 PM
Simple amplifier	hobbit	Solid State	15	27th January 2008 06:30 AM
pcb for Crescendo power amplifier / 2 pair of mosfets	Dusk	Elektor	1	23rd December 2006 12:08 PM
simple amplifier	sakis	Solid State	12	26th January 2006 11:18 PM
Lynx amplifier using Mosfets	Dusk	Solid State	7	16th July 2005 11:26 PM

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

12th March 2004, 02:43 PM	#3
Charles Hansen diyAudio Member Join Date: May 2003 Location: Colorado	Also, it will be very difficult to achieve the correct biasing. One solution would be to use a Vbe (or Vgs) multiplier, mounted to the heatsink to allow thermal tracking. Good luck!

12th March 2004, 03:28 PM	#4
traderbam diyAudio Member Join Date: Jan 2002 Location: Earth	The biasing difficulty is this. With no input signal you, ideally, want the source voltage to be half the power rail, or 43V. The voltage across R2 will be about 0.6V. So the voltage across R3 will be 42.4V and the current through R3 will be 42.4/860k = 49uA. This current must be the sum of the current through R2 and the base current of Q3. R2's current is about 0.6V/56k = 11uA, so Q3 base current must be 38uA. So Q3's base current is dominating the current through R3. Now, Q3's collector current needs to be about 43/10k = 4.3mA. Knowing Q3's base and collector currents we can estimate it's beta to be 4.3m/38u = 113. So the source voltage will sit at half the supply voltage only if the beta of Q3 is 113. The trouble is that the beta characteristic varies from one device to the next, often 50% or more, and within a single device it varies with Ic and temperature and Vce. So the odds of beta being 113 are very slim. If the beta was 50% out so would be your source voltage. You need a bias arrangement that is relatively insensitive to beta.

12th March 2004, 10:00 PM	#5
Nelson Pass The one and only Join Date: Mar 2001	Ditto on the above. Also: I would consider splitting R1 into two values and attaching the + terminal of the 100 uF cap between them for bootstrapping, which is what might have originally been intended.

13th March 2004, 03:32 PM	#8
Charles Hansen diyAudio Member Join Date: May 2003 Location: Colorado	Per Anders is right. Use a Vbe multiplier as I suggested above, and mount it on the heatsinks. To address the bias stability of the input transistor as Traderbam noted, add an emitter resistor on the input transistor.

13th March 2004, 03:52 PM	#9
millwood Account Disabled Join Date: Apr 2003 Location: US	alternatively, you can use a pot in place of the three diodes to get this thing going. It wouldn't be able provide any thermal stability, which is critical in your design as there are no source resistors on Q3/Q4. the voltage drop on R4/R7 is about 40v (86/2-3v for the Vgs drop on Q3). and you need about 6v drop on between the gates of Q3/Q4 so you need a pot about 6/40*(3300+3300) = 990ohm so a 1.5k pot should do. I would strongly suggest a pair of source resistors.

13th March 2004, 04:45 PM	#10
Michael Rothacher www.audiohobby.com diyAudio Member Join Date: May 2001 Blog Entries: 3	Have a look at SuperFlea: http://www.diyaudio.com/forums/showt...8376#post68376 Just look at one half. Of course the ultimate in simplicity for a complementary design is the ZV5

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