Vertical MOSFETs for a PA amplifier OP - diyAudio
Go Back   Home > Forums > Amplifiers > Solid State

Solid State Talk all about solid state amplification.

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
Reply
 
Thread Tools Search this Thread
Old 18th June 2012, 08:39 AM   #1
effebi is offline effebi  Italy
diyAudio Member
 
effebi's Avatar
 
Join Date: Nov 2010
Location: Milan
Default Vertical MOSFETs for a PA amplifier OP

Hi all,
Many people praises Mosfets for OP stages.
Let's speak about a middle size PA amplifier (500-1000 W on 4 Ohm).
Are there any advantages to use Vertical Mosfets in a class AB design over BJTs?
Please let Laterals and other exotic parts out of this discussion.

I see some disadvantages:
Less watts due to bigger losses, more difficult to select perfectly paired devices, more HF stability issues.
Also there is not perfectly P-N part equivalence, but for this a "quasi" topology would probably do for PA use.

Part cost is about the same as good BJTs , depending where you buy them.

On the good side it seems Vertical MOS are more "robust". But since they are also prone to thermal runaway I would like to hear more practical experience about that. What I can se is that there are parts rated for very high currents and with decent SOA, but the power is ultimately limited by the package, so it seems you will have to use the same number of devices as BJTs at the end.
So, why MOS?
  Reply With Quote
Old 18th June 2012, 11:00 AM   #2
Ar4 is offline Ar4  Latvia
diyAudio Member
 
Join Date: Apr 2010
Location: Latvia
Send a message via Skype™ to Ar4
You might want to read this thread Bob Cordell Interview: BJT vs. MOSFET
I personally prefer them (hexFets) in simple circuits with high bias
  Reply With Quote
Old 18th June 2012, 11:06 AM   #3
AndrewT is offline AndrewT  Scotland
diyAudio Member
 
Join Date: Jul 2004
Location: Scottish Borders
There is no doubt that good performance can be obtained using vertical mosFET output stages.
These have become ever more popular due to the relatively cheap availability of switching mosFETs.

For PA duty, I would tend to go towards cheapest, rather than best.

For best, I would aim for mosFET driver, feeding a wide bank of BJTs.
__________________
regards Andrew T.
  Reply With Quote
Old 18th June 2012, 08:13 PM   #4
Bensen is offline Bensen  Belgium
diyAudio Member
 
Bensen's Avatar
 
Join Date: May 2003
Location: Belgium, Limburg
Quote:
Originally Posted by AndrewT View Post
For best, I would aim for mosFET driver, feeding a wide bank of BJTs.
Wow, that you are telling this, takes all my doubts away to go further with a current new amp design, in which I'm driving Sanken 2922/1216 with 2sk2013/2sj313 mosfet's.
  Reply With Quote
Old 18th June 2012, 08:35 PM   #5
diyAudio Member
 
Lazy Cat's Avatar
 
Join Date: Sep 2007
Quote:
Originally Posted by AndrewT View Post
For best, I would aim for mosFET driver, feeding a wide bank of BJTs.
Hehe like my BIGBT-s
  Reply With Quote
Old 18th June 2012, 09:17 PM   #6
CBS240 is offline CBS240  United States
diyAudio Member
 
CBS240's Avatar
 
Join Date: Aug 2005
Location: K-town
With EC, matching P-ch to N-ch is not so much of an issue if you choose two that have relatively equal transconductance. The Rds-on figure is irrelevant for this type of application. The P-ch will have a greater Ciss than the N-ch for equal Gm, but if you use a complementary source follower Ciss is not so much of a problem because it is bootstrapped from the output node. It is important to clamp the gate drive so it does not excessively charge the gates when driven above Vd. Don't forget the gate source Zener protection diodes. Coss is more of an issue.

I am fond of the planer stripe architecture. I find them to be a bit more robust and able to take significant abuse when used for analog applications. Stay away from Trench fets, U-fets, and other types with similar architecture. They are not suitable for analog apps. Vds breakdown is not so consequential as long as it is higher than 2Vdd. For example, FQP44N10 (Vds=100V)and FQP47P06 (Vds=60V) would be good complements for rails +/-25V or less, Gm is similar. Local frequency compensation is a must and a small value gate Zobel from gate to drain as close to the device pins after the gate stopper resistor on each device is a great way to dampen the tanking components, those being the internal capacitances and lead/bonding wire inductances.

Although these devices create lots of distortion due to the exponential transfer (great for switching apps tho) there is no reason why a class AB Hi-Fi amplifier cannot be made. Bob's paper is a great example. In fact, I use the concepts discussed to build this amp. It has shown me and to my surprise how rugged those devices can be made to be. I recommend you use about 20mA per device pair in the driving stage. This assures the drivers always remain in class A operation during charging/discharging of the gates. Bias current should be at least 100-200mA. This helps to traverse the low Gm operating region at around cut-off. Because of the low Gm region at around bias, vertical mosfets are more forgiving as far as thermal compensation is concerned but they still require it. Here is another amp I previously made using the planer stripe vertical mosfets.
__________________
All the trouble I've ever been in started out as fun......

Last edited by CBS240; 18th June 2012 at 09:25 PM.
  Reply With Quote
Old 19th June 2012, 04:32 PM   #7
effebi is offline effebi  Italy
diyAudio Member
 
effebi's Avatar
 
Join Date: Nov 2010
Location: Milan
Sorry for the late answer, there is so much to learn!
I am reading the Bob's paper as well.
For AndrewT: Thanks for the hints.
So, forgetting price and sound quality, what you think about reliability?
Do VMOS bring any advantage over good (say ONsemi stuff) BJTs?

CBS 240:
Thanks a lot for the very useful infos.
However:
Quote:
Originally Posted by CBS240 View Post
With EC, matching P-ch to N-ch is not so much of an issue if you choose two that have relatively equal transconductance. The Rds-on figure is irrelevant for this type of application. The P-ch will have a greater Ciss than the N-ch for equal Gm, but if you use a complementary source follower Ciss is not so much of a problem because it is bootstrapped from the output node. It is important to clamp the gate drive so it does not excessively charge the gates when driven above Vd. Don't forget the gate source Zener protection diodes. Coss is more of an issue. .
Agree, but some said complementar pairs do not even have the same Id Max. That's why the Quasi way looked simpler to me.

Quote:
Originally Posted by CBS240 View Post
I am fond of the planer stripe architecture. I find them to be a bit more robust and able to take significant abuse when used for analog applications. .
Would you elaborate on why you like more planar?
Could you please quote some part number in the, say, 250W package range?
And, BTW, your proposed 25 V rail value is too less in PA.
I am looking at something that can do 200V Vds at least....
  Reply With Quote
Old 19th June 2012, 05:35 PM   #8
AndrewT is offline AndrewT  Scotland
diyAudio Member
 
Join Date: Jul 2004
Location: Scottish Borders
Have you looked at Quasi's Threads?
Or Apex's Threads?
There are some mighty high power amps in there.
__________________
regards Andrew T.
  Reply With Quote
Old 19th June 2012, 07:38 PM   #9
diyAudio Member
 
Join Date: May 2010
Location: was Chicago IL, now Long Beach CA
Paraphrasing Nelson Pass: MOSFETs are generally much faster than BJTs, so they can tolerate more feedback without increased TIM, so despite initially worse distortion they can end up better after feedback is applied.

And who said MOSFETs are prone to thermal runaway? My MOSFET PA amps only have some protection at the power supply; if you short the output the house lights dim, the fans go on high, the shorted speaker wire strand glows red and burns out and then the house light come back up and the fans quiet down and the show goes on LOL. And I hear that if I ever do blow one, they don't tend to put the DC rail thru your speakers.
  Reply With Quote
Old 20th June 2012, 06:46 AM   #10
effebi is offline effebi  Italy
diyAudio Member
 
effebi's Avatar
 
Join Date: Nov 2010
Location: Milan
Quote:
Originally Posted by AndrewT View Post
Have you looked at Quasi's Threads?
Or Apex's Threads?
There are some mighty high power amps in there.
Yes, and Yes.
While Quasi seems to prefer the MOS, for what I gathered from Apex he abandoned MOS for BJTs in his excellent PA designs.
So the question is still open, if the VMOS brings any real advantage over MODERN BJTs. Myself I have long term experience of BJT high power stuff
(on the 1-2 kW range, with TO3s) but none with MOS, so I am trying to make up my mind before eventually starting to play with them (it is going to be an expensive game).
  Reply With Quote

Reply


Hide this!Advertise here!
Thread Tools Search this Thread
Search this Thread:

Advanced Search

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are Off
Pingbacks are Off
Refbacks are Off


Similar Threads
Thread Thread Starter Forum Replies Last Post
QUAD 405 revisited (TL071 & vertical MOSFETs) steph_tsf Solid State 12 16th April 2010 09:23 PM
QUAD 405 revisited (AD844 & vertical MOSFETs) steph_tsf Solid State 4 26th March 2010 12:26 AM
Vertical amplifier setup Thomasjunta Everything Else 3 13th May 2009 05:17 AM
help about power amplifier mosfets milad-k800 Car Audio 1 23rd August 2008 09:26 PM
2SK/irfp lateral vertical mosfet amplifier HELP ?? medogrizli Solid State 53 2nd November 2006 12:47 AM


New To Site? Need Help?

All times are GMT. The time now is 04:47 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