MOSFET Amplifier IRFP240/IRFP9240 - Page 9 - 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 6th March 2010, 08:38 AM   #81
diyAudio Member
 
Join Date: May 2009
Quote:
Originally Posted by krachkiste View Post
The relevant capacitance is the gate-drain capacitance C_gd. The datasheet sais C_gd < 300pF @ V_gs>10V (and almost disappears for higher V_gs) and C_gd_max=1250pF (@0V V_gs)
there's an error, it's supposed to say:
C_gd < 300pF @ V_ds>10V (and almost disappears for higher V_ds) and C_gd_max=1250pF (@0V V_ds)
  Reply With Quote
Old 6th March 2010, 08:48 AM   #82
diyAudio Member
 
steph_tsf's Avatar
 
Join Date: Mar 2008
Quote:
Originally Posted by krachkiste View Post
According to the datasheet the C_gs of an IRFP240 is max 1.2nF. But the V_gs is only changing about 0.5V while the output swings 75V. So the effective C_gs presented to the VAS is about 0.5/75*1n2=8pF and can therefore be neglected.
Very interesting consideration. I partially agree with you.

One can see the circuit from the devil's advocate, worst case analysis.

Consider that the load (the loudspeaker) is 4 Ohm. This is a very small impedance compared to what we expect in the collector of the VAS.

I don't say here that it shows on the collector of the VAS.

But you will agree with me that the VAS gets a path to it (to the load) through the C_gs capacitance. Which is 1200pF maybe as you wrote.

So now, look what's happening when you need to swing a large voltage on the load, which means a large output current. You'll need to go from a Vgs voltage equal to maybe 2V (the MOS was OFF), to a Vgs maybe equal to 4V (the MOS conducts some current). When the MOS conducts, the current flows in the load, and the voltage greatly increase, maybe a 40V swing. We thus need an increase of 2V on Vgs, which then (automatically ?) commands a 40V swing on the load.

A voltage of 40V on the load means 10 Amp in a 4 Ohm load.
A 10 Amp in the load, commanded by a Vgs increase of 2 V, makes an overall transconductance equal to 5 Siemens.
I think we are okay with these parameters. The orders of magnitude seem correct.

With the above example, can we thus conclude that seen from the VAS collector, the parasitic capacitance is not equal to 1200pF, but equal to 1200pF divided by 20, which is the ratio between the Vout swing (40V) and Vgs swing (2V) ?

If this is a valid devil's advocate, worst case analysis, is is valid to say that the effective C_gs presented to the VAS is about 1200pF/20 = 60pF ?

Instead of the 8pF you have quoted.

I wrote "We thus need an increase of 2V on Vgs, which then (automatically ?) commands a 40V swing on the load."
If there is any glitch about the idea of "automatic", then we face the plain and ugly situation of the VAS stage seeing a 1200pF parasitic capacitance, not connected to ground, but connected to the 4 Ohm load, which for the VAS point of view is a ground, roughly.

Are there mechanisms making the above process less than "automatic" ?
MOSFETs are not infinitely fast.
However, with their 50ns turn-on time, rise time, turn-off time, we may assume they are infinitely fast for audio frequencies.

1. From the devil's advocate, worst case analysis, are there delays or other effects to be taken into account, in the s range, that cannot be neglected for audio applications ?

2. Don't you think that the 5 Siemens transconductance that we have taken into consideration, may be much lower when the amplifier is operating at low power (below 250 mW - home HiFi use), setup with a low bias current (maybe 25 mA in each MOS) ? The transconductance may be 3 times smaller, hence the effective C_gs presented to the VAS to be multiplied by 3, making it equal to 60pF * 3 = 180pF. Per MOSFET !

Any comment welcome.
Very interesting discussion.
Attached Files
File Type: pdf irfp240.pdf (170.9 KB, 598 views)
  Reply With Quote
Old 6th March 2010, 09:00 AM   #83
diyAudio Member
 
Join Date: Mar 2010
Dear Krachkiste,

Your are absolute right. Even some amater audio designe have output stage with Power Mosfets they all have generally problem, the parasite capacitance Cgd - Crss (in attachmet file).
Attached Files
File Type: pdf IRFP240[1].pdf (100.7 KB, 373 views)
  Reply With Quote
Old 6th March 2010, 09:51 AM   #84
diyAudio Member
 
Join Date: May 2009
Quote:
Originally Posted by steph_tsf View Post
So now, look what's happening when you need to swing a large voltage on the load, which means a large output current. You'll need to go from a Vgs voltage equal to maybe 2V (the MOS was OFF), to a Vgs maybe equal to 4V (the MOS conducts some current). When the MOS conducts, the current flows in the load, and the voltage greatly increase, maybe a 40V swing. We thus need an increase of 2V on Vgs, which then (automatically ?) commands a 40V swing on the load.
Thats the thing with the human mind. We tend to simplify things down to an almost absurd level so we can get a glimpse of understanding (just as I did before), but then we already might have overlooked some important things, or our simplified modell no longer applies to the real world.

We also tend to tweak the assumptions that go into our simplified model (the devil's advocat) so that we get the result we expect. (Just as I did before)
Quote:
If this is a valid devil's advocate, worst case analysis, is is valid to say that the effective C_gs presented to the VAS is about 1200pF/20 = 60pF ?

Instead of the 8pF you have quoted.
At the end of the day one "can" calculate the amount of the capacitance C_gs that contributes to the capacitance presented to the VAS as
C_p(V_out) = C_gs(V_ds)/(g_m(I_d)*R_Load). (One can get the idea.)
Note all the dependancies, that cannot be neglected.

But I would only rely on an actual measurement, or maybe on a spice simulation, if properly conducted.
Quote:
If there is any glitch about the idea of "automatic", then we face the plain and ugly situation of the VAS stage seeing a 1200pF parasitic capacitance, not connected to ground, but connected to the 4 Ohm load, which for the VAS point of view is a ground, roughly.
You forget to mention the gm of the FETs that is also connected to the load and C_gs. I'm sure you had it in mind.
It's maybe like saying, its hard to press the gas pedal as I (the VAS) will have to move the whole car, but forgetting that theres an engine (g_m) under the hood helping me with that.
I know that is not what you are saying, but i found the image quite interesting to stress the importance of the g_m.

PS:
And actually this whole discussion is one of the reasons (maybe there are more, I don't know yet) I will use an emitter follower driver stage before my FETs

Last edited by krachkiste; 6th March 2010 at 09:57 AM.
  Reply With Quote
Old 6th March 2010, 11:30 AM   #85
Account disabled at member's request
 
Join Date: May 2006
I still dont see anything special about the cfp vas, has been used for a very long time, I have them in amps I built 15 years ago, theres nothing unique about it, works well because it has high input impedance.

Ideally those mosfets should be driven by a emmiter follower for good performance, even better use them in a triple configuration.

Krachkista if you go through Bob Cordells thread BJT vs Mosfet I think it was youll find many more reasons to use a emmiter follower than just the parasitic capacitances.

Steph tfs, I find it amusing you say that those bc parts are better than the toshiba parts, you obviously havent done a thorough analysis like you did with the rest of the amp. The toshibas have far less noise (think of the effect of rbb, especially when using large degeneration) and the beta easily match or better those of the bc parts, their Vce is more than adequite for the voltages of this circuit. Also theres other factors to consider, once again capacitances, early voltage ......
  Reply With Quote
Old 6th March 2010, 01:20 PM   #86
diyAudio Member
 
Join Date: Mar 2010
As experience audio designer to explain to audience why this schematic amplifier is not good, is not very simple.
The practical explanation is sometimes simplest: Try to put this amplifier with output mosfet IRF240/IRFP9240 in clipping (high input signal) but without load (loudspeakers).

The amplifier will burn. Before you do that have on mind that I already do that in the past. The only Mosfet at output stage which will not burn is form 2SK/2SJ audio series but if you look their characteristic carefully they have more common with transistors then with POWER MOSFETS.

The conclusion form respect Mr. T.Leach that mosfet is not suitable for output audio stages are true.
  Reply With Quote
Old 6th March 2010, 03:37 PM   #87
diyAudio Member
 
apexaudio's Avatar
 
Join Date: Feb 2010
Location: Serbia
Quote:
Originally Posted by AUDIOMEN View Post
As experience audio designer to explain to audience why this schematic amplifier is not good, is not very simple.
The practical explanation is sometimes simplest: Try to put this amplifier with output mosfet IRF240/IRFP9240 in clipping (high input signal) but without load (loudspeakers).

The amplifier will burn. Before you do that have on mind that I already do that in the past. The only Mosfet at output stage which will not burn is form 2SK/2SJ audio series but if you look their characteristic carefully they have more common with transistors then with POWER MOSFETS.

The conclusion form respect Mr. T.Leach that mosfet is not suitable for output audio stages are true.
The amplifier will not burn at clipping level without load.
  Reply With Quote
Old 6th March 2010, 07:17 PM   #88
diyAudio Member
 
steph_tsf's Avatar
 
Join Date: Mar 2008
Quote:
Originally Posted by krachkiste View Post
At the end of the day one "can" calculate the amount of the capacitance C_gs that contributes to the capacitance presented to the VAS as
C_p(V_out) = C_gs(V_ds)/(g_m(I_d)*R_Load). (One can get the idea.)
Note all the dependancies, that cannot be neglected.
But I would only rely on an actual measurement, or maybe on a spice simulation, if properly conducted.
I agree on the formula !
Where can I find a vertical power MOSFET spice model, providing decent accuracy regarding the g_m dependency in function of I_d ?
  Reply With Quote
Old 6th March 2010, 07:56 PM   #89
diyAudio Member
 
Join Date: May 2009
Good hint Homemodder. Thank you.

Bob Cordell Interview: BJT vs. MOSFET

@ steph_tsf
take a look at post 24.
On the IRF page you can download subcircuit models of their FETs.
I don't to how detailed or accurate they are though.
  Reply With Quote
Old 6th March 2010, 08:50 PM   #90
wahab is offline wahab  Algeria
diyAudio Member
 
Join Date: Nov 2009
Location: algeria/france
Quote:
Originally Posted by steph_tsf View Post
I agree on the formula !
Where can I find a vertical power MOSFET spice model, providing decent accuracy regarding the g_m dependency in function of I_d ?
.subckt IRF640 D G S
.model mosfet NMOS( LEVEL=17 VTO=3.869 RS=0.01909 KP=7.151 RD=0.0981 TC1RD=0.0126 RG=6 IS=1e-36 LIMIT(VDS)=(*, 200)
+ CGDMAX=4.00E-09 CGDMIN=1.00E-11 XG2CGD=0.5 XG1CGD=0.1 CBD=3.38E-10 VTCGD=0.8)
.model diode D( IS=2.10e-12 RS=0.0125 TT=4.673e-07)
M1 D G S S mosfet
D1 S D Diode
Cgs G S 1.19E-09
.ends



.subckt IRF9640 D G S
.model mosfet PMOS( LEVEL=17 VTO=-3.608 RS=0.0724 KP=8.645 RD=0.3095 TC1RD=0.0128 RG=13.171 IS=1e-36 LIMIT(VDS)=(-200,*)
+ CGDMAX=2.30E-09 CGDMIN=2.00E-11 XG2CGD=0.3 XG1CGD=0.1 CBD=3.34E-10 VTCGD=2)
.model diode D( IS=1.17e-28 RS=0.0634 TT=4.617e-07)
M1 D G S S mosfet
D1 D S Diode
Cgs G S 1.07E-09
.ends
  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
IRFP240&amp;IRFP9240 power amplifier schematic mos Solid State 31 26th June 2012 08:21 AM
IRFP240 / IRFP9240 MOSFETs danieljw Swap Meet 8 4th July 2009 01:51 AM
Wtb: Irfp240 / Irfp9240 folkeb Swap Meet 1 10th March 2009 10:18 AM
IRFP240 + IRFP9240 -- pairs jackinnj Swap Meet 1 2nd June 2006 02:22 AM
IRFP240 and IRFP9240 Wanted richeros Swap Meet 1 9th November 2002 12:38 PM


New To Site? Need Help?

All times are GMT. The time now is 05:50 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