F4 power amplifier

Jump to Latest
B

Blitz

Member
Joined 2002
Another question: Is the recommended bias per device now 0.53 ampere or 0.43 ampere ? In the latest service manual it is 0.53, but I have have as well a version with 0.43 in it?

And second question: If I measure across the source resistors (0.5 ohm) of the 9240 and of the 240, I get two different values, on typicically is 0.200V, while the other is 0.220 V. Do you found the same true ?
 
G

giacomo_pagani

Member
Joined 2007
Blitz said:

And second question: If I measure across the source resistors (0.5 ohm) of the 9240 and of the 240, I get two different values, on typicically is 0.200V, while the other is 0.220 V. Do you found the same true ?
Click to expand...

I'm very interested in this one too. I read that the correct BIAS should match with 250mV across those resistors... Then I set it that way, and I have very HOT heatsinks (not measured yet). Now Blitz' question is makin' me think about it: is 250 mV wrong, should I stay within 200-220?

Giacomo
 
Zen Mod

Zen Mod

Official Court Jester
Joined 2003
zenmod.in.rs/
Paid Member
Blitz said:
Another question: Is the recommended bias per device now 0.53 ampere or 0.43 ampere ? In the latest service manual it is 0.53, but I have have as well a version with 0.43 in it?

And second question: If I measure across the source resistors (0.5 ohm) of the 9240 and of the 240, I get two different values, on typicically is 0.200V, while the other is 0.220 V. Do you found the same true ?
Click to expand...


giacomo_pagani said:


I'm very interested in this one too. I read that the correct BIAS should match with 250mV across those resistors... Then I set it that way, and I have very HOT heatsinks (not measured yet). Now Blitz' question is makin' me think about it: is 250 mV wrong, should I stay within 200-220?

Giacomo
Click to expand...

question for current - as much output mosfets are happy ........ meaning on excessive temperature of die , or dissipation ..... what comes first ;

considering writing of you both, in your case(s) temperature will come first .

so - taking in account Papa's "skin burning test" ........ fry them 'till fingers can stay on heatsinks ; then stop , or even back up a little .

nota bene - if Magura try to tell ya something about these issues - do not even read ..... he like them toasty :rofl:

regarding 23V6 vs. 22V ............ naah ....... don't even bother .... ;)
 
Tarasque

Tarasque

Member
Joined 2002
Blitz said:
Another question: Is the recommended bias per device now 0.53 ampere or 0.43 ampere ? In the latest service manual it is 0.53, but I have have as well a version with 0.43 in it?

And second question: If I measure across the source resistors (0.5 ohm) of the 9240 and of the 240, I get two different values, on typicically is 0.200V, while the other is 0.220 V. Do you found the same true ?
Click to expand...


I will not comment on the bias per fet. As I typically use heatsinks from the scrapyard, and these can vary a bit wrt thermal properties, I tend to use a bias that gives me a heatsink temperature of ~45 - 50 degrees C.

As source resistors have a tolerance as well. Did you check the resistor values? 5% tolerance and 2 resistors on the extremes can give you the situation presented.
 
S

Samuel Jayaraj

Member
Joined 2001
Blitz said:
Just a short update:
- Is the "R" in the PSU-filter only there to reduce ripple or does it as well do other stuff (it obviously enhance the impedance of the PU, which is normally not a good thing) which impacts the sound positively ? I have as well some nice chokes here around, but I ask myself if that makes any sense when you simply don't have any ripple to fight ?
Click to expand...

The RC has some time constant and works like a high pass filter at very low frequencies. Generally, the sound is cleaner upto about 500Hz, as a result of CRC filters. They also seem to remove the bright edge in the higher frequencies. Finally the caps on board the amp-boards play a very important role in "voicing" the amp.
 
MRupp

MRupp

Member
Joined 2003
Now I have changed to Jensen's 4 Pole. I read in the data sheet that the Jensen have a kind of build in CLC, so I simply did not use a CRC, withdrew the resisor from the equation and have now four 25/47000uF Jensens playing per channel.
Click to expand...

Did Jensen state what value the L is so that you can calculate the filter? If anything it must be much smaller compared to a discrete CLC filter, probably by several orders of magnitude.

I believe for all practical purposes you should consider the Jensens to be single caps, so you have now converted from a CRC filter to a simple C filter and are comparing the two, this may explain most of the sound difference. What you could off course try is building a CRC with the Jensens in the second position.
 
B

Blitz

Member
Joined 2002
Samuel Jayaraj said:


The RC has some time constant and works like a high pass filter at very low frequencies. Generally, the sound is cleaner upto about 500Hz, as a result of CRC filters. They also seem to remove the bright edge in the higher frequencies. Finally the caps on board the amp-boards play a very important role in "voicing" the amp.
Click to expand...

OK...that is a lot sounding like what I am missing right now, some of the charme / smoothness is gone and some bright egde in the trebles are introduced...
 
T

tab30

Member
Joined 2001
diferent Mosfets

Hello,

interposed question:
I have a lot of IRF9540 and IRF540. Can I use this devices instead of IRFP9240 / IRF240 in the F4 ?

( I'm not sure but I belief somebody said that the Devices even have the same DIE inside with just another package)

Greetings from Germany,

Andreas
 
jacco vermeulen

jacco vermeulen

Member
Joined 2004
At a little over 13W of dissipation each, i've been a bit surprised as yet no one has posted IRF630/9630.

I've used Fairchild's TO220 for the A75 (shivers, that's fifteen help me gawd years ago), "papa's got the same die in a brand new pig bag" as the intended IRF230-9230/31 TO3 combo for Pyjama-P and Spaced-Out Normy's amp.
630/9630s are like 25cts each nowadays.
 
Zen Mod

Zen Mod

Official Court Jester
Joined 2003
zenmod.in.rs/
Paid Member
jacco vermeulen said:
At a little over 13W of dissipation each, i've been a bit surprised as yet no one has posted IRF630/9630.

I've used Fairchild's TO220 for the A75 (shivers, that's fifteen help me gawd years ago), "papa's got the same die in a brand new pig bag" as the intended IRF230-9230/31 TO3 combo for Pyjama-P and Spaced-Out Normy's amp.
630/9630s are like 25cts each nowadays.
Click to expand...


naah .........

even 530/540 with cousins are good enough

there are numerous possibilities around these numbers ........ if someone is cheapskate as Jaccovitty and ZM ....... :rofl:
 
You must log in or register to reply here.
Top Bottom