F5 Listening Impressions & Discussion

[CameronTran]

Yeah, I definitely cooked R1 and R2. Will replace the IRFs and .47 ohm resistors. What worries me though is why I cooked them. DC offset had drifted a little high. Oscillation in the right channel (but not the left) with inputs open. So I shorted the inputs, oscillation went away and I successfully adjusted the left channel. Disaster during biasing of the right channel. I was getting close to .59 V across R11 when R12 somehow went over .8 V and I smelled something cooking. Shut it all down but it was too late.

Thoughts?

Also, I'm pretty sure I posted my question in the long F5 thread, but I think the admins moved it here. Or I could have posted it in the wrong thread too; it was a very late night.

Thanks,
C

[woodturner-fran]

Could yon have biased it too quickly?

Second: I had a similar problem in mine - turns out there was a metal filing caught in the insulating pad and it was shorting the irf when hot. The other thing is that I believe that the irf are very tough so its kinda unlikely that they are damaged - most likely casualties are the 0.47r resistor.

Are your 0.47r high enough wattage?


Fran

[Tea-Bag]

I have shorted a Mosfet before. The resistors started smoking, really glowing on that get, then it moved to the others..Not pretty.

You can test the mosfet by putting over 4 volts across it, with a larger source resistor. Look for some links on testing mosfets.

As far as shorting with the heatsink filings, that is more common, or at least shorting the Fets to the heatsink in the first place. What I do is not connect the Common grounds when first mounted to the chassis. If the metal tab on the Fet gets continuity to ground, don't power it up. Good Luck. Although this problem really should go to the primary F5 thread...

[hags]

I bulged/blistered and smoked a .47R Panasonic 3W monkeying around with the bias.

I shut it down, measured resistances and powered up again, all apparently with no damage to anything.

I'm using Toshiba 2SK1529/2SJ200 MOSFETs.

Excellent quality and just the right quantity of low frequency info. Clear, smooth highs, not bright or in your face. For example, there is great stick definition on ride cymbals with the corresponding wash sounding tonally perfect. I can almost tell whether or not the sticks have nylon or wooden tips.

[Renron]

Hey guys, Sorry to ask this here but it is relevant to the sonic output of the F5.
I have searched, and read, and read, and read, but I'm not finding what I was hoping for.,

Which resistors values were changed to "upgrade" the sound? I think Papa wrote about it but for the life of me I can't find it!

When speaking of changing R5&8 to 100 Ohms is that 2 / 100Ohms in ?? or 100 ohms in total?

Thanks for the help.

Ron

[alazira]

Originally Posted by Renron

Hey guys, Sorry to ask this here but it is relevant to the sonic output of the F5.
I have searched, and read, and read, and read, but I'm not finding what I was hoping for.,

Which resistors values were changed to "upgrade" the sound? I think Papa wrote about it but for the life of me I can't find it!

When speaking of changing R5&8 to 100 Ohms is that 2 / 100Ohms in ?? or 100 ohms in total?

Thanks for the help.

Ron

Hi Ron,

R5&R8 are feedback resistors. Standard configuration is 50ohm for R5&R8 (leaving out R6&R7). Changing the values to say 100R will certainly change the sound. Can't say "upgrade" as it becomes a matter of personal taste. If you do change it, try to keep the ratio of R5:R1 approximately 5:1 (i.e. if R5=100R then R1=20R).

[Renron]

Alazira,
Right, Modification not upgrade.........too much time on the inside of the computer lately.
thanks, that's exactly what I wanted to know!
I'm running a 50R 5W wire wound in R5 & R8 so 1:5 ratio is 10R for R1 & R2.

thanks very much.
Ron

[ichiban]

Originally Posted by Renron

Hey guys, Sorry to ask this here but it is relevant to the sonic output of the F5.
I have searched, and read, and read, and read, but I'm not finding what I was hoping for.,
Ron

Ron, these may help, from my notes............

Post #3731

Originally posted by JC951t
I know I'll get shot for this. If sound is too clinical or clean, stay away from metal film resistors.
=================================================
NP > Alternatively, you can increase the value of the Source resistors to 1 ohms,
and increase the value of the feedback resistors from 2 X 100 ohms and 10
ohms to 1 X 100 ohms and 22 ohms.

The lower amount of feedback will warm the sound up, and you can keep
increasing proportional values in the feedback loop if you wish.
=================================================
rob lenk f5 feedback mod per NP Post #3843

Just thought I'd post my initial impressions of having modified the feedback network relative to Nelson's suggestion( R11 & 12 to 1 ohm/
R5 & R8 to 100 ohms and R1 & R2 to 22ohms-pass site)
Yep,Yep amp runs a little cooler(perhaps 8-10 degree farenheit) and the "tight fist" now has a velvet glove!
I found the original spec a bit dry and etched in sound relative to my dynaco st70- accurate and detailed but lacking a bit of body or warmth. The new feedback seems to be a definite step in the right direction.

=================================================
Pillendreher post#4215 page 141

Hello there,
i also tried the alternative feedback a´la:
Alternatively, you can increase the value of the Source resistors to 1 ohms, and increase the value of the feedback resistors from 2 X 100 ohms and 10 ohms to 1 X 100 ohms and 22 ohms.
But now i cannot set the bias to more than 1.1A! If i set the dc offset to 0 the bias levels at 0.9A.
Is it correct to increase the value of P1/P2 to 10k?

bobodioulasso
It should be simpler to increase R3 and R4 to 2K7 or 3k3

Pillendreher
You are right....Thank you!

f5 problem with amp draw Post #4218
Pill
I had a similar problem at one point in the mod process(ie could not bring the bias up on one channel) and found a cold solder on a ground lead to be the problem.
I am currently(no pun intended) running my f5 with the revised feedback and no current limiting or thermister and have no problems setting bias- I did make sure to set the bias around .4vdc until the amp had warmed up for about an hour then reset to .59vdc and watched and micro tweaked for the next hour.
with my heatsinks and fan arraingment my usual ambient temp directly above the mosfets is about 20 degree ferinheit above room temp.
Hope this helps
rob
Pillendreher [Find more posts by Pillendreher]
diyAudio Member

Registered: Jan 2007Location: RTGermany Post #4219
Rob, do you mean you run it with 1R source resistors and set the voltage to 0.59V?
=================================================
Regarding the 10 ohm resistors that sit between the input jfets.

I have observed that if the the source resistor of 2sk170 is slightly higher than the drain resistor of 2sj74 (eg 10.2ohms and 9.8ohms respectively) then the 2nd harmonic content increases.
If the source resistor of 2sk170 is slightly lower than the drain resistor of 2sj74 then the 3rd harnonic content increases (eg 9.8ohms and 10.2ohms respectively).
=================================================

[osscar]

Another option is to change the limiter resistors R21 and R22 - 22K instead of 10K. At least the version proposed by Papa. I did it. I can not compare the sound with 10K because i stayed at 22K from beginning.

[ichiban]

Originally Posted by ichiban

Ron, these may help, from my notes............

The post#'s referrenced are to thread
F5 power amplifier
the F5 power amplifier thread