Initial setting of F5 bias--help, please!

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Member
Joined 2019
Paid Member
I'd recommend reading 6L6's build guide - the functions of all three pots and how to set them prior to the first power-up will save you some time and effort. It's all in the first page or two. If you don't set the bias pot properly prior to power up... you could blow another set of FETs.

Edited to add... and what ZM said :D
 
I have built a F5 turbo for another with purchases from the Diyaudio shop. Higher quality components and matched jFETs We’re used. But I can’t get the amp brds to bias.

Using a variac with only 40VAC going into the amp for 2.65VDC on the brds the 0.47 resistors have 1V across them. It doesn’t matter which of the two trim pots (10 turn) are on full anti-clock or fully clockwise turned in any or all combinations the bias doesn’t change. Also there is 0.33V DC offset stress the speaker posts. The resistors are slightly warm.

I have checked all components resistors and jFETs a number of times and all appear to be correct. At assembly each component was quad checked against the schematic at assembly.

Any clues why I cannot bias off the power FETs. They appear to be fully on and conducting? Anyone else experienced this?
 
See above for the voltages DC and AC.
 

Attachments

  • DB7E93CE-06CB-42A2-8C7B-E05BE4FD3877.jpg
    DB7E93CE-06CB-42A2-8C7B-E05BE4FD3877.jpg
    863 KB · Views: 314
  • 6F45CB6D-9D49-41AA-86A6-460C4B60C6CE.jpg
    6F45CB6D-9D49-41AA-86A6-460C4B60C6CE.jpg
    701.3 KB · Views: 309
Read above to see my problems with my F5 turbo. I sent the brds. off to an interstate friend who has built many more class A SS amps than me. Also he has a pro. desoldering station. He removed the driver trannies and deemed them the right ones in the right place. But when he pulled the power FETs he said they were the wrong ones. I checked them at least a dozen times and would bet my life on the fact I had them right.

He claims according to the markings you would think they were correct and in the right place. Only when placed on a smart meter did it show I had them the wrong way round. Brds. have been re-assembled and he is going to test them over the weekend.

I’ll update this thread when they return.
 
JFET failures?

I too decided to build an F5 as a pandemic project. I started firing it up today, one channel only. I believe it had been powered for a few minutes without issues and the bias pots were in the fully shorted configuration. As I started adding bias to one of the MOSFETs I caught a slight odor that I thought was that output device heating up. However, I soon saw fumes coming from the board. I think it was both JFETs and they're probably toast.

I verified the JFETs are in the right spots and polarity. Unless the gates are being driven, I can't figure out how they wouldn't self limit. The only way I can see that could possibly happen would be if the output DC, say, rises enough to force enough current through R4--with Q4 off, effectively driving Q2. Could this be happening enough to exceed Q2's SOA?

Alternately, the 18VAC secondary windings on my P.S. seem to be exceeding 25V, but only by a fraction of a volt. I understand the LSJ74 should tolerate this and the FET will see less once bias is set.

(Someone pointed out the input needs to be shorted, which it was, but I don't see how that should matter, as R1 and R2 provide path to ground.)

I'm ordering some small test clips to better monitor this stuff, but has anyone encountered this or know what's going on?

Thanks.
 
@ Aleph5,
please explain "the 18VAC secondary windings on my P.S. seem to be exceeding 25V".
do you really get 25VAC from xformer or do you mean 25VDC from unloaded psu?

I once fried a pair of jfets by accidently swapping positive and negative rails.

use a lightbulb in series for first power on.
double check for shorts.

good luck
A.
 
I have a quick question (I hope) about biasing...

I have both R/L channels fluctuating around 0mV and am currently targeting 550mV measured at the resistors. On the left channel, I have the pots set so the measured values are -549mV and 551mV, meaning there is roughly a 2mV spread measured at the resistors to achieve 0mV at the output. On the right channel, The pots are set so the measured values are -545mV and 555mV with 0mV, which is roughly a 10mV spread.

In both cases the average voltage is 550mV, and I may leave it at that bias setting. I don't imagine the variance in spreads is going to be too much of an issue, but the v2 guide says if you have decently specced parts (I used the F5 parts kit from DiyAudioStore), then the + and - voltages for each channel should be equal when measuring 0mV at the output.

So which components on the amp boards would account for the voltage spread seen on my R channel? The output resistors? FETs? Something else? And is it worth trying to do better than this by swapping something out?

Thanks in advance!
 
What about variation in the 0.47ohm resistors?

That was my initial guess since that's what I'm measuring the voltage access. It would be nice to be able to measure them without removing them from the circuit, but that's probably not going to be very accurate. Then again, maybe whatever I measure with them in circuit could be telling... I'll give it a shot tomorrow.
 
I measured the output resistors in circuit, which I know isn't very useful because other parts will be contributing to the measurement, but all 4 measured the same, fluctuating between 0.54 and 0.55.

I can try lifting one leg from each to get an exact measurement, but there would seem to be many other possible causes -- R5, 6, 15-18, 21, 22, etc. Also thermistors.

Is this worth chasing down or is it fine to have ~2% voltage difference in one channel between R7 and R8 when output is 0mV?

I have a quick question (I hope) about biasing...

I have both R/L channels fluctuating around 0mV and am currently targeting 550mV measured at the resistors. On the left channel, I have the pots set so the measured values are -549mV and 551mV, meaning there is roughly a 2mV spread measured at the resistors to achieve 0mV at the output. On the right channel, The pots are set so the measured values are -545mV and 555mV with 0mV, which is roughly a 10mV spread.

In both cases the average voltage is 550mV, and I may leave it at that bias setting. I don't imagine the variance in spreads is going to be too much of an issue, but the v2 guide says if you have decently specced parts (I used the F5 parts kit from DiyAudioStore), then the + and - voltages for each channel should be equal when measuring 0mV at the output.

So which components on the amp boards would account for the voltage spread seen on my R channel? The output resistors? FETs? Something else? And is it worth trying to do better than this by swapping something out?

Thanks in advance!
 
I personally would be ok with that difference.

Yeah, I think I will leave it alone. I managed to lift a leg on each of the four 0.47 3W (5%) resistors to check and they showed 0.53Ω, 0.54Ω, 0.55Ω and 0.56Ω isolated from the rest of the circuit. The two lower were on one channel and the two higher on the other, so it doesn't seem like these are the cause of the discrepancy anyway.
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.