F5 power amplifier

Jump to Latest
jackinnj

jackinnj

Member
Joined 2002
tech-diy.com
sangram said:
@Jack: Thanks for that. I am trying to bias it near the 600mV point in the article but yes, it's a bit tricky. Do you think it'll help to go to the 50ppm Mills Audio resistors? they're spendy, but maybe I'll do it for my next build. My heatsinks are also just within the thermal spec - I'm probably doing 22-24 degrees above ambient.
Click to expand...

I wouldn't bother with the Mills.

Here's a graph of the impedance of a 100 ohm Panasonic:

An externally hosted image should be here but it was not working when we last tested it.


and a 50 ohm TO-220 Caddock:

An externally hosted image should be here but it was not working when we last tested it.
 
Sangram

Sangram

Moderator
Joined 2002
Paid Member
Jack: I have the other kit - no Panasonics, but the vertical 0.47 Ohmites and Dale 50 ohm 6 watt in the feedback position. I see very similar performance between the two resistors you're testing though, so I get your point. Would you have a similar graph for temperature testing, and would it be possible to see if the Ohmites are the issue?

As you can see from the pics, both of my power resistors *had* to be soldered to the back of the board, which now faces the heatsinks and gets the full monty from the sink's flat surface. Maybe that's the reason why, and I'll have to figure out a way of cooling those resistors down.

@Zen Mod: Yes, I twisted the wires together in the first pic, and I use a shorting plug for adjusting it once the jacks are soldered on...
 
M

massimo

Member
Joined 2002
sangram said:
I finally had to rip the FETs off the board because I could not stick to the original plan of opposed heatsinks. They are now mounted with stranded 15AWG copper wire for the Drain and Source, and single-strand wire for the Gate. So obviously it does not look as pretty anymore: (
Click to expand...

In this way the "grid stoppers", the 47 ohms resistors R13 and R14, are far away from the mosfets. They are supposed to prevent oscillation. You can try to desolder them from the pcb and connect directly to the gate pins.
 
Sangram

Sangram

Moderator
Joined 2002
Paid Member
OK, I'll try that - though desoldering them is impossible. I'll short them out on the PCB pads and wire in a 100ohm stopper directly to the gate.

I have to show off the amp to some friends on Saturday, so I'll get on this next week, looks like I got a lot to do.

Would it help if I bonded some heatsinks to R11/12? To keep them a little cooler?
 
M

massimo

Member
Joined 2002
sangram said:
OK, I'll try that - though desoldering them is impossible. I'll short them out on the PCB pads and wire in a 100ohm stopper directly to the gate.
Click to expand...
You can leave the 47 ohms resistors on the pcb and move the red cable to the R15+Q collector node (R14+Q6 on the other side) if you like. Then solder two 100 ohms in // (better) at the gates.

sangram said:
Would it help if I bonded some heatsinks to R11/12? To keep them a little cooler?
Click to expand...
Yes, try to find out some heatsinks with clip (generally for TO220) to avoid glue.
 
Last edited:
Sangram

Sangram

Moderator
Joined 2002
Paid Member
:|

I managed to get the amp thermally stable, and the offset is now predictable. Basically, I rewired everything using some Wonder Wire I had on hand, getting rid of the ugly flyouts. I added some heatsinks to the source resistors, and the feedback resistors bond to it as well - I use thermal tape - all resistors now run a bit cooler. Also put the thermistors directly on the MOSFETs, using thermal compound and hot glue. They are wired to the amp board with some flyout wire.

Basically, when I started up in this configuration, the bias dropped significantly, indicating the thermistors were working properly (lower resistance) due to the better thermal contact. The bias moves predictably and the offset is kind of stable - I think it will be OK after a couple of hours of running, in the second round of adjustments.

The only thing I haven't done is use additional stoppers - the 100ohm resistors I have on hand are of dubious quality sonically and in material and construction.

An externally hosted image should be here but it was not working when we last tested it.


Unfortunately, it sounds a lot worse now. :(

The highs have been pushed back a bit, the midrange is raspy and harsh - the bass is similar to what it was before, but the amp is now not very nice.

I suspect the amp is oscillating and needs stoppers immediately. However, I don't have any known quality components on hand. The other suspect is the wire used for the flyouts for the thermistors. I was using the same wire for the gate connections earlier, and noticed a touch of harshness, but I put it down to the revealing nature of the amp and harsh recordings. Maybe it's plated steel. No way to find out, so I'll try changing that out and see if it helps.
 
Last edited:
Sangram

Sangram

Moderator
Joined 2002
Paid Member
Unfortunately, it sounds a lot worse now.
Click to expand...

And we're back to proper operation, sort of.

I had to dispose of the thermistors, rewire everything with generic 1.5mm thick electrical copper wire, clean all the speaker connections (including the spade lugs), and the F5 sounds like it originally did when I built it, which is very nice. Offset is rock-solid, though without the thermistors warm-up is indeed longer. At least two hours till the bias gets to the ~600mV it is set at.

I tried some gate stoppers from 100ohm resistors I had laying around, and they only made matters worse, they were only 1/8W rated and were audibly distorting when the volume was turned up. I then rewired the gates with some of the generic copper, and the improvement was enough to make me consider rewiring the FETs.

The PCBs did not like so much repeated soldering and desoldering, so this pair of F5 PCBs is pretty much destroyed (though it works). It is not going to survive the transplant to a chassis. Oh well - live and learn, and I'm leaving it like this. For the next build, I'll try putting it in a chassis *after buying it*.
 
You must log in or register to reply here.
Top Bottom