F4 in Burning Amp Chassis – Wondering What To Do Next?


2006-12-13 1:51 am
Hi Everyone,

At long last I got around but finished the build of an F4 in one of my Burning Amp chassis. I actually started right after Nelson gave them away in, what was that, 2008? Kinda slow, huh? As it stands now, it works but I’m at a loss on where to take it from here. Here’s the story…

I had decided to build an F4 variant in a 4-deep, IRF240/9240 Burning Amp Chassis. I built it as follows:

1. Dual mono power supply.
  • a. 2 Avel 18+18V 250VA transformers
  • b. 2 pairs of Peter Daniel’s Universal PS boards (CRC)
  • c. 44,000 µF/rail, 176,000 µF total (8 22,000µF Panasonic TS-HA)
  • d. 4 discrete diode rectifiers, 16A/600V diodes
2. Peter Daniel’s F4 clone boards (using driver portion and wiring to MOSFET panels.
3. Replaced Source resistors from 1.8Ω to 0.47Ω.
4. Replaced Gate resistors from 90.9Ω to 150Ω.

As you can see from the pix, the driver boards are mounted on a cool-dude piece of aluminum-backed phenolic that’s attached to cooler-dude transformer pedestal – all courtesy Bill Berndt. The wire routing is dictated by the limited space in the short chassis. Despite the close quarters, the amp is extremely quiet with the only noise being a little bit of mechanical hum from the trafos. The sound is amazingly detailed with good imaging and what I would call a gigantic sound stage. You can see it in action installed in Bill’s rig in the pic.

What’s missing is that the bass is pretty flimsy, lacking punch, and the amp feels like it gets bogged down in complex, high-volume passages. Overall, it seems slow. it makes the me think that the PS isn't up to par. Also, even with the volume control near to the top of the range, it doesn’t get very loud. I’m not sure what the gains are for Bill’s or my test preamps are but I guess they just don’t generate enough voltage swing.

After things get warmed up, it’s getting about +/- 24V on the rails and the deepest bias I can get without burning my fingers with the 5-second rule is about ~490mA.

So I’m wondering what I might be able to do about the sound. Or is this just the way this amp sounds? That seems unlikely but what to do?

Any help would be appreciated.



  • IMAG0180.jpg
    431.8 KB · Views: 455
  • IMG_0308.jpg
    554.5 KB · Views: 434
  • IMG_0312.jpg
    724.2 KB · Views: 425
  • IMG_0313.jpg
    620.7 KB · Views: 396
  • IMG_0314.jpg
    662.1 KB · Views: 380
Fwiw, the speakers we are driving are the audio kinesis jazz modules, that I brought to BA this year. We are comparing Kurts f4 build to my 6v6 pp amp that is my daily driver. The 6v6 amp is less than 10w. It has better control in the bass region. The f4 has a wider deeper more resolving image and would wipe the floor with the tube amp except for the timid bass presentation. The speakers are 93db efficient and have an impedance balancing XO that rises to 16 ohm in the lower registers, being designed for atmasphere otls.
- I think you have less than optimal capacity in PSU , considering 4 deep meaning 490mA x 4 Iq ?

- are those signal electrolyts bypassed with anything ?

- are these PSU caps bypassed with - say- solid 4u7 ?

- with F4 you really need proper pre

try - even if just for laugh - your proper tube amp , loaded with 8-16R dummy load - as pre for F4 ; I can bet that - if tube amp have proper bass - F4 ( if properly made ) will not ruin that

then you'll have point from which you can build further


2006-12-13 1:51 am
And who could blame him, Mr. Mod?!?!

Nelson, yes, that's ~230mV across each 0.47Ω resistor.

ZM, I was thinking the same thing about the capacity of the PS. The PS has two 22,000 µF filter caps separated by 4 0.47Ω resistors per rail (dual mono style). The second filter cap is bypassed by a 10 µF Panasonic film cap. What is confusing to me is that the PS parts list for the F4 clone that I've found here on diyaudio is 8x 15,000µF caps for 12 MOSFETS v 8x 22,000µF caps for 12 MOSFETS that I have. I thought that should put me in the zone though they are configured differently.

The signal electrolytics are Silmic IIs and are not bypassed.

Roger on the proper preamp thing. But it seems like the base shouldn't be weak at lower volumes, right? Also, we did try to 6V6 as a in put to the F4. Unfortunately, it the tube amp motorboated so we stopped the proceedings.



sorry - blame my eyes , or blame my speed :rofl: - I didn't saw that you have two levels of caps ;)

that's fine , certainly ;

you didn't wrote what preamps you tried as source for F4 ;

in occasion when you tried tube amp as feed for booster F4 amp - what were dummy load resistors for tube amp ?

regarding low level bass quality - take this as comparison - this test is same as you are trying various headphone amps - doesn't matter what level is - bad headamp always have mediocre sound ; good one always have good sound - in whisper and in thunder state .

in any case - my crystal ball is somewhat muddy lately , so I'm operating strictly with data you laid here . I'm resuming that your F4 is made by the book ; what's current through input Jfets (voltage across R3 and R4 ) ?


2006-12-13 1:51 am
Yes, the F4 is by the book with the only differences being that I have 4 output devices per side instead of the original 3 and the PS is different as described above.

The voltages for R3 and R4 (10Ω resistors) are as follows:

Right Channel:
R3: 7.6mA​
R4: 7.6mA​
Left Channel:
R3: 6.7mA​
R4: 6.7mA​

For preamps, my test preamp is an old Audio Alchemy DLC with a fancy PS. The specs for its gain are enormous. Bill's pre is a deluxe DIY EF804 (or maybe EF86) preamp. The gain hasn't been measured to my knowledge but my calculated gain is 17dB. I do not, nor should anyone else, trust this number. He drives the 6V6 with it as noted above but has also successfully driven at least one other SS amp with it. Bill, please correct me as required. Bill's sounds better than the DLC, of course, but the DLC is ok. For each of them, bass not a problem with other equipment.

I'm saving the best for last...
We didn't add the dummy load on the 6V6 on the output. :eek: We'll run that one again soon.

Nelson Pass

The one and only
Paid Member
2001-03-29 12:38 am
A couple of thoughts,

1 490 mA is light on bias and will slightly alter the bass

2 If you are using it as a follower for a flea-watt tube amp
then some resistive loading for the tube amp is appropriate,
usually 8 to 22 ohms.

3 It is possible that the flea-watt's output impedance is
somewhere between 1 and 4 ohms, in which case you might
want to put some resistance in series with the output of the
F4 to duplicate that for your speaker.

A couple of thoughts,

1 490 mA is light on bias and will slightly alter the bass



considering this :


Nelson, yes, that's ~230mV across each 0.47Ω resistor.


it seems that overall bias is (0,23/0,47)x 4 = 1,96A

considering that F4 "by the book" have approx. same bias per pair ....

Output device bias is approximately .53 amps per device, which measures 0.25
volts across the .47 ohm source resistors.

dunno ..... ZM naively thinks that additional (20mV/0R47)x 4 , will not bring Nirvana ...

so - it seems that - either OP made some systematic mistake in both channels and loose bass control ...... or amps are just fine - just like you said , and he prefer fluffy tube bass
(not saying that tube amps have fluffy bass , just that his tube amp possibly have more fluffy bass than F4 ) :rofl:


2006-12-13 1:51 am
Thanks for the thoughts guys. Much appreciated!

ZM I'm with you in thinking I must have made some sort of mistake in the driver section. Where the current values for R3 and R4 within range? Are there other references that I could check? I'll also do a visual check of all of the components of course.

If I haven't made a mistake ( fingers crossed that i have!) and bias can indeed deliver nirvana, I would have to find a way to dissipate the heat since the current chassis max'd. That makes me wonder if I should consider a BA-2.

Very interested in answers to ZM's questions.

Thanks again guys!

Kurt, I have the same situation with my F-4's but I am quite sure that it is because I calc'ed my bass box to be highly damped and then used a 15" woofer that was very highly damped also. I ended up with everything sounding fantastic, but the opposite of boomy bass: incredibly defined, but not as much, not as present...

Trouble is I have gotten used to it and might suffer with less (dare I say it) tight bass..
but I am planning to experiment with other cabinet options. Someday soon...

Just interesting that we seem to have the same situation but maybe for different reasons.


2006-12-13 1:51 am
I'm glad you weighed in, Variac. I was curious about your setup.

What you describe sounds vaguely familiar. Have you tried your F4 with other speakers? What preamp are you using? You did a standard 3-deep F4 right? What bias are you able to run? While I'm at it, what is your PS like?

I am trying to get some time this week for a deep dive on the driver boards to see if I've made any mistakes in construction. Then want to try with Bill's 6V6 with load resistors this time. Any maybe get out some fans and see what happens if I can managed the heat produced by deeper bias.
Hey Kurt,

As always, its a mistake to assume that I have great sophistication about things ampish. I'm more of a typical DIYer wandering in the woods, hoping for good results and getting lucky usually, thanks to the great designs that we have available from our advanced members!

However, to complicate things, my power supplies ( I have two 2 ch F-4's) aren't stock.

For each 2 ch amp I have one supply:
My transformers are supposedly 1KVA each at +-37v, although they don't seem quite that big physically. . I run them with the primaries set for 220v so they put out about +-18.5v,using 110v and I guess 500VA but with the more robust construction of a somewhat bigger transformer.

I have C-L-C on each rail. Each cap is 30,000uF, 63v. My inductors are I think about 5 mH or maybe more. They are coathanger thick copper wire on toroids and are each about the size of a 400 VA transformer.

So I have 30,000uF-inductor-30,000uF on each rail with a 3.5uF and a .1uF film cap in parallel with the final cap.

I have the stock 3+3 MOSFETS, 6 total each channel, and they are biased stock.

My speakers are Basszilla's BUT with a highly damped JBL 15" bass driver that was used in speakers sold in Asia, not the one Dick Olsher recommends. It sounded a bit thin so I messed with the crossover a bit.. and that helped a lot, but still, not a ton o' bass.

At Burning Amp One I used an amp by Wrenchone I think for bass, which I believe he said had almost no damping factor. Much bassier. Otherwise I don't have that much experience with lots of other gear. I biamped with a chip amp on my bass driver for a while and it was about the same, but a bit more robust I think. I'm now using SY's Impasse pre and it's great, but bass wise no real difference than running the amps straight out of my CD player-except a hell of a lot louder and more authoritative of course! While the Impasse is a tube pre, it doesn't have soft uncontrolled bass - SY insists on linear, low distortion designs..

So it seems that the F-4 has a better grip on the bass drivers than you might think based on typical Pass Labs damping Factor numbers. But Nelson has told me that he doesn't see a good correlation between DF and bass in his experience.

Certainly perceived DF is better than many tube amps. But again, I haven't played around with it much. It sure is to die for on the mids and treble, and I think that if I got different woofers and/or revised my box, would get a bassier sound.. However the bass definition is awesome...

Last edited:


2006-12-13 1:51 am
Testing Update

Hi again!

First of all, thanks for the rundown Variac. That's that's very helpful to know.

Bill and I did some testing last night over at his place with the 6V6 running into the F4'. This time we had an 11Ω load resistor across each 6V6 output. There was no motorboating this time.

There was also no bass. :sad: The positive characteristics of the F4' were still present but the bass was weak. Also, the sound had an overall distorted character.

We ran it again for a while with the pre directly into the F4' and concluded that we liked the sound of this setup better.

I'd love to hear what you guys think I might want try next but I was thinking of trying to get another few mV's into the bias without cooking anything. Then I was thinking I might try the 33,000 µF cans that were originally in the two BA chassis across final C in the CRC PS chain for each rail. If you guys think that's a waste of time, please let me know.

Any and ideas welcome.