VFET amp Bourns trimmer exchange

Official Court Jester
Joined 2003
Paid Member
process is simple - swap VFets between channels, re-set drain voltage

if problem moves, following now suspicious VFet, you know the culprit
if problem stays, change opto and all thingies on problematic pcb

you'll solve it , one way or another, only working on amp, not here in thread

If I know that it would help, I'll write (it'll be gone in 36hrs"

but I don't know that :)

so - just dismantle it .......
 
Swapping VFETs is a major task. Opto and or zeners a minor task.
Think I will need a de-solder station to swap VFETs to be sure not to damage them (could be a good excuse to buy a de-solder station.....but which?).

But these spikes I see......if I measure on VFET gate.....and spikes are still there.....or not there.....could that make a conclusion? .....or is there another "secret" place to measure to be able to conclude something? .....measurements so far are all on speaker out.
 
Prob that comes and goes... hmm, hot weather currently, could it be temp related?

Thinking of an average connection, or not perfect solder joint that seems to ensure continuity "most of the time" but not really properly and when conditions change the noise comes and goes? Jumps are quite odd otherwise...

Just thinking out loud, having had a similar intermittent small failure, in from of random drops "sometimes", on a DVD player last week... your case could be of course very different

Good luck either way

Claude
 
I have ruled out bad solder joints. It seems not to be sensitive to mechanical knocks. What I don't understand is the quite low frequency content of the noise. The largest spikes can be +-8 to -+ 10 mV pp. There can be periods of several seconds where there is silence. Then some seconds with many spikes.

I can get some liquid nitrogene and cool VFET down to -196C and see what happens.....and give "him" a shock..... :) .....it may reveal the bad component....something I would never do with tubes......
 
No, I did not try that but I remember I did consider to be brutal and short the Edcor output. But I could not "hear" any noise at the FE-output / VFET board input when using my special amplifying device (battery driven amp with built-in speaker). After this I concluded that it was not the FE-board. But scope will show what happens at a more "technical" level......and it will hopeful add something to my "education".....
 
I have performed some additional scope measurements inside the amp.
I found that -D input (output from FE-board) is "clean" but both sides of R4 (2SJ28 gate) is noisy (same type of "spikes" as on speaker out).

Then C2+ side of R7 is noisy but other side of R7 is "clean".

Probably difficult to conclude much......other than maybe VFET is ok?
If gate change then ok that output change......the question is who comes first?
 
Next will probably be 4N35. I have the Toshiba variant (white plastic) which I used for M2X. But I better get the LiteOn which are used for VFET.
I noticed that pin3 (NC) and pin6 (Base of photo transistor) was not soldered and I must admit I did solder these when I mounted all the other components. I guess it is ok. Base is just floating....also on PCB of course....even if soldered. But maybe better to cut pin6 close to house so there is no chance it can pick up noise?
 
LiteOn 4N35 ordered. A tube of 65. They were quite cheap. Seems these are used a lot in Pass DIY. It seems Base (pin6) can be used to adjust sensitivity of transistor by applying a quite large resistor to Gnd and then a small parallel cap to suppress noise. But we let it float.....
18V Zeners in DO-35 also ordered just in case.....OnSemi types which had the best max. ratings (-65 - 200 C). 200 pcs.....also very cheap.

I looked at thermal pads for TO-3 but they were at back-order....so they will wait (some silicon glass fiber types which should be difficult to penetrate).
 
VFET amp 4N35 discussion

While I wait for new 4N35 IC's I looked at the "floating" base of the photo-transistor. I wondered if the "noise issue" could be caused by the floating base if power supply noise could trigger the noise pulses I see.

I looked at these threads:

opto isolator - Optocoupler with phototransistor base lead - Electrical Engineering Stack Exchange

Noise Phenomena of Photocouplers / Optocouplers | Renesas

It seems it is good practice not to have the base floating. Noise could go via PSU and then via the "invisible" Ccb (capacitance between C and B of photo-transistor) and trigger a noise pulse (however the 1000uF cap should filter such spikes)?

My idea is to solder a 100nF ceramic cap between B and E and check if it makes a difference or not......before I put in another 4N35....as next step.
 
I have found a new path I need to investigate. I noticed something that looks like "self-oscillation". Signal overlaid by a very high frequency signal in MHz area but only for a couple of seconds then clean signal for a couple of seconds and so on. I see this on both channels but only right channel has this "jumping" up and down few mV randomly. When I measure on PSU (disconnected from amp) I see the same on/off HF-noise. It is like a clock....approx. 2 sec on and then 2 sec off. I have ground from probe at the shell of connector and I see the on/off HF signal on each pin. Then I also see on two of the pins the "jumping" up and down randomly (I measure AC). So it some fast DC-shift (pulses) in mV area. What I don't understand is why the jumping up and down is only visible on R-channel and not on L-channel......very strange. PSU is common for both channel.
Will try.....when I get time......a lab supply if I can find a way to connect it. Not easy with crocodile clips. Also need to be sure I don't reverse the voltage.....
Maybe these SMPS just performs like this and there is nothing wrong.....and R-Channel is more sensitive to this than the other.....

Was there a part no. for the male PSU connector?
 
Now I have ordered some male connectors at Mouser.
Then it should be possible to connect another PSU (linear) quite easy.....for test purpose.
If it is the the PSU that makes this......I will see how far I can throw a Meanwell PSU. But first I will cut the cable to I can use it for another PSU.
 
I'm still saying that only way is changing VFets between channels, or to change everything on pcb but VFet

49 posts, and you practically changed just trimpot

Yeh, I completely agree. Man up and pull the VFets out and test them.

That's what I did with my Sony VFets in my PP CSX1 - the utmost importance of testing the most reactive component in the whole design. The transistor itself. The job of resistance is too simplistic that we're splitting hairs on how good resistance can be.

Did I say there's too much resistance? :rofl:

I'm afraid to say, the lottery SE VFet amp sits under covers. My ears much prefer the sound of the VFets in constant source PP with balanced input tranny. It's cleaner, more character, more exciting, more bass, i'm sorry to say. :confused: