I was able to use my lab-supply as PSU for the VFET amp by using some mini-clips just for a test at 30 VDC. The lab-supply is a regulated linear supply. Using this supply both channels seems to behave well. Therefor I think the stange noise I see/hear comes from the SMPS. Why it seems only to be audile/visible in R-channel I can't explain but SMPS has some very high frequency noise (in MHz area....according to scope) and that it also makes some strange random spikes ("mV area").
The lab-supply has some 50 Hz ripple (20 mSec period) and this ripple is very visible at VFET output (with grounded input). I think this is expected for a SE-output stage. Before exchanging any components I need to try out a "clean" PSU. I wonder if my SMPS is inside specs. I could get another SMPS and see if this will fix it.
The lab-supply has some 50 Hz ripple (20 mSec period) and this ripple is very visible at VFET output (with grounded input). I think this is expected for a SE-output stage. Before exchanging any components I need to try out a "clean" PSU. I wonder if my SMPS is inside specs. I could get another SMPS and see if this will fix it.
I ordered a new 36 VDC PSU for the VFET amp:
CONNEX SMPS2000RS Switching Mode Power Supply Module 2000W +36V - Audiophonics
I expect it to be reasonable quiet when it is only loaded about 5% of max. power. Then we will see if it makes a difference. Under all circumstances it should be an upgrade from the small Meanwell. I need to find a chassis for it and install some fuses to avoid the large amount of energy if something goes wrong.
CONNEX SMPS2000RS Switching Mode Power Supply Module 2000W +36V - Audiophonics
I expect it to be reasonable quiet when it is only loaded about 5% of max. power. Then we will see if it makes a difference. Under all circumstances it should be an upgrade from the small Meanwell. I need to find a chassis for it and install some fuses to avoid the large amount of energy if something goes wrong.
I suppose if one was serious in the market for a new power supply, you can't go wrong by spending more than the rest of the amplifier components. 
That's a bold move and as mentioned before, there are ways to narrow down this minimal noise problem. But at over €200? What if this does not solve the problem?
If it does not solve the problem I have a more quiet PSU when problem is solved. It seems that with lap-supply which is a cheap one from China (2 x 30 VDC, 5A with current limiter etc.) problem seems solved. But this supply has some ripple even that it is regulated so not much sense using it for a test with music. But on scope I did not see anything unusual other than ripple at amp speaker out with grounded input. So now I go for a PSU with almost no ripple and very little noise.
I have ordered a chassis for it and I got the special 4-pole connectors so I can make a wire from PSU to amp. These male connectors has really many parts....8 parts. I need an instruction manual to assemble it......
I have ordered a chassis for it and I got the special 4-pole connectors so I can make a wire from PSU to amp. These male connectors has really many parts....8 parts. I need an instruction manual to assemble it......
Pass DIY Addict
Joined 2000
Paid Member
OCD = Obsessive-Compulsive Disorder?
I think this is "spot-on"
But I don't feel alone with this in this forum
I'm a bit late to this part of the conversation, but I thought you guys might appreciate the T-shirt that my wife got for me a while ago...
Attachments
No, not sure about this but want to give the VFET a "last chance".
I did not see anything unusual using lab-supply and I did see some strange noise on Meanwell supply so new supply will be an upgrade to the amp.
If one VFET is noisy I must decide to either live with it, try to get a new pair (is that possible?) or "upgrade" to THF-51 (I got some of these). There will be such an output stage in the future for this amp? .....also running at 36 VDC?
I did not see anything unusual using lab-supply and I did see some strange noise on Meanwell supply so new supply will be an upgrade to the amp.
If one VFET is noisy I must decide to either live with it, try to get a new pair (is that possible?) or "upgrade" to THF-51 (I got some of these). There will be such an output stage in the future for this amp? .....also running at 36 VDC?
It has no. 33.....same no. as Max Verstappen in Formula 1
I assume/hope the new SMPS PSU will solve the problem. Until then I have the fun "building" the PSU. I may add small bank of caps if space inside the chassis allows it. The quite powerful PSU can probably handle a "few" uF at the output.
I assume/hope the new SMPS PSU will solve the problem. Until then I have the fun "building" the PSU. I may add small bank of caps if space inside the chassis allows it. The quite powerful PSU can probably handle a "few" uF at the output.
I did not see the "request for news" until now where there is a little bit of "news".
The new SMPS module arrived after two months last week and I then built the module into a chassis and added a couple of filter caps at the output (4-poles) as you can see. It reduced the ripple noise a bit (from 9 to 6 mV when just measured using a DMM).
Also added an output fuse (4A F) and in IEC inlet there is a 1A T fuse to avoid any disaster from the 2 kW module. Without additional heatsinking it run about 62 C where it is hottest (when loading amp (3.2A)). The thermal fuse kicks-in at about 90C according to spec. So I think it is ok to let it run "as is". When chassis is closed it will probably be a little "hotter". I guess it is ok as I measured the temp direct at component side. Heatsink is a bit cooler. Module has soft start both at input and output and has no problems with the filter caps.
But noise problem is still there. I get these "noise bursts" from time to time (20 mV when it peaks) at speaker out. Other times it is 0.0 mV and can be that for at long time. Other channel is always 0.0 mV AC.
I made the big decision to de-solder VFETs and swap them. Until now only de-soldered and as you can see three solder pads also came off. But I assume these are "stand-alone" and it will only make it easier to de-solder next time ...still one pad in place....
I need to clean the TO-3 PINs and insert again. I have marked the "suspicious" one
The new SMPS module arrived after two months last week and I then built the module into a chassis and added a couple of filter caps at the output (4-poles) as you can see. It reduced the ripple noise a bit (from 9 to 6 mV when just measured using a DMM).
Also added an output fuse (4A F) and in IEC inlet there is a 1A T fuse to avoid any disaster from the 2 kW module. Without additional heatsinking it run about 62 C where it is hottest (when loading amp (3.2A)). The thermal fuse kicks-in at about 90C according to spec. So I think it is ok to let it run "as is". When chassis is closed it will probably be a little "hotter". I guess it is ok as I measured the temp direct at component side. Heatsink is a bit cooler. Module has soft start both at input and output and has no problems with the filter caps.
But noise problem is still there. I get these "noise bursts" from time to time (20 mV when it peaks) at speaker out. Other times it is 0.0 mV and can be that for at long time. Other channel is always 0.0 mV AC.
I made the big decision to de-solder VFETs and swap them. Until now only de-soldered and as you can see three solder pads also came off. But I assume these are "stand-alone" and it will only make it easier to de-solder next time
...still one pad in place....I need to clean the TO-3 PINs and insert again. I have marked the "suspicious" one
Attachments
@ Meper
Such courage!
Fantastic you don't give up and make your all possible efforts.
Good luck for perfectly perfect amplifierCheers
Yes, let us see if I end up with a perfect amp
You often show pictures of boxes full of these devices?
Do you need them all?