Try without any bypass first as I say. This is a 2.5MHZ amp as I witnessed on George's scope. Small resonances as caps parallel can be more evident than in your fastest amp till now. To get a completely resonance free parallel system it takes 3 consecutive SMD caps damped with SMD resistors as cells with miniscule vias inductance tuned for specific electrolytic part number and fixed with lab gear by an expert.
Hi Salas,
Having some trouble visualising ur words . . . . any chance of a hand drawn diagram to clarify ?
Excellent research Salas - Well Done !
I have adopted a similar scheme in my supplies but using those nice small epcos open frame 4.7uF caps paralleled and bigger resistors.
My values were only tested in spice not in real life but subjectively seem to work very well. Based on your research presented here I think I could refine my approach.
I think using paralleled bigger 4.7uF caps brings some good benefits to supplies. With a total of 14.1uF I think the film caps begin to substantially smooth out the characteristic "sound" of the electrolytics and also the bigger caps need smaller resistors to achieve a smooth response.
I have adopted a similar scheme in my supplies but using those nice small epcos open frame 4.7uF caps paralleled and bigger resistors.
My values were only tested in spice not in real life but subjectively seem to work very well. Based on your research presented here I think I could refine my approach.
I think using paralleled bigger 4.7uF caps brings some good benefits to supplies. With a total of 14.1uF I think the film caps begin to substantially smooth out the characteristic "sound" of the electrolytics and also the bigger caps need smaller resistors to achieve a smooth response.
Last edited:
Although, it's my birthday, I can't say it's my lucky day. I can't power it up. I have an intuition it might be ground referencing that trips the soft start and is not delivering power.
I have a light bulb tester wired in series, soft start from Sonny, 2x300 VA transformers, double mono PSUs.
When not connecting PSU to the modules, everything works smooth. Correct Vout voltages on PSUs, etc. Problems start when I try to power up the amp modules. The light bulb flashes bright, soft start is not turning off right after turning up. same with separate channels.
I've made some pictures for the guys who used same arrangements, maybe they can spot something:
Full picture, sorry the light bulb not visible.
Heatsink mounted channel.
PSU Vout
Transformers secondaries connected to PSU.
Chassis and primaries connection hub.
Top pins connects yellow/green "0" wire from L&R transformers>Chassis>Mains Earth
Second pins from top connect one leg of primaries from L&R tranformers> to Soft start AC out
Third pins does the same with another leg of primaries.
I'm going to take a walk outside, just to releave the head on a beautiful day. Sonny could help the best, but he's a very busy man. Any suggestions welcomed in the meantime.
I have a light bulb tester wired in series, soft start from Sonny, 2x300 VA transformers, double mono PSUs.
When not connecting PSU to the modules, everything works smooth. Correct Vout voltages on PSUs, etc. Problems start when I try to power up the amp modules. The light bulb flashes bright, soft start is not turning off right after turning up. same with separate channels.
I've made some pictures for the guys who used same arrangements, maybe they can spot something:
Full picture, sorry the light bulb not visible.
Heatsink mounted channel.
PSU Vout
Transformers secondaries connected to PSU.
Chassis and primaries connection hub.
Top pins connects yellow/green "0" wire from L&R transformers>Chassis>Mains Earth
Second pins from top connect one leg of primaries from L&R tranformers> to Soft start AC out
Third pins does the same with another leg of primaries.
I'm going to take a walk outside, just to releave the head on a beautiful day. Sonny could help the best, but he's a very busy man. Any suggestions welcomed in the meantime.
Great call, as always, Salas. Removing bypass caps definitely opened things up. Piano has a more natural tone to it now. Little more body. May even argue that the depth of soundstage improved as well. Based on Lynn Olsens comments on his OTL horn rig, I can see how this would sound great on Vgeorges horn system. Pace and timing are wonderful. Will have to take it brothers house to see how it handles vinyl. I think I can say it is at a point where you could only improve with minor personal tweaks. I would add a little more 2nd harmonic to give it a little more height(bloom) in its presentation. Then again, on my upcoming horn system, this may not be a problem. Also have to listen with 6v6 pre, as this may open it up a little as well. Without a doubt, Sonny, it will be a permanent member of my audio family. Excellent job.
Will probably get Caddock TF for main through holes and Vishay TNPU or Susumu RG, both which are low noise, 5ppm SMD's.
What about 100nF on input and c12 and 13, on transistors.
Will probably get Caddock TF for main through holes and Vishay TNPU or Susumu RG, both which are low noise, 5ppm SMD's.
What about 100nF on input and c12 and 13, on transistors.
Attachments
Last edited:
I was not aware of that problem. I have used them in several amps. They are really thin to have as low W/K as possible. And for that reason the smallest alu grain will punch through them.
But i will check!
I have written to Tomas. I am busy but i try to answer all questions.
It could have been my implementation, but that was definitely it for me. With speakers against the wall, you get a taller presentation. Really wonderful amp. I personally don't think it needs more OLG or feedback. It is already a very clean and clear amp. Feels very fast. I think with right speaker combo, it could be a stopping point for anyone interested in HiFi. I am psyched about trying with tube pre. Should make for a great combo.
There is 100nF in series with a resistor from output to ground, that one is alright. For C12, C13 I had got Smica.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.