The .22 value comes from a conversation with Wayne Colburn of Pass Labs who said that it was the sweet spot -- just one .22 across the AC worked as good or better than individual caps across each diode.
Regardless, using the .1 across the AC is still going help. 🙂 Go ahead and use them, I like using what's on hand as well!
If you have a scope, clip it in and take a look. If you don't have a scope, I highly suggest getting one. It's awfully useful for this hobby.
Regardless, using the .1 across the AC is still going help. 🙂 Go ahead and use them, I like using what's on hand as well!
If you have a scope, clip it in and take a look. If you don't have a scope, I highly suggest getting one. It's awfully useful for this hobby.
I actually did pick up a Rigol scope a little while ago. Last used it when I was tweaking my F5 so this will be a good excuse to fire it up. Umm... what exactly would I be looking for?
I'll try the .1, then 1.0, and whatever else I might be able to find in that range. Try to find my own sweet spot. 😉
I'll try the .1, then 1.0, and whatever else I might be able to find in that range. Try to find my own sweet spot. 😉
I think you can get too big... Most snubbers you see out in the world are 0.1uf or thereabouts. 1.0uf seems big, glancing through forum discussions on the topic.
Interesting read: http://www.hagtech.com/pdf/snubber.pdf
From here: What are the advantages of full wave bridge rectifier with capacitors parallel to diodes? - Electrical Engineering
And here: http://www.diyaudio.com/forums/solid-state/134243-what-does-01-cap-bridge-rectifiers-do.html
I think I have some smaller Kemet caps used as in the last link on my beta24 (of which I still have to trouble shoot one channel). Might have those lying around.
To answer my own question above, I assume I'm just looking for amount of ac that gets through the psu while the amp is warmed up and playing music?
From here: What are the advantages of full wave bridge rectifier with capacitors parallel to diodes? - Electrical Engineering
And here: http://www.diyaudio.com/forums/solid-state/134243-what-does-01-cap-bridge-rectifiers-do.html
I think I have some smaller Kemet caps used as in the last link on my beta24 (of which I still have to trouble shoot one channel). Might have those lying around.
To answer my own question above, I assume I'm just looking for amount of ac that gets through the psu while the amp is warmed up and playing music?
no, its not a less than critical part, wrong snubber value on a fast rectifier part could easily mean resonance and resonance on the supply rails can mean oscillation
Fair enough! I was just thinking that getting in the ballpark would be beneficial even if not totally optimized. And since I can't measure stuff as in the article I guess I'll just have to try things out. Seems like I can probably start with values a bit less than 0.1uF and move up to from there to see if I can get at least a local optimum.
how critical it is will depend on the exact rectifier parts you are using, if you are using SIC zero recovery parts then it becomes quite important and without measurement I wouldnt recommend just trying stuff out. the AX can be a bit touchy already really cant it? I would say anything between 10-100nF would be a good place to start, 1uf is too large IMO
On my beta24, I have four caps from the ac to dc terminals as opposed to one across the ac terminals. Any benefit to that configuration?
Found 0.047, 0.1, and 0.22 in the parts bin. But I have to place an order for the rectifiers anyway so I guess I might as well get some others too.
She lives! I get about 21VDC under load measuring 18mv of AC ripple with my fluke 45 bench top DMM. I started with the 0.1uF film caps across the rectifier AC tabs. Will try other values once my other rectifiers come in.
Not sure if its working correctly yet but I do get music out the end. I might try eliminating the input caps as well. Need to measure my preamp to see if there's DC there.
Not sure if its working correctly yet but I do get music out the end. I might try eliminating the input caps as well. Need to measure my preamp to see if there's DC there.
I've noticed the heat from the output MOSFETs is much different from the current source mosfets. Is that normal?
must be the amp... My noticing is flawless. 😛
I thought I measured the bias across the resistors to be the same across all of them... but I could be mistaken. What else can I measure to help trouble shoot?
I thought I measured the bias across the resistors to be the same across all of them... but I could be mistaken. What else can I measure to help trouble shoot?
Last edited:
with signal , or without ?
what's physical layout , meaning - how you grouped mosfet banks ?
maybe hotter ones are just on naturally hotter area of hsink
dunno , without pics
what's physical layout , meaning - how you grouped mosfet banks ?
maybe hotter ones are just on naturally hotter area of hsink
dunno , without pics
Physical layout is the same across all four banks. Each bank has four mosfets each. They are on their own heat sink (you can see from post #112 above that the four heat sinks are 2x2 front to back.)
This is likely with signal and without signal. I noticed it initially without signal but it's still happened with signal. Although I didn't let it go very long.
I'll take some pics and some more thorough measurements tonight.
This is likely with signal and without signal. I noticed it initially without signal but it's still happened with signal. Although I didn't let it go very long.
I'll take some pics and some more thorough measurements tonight.
Well, the pics aren't going to do much good. They all look exactly the same... 4 MOSFETs per heat sink. I measured all 16 resistors (0.47R) and they are all at 450-460mV and thus, biased at about 1A. Does that not indicate they are all biased the same? Or am I missing something entirely?
Also, for what it's worth, I'm drawing about 1.75V across each of my 1R resistors in my CRC (5 per rail) so if V=IR, I'm drawing about 8.75A per rail which seems about right.
And still, the heat across the CS and output MOSFETs is significantly different.
It plays music... but not that well based on my ears but it's only one speaker so a bit hard to tell for sure. My guess is that something is wrong, but there's also the 4.7uF input caps that may be causing an issue. I'm going to try bypassing them to see if there's a difference... my source doesn't have more than a couple hundredths of a mV of DC so I'm not worried at all.
Also, for what it's worth, I'm drawing about 1.75V across each of my 1R resistors in my CRC (5 per rail) so if V=IR, I'm drawing about 8.75A per rail which seems about right.
And still, the heat across the CS and output MOSFETs is significantly different.
It plays music... but not that well based on my ears but it's only one speaker so a bit hard to tell for sure. My guess is that something is wrong, but there's also the 4.7uF input caps that may be causing an issue. I'm going to try bypassing them to see if there's a difference... my source doesn't have more than a couple hundredths of a mV of DC so I'm not worried at all.
ok - I presume that DC offset between outputs ( + and - ) is near zero ;
what's offset between outputs and gnd ?
what's offset between outputs and gnd ?
Yes, near zero. Output to ground was about 10V at startup but I haven't let it warm up fully before taking measurement. I'll try that today and also play with the vr2.
On a separate topic, is there a guide somewhere as to what conditions to do function generator testing? I hope to have one soon so just wanted to see what freq and amplitude scenarios I should plan on measuring with the scope.
On a separate topic, is there a guide somewhere as to what conditions to do function generator testing? I hope to have one soon so just wanted to see what freq and amplitude scenarios I should plan on measuring with the scope.
- Status
- Not open for further replies.