Optimal supply design for UCD and Zappulse modules

Status
Not open for further replies.
Sorry to be a pain, but can you point me to a couple of posts where Bruno explains his reasoning behing this. The way I understand it, the twin/twin setup should make hum problems even less likely. I can see possibly how using monoblock type construction off a single transformer could invite hum, but not the twin/twin setup. Thanks
 
Hi,

He's stated it several times when asked about PSU options. I think you'll find it several times in the UCD180 thread (you understand right away my reluctance to have a scan for you).

I recall in one response, he brought up inviting hum as the reason, in others he simply said don't do it.

I know, it is contrary to what everyone else believes, but I think Bruno's opinion carries a little weight, no? So like I said, personally I'd probably try it anyway and see what happened.

Sorry, I tried to use google to find it but it just isn't finding things in the diyaudio archieves like it used to, and that's one hell of a loooong thread to fish through manually.

Regards
 
Well, yeah.... nightmare to say the least.

One trick if you're ever hellbent on finding it, just scan all posts by Bruno via his profile.

This is what I did to come up with this example:

http://www.diyaudio.com/forums/showthread.php?postid=494606#post494606

It's one from the ucd400 thread, and sadly is just one of the examples where he say's don't do it. Actuallly this time he just said "don't bother".

The "only" time of all the times he's said don't do it, that I'm aware of, for which he really gave any reason I still think lies in the ucd180 thread, the reason was something like "invites hideous hum" and there was no "reasoning" along with it, as in why it might hum. So don't bother scanning expecting to find any as far as I know such an explanation doesn't exist.

Anyway it took me scanning through about 75 of his posts to come up with that much of an example for you.

So... yeah, he doesn't like it anyway and he's always said don't do it, I posted this for purpose of controversy 🙂

I believe there's extensive explanations in the SS forum as to why/how it is beneficial.... (don't ask for a link) everyone else likes it, only Bruno say's otherwise. Interesting isn't it?

To that end if anyone has ever had a bad experience with the dual rectifier method please be sure to let us know.

Cheers.
 
Re: i found it

ackcheng said:

Well done, but that's the one I posted already lol (see above).

If you find the one that mentions hideous hum or "invites hideous hum" (something like that) let us know. Don't waste too much time on it though, that's all it says and it will take you hours or weeks to find it.

I for one won't be looking any further 🙂

Regards
 
Thanks to both of you for trying to look. The thread sizes are a nightmare. I have read every post on all 3 or 4 of the threads as they have been growing, but trying to go back to check a detail I know is in one of them somewhere is such a nightmare. Looking up an individual user's posts is about the only method I have any success with as well.

I spent over an hour last night looking for a post where Lars and Bruno were discussing what would happen if the load were suddenly disconnected with an output signal, but couldn't find it anywhere :bawling:

I myself believe (you may have seen my recent response to Eric Peter's post on the supply thread) that the twin/twin setup is better -- especially with respect to hum -- which is why I'm very interested in seeing any evidence or information to the contrary.
 
Yeah, sorry, like I said there never was much of an explanation to it, and the hum I meantioned was really the only reason ever given.

You said you never found the post you were looking for last night? I know for a fact I tripped over it today while I was looking for that/those links on this rectifier subject. So if you need a tip you'll find it in one the first 75 posts of Bruno's, or rather likely the last 75 (I think it lists them from the most recent ones).

I think it had to do with the different feedback schemes and loading of it, take the load off and it became possible to oscillate at the filters resonant frequency, ring a bell?

*Censored*'s me off to be honest with you, a short time ago all you needed to do was remember one or two key words of the post in question, hit google and and type "(keywords/phrase here) site:www.diyaudio.com" but for some reason unknown it's become totally useless.

Maybe they'll fix up the forums own search functions sometime soon.

Regards
 
I think the best way to search a thread is to click the "Print" option and then the link to list all posts on one page. Then you can use the browser function to search for stuff and keep fflicking through the thread.

Most of the UCD stuff is in one of three threads, so it's not too hard to search this way
 
Heh, was wondering if you had found the link discussed to the hum issue of using dual rectifiers. If you don't have it though don't worry, as I recall that's all it had to say on the matter.

So, will you be using 1, or 2?

Alot of people have gone against Bruno's advice here and used 2, and have reported excellent results.

For the cost of a rectifier and some caps, you should be able to test both methods easily enough and see how it works either way.

Find yourself any caps yet? Still researching?

This seems to be a good place to get them from, at a decent price:

http://www.thlaudio.com/indexE.htm

Check out their BHC caps, 100V/15000 uf

Ordering process seems like a pain though.

Cheers
 
I found some Philips Caps 10.000uF/80V but it was for PCB mounting. I don't want that, I want caps with screw connectors (when I asked for that, the part guy gave me a weird look and told me that doesn't exist anymore... audio and computer grade, he didn't understand that either)

What do the ppl in here think about Philips Caps?

I'm still worried about the DC protection/relay issue - I hope I won't lose all my UcD sound quality because of the relay. The PCB's speaker traces are a little thin too... I'm going to solder 2.5mm2 on the bottom side to help things out.

I'm going to try 2 rectifiers, it feels "better" because the transformer seems more "disconnected" that way, Maybe it's all just in my head.

Yves
 
Yves Smolders said:
I found some Philips Caps 10.000uF/80V but it was for PCB mounting. I don't want that, I want caps with screw connectors (when I asked for that, the part guy gave me a weird look and told me that doesn't exist anymore... audio and computer grade, he didn't understand that either)

What do the ppl in here think about Philips Caps?

I'm still worried about the DC protection/relay issue - I hope I won't lose all my UcD sound quality because of the relay. The PCB's speaker traces are a little thin too... I'm going to solder 2.5mm2 on the bottom side to help things out.

I'm going to try 2 rectifiers, it feels "better" because the transformer seems more "disconnected" that way, Maybe it's all just in my head.

Yves

Funny. I once had a guy at Radio Shack tell me IR detectors are obsolete and no one makes them anymore, I think I started blowing bubbles and foaming at the mouth, next time I'll just ask if he's got a charger for solar cells, see if he tells me the same.

BC Components makes Philips caps now (BCC, from Vishay I think), since 2002 I think. So if the caps you found are actually marked "Philips" I'd suspect the shelf life of them, might need reforming. BCC seem to have some good caps, I'm unable to comment on the "sound" on them.

Have a look here, they give Philips caps a good grade:

http://www.tnt-audio.com/clinica/ssps3_e.html

I also like the idea of screw terminals, and high temp computer grade for long life.

Regards
 
OK.

Time for another electronics shop, where they don't treat n00bs like me like sh*t.

I have another Q about slow start. I'd like to use resistors for a few seconds to slowstart my amp. I've seen schematics with 4 times 10 ohms/5W in series. Will this do?

I'm a bit puzzled on how to calculate the right resistance.

Yves
 
The resistance is not that critical. You need to make it high enough so that the current is limited enough, but not too high so that the caps etc are hardly charged when the delay is over so you get another inrush. 50-100 ohms seems to be what most people use. You will probably need to series resistors to meet the required voltage capability as the full mains will be across them at the start.
 
Status
Not open for further replies.