Head-Fi.org › Forums › Misc.-Category Forums › DIY (Do-It-Yourself) Discussions › TREADish based psu question
New Posts  All Forums:Forum Nav:

TREADish based psu question

post #1 of 7
Thread Starter 
So I'm building a sort of variation of the Tangents TREAD psu using a non adjustable VReg, and was wondering if there is a max recommended limit the value of the input smoothing capacitor - C5 on the TREAD design?

Essentially I have stacks of 1000-4700uF electrolytics lying around, so thought I might put them to use.

I'm wondering whether an excessively low Xc might cause too much of a surge at power up, and be a problem for my traffo.

I'd like to get the ripple as low as possible as its for a Szekeres amp which doesn't like a noisy supply.

Also would you stick some on the amp rail side?

Cheers.
post #2 of 7
Unless you're using a really small transformer -- and I don't see how you could get away with that with a Szekeres -- you probably won't melt anything with a huge cap bank.

Not that it's a good idea.

First, it means you have a fuse problem. One alternative here is that you have to oversize your fuse to survive the inrush, which then means it will be less likely to protect the amp if the amp goes crazy. That is to say, the inrush might be many amps, whereas the amp can destroy itself on less than an amp. The other alternative is to get clever with inrush current limiting so you can get away with a smaller fuse, which then is more likely to blow if something goes wrong after power-up. I gather from your description that you're looking for "cheap and simple", so inrush current limiting is out of the question. Even if you were willing to consider that, ICL fixes can cause their own problems.

Second, it's a poor use of parts. Any properly-implemented linear voltage regulator will buy you many orders of magnitude of ripple rejection. Doubling the filter cap bank size merely doubles the ripple rejection ability. You'd have to increase the filter cap bank by 10x to get a result at the output that's in line with the sort of improvement you can get with a better regulator, which will fit in less space and maybe cost less, too.
post #3 of 7
Thread Starter 
Hey thanks.

Quote:
Originally Posted by tangent View Post
Unless you're using a really small transformer -- and I don't see how you could get away with that with a Szekeres -- you probably won't melt anything with a huge cap bank.
No I do have a substantial transformer.

Quote:
Originally Posted by tangent View Post
Second, it's a poor use of parts. Any properly-implemented linear voltage regulator will buy you many orders of magnitude of ripple rejection. Doubling the filter cap bank size merely doubles the ripple rejection ability. You'd have to increase the filter cap bank by 10x to get a result at the output that's in line with the sort of improvement you can get with a better regulator, which will fit in less space and maybe cost less, too.
Ok, so what would you recommend as a better regulator? I'm not bothered about cost or complexity, in fact I'd quite like to build something more challenging. However, it seems to me that most of the DIY offerings are designed with lower OP current in mind. Any suggestions?
post #4 of 7
Since you have the caps:

Capacitance Multiplier Power Supply Filter

targeted for class A so it should work for you.
post #5 of 7
I'm too focused on my own niche to pay attention to high-current designs. I do seem to recall that one of AMBs supplies is basically a YJPS redesigned for high output current.
post #6 of 7
Thread Starter 
Ok thanks for the ideas guys, I will have a look araound.
post #7 of 7
the salas simplistic shunt regulator is higher current. check it out on DIYAUDIO, but it generates quite a bit of heat and therefor some serious heatsinking is required. you also need to account for a larger voltage drop across the regulator with shunt regs at higher current.
New Posts  All Forums:Forum Nav:
  Return Home
Head-Fi.org › Forums › Misc.-Category Forums › DIY (Do-It-Yourself) Discussions › TREADish based psu question