|
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
58.110.34.49
In Reply to: RE: What do you need? posted by Russ57 on September 21, 2014 at 09:52:08
Sorry for not making my needs clear. I am building a 1200 volt supply with 200uf each channel on the final cap. I am not too skilled with psu designer but 200uf on the last cap looks the best for my needs A big se amp. 805 or 4e27a tubes in se a2 and a for the 4e27a . I would like 2kv ratting for safety and long life. I read that motor run caps are not really designed for for constant dc. Thats why I decided to increase the rating Currently 3 times 450 ac rated caps for 1800 volts rating.
Follow Ups:
The power supply needs to provide ripple reduction, energy storage, and low impedance return at audio hertz.
The ripple reduction is best handled by a combination of caps and chokes. Chokes store energy as well as caps. Joules per audio output watt is what you should be looking at. Voltage is the bigger factor there so you don't need a lot of uF (and remeber it is total uF not just the last cap). I think 2 joules per watt is a lofty goal (and total overkill in many's opinion but I like overkill).
Consider what is driving the market and take advantage of it. Because some areas use 50 hertz and some 60 hertz....and because some use 3 phase 480 VAC and others 3 phase 575 VAC...and some designs use half wave and some full wave....the market designs for 600 VAC and 120 hertz.
Now consider that 600 VAC swings positive and negative.....so actual peak to peak voltage is 2.8 times 600 = 1680 VDC rating. Because of my job (and budget) I have been using motor run caps for a very long time. I can assure you the 440 VAC rated (American) caps used to be dual rated for 1,000 VDC also. These caps were designed to work with complex reactive loads and carry 10, 20, 30 amps.
I would say you are wrong when you say they aren't happy with DC voltage as that is a much easier load. The currents we deal with are much lower. The only issuse might be high hertz but one can bypass with a small (1 uF) film cap.
Again the market is driven by the needs of the UPS and VFD area. These units are taking 480 to 575 VAC, turning it into plus and minus DC rails, and cranking out tens to hundreds of amps. Childs play to get a decent 100-200uF polypro cap rated for 1,200 to 1,600 VDC for our needs....all for around 50 bucks. You don't need to derate these caps. I see them last for decades in much more hostile environments.
The last thing you want to do is string a few 450 VDC electrolytics in series and hang resistors across them. Very old school, costly, labor intensive, space hogs, and prone to going to hell in a hurry if a single cap in the string goes bad. Even under the best circumstances I wouldn't trust them over ten years of age. As far as motor run type caps go....my facility was built in 1947 and I still have original GE and CDE caps in DC power supplies going strong. Always been a fan of the GE 27F and 97F series.
So work backwards based on output audio watts. Determine total joules needed. Decide how to split that up between the caps and chokes. Use an on line design program to make sure your power supply is well designed. Way too much nonsense around here these days on ultra low DCR and such.
I don't want to nitpick, especially since there's a great deal of good advice in your post. However, I think it might be useful to consider two of the points you made from a different perspective.
"The power supply needs to provide ripple reduction, energy storage, and low impedance return at audio hertz."
This has been the common wisdom for years, but one of the "requirements" in this statement is actually a superfluous function. It can't be easily quantified as a singular entity by most of us, nor does it need to be. I'm referring to "energy storage." This may seem intuitively to be a separate function, but it is fully defined by the other parameter, "low impedance at audio hertz." The reason for this is that "low impedance" is a relative term, and in this case, it is relative to the lowest impedance presented by the load (amplifier). If the power supply represents only a small fraction of the lowest Z presented by the load (presumably occurring at full power output) at the lowest frequency of operation, it does in fact - automatically - exhibit sufficient energy storage. Energy storage is not a separate quantity that must be somehow designed into the supply by calculating numbers of electrons. The designer only needs to calculate (or model) the lowest Z of the load in ohms and compare it to the Z of the power supply over the frequencies of interest.
"Chokes store energy as well as caps."
The problem here is that unlike capacitors, chokes don't "give up" their energy when the load needs it. Just the opposite. When the load attempts to draw more current in order to output more power, the output voltage of the choke goes down. It's important to remember that chokes act as high Z current sources, and a large one can swing the voltage wildly when the load varies. A series choke in a power supply must always be terminated with a capacitor, and the final capacitor must be made larger if it is preceded by a choke.
Just my $0.02. Keep up the good work!
--------------------------
Buy Chinese. Bury freedom.
If it isn't, it is called a plate choke. :-)
Post a Followup:
FAQ |
Post a Message! |
Forgot Password? |
|
||||||||||||||
|
This post is made possible by the generous support of people like you and our sponsors: