|
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
192.91.75.30
In Reply to: psud2 question posted by Vinnie on April 16, 2007 at 03:35:05:
PSUD2, or any other simulator, will probably not lie. The question really is, 'are you simulateing the conditions correctly?'Here are some things to consider.
Are you using semiconductor or tube diodes? Remember that indirectly heated tubes take a long time to warm up and begin to draw current. If you are using semconductor diodes, there will be a period of time during which your power supply will 'see' a load current of only that drawn by the bleeder resistor(s). Be sure to run PSUD under these (power up) conditions. Quite likely, the voltage on the caps will be considerably higher after initial turn on than when the tubes warm up and draw nominal current. This is also somewhat true if you use a directly heated rectifier (5U4, 5V3, etc) and indirectly heated output tubes. This is one reason I prefer to use and indirectly heated rectifier, such as the 5AR4.
Remember that you line voltage is unlikely to be truly constant. Have you allowed for the maximum line voltage? How about spikes in the line voltage caused by turning off inductive loads?
I'm sure that others can add to the list of considerations ...
Bottom line, yes, I think that using 1200VDC caps on a 980 VDC (nominal) supply is cutting it too close.
Follow Ups:
Very interesting. I am using a ss bridge into a CLC filter. I have 3 390k/1w resistors across the 3 330uf/400vdc caps I have in the ps (one set for each "cap" in the clc). So far I have only brought it up with the variac and it has worked ok, but I know a full blown turn on will not be so gentle.
How does one figure the resistance of resistors in series but paralleled with caps? Sounds like that is what I should plug into psud2, huh?
For your purposes, just ignore the paralled caps and consider the resistors to be in series. For simultion, your load before the tubes warm up is 2*980/1170 mA = 1.7 mA. Not much.
Using that criteria the caps would most likely blow. Start up voltage doesn't show a spike as such, but it goes on up to 1200. Bye bye caps. Well, back to the drawing board.
If you want to have a choke input filter (that acts as such) then it needs to meet "critical inductance". This requires a certain minimum current draw. Bleeder resistors were commonly used. You can cut corners a bit if you insure tubes are warm enough to conduct before B+ comes up.So if you either delay application of B+ until tubes are warm and/or add a bleeder resistor to meet "critical inductance". Then your B+ max will be 0.9 times the PT secondary voltage....if not it goes to 1.4 times secondary....and caps suffer. BTW, best to use some poly in oil motor run type caps. They have ones that are rated high enough. Forget them series caps. But DO have enough bleeder current:)
In PSUD learn how to do a step function on a current tap. Focus on what happens in the small window before the current tap kicks in and after. Are you over or under dampened?
Russ
P.S. It is best to get a firm handle on what determines max voltage before you fire it up. It is equally nice to design something that can handle a "no-load" condition. In your case....are your coupling caps over 1kv??
The problem is it is going to take way more uf's than I can find in a oil caps. At least that is the picture I am getting for a cap input filter system. I have been looking at choke input too, but that has it's own set of problems. I plan on doing a lot more experimenting with psud2 now that I have some guidance how how to best use it for designing a ps. Should be most educational and interesting.
This post is made possible by the generous support of people like you and our sponsors: