Brute force approach?
400 amp Hexfred diodes.
(they're small...)
ok, so you still need to limit the intial turn on current in most cases, but you won't blow the diodes if you have a stiff enough line to run the amp...
_-_-bear
PS. I turn on ~500,000ufd in my big amp with a pair of relays that connect the Mains --> resistor --> then shunts the resistor with another big *** relay (multi- contact) run off a lowly 555. Dumb, simple & reliable.
400 amp Hexfred diodes.
(they're small...)ok, so you still need to limit the intial turn on current in most cases, but you won't blow the diodes if you have a stiff enough line to run the amp...
_-_-bear
PS. I turn on ~500,000ufd in my big amp with a pair of relays that connect the Mains --> resistor --> then shunts the resistor with another big *** relay (multi- contact) run off a lowly 555. Dumb, simple & reliable.
Zero Cool said:What about a soft start zero crossing solid state relay? at least to power up the transfo????
Zero
Zero crossing switching is the worst thing you can do with a transformer. If you are able to control the moment of switching, then switch at the input voltage peak.
http://relays.tycoelectronics.com/app_pdfs/13c3206.pdf
Steven
To peranders : By switch on is resistor of this value ( 33 R ) manytimes overloaded. Round ones have wire spiral around all body, while square ones have this spiral smaler and inside body, so short - time overload is by round ones bigger. I had try on this position both types and " white one " had lower reliability, belive me .
.
janneman said:
Hi Jan,
I would say: yes. The flux B is 90 degrees lagging wrt the applied primary voltage. If voltage is at maximum, the B is zero. This is importanty for both switch on and switch off. If switched off during zero crossing of the voltage a strong residual flux will remain, up to 80% of Bmax. If then switched on at the wrong zero voltage crossing, saturation will be even more severe. Worst case: switch off at negative zero voltage crossing, switch on at positive zero voltage crossing (or the other way around, i.e. the crossing in the opposite direction).
Switch off at zero current would be close to zero flux, although a loaded transformer will not be purely inductive.
Steven
Upupa Epops said:To peranders : By switch on is resistor of this value ( 33 R ) manytimes overloaded. Round ones have wire spiral around all body, while square ones have this spiral smaler and inside body, so short - time overload is by round ones bigger. I had try on this position both types and " white one " had lower reliability, belive me
This is correct - surface cooling area is considerably larger for the "round ones". The "squared ones" have smaller body encapsulated by ceramic housing and their peak current overload capacity is lower.
alvaius said:
there are a couple application notes which go with this part (which costs about $1.70 from DK here in the states) AN1600 and AN1602 on ST's website.
the STIL04-P5 requires an auxiliary supply -- like synchronous rectification
one thing which is interestingfrom AN1600 is that the EMI inductor filter is used to smooth the current into the bulk capacitor -- (this is also inrush current limiting!)
it's an interesting idea.
Going back to passive inrush current limiting:
I put back the 1mH choke (from the transformer center tap to ground) -- it really does the trick, but the choke itself weighs 10 pounds:
certainly not the most cost or space-conscious way of doing things
I put back the 1mH choke (from the transformer center tap to ground) -- it really does the trick, but the choke itself weighs 10 pounds:
An externally hosted image should be here but it was not working when we last tested it.
certainly not the most cost or space-conscious way of doing things
This is maybe true but I doubt that this fact really is important for a 0.3 sec pulse. Anyway I have used 4 or 5 Watts, 100 ohms starting 600 VA/2 x 22000 uF/+-60 V since 1989. The amp has been switched on almost everyday since that year. The only thing that have gone broken was my 470 ohms in series with the "big" cap. This resistor was a 0.6 W metall film. I have changed this to four 120 ohms in series.Upupa Epops said:To peranders : By switch on is resistor of this value ( 33 R ) manytimes overloaded. Round ones have wire spiral around all body, while square ones have this spiral smaler and inside body, so short - time overload is by round ones bigger. I had try on this position both types and " white one " had lower reliability, belive me
Yes, this is correct, by 100 R value will be overload not so big. But look at DC voltage on caps during the process : probably serial resistor is shorted by relay in time, when DC voltage isn't full, so you get still relatively high inrush pulse. One example by my experience : two toroidal transformers 850 VA each, total capacity of PS 160 G / 80 V, serial resistor 40 Ohm / 50 W, delay of switching cca 1.5 sec. . Average " first " inrush current was cca 6 A ( measured on mains side by relatively " fast " pointer ampermeter ), " second " inrush current 6 - 8 A. Voltage on elyts was at the moment of switching cca 90 % of full voltage. In your case must be voltage lower, so you get higher " second " inrush and it is not good for reliability.
Here is example of this circuit, which I'm using many years. Delay is make by counter which count mains frequency, so delay is independent on tolerance of devices. Diodes at rectifier are 1N4002, other are 1N4148. All transistors are BC 546 or similar. Transformer is 3 VA, 15 V on secondary. On PCB are also mains fuses for two mains transformer. Circuit is connected between mains switch and transformers. Fuse 2A protect R 10 against long time overload ( by disorder of delay circuit or relay ).
Unhappy Diode
I was experimenting with MBR20100's -- cycling the power supply -- this is the result (before choke installation):
not only did it blow the diode apart, it also blew the leg off the bleeder resistor.
I was experimenting with MBR20100's -- cycling the power supply -- this is the result (before choke installation):
An externally hosted image should be here but it was not working when we last tested it.
not only did it blow the diode apart, it also blew the leg off the bleeder resistor.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.