Hello!
and sorry for the title, but the website is buggy ATM...



I am curious to as if anyone has tried the attached image circuit in practice.?
I would be using it only on 110AC power. The intention is to power up a large torroidal transformer that can knock out a circuit breaker with it's large surge upon power connection. In this case, I wonder if initially connecting the primary windings in series will reduce the surge, then connecting them by dpdt relay into parallel after a short delay. Most circuits involve resistors and other methods and I haven't yet seen such a circuit for a 110V wall supply!


Thanks in advance for your critique and suggestions!

 

Are the secondaries feeding a DC supply? If so delay the capacitor connections until after switch on.
Max.

 

Hi and thanks!

Well yes they are, but does this mean that my idea is wacky? Could it work?

 

I built a "soft-start" delay into the 4000V high-voltage supply in a home-brew Kilowatt RF amp. It had a 240V primary. The goal is to prevent blowing up the high-voltage rectifiers by limiting the inrush current required to initially charge the filter capacitor.

Here is the gist of the idea: R1 is chosen to limit the in-rush current. The drop across R1 an R2 prevents the relay from pulling-in until the voltage across R1 drops because the filter capacitor is partially charged, and the transformer primary current drops. When the relay does pull-in, its contacts short out R1, allowing the primary voltage to come up to the full line voltage. In the steady state, R2 keeps the coil voltage of the relay at its rated voltage.

AFAICR,I used a 120Vac relay so that in the steady-state, R2 needs to drop about half the 240V line voltage. I think that R1 was about 20Ω at 25W.

 

I built a "soft-start" delay into the 4000V high-voltage supply in a home-brew Kilowatt RF amp. It had a 240V primary. The goal is to prevent blowing up the high-voltage rectifiers by limiting the inrush current required to initially charge the filter capacitor.

Here is the gist of the idea: R1 is chosen to limit the in-rush current. The drop across R1 an R2 prevents the relay from pulling-in until the voltage across R1 drops because the filter capacitor is partially charged, and the transformer primary current drops. When the relay does pull-in, its contacts short out R1, allowing the primary voltage to come up to the full line voltage. In the steady state, R2 keeps the coil voltage of the relay at its rated voltage.

AFAICR,I used a 120Vac relay so that in the steady-state, R2 needs to drop about half the 240V line voltage. I think that R1 was about 20Ω at 25W.

Brilliant,
I've been thinking about something similar. What's nice about an intrinsic delay like that is, besides fewer parts, that it only operates when it's actually needed! If my design works, I will most likely attempt the same delay idea as similar to yours.

Thanks! .

 

Well,
This idea is for an amplifier project that currently uses a 750VA transformer. Right now there is no "soft start" circuitry as it stands, just a switch to control the power on/off. If I power it on cold, the lights in the room will blink, but if I then turn it off and back on again in a short amount of time, there is no issue of surge (most likely in relation to Max's reply above) in firm assumption that the large caps are retaining their charge. In audio however, relays are shunned like a capacitor in the chain. If I can keep it on the mains side of the transformer it would be best.

At best, what could be an issue with such a prospect? I'm about to get some parts and try it out!!! I get shocked by mains potential all the time, it keeps me current! lol

 

How about an inrush current limiter.

CL-30 handles 8 amps. Goes from 2.5 ohms to less than an ohm.

Couple bucks. One part.

I don't know how to size them properly. Keep a few sizes around for trial and error.

 

.... Keep a few sizes around for trial and error.

Trial and error is my kind of science!
How about this schematic attachment?

 

Is that to decrease inrush even more?

Seems too much, as going from parallel to series connection, inrush should follow impedance. Factor of 4 already.

 

Use a relay fed by a zener off of the DC to switch from series to parallel after the initial charge up time, you should not need the diodes or timer?
It will need a bit of initial testing to find the zener value.
Max.

 

INWO's idea is easiest if you know the size of the caps in your circuit. Single part in series with the line. Since you can replug it without a problem it should have time to cool off in between power cycles.

http://www.mouser.com/ds/2/600/SL2210008-238039.pdf

 

Is that to decrease inrush even more?

Seems too much, as going from parallel to series connection, inrush should follow impedance. Factor of 4 already.

Well in this case (the second) the primaries are already in parallel, however, with the diodes, each primary would see 1/2 duty on opposite voltage swings until the relay closes.

"Use a relay fed by a zener off of the DC "

Where would you get DC off this circuit without a diode?
I would like it self powered if possible ;-)

When I get a chance I may just try it out with a DPDT switch.

 

INWO's idea is easiest if you know the size of the caps in your circuit. Single part in series with the line. Since you can replug it without a problem it should have time to cool off in between power cycles.

http://www.mouser.com/ds/2/600/SL2210008-238039.pdf

Thanks and yes I agree,
The capacitors are 33,000microF. Big and expensive cans

 

"Use a relay fed by a zener off of the DC "

Where would you get DC off this circuit without a diode?
I would like it self powered if possible ;-)

.

The relay would be fed off of the DC after the bridge, the relay would only energize at the peak value of the lower DC value due to the series connections.
For a relay of that voltage the current will be minimal, and there should be no problem on maximum voltage.
You will find that it will pick up well below the rated voltage with the suitably selected series zener.
Max.

 

Good thing this isn't rocket surgery,
It works, but now there's two smaller transient blinks of adjacent light fixtures, heh.

After searching around, it appears that many use a relay in parallel with thermistors, activating the relay after a short delay. My only real concern was in event of the relay failing or other control mechanism, that the resistive device used for the inrush limiting would be on full time, generating heat etc.. This is my reasoning for the transformer winding switching, in the event of a failure, no problem.

Thanks again for the ideas
Scott

 

The thermistors probably won't like caps that big (think fuse). The resistor on either the primary or secondary side is probably the best solution. What is the voltage and current on the secondary side?

 

The thermistors probably won't like caps that big (think fuse). The resistor on either the primary or secondary side is probably the best solution. What is the voltage and current on the secondary side?

50VAC @ 7.5A X2

Wouldn't the CL90 limit the initial current to about 1A?
Perhaps this all leads back to a ceramic resistor circuit! heh

Thanks
Scott

 

Cl's go to fraction of an ohm.

They can be shunted after time delay.

The classic power supply at one time was not capacitor input but choke input.

Here is a couple CL diagrams from the web.

 

This one would do it I think.

http://www.mouser.com/ProductDetail/Ametherm/MS32-7R015/?qs=PVyXUKBeRH1OvrcFIWTAZg%3d%3d