Page 1 of 3 123 LastLast
Results 1 to 10 of 22

Thread: Power supply interrupter

  1. #1
    Join Date
    Feb 2011
    Location
    New Hampshire
    Posts
    3,513

    Default Power supply interrupter

    I am assembling a very large capacitor charging power supply. The medical dye lasers that I am converting both had defunct switch mode power supplies. I intend to replace them with a fairly simple variac controlled gang of microwave oven transformers, bridge rectified to approx 4,000V and current limited by selecting a sufficient number of transformers to self limit to 20-30A @ 240V. This is very scalable and inexpensive (and of course dangerous!). The system repetitively drives xenon flash lamps and is triggered by a saturating series trigger transformer. The trigger is opto-isolated and controlled by a 12V DC pulse generator that drives a small LED (thanks Steve for the advise). So far, so good. The problem I need to address it that as the capacitors dump into the lamps the voltage drops and eventually the discharge terminates...in theory, but the continuous HV supplied from the power supply continues to maintain the discharge for a short sometimes not so short time. I need to briefly block/short/switch off the supply so that the discharge ceases, after which the lamps will easily hold off a 4,000V potential.

    Because the pulse rate is low (less than 6Hz), holding off the supply voltage for a 1/2 or a full cycle should not significantly effect power delivery and would be generous enough to allow the flash lamps to stabilize. Therefore I assume it might be easiest to interrupt the 240V supply to the MOT's because even their stored energy will dissipate in that kind of time. The 12V trigger pulse can be adjusted from 1-100% of the pulse width. Now, I'm not too slick with circuit design and I am looking for something simple, cheap and SPECIFIC. If you start throwing terms like "well obviously you need to assemble several MOL-MOSFET Q1000's in a quasi full bridge" at me, then I'm going to feel like I'm loosing ground here.

    Any help appreciated. Thanks

  2. #2
    Join Date
    Dec 2010
    Location
    DC/VA metro area, USA
    Posts
    554

    Default

    Take a hard look at solid state relays. Super simple; they accept a small DC input for switching. Minimum off time for the first one I checked was 500 uS. You'll need a heat sink. Lots of combinations available, zero-crossing switching, immediate switching, some specs I don't know what they do:

    http://www.digikey.com/product-searc...048664?stock=1

    http://www.crydom.com/en/Products/Catalog/s_1.pdf

  3. #3
    Join Date
    Feb 2008
    Location
    Northern Indiana
    Posts
    915

    Default

    Trying to understand your circuit.... Are you wanting the capacitors to set the discharge time of the lamp? So, basically you will be partially discharging and recharging the caps 6 times per second?

    Also, what do you mean by limiting the current to 20-30A? Is that peak inrush current? Depending on the series R L C of the circuit that may be hard to do.

    How much capacitance are you charging?

  4. #4
    Join Date
    Feb 2011
    Location
    New Hampshire
    Posts
    3,513

    Default

    The circuit is super simple. The dumb, HV supply is current limited simply by the resistance of the primary windings. The average current is determined by the pulse frequency of the caps (300uF). The variac/mot determines the charging voltage and this is held until the flash lamps discharge. I will look through the Digikey list and see if I can get a relay that shuts OFF with a positive pulse.

  5. #5
    Join Date
    Dec 2010
    Location
    DC/VA metro area, USA
    Posts
    554

    Default

    If you look for a "normally closed" relay, it will "open" upon application of the signal. A "normally open" relay will "close" upon application of the signal. E.G.,

    http://www.digikey.com/product-searc...=0&pageSize=25

    Still, worth shooting an email to the vendor to confirm this is what you're looking for. There may be other vendors offering cheaper parts.

  6. #6
    Join Date
    Feb 2008
    Location
    Northern Indiana
    Posts
    915

    Default

    My first thought was you may not be able to charge the caps fast enough to get 6hz rate. But if that circuit is already functioning then..... Just add the triac relay and drive with the trigger pulse and adjust the PW for the off time needed. Beware the inductive kickback of the MOT's will be giant. The solid state relays will have some built in snubber circuit but a 250V MOV (or several in parallel) may be needed in addition to whats in the SSR to be safe.

  7. #7
    Join Date
    Feb 2011
    Location
    New Hampshire
    Posts
    3,513

    Default

    tribble,
    Got it, thanks.

    log,
    Ok. Just to be safe, is there any way to estimate the varistor capacity that would be appropriate? Also, how would these act to deal with the kick back? I'm not arguing. I simply need to understand. If the primary is suddenly shut off from the variac then what is the source of the energy and am I looking to protect the relay, or the line etc?

  8. #8
    Join Date
    Feb 2008
    Location
    Northern Indiana
    Posts
    915

    Default

    The kickback comes from the magnetic field collapsing in the primary. Theoretically the voltage would go to infinity. Capacitance and resistance keep that from happening.

    The MOV has a high resistance below its clamping voltage. Above the clamping voltage the R goes very low effectively shorting its leads. This is meant to protect the relay circuit and keep noise spikes out of other parts of the circuit.

    I'm just a lowly EET. So, the exact calculation is a bit out of my pay grade I know that the kickback voltage is E = L (dI/dt). Problem is with an AC signal you dont always know where in the cycle it is going to turn off.

    I was thinking about your circuit earlier while eating lunch. Using the variac to adjust the voltage during steady state operation sounds fine. My concern would be the initial cap voltage before your first trigger pulse. Its possible the caps could be charged several percent more than during steady state operation. This could give a huge first pulse.

    Edit: I guess you are probably starting the trigger pulses then slowly turning up the variac voltage. This would solve the possible big first pulse issue.

  9. #9
    Join Date
    Dec 2010
    Location
    DC/VA metro area, USA
    Posts
    554

    Default

    I'm no EE, but... With an SSR, it's always going to stay on until the next zero-crossing of the controlled waveform. Therefore, it stands to reason that the voltage is always zero when it shuts off. Of course, this is during maximum dV/dt and therefore maximum current, and therefore maximum B field in the transformer, correct? So you're always going to get the biggest spike possible back in to your snubber. Yay.

    http://www.digikey.com/Web%20Export/...f?redirected=1

  10. #10
    Join Date
    Feb 2008
    Location
    Northern Indiana
    Posts
    915

    Default

    Yes, SSR will always stay on till zero crossing. Flux density should be at maximum. The collapsing field should cause a spike in the secondary as well. I am not sure how to calculate that.

    Another potential problem could be inrush current when the primary is switched back on. Residual magnetism in the core could give a big initial current spike (probably why crydom says to not use zero crossing model with inductive loads) plus what is reflected back from the load could create large inrush over a few cycles. Its hard to say how low the cap voltage will be after each discharge

    Basically, what seems like a simple circuit is very complex and requires a lot of analysis to figure exact numbers.

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •