Results 1 to 8 of 8

Thread: Operating 3 Phase ion laser PSU off VFD with square-wave output - problems?

  1. #1
    Join Date
    Jan 2006
    Location
    Charleston, SC
    Posts
    2,147,489,296

    Question Operating 3 Phase ion laser PSU off VFD with square-wave output - problems?

    Hey everyone (and especially Steve Roberts!)

    I've been chatting with a fellow laserist who loves old-school ion lasers and is in the process of resurrecting some older gear. The goal is a tandem pair of I-90s for some lovely ion white. The plan is to fire the I-90's using a pair of Lexel 95 PSUs. He understands that they probably won't be able to drive the I-90s to absolute max current, but that isn't an issue as peak power isn't the goal here. Longevity is.

    To that point about longevity, I have a question: What is the failure mode for an ion laser power supply that is run off a VFD that has a square-wave output? And what other problems can one expect as a consequence of using square-wave AC power instead of true sine-wave power?

    I've heard talk of issues with cathode sag, but I can't really wrap my head around WHY that would be caused by square-wave power.

    The reason this is relevant is that his shop does not have 3 phase power. He does have a nice VFD that will produce 3 phase output from single-phase input, however, and it's large enough to be able to power both Lexel 95 power supplies at the same time. But if this is going to shorten the life of the tubes or the supplies, then it might not be the way to go. Rotary phase converter construction is also an option, as is the purchase of a surplus genset (for the right price, of course!).

    But before any of this takes place, I'd like to know what the real issue with square-wave power is, and how bad it actually is...

    Your thoughts?

    Adam

  2. #2
    Join Date
    Nov 2007
    Location
    SoCal / San Salvador / NY
    Posts
    4,018

    Lightbulb Umm...

    Hey Adam -

    Quote Originally Posted by buffo View Post
    ...The plan is to fire the I-90's using a pair of Lexel 95 PSUs.
    I assume this 'plan' is because the Lex 95 supplies are already 'in hand' / avail, within budget, whatever? My thought would be.. Ditch the Lexxy supplies (there was someone, in here, not that long ago, iirc, that was hunting Lex 95 supplies, if not whole sys (?).. I dunno, prolly getting old.. and get a couple 1 / 3 American 909 'switchers' - They'll run I-90s to 35A+, off 1 (assuming you've got the Amps, of course..) and they're plenty-robust, easy to work on, and provided you keep an eye on your cath-xformer V, will keep an I-90 Cath quite happy.. I've ran those across Several I-90 installs / mods, both in Ar/Kr configs, and 'fussier' WL-tube configs, and.. Nary a hiccup. Then, his whole entire 'VFD vs Rotary vs Genny issue' goes away, and I am Quite sure, $$-wise, he will be better off, in the long run.. Plus, the 95 supplies are Beast, weight-wise, compared to the 'svelte' 909-supplies (see pix, below..)

    ..In fact, there's a longtime 'supplier', out here, that is moving-shop (not me, I've already passed thru Dante's Inferno, yet I'm sill burning.. and Liquidating - and I mean, Liquidating - 90% of his Ion-stock, and Parts, so.. I could check with him on Am 909 switcher-stock, and/or see what he'd charge to get a couple ready (the 'Mods' are not bad - just a few wiring changes, mostly, ie: I-90 style Plug, and accomodating the starter In the head, vs PS; magnet, couple interlock-tweeks, etc.. Not bad, really..) Anyway, that's my (seasoned) .02, Fwiw..

    Click image for larger version. 

Name:	GX5W_PSU_1.jpg 
Views:	12 
Size:	66.0 KB 
ID:	52132 Click image for larger version. 

Name:	GX5W_PSU_2.jpg 
Views:	16 
Size:	84.1 KB 
ID:	52133 Click image for larger version. 

Name:	GX5W_1o.jpg 
Views:	11 
Size:	61.8 KB 
ID:	52134
    (Yes, this one was 'wired' for 3, but all you had to do was hook yer 2 Hots / Gnd (thus, the 1-3 indicators in the rear..) and you'd be off..


    ..now The P'fessor will come in here and tell you why I'm all washed-up..

    Cheers..
    j
    Last edited by dsli_jon; 05-17-2017 at 23:29. Reason: sp
    ....and armed only with his trusty 21 Zorgawatt KTiOPO4...

  3. #3
    mixedgas's Avatar
    mixedgas is offline Creaky Old Award Winning Bastard Technologist
    Infinitus Excellentia Ion Laser Dominatus
    Join Date
    May 2007
    Location
    A lab with some dripping water on the floor.
    Posts
    9,665

    Default

    Considering the Lexel has a direct input autotransformer, you could get away with the VFD, provided you lock it at 60 Hz and have the duty cycle set correctly. Its going to get a little warmer from harmonics, and the tap settings may change. I know one fellow who bought a 25 horse VFD and is very happy.

    !
    The autotransformer usually will hum/moan quite loudly if its unhappy with its input power. Some lamination singing is normal when unloaded, but if it really starts to hum/scream/radiate loudly when loaded, stop.
    !
    If it blows up, which I really, really doubt it will, I have one spare 95 input transformer in my barn. Whatever you do, DO NOT attempt to regulate the laser power by constantly tweeking the VFD. You will have to check the cathodes with an amp clamp and voltmeter. Your "ABC" switch ranges will be a bit different. Depending on which cathode is installed, they are either 75 watt or 125 watt cathodes.
    ~
    Other psus, expecially Innova 70 and Innova 90, and newer SP products, will not be so forgiving. Most other systems use a phase fired SCR rectifier bank and really are set up for sine wave.
    ~

    So Lexel, probably, depends on the quality of the VFD's output.
    ~
    HGM/ALC three phase supplies are basically the same for three phase and single phase, and they are switchers. The first thing after the contactor is a bunch of water cooled rectifiers, the only difference being a few extra diodes for the three phase models. So for HGM, just make sure you rewire the cathode transformer, magnet supply rectifier and low voltage transformers, and interlocks for single phase operation. HGM checks for two phases connected by placing the low voltage transformer on a different phase, that's it. They then run just fine off a single phase line once you make all the standard adjustments on the control board. Then ensure you're not sagging the power line down low enough that the cathode gets too cold. The difference between a 9 series and 20 series is the amount of filter capacitance and passbank transistors HGM used. I've done this with HGM systems.
    !
    The one thing to NEVER do, is run a Lexel off a single or three phase SOLA transformer /power line filter that is designed to regulate its output voltage using Ferro resonance. I had a client do that, and the PSU was a mess inside. I repaired it on site, and tried it again, and again, the harmonic currents started to rip the PSU apart. . Customer refused to pay me for the service call, however, because the local surplus place told him the Sola step down regulator would be one third the cost of the 480 to 220 transformer I specified. He bought it, could not return it, and got burned. Club owners can be so stubborn and stingy some times, well all the time.. Since those only exist in computer centers and nuclear bunkers, I'm not too worried about it happening again.

    !
    Sir Steven
    Last edited by mixedgas; 05-18-2017 at 06:26.
    Qui habet Christos, habet Vitam!
    I should have rented the space under my name for advertising.
    When I still could have...

  4. #4
    Join Date
    Mar 2015
    Location
    Cambridgeshire, England
    Posts
    478

    Default

    Beware of VFDs that have sensorless vector control or similar control technology, these rely on the back EMF of the motor and will do bad things on a non-motor load.

  5. #5
    Join Date
    Jan 2006
    Location
    Charleston, SC
    Posts
    2,147,489,296

    Thumbs up *GREAT* replies everyone!

    Quote Originally Posted by dsli_jon View Post
    I assume this 'plan' is because the Lex 95 supplies are already 'in hand'
    Hi Jon;

    Yes, exactly. He's got a few other supplies as well (including a few SP 265's I think), but the fact that the Lexel has the buck-boost taps on the input transformer makes it a better choice I think.

    get a couple 1 / 3 American 909 'switchers' - They'll run I-90s to 35A+, off 1
    Wow! Had no idea they could run that hot... Given that the 909 is a much shorter tube than the I-90 I'm surprised the 909 PSU has that much extra headroom. But yeah, if he could source a pair of those supplies, that would solve everything. He has 100 amps of 220 single phase currently available, although if he needed a bit more to run a pair of those supplies it would be much easier (and cheaper!) to increase the single phase supply to the shop vs any of the options for 3 phase we're considering.

    the 95 supplies are Beast, weight-wise, compared to the 'svelte' 909-supplies
    Guessing that's due to the weight of that huge transformer in the 95 supply. I had a couple myself, and I agree they were heavy as hell. But they were pretty sweet supplies nonetheless. (Ran my HGM-20 up to over 11 watts!)

    there's a longtime 'supplier', out here, that is moving-shop
    I could check with him on Am 909 switcher-stock, and/or see what he'd charge to get a couple ready
    That would be cool. Please post back in this thread if you get a positive response from him.

    the 'Mods' are not bad - just a few wiring changes, mostly
    If this becomes an option, we'll no doubt post pics here in the thread during the modification process for comments on what we've changed so we don't screw it up!

    this one was 'wired' for 3, but all you had to do was hook yer 2 Hots / Gnd (thus, the 1-3 indicators in the rear..) and you'd be off..
    Just to confirm: this is similar to the old HGM-5 medical supplies that would run single phase or three phase, right? You get the error light on the third leg until you switch the PSU mode to single phase.?.

    Quote Originally Posted by mixedgas View Post
    Considering the Lexel has a direct input autotransformer, you could get away with the VFD, provided you lock it at 60 Hz and have the duty cycle set correctly. Its going to get a little warmer from harmonics, and the tap settings may change. I know one fellow who bought a 25 horse VFD and is very happy.
    Hi Steve;

    He's actually powered the PSU up on the VFD once, but he didn't try to light the tube on it. He definitely heard the harmonics in the input transformer though. (It was singing!) Heat is also something we're concerned about, both for the transformer and the filter caps.

    Was talking with Aron Bacs about this, and he cautioned us about the fact that since square-wave power has essentially infinite harmonics, and a lot of those frequencies will make it through the transformer to the filter caps, we need to worry about how the ESR on those caps will change with frequency... Some caps really go to shit at higher frequencies, and if the ESR goes up, the caps will get hot. Heat + electrolytic caps usually equals BAD THINGS HAPPENING...

    Some lamination singing is normal when unloaded, but if it really starts to hum/scream/radiate loudly when loaded, stop.
    Good advice!

    DO NOT attempt to regulate the laser power by constantly tweeking the VFD. You will have to check the cathodes with an amp clamp and voltmeter. Your "ABC" switch ranges will be a bit different.
    With everything unloaded, he checked the output voltages and they were OK. He had intentionally tapped a bit on the low side, but the resulting output voltage was right on. The harmonics are a concern though.

    Other psus, expecially Innova 70 and Innova 90, and newer SP products, will not be so forgiving.
    This was my thought as well.

    Lexel, probably, depends on the quality of the VFD's output.
    Not sure exactly how bad the output is. Initially we assumed it was pure square wave. Now it looks like it might be a stepped square-wave. (Pseudo-sine) Need to put a 'scope on the output and check it... It's supposed to support 3 Khz switching, so to me that sure sounds like a stepped square wave.

    HGM/ALC three phase supplies are basically the same for three phase and single phase, and they are switchers. The first thing after the contactor is a bunch of water cooled rectifiers, the only difference being a few extra diodes for the three phase models.
    This is what Jon was suggesting above. I just didn't realize that the 909 supply would handle a long tube like an I-90... Surely an HGM-20 supply would work too. Maybe an HGM-8?

    So for HGM, just make sure you rewire the cathode transformer, magnet supply rectifier and low voltage transformers, and interlocks for single phase operation.
    Hmmm... I thought I remember a switch on the back that selected single or 3 phase operation? Or was that only on the HGM-5 supplies? (Damn, it's been too long since you and I worked on all this crap years ago, and I've forgotten most of it!)

    HGM checks for two phases connected by placing the low voltage transformer on a different phase, that's it. They then run just fine off a single phase line once you make all the standard adjustments on the control board. Then ensure you're not sagging the power line down low enough that the cathode gets too cold. The difference between a 9 series and 20 series is the amount of filter capacitance and passbank transistors HGM used. I've done this with HGM systems.
    So of the two, an HGM-20 supply would likely have more filter caps and a heftier passbank, right? (Seem to remember that the HGM-20 tube maxed out at 50 amps or something...)

    Quote Originally Posted by NobleGas View Post
    Beware of VFDs that have sensorless vector control or similar control technology, these rely on the back EMF of the motor and will do bad things on a non-motor load.
    Did not know anything about this. Would there be something about this in the manual? What's a good way to check?

    Thanks to everyone for all the input! And I do apologize for the late reply. Life has gotten rather complicated for me recently, and I haven't had much free time over the past 6 weeks or so. Hoping things will calm down soon so I can get back to playing with cool toys!

    Adam

  6. #6
    mixedgas's Avatar
    mixedgas is offline Creaky Old Award Winning Bastard Technologist
    Infinitus Excellentia Ion Laser Dominatus
    Join Date
    May 2007
    Location
    A lab with some dripping water on the floor.
    Posts
    9,665

    Default

    A HGM 8 and HGM 20 are the same chassis with a few more passbank transistors on the 20, plus the three phase bridge on the 20. A HGM 5 is a condensed version of a 8, modified for air cooling by the addition of fins. I have a hard time telling a 20 and a 8 apart. The circuit cards are nearly identical, although some older versions of HGM5 use a bit different control chip, however the cards are nearly interchangeable. There is no phase indicator lights, all they do on the three phase models is add a transformer and relay to sense if any two legs are there, and the third leg is sensed by running the water flow solenoid off it.

    If you pop open an ALC 909 or 920 rack supply, you will find a HGM module in it, they had the same parent company. 909 is basically an eight, and 920 is basically a HGM 20.


    Just beef up the filter caps on the 95 supply to deal with the harmonics. Once the plasma is lit and the inductor is saturated, most of those harmonics turn into plasma current. I would double up on the filter caps. If a 95 is struggling, a oscilloscope on the light sensor output will show a triangle wave, if the depth of the filter cap failure is immense, about 25% of the light voltage will be a big triangle wave, and you'd see it in the tube current too. . (If you need to how I know, professor ignored my advice to replace a cap, when he needed a critical experiment running. The tube failed 48 hours later)

    Steve
    Qui habet Christos, habet Vitam!
    I should have rented the space under my name for advertising.
    When I still could have...

  7. #7
    Join Date
    Mar 2015
    Location
    Cambridgeshire, England
    Posts
    478

    Default

    Quote Originally Posted by buffo View Post

    Did not know anything about this. Would there be something about this in the manual? What's a good way to check?

    Adam
    Yes, it will be on the feature list, sensorless vector tuning, auto tuning, optimised control are all terms that are used for similar technologies; basically if it doesn't mention any advanced drive technology, it's likely to be fine.

  8. #8
    Join Date
    Nov 2007
    Location
    SoCal / San Salvador / NY
    Posts
    4,018

    Lightbulb

    Heya Chief -

    Quote Originally Posted by buffo View Post
    Wow! Had no idea they could run that hot...
    Yep, IIRC, they would usually top out at about 35-36A, but usually were 'happiest' at about 32-35.. not much difference in Power, really.. Even the I-90 supplies (or I-70s / their supplies..) that could run them to 40A, there was not a 'Uuuuge' difference in Power, between 36-40A, anyway, so.. Unless your goal is 'Moon-shots' you'll prolly be fine with the 909's top-end..

    ..I recall getting Easily 4-6W of WL (depending on gas / 'color balance'..) out of an Ar/Kr, up to ~7-8W of Ar, and ~2-2.5W of Kr, out of a standard I-90 MRA tube in a 909-res, with the 909-supply, as-shown above, at those currents, so.. Should be, I'd think, more than 'good enough' (?)

    Quote Originally Posted by buffo View Post
    ...I agree they were heavy as hell. But they were pretty sweet supplies nonetheless.
    Click image for larger version. 

Name:	DSC04628.JPG 
Views:	10 
Size:	110.3 KB 
ID:	52280 ...Ayep, those things were a tank 'n a half, but yes, nice (and nice controls.. Click image for larger version. 

Name:	DSC04637.JPG 
Views:	7 
Size:	91.5 KB 
ID:	52281 ..The 909-supply is Much simpler (but simpler to work on, as well..) and the Controls are more than sufficient, for 'light show-caliber needs' (vs 'Lab', etc..)

    Quote Originally Posted by buffo View Post
    Just to confirm: this is similar to the old HGM-5 medical supplies ..You get the error light on the third leg until you switch the PSU mode to single phase.?
    Uh, well.. In the 1 / 3-sense, yes, but the 909 doesn't have any 'phase-sensing' / feedback circuitry in it, it's pretty 'comparatively dumb', but.. I just added those pre-breaker neon 'phase lights', for convenience (and one 'breaker On' neon, in-front..) That way, you could see, at least, that you Had positive 1, 2 or 3 connex, and if your legs (in your distro, etc..) were 'sequential' or not.. It didn't 'care' about order, etc..

    Anyway, will report any info, as-relevant..
    j
    ....and armed only with his trusty 21 Zorgawatt KTiOPO4...

Posting Permissions

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