Speaking of high frequency transformer stuff...
in [Design]
|
Prev: Noise in electrolytic capacitors
Next: LTspice issues.
From: Joerg on 21 Feb 2010 18:50 Tim Williams wrote: > How would you tackle the problem(?) of 1MHz, 100V common mode on > transformers or whatever coupling you're using? Think high side drive. Is > an ordinary gate drive transformer suitable, does it need special > construction or drive? Or would it be better to use a big fat CM choke to > bring things down to earth? > > I don't think optical coupling would be very useful. I've already had poor > results with 6N136's at 20kHz. I know there are fancier parts available, > too, some with built in gate drivers. I've seen it before where fast edges > will couple through the pri-sec capacitance of a GDT, even when driven hard. > Not sure what you are concerned about but if it's the secondary winding capacitance to whatever you could make that winding out of coax and hang the shield onto the source of the upper FET. That's how we also get rid of E-field stuff leaking into RF transformers. It doesn't have to be any fancy coax. Thin audio line might do if it can take the voltages and temperatures. -- Regards, Joerg http://www.analogconsultants.com/ "gmail" domain blocked because of excessive spam. Use another domain or send PM.
From: Tim Williams on 21 Feb 2010 21:20 So, here's a possible optoisolated solution: http://webpages.charter.net/dawill/Images/1MHz_Induction_Heater.gif Notice the feedback circuit has time limited positive feedback, so shoot-through is impossible, high and low side on-times are equal (within component variation), and if it stops oscillating, it just stops (accordingly, something needs to be provided to kick it moving in the first place, which isn't shown here). Not sure if everything is quite fast enough. There's a good ~100ns of delay in the loop, which is wholly 1/5th of the half cycle time. I'd rather not spend more bucks on an LT1016 or something like that. Incidentially, Mouser seems to have dropped LT from their catalog?? The timer is controlled for phase lock on the resonant tank, and either phase or supply voltage controls output power. (Someone mentioned PFC SEPIC recently, something like that could be handy here for generating 0-200V from a 90-265VAC input. No need for postregulation in an app like this, 120Hz ripple lets you know it's running.) Regarding this thread, I've kept the application general, since 1MHz power transformers work just as well for induction heating as for generic switchers. Specifically, I am making another induction heater, but lessons learned here will do just as well anywhere else (like D from BC's 1MHz switcher with too much transistor-heatsink capacitance). Tim -- Deep Friar: a very philosophical monk. Website: http://webpages.charter.net/dawill/tmoranwms "Tim Williams" <tmoranwms(a)charter.net> wrote in message news:hlpo5u$re4$1(a)news.eternal-september.org... > How would you tackle the problem(?) of 1MHz, 100V common mode on > transformers or whatever coupling you're using? Think high side drive. > Is an ordinary gate drive transformer suitable, does it need special > construction or drive? Or would it be better to use a big fat CM choke to > bring things down to earth? > > I don't think optical coupling would be very useful. I've already had > poor results with 6N136's at 20kHz. I know there are fancier parts > available, too, some with built in gate drivers. I've seen it before > where fast edges will couple through the pri-sec capacitance of a GDT, > even when driven hard. > > Tim > > -- > Deep Friar: a very philosophical monk. > Website: http://webpages.charter.net/dawill/tmoranwms >
From: Winfield Hill on 22 Feb 2010 12:52 Tim Williams wrote... > > How would you tackle the problem(?) of 1MHz, 100V common mode on > transformers or whatever coupling you're using? Think high side > drive. Is an ordinary gate drive transformer suitable ... My favorite Intersil HIP4080 series of high-side/ low-side driver ICs easily goes to 1MHz. Check 'em out. No transformers, no optical couplers, nada. Ahem, HIP4081A in stock at DigiKey. Rated at 80V (plus 15V swinging gate-drive), but hey, go ahead and push it to 100V if necessary, or adjust your turns ratio. -- Thanks, - Win
From: Jon Elson on 25 Feb 2010 17:11 Winfield Hill wrote: > Tim Williams wrote... >> How would you tackle the problem(?) of 1MHz, 100V common mode on >> transformers or whatever coupling you're using? Think high side >> drive. Is an ordinary gate drive transformer suitable ... > > My favorite Intersil HIP4080 series of high-side/ > low-side driver ICs easily goes to 1MHz. Check 'em > out. No transformers, no optical couplers, nada. > Ahem, HIP4081A in stock at DigiKey. > > Rated at 80V (plus 15V swinging gate-drive), but hey, > go ahead and push it to 100V if necessary, or adjust > your turns ratio. > > I built a servo amp with the 4080 over a decade ago. The chip is rated for 80 V, but the applications manager eventually admitted to me "Oh, you're doing REALLY good to get them to run at 59 V, none of our other customers ever got them to last above 54 V!" Oh, that was a REAL NICE admission! So, I eventually redesigned the whole thing to use the IR2113 half-bridge driver chip. They have been VERY reliable, way above the voltages I normally run at. The one gotcha is that the common point between the two transistors cannot be allowed to go negative, so I had to put an ULTRA-fast diode across the low-side transistor. (The body diodes in the FETS are incredibly slow to turn on.) Unless the HIP4081A is on some improved process from the 4080 (I doubt it) you'll never get close to 80 V, even. Winfield, have you ever run the 4081A up above 40 V or so? Jon
From: Winfield Hill on 25 Feb 2010 21:50
Jon Elson wrote... > >Winfield Hill wrote: >> Tim Williams wrote... >>> How would you tackle the problem(?) of 1MHz, 100V common mode on >>> transformers or whatever coupling you're using? Think high side >>> drive. Is an ordinary gate drive transformer suitable ... >> >> My favorite Intersil HIP4080 series of high-side/ >> low-side driver ICs easily goes to 1MHz. Check 'em >> out. No transformers, no optical couplers, nada. >> Ahem, HIP4081A in stock at DigiKey. >> >> Rated at 80V (plus 15V swinging gate-drive), but hey, >> go ahead and push it to 100V if necessary, or adjust >> your turns ratio. > > I built a servo amp with the 4080 over a decade ago. The chip is rated > for 80 V, but the applications manager eventually admitted to me "Oh, > you're doing REALLY good to get them to run at 59 V, none of our other > customers ever got them to last above 54 V!" Oh, that was a REAL NICE > admission! So, I eventually redesigned the whole thing to use the > IR2113 half-bridge driver chip. They have been VERY reliable, way above > the voltages I normally run at. The one gotcha is that the common point > between the two transistors cannot be allowed to go negative, so I had > to put an ULTRA-fast diode across the low-side transistor. (The body > diodes in the FETS are incredibly slow to turn on.) > > Unless the HIP4081A is on some improved process from the 4080 (I doubt > it) you'll never get close to 80 V, even. Winfield, have you ever run > the 4081A up above 40 V or so? Whoa, I've been a fan of the HIP4081A for more than 15 years. The "A" variant, I must admit. In that time I never heard anyone 'dis them. My own designs use the parts (purchased repeatedly over a 7-year period) up to and over the maximum voltage spec rating, 80, 90 and 100V, and to and over 1MHz. In my bench prototype testing I've had some ugly setups, creating substantial V = L dI/dt voltages, that get out of control in a real hurry (try 50nH*4A/10ns = 20V, be careful), yet I only had one failure, during testing, which quickly reverted to reliable operation after I tightened up critical wiring to reduce the current-loop area. I have decades (OK, 1.5 decades) of successful 24-7 operation of my designs (the guys leave everything running continuously) using these parts at or near their limits. Recommended. BTW, I'm also a fan of the IR half-bridge drivers, but they are relatively slow parts, meant for another kind of design. Their slow speeds mean you won't so quickly run into wiring issues, because for them L dI/dt has a much slower dt term. I use these to make elegant 400V pulse generators (posted on s.e.d.), but definitely not to run at MHz frequencies. -- Thanks, - Win |