isolated gate drives / General Science and Electronics / Forums

Forums

4hv.org :: Forums :: General Science and Electronics

« Previous topic | Next topic »

isolated gate drives

first previous 2 3 4 5 next last

Move Thread

LAN_403

Avalanche

Wed May 17 2006, 11:04AM

Registered Member #103 Joined: Thu Feb 09 2006, 08:16PM
Location: Derby, UK
Posts: 845

To get your isolated power supplies for each driver, couldn't you use a big capacitive divider (with say 4 caps) and feed it with 100v+? I haven't thought this through properly but the thought just popped into my head...

ragnar

Wed May 17 2006, 10:17PM

Registered Member #63 Joined: Thu Feb 09 2006, 06:18AM
Location:
Posts: 1425

A capacitive divider is an interesting idea for AC, but I feel the distributed power / isolation transformer is more appropriate for varying input voltages, etc =)

Surely, EVR, you've worked with MSOP powerpad packages? (With the ground/die exposed at the bottom of the chip, which you reflow solder to the board to take heat away? =)

...

Wed May 17 2006, 10:58PM

Registered Member #56 Joined: Thu Feb 09 2006, 05:02AM
Location: Southern Califorina, USA
Posts: 2445

I have never worked with said packages, but in the one board I have made...
I had a voltage regulator in a little package which I needed to dissipate a little over a watt with, so I just put about 1cm by .5cm copper area on the tab and put a few vias (hole with piece of wire soldered on either side) around it into a similar sized piece of copper on the back. It seemed to work just fine, both sides got equally as warm...

I would say for these chips you should use a H shaped copper fill so you have tabs on either side of the chips, and from put a bunch of pins through to the bottom of the board... I suppose if you wanted to keep the flat format of your boards you could use some copper foil on either side to get the heat down to the bottom of the board.

As to the supply... I would have to say the best method would be to use a little ferrite core powered by a mazzilli driver or just a simple fixed frequency oscillator... Then wind a secondary for each board and have fun. Even normal magnet wire should be able to handle the <500v that would ever be across the bridge...

HV Enthusiast

Wed May 17 2006, 11:11PM

Registered Member #15 Joined: Thu Feb 02 2006, 01:11PM
Location:
Posts: 3068

Yes, I have worked with MSOPs, although i never hand soldered them. They would definitely be an improvement over DIP packages as with a DIP package, the only heat path is through the pins themselves which is pretty poor.

ragnar

Wed May 17 2006, 11:16PM

Registered Member #63 Joined: Thu Feb 09 2006, 06:18AM
Location:
Posts: 1425

Steve Conner used a ZVS as the sinewave wouldn't induce as much noise as sharp squarewaves... IIRC. I'll be doing something similar, it's the best way.

Now I don't have to execute my "UCCs under oil" plan =P

Steve Conner

Thu May 18 2006, 09:07AM

Registered Member #30 Joined: Fri Feb 03 2006, 10:52AM
Location: Glasgow, Scotland
Posts: 6706

That's exactly what I'm using for my isolated gate drivers, the Mazzilli oscillator circuit. I wanted sine waves for quite a long and complicated reason.

I designed my isolated drivers as four drive boards that mount very close to their IGBTs, and plug into the main PLL box with Cat5 network cables. So I had to feed the HF AC power down this cable along with the gate drive signal. I thought if I used a square wave, the edges might couple capacitively over into the gate drive signal.

In practice I found that when I rectify the sine wave on the gate driver board, the diode recovery spikes add lots of crud anyway.

But the Mazzilli circuit is real nice and simple. I found that if I used a choke input filter on the rectifier, the voltage regulation was good enough that I could get away without regulator ICs on each board. I just used a zener to draw some current, since the voltage skyrockets if you draw less than a critical value.

I used a ferrite E core with airgap for the mazzilli circuit, with a single output winding that fed a little ferrite toroid 1:1 isolating transformer on each board. I just did that so the different cores on the Cat5 cable wouldn't have high voltages between them. If you wanted to save some hassle, it should be fine to combine the oscillator and isolation transformer into a single component with 5 windings, the way Steve Ward etc. did. Just make sure you remember what is grounded and what is floating on your DC bus voltage.

I haven't time to think through the capacitive divider thing that Avalanche mentioned, but it just sounds plain wrong. :(

Marko

Thu May 18 2006, 05:08PM

Registered Member #89 Joined: Thu Feb 09 2006, 02:40PM
Location: Zadar, Croatia
Posts: 3145

Wouldn't it be simplest to use one transformer like steve ward's big DRSSTC (a bunch of CAT5 wrapped around flyback core), possibly rectified at the place and fed to each driver?

Desmogod

Fri May 19 2006, 07:46AM

Registered Member #139 Joined: Sat Feb 11 2006, 11:01AM
Location: Perth, Western Australia
Posts: 358

Firkragg wrote ...

Wouldn't it be simplest to use one transformer like steve ward's big DRSSTC (a bunch of CAT5 wrapped around flyback core), possibly rectified at the place and fed to each driver?

I think you are missing the point of the exercise.

ragnar

Sat May 20 2006, 02:38PM

Registered Member #63 Joined: Thu Feb 09 2006, 06:18AM
Location:
Posts: 1425

It was a dark and stormy night... the rain fell in torrents...except at occasional intervals, when it was checked by a violent gust of wind which swept up the streets (for it is in Sydney that our scene lies), rattling along the housetops, and fiercely agitating the blue glow of the fluorescent lamps that struggled against the darkness...

And deep in the laboratory... I started pulling isolated-gate-driver circuit boards out of my arse.

The 'ghosting (pun intended)' you see is from the nosecone rubbing oil over the surface of the board as it engraves.

ragnar

Sun May 21 2006, 10:39AM

Registered Member #63 Joined: Thu Feb 09 2006, 06:18AM
Location:
Posts: 1425

OK, since none of you buggers replied to my prose... I'm going to... post more pictures. (These ones are better lit since I wasn't working at midnight)

Boards had until now been conceived, designed, flipped, engraved.

Today I drilled all holes (73 holes per board, a total of 730 holes over ten boards) using 0.8mm solid tungsten carbide drills by UNION TOOL, which performed very well after I biased the axial slop in my cheap drill press to the left with an ockey strap.

These were previously 3oz double sided copper boards. I'd removed one side of the copper by peeling. Now with my excellent co-ordination since they were separated, I weeded out the groundplane from all boards (since there were artefacts from engraving, copper shavings and whiskers, and general distance issues etc) so it's now just the copper tracks.

Now I ground all ten boards' edges to size, gave them a quick acetone clean, and they are ready for tinning! =)

Stuffing will be the fun part, because then I can see how my isodrives actually perform!

Comments?

first previous 2 3 4 5 next last

Moderator(s): Chris Russell, Noelle, Alex, Tesladownunder, Dave Marshall, Dave Billington, Bjørn, Steve Conner, Wolfram, Kizmo, Mads Barnkob