capacitor gets hot in a HF-HV resonant circuit

I don't see a schematic, just white.

Also, ugh, Tinypic sucks. Can't you use Imgur or something?

Tim

For what purpose? How much current? Does it have to deliver that into an impedance?

It better be a pretty big transformer (Ae > 265 mm^2).

A couple turns? So the MOSFET side sees a couple hundred Vpp??

I'm impressed your transistor hasn't simply poofed yet..

Quite possible. Start at the beginning. What is it for, what is it driving, if anything?

Tim

On Tue, 27 Oct 2015 15:28:23 -0500, "Tim Williams" Gave us:

Wima makes pretty good caps for stuff like this.

That sounds right.

Looks like the ringing oscillation circulates lots of current between the coil and the cap. You could Spice that and maybe assume equal losses in the coil and the cap, and estimate the power dissipation in the capacitor.

But if the cap is self-heating to 48C or some such, it's really fine. That's not very hot. Paralleling smaller caps, or buying low-ESR RF caps, would help if you do want to reduce dissipation in the capacitors.

Are they seriously thinking engineers in hi-rel places (the ones usually needing such info) can download and install software just like that?

Loss factor curves and such belong in datasheets. Or at least in supplementing documents that can be retrieved as data files in a safe format. Not executables and such.

Well, fine, military designs will get stuck in the '70s, using '70s era paper databooks and metal-can bipolar ICs. As they are. If they don't want to innovate, that's their problem. Well, sooner or later it's our problem, but nevermind.

For everyone else, just install it somewhere. Get a portable. eeePC, even VM on iPad or something I bet. Take a screenshot and print it out if you have to...

I haven't worked at a single place where I didn't have administrative rights to the machine I work on. One place tried to "fix" that, I GTFO'd. It's my understanding they're still floundering.

On a related note... Doesn't *everyone* have that cute little "not for life support" and etc. disclaimer, nowadays? Right in the datasheet? And "subject to change at any time -- uncontrolled" and such? So aren't all those datasheets useless to you? You have to work through sales to get written approval of that anyway, so the "document" format doesn't even matter.

Tim

What is so difficult to print a loss factor in a datasheet? The high cost of toner? Oil price? Well, there's PDF.

I usually move on. To manufacturers who have more complete datasheets. Less fluff, more data.

I don't do life support designs. In the world of life support it is not uncommon that companies have even their transistors custom-made or make those themselves. After all, you don't want a pacemaker in your chest to go beeeeep ... tchwock. At least I wouldn't.

At a previous job, the EE and software people had admin on all their machines. The informal agreement with the official IT people was that IT would cheerfully replace broken hardware and assist with company- standard software like Orifice, Outhouse, etc. If you needed anything else, you could install it, and if you broke it, you fixed it. If you had hardware that wanted to speak IP to other hardware, but didn't need the Internet or company intranet, you could get switches and Ethernet cables and create a little LAN for it all, no problem.

That was when they were the US arm of a European company. The same European company had another US arm that did defense stuff. They decided to put the place I worked underneath the defense company. The initial proposal was to lock everything down; the engineering manager asked if the IT department was prepared to have dedicated IT staff to install and change things for the engineers.

After a little discussion, the usual thing happened: all the engineers got two PCs on their desks. One had Orifice and Outhouse and intranet and Internet, but was locked down and otherwise useless. They could install whatever they wanted on the other PC and use it to get work done.

Matt Roberds