Quote:
Originally Posted by Xringer
Since I have a dry contact from the controller (closed with calling for heat, open when the tank is hot enough),
I could use it to switch +12V into the Rin..
When the contact opens, Rgs would pull the gate down to Zero volts...
I'll check out the specs on the FETs to see what resistors will be needed to
develop the gate voltage needed for full turn-on..
The PV panel array can't deliver any more than 8Amps. It impossible for
PV to exceed it's amperage rating, since it's limited by sun light.
Only when you get certain solar-cloud-effects, will you get more power..
And that will be on the order of 2 to 20% more..
It would take a supernova to get much more than 20%.. 
I can short out the PV inside the house and never blow the 10A fuse on the array..
Which is why it's a waste of time to put fuses on solar panels.. 
Not sure why those DC SSRs suck so bad. I think it might be the DC signal source I'm using. Spikes maybe..
I'll have to think about a way to pull down the DC signal level quickly in case of a grid failure.
Don't want the gate voltage to drift down slowly, after the power supply input dies.. Need to load it down..
I figure the (AC powered) controller contact will fall out fast enough to prevent much over-heating of the SSR, if the AC failed.. |
I ẃould definitely rig up some kind of gate driver circuit to the mosfet. Since you have experience with vacuum tubes, you know how sensitive the grids are to interference and oscillation if you don't tie them down. Mosfets are the same way. Where a tube will redplate or arc if it's not happy, a mosfet will just die. They are not forgiving at all. All it takes is a couple of milliseconds of something nasty to ruin a mosfet.
Relays will have some "bounce" in the contacts when they open or close, which could cause the mosfet to leave saturation. At 100VDC, driving a 13 ohm load, the mosfet would only survive a fraction of a second if it rose above a couple of ohms above its minimum resistance. Even on a big heat sink. A schmitt trigger or pulse stretcher are a few ideas that come to mind. Anything that makes sure the fet gets a nice discrete transition when changing states should make it last a long time.
You definitely have something strange going on there. The IRF250 transistors you are using are heavier than what I would have started out with. They are pretty tough as far as power mosfets go. The only circuits I have seen these die in, they have been driving big flyback transformers. They throw lightning for a while before the mosfet finally dies.
Another product I should mention for pipe joints is this stuff:
I originally used this stuff in the Navy on joints that had to be watertight but easily separated for maintenance. Naturally, they had jars and buckets of the stuff, and it wasn't blue, but gray. Since then, I have used it on more surfaces than I can remember. It is a lot like the nylog sealant used on flare joints, made specifically for sealing machined surfaces, and for holding o-rings and gaskets in place during assembly. On threaded fittings, it works in places where pipe dope or thread tape leaks. OEM's like Jaguar, Harley Davidson, NASCAR, and Rolls Royce use it all the time.
Like I have said before, I don't trust teflon thread tape. On pneumatic fittings, I will use it if nothing else is on hand. For anything hydraulic, I would rather run to the store.