SSR? a Dump Load for PV array

[Xringer]

When I attempted to use a DC SSR, it was in series with the + line to the heating element.
http://i46.photobucket.com/albums/f1...psc5151ccc.jpg

It worked for a while then failed. Then I tried to use an AC SSR.
Also a failure..

Now, I'm testing with a real relay..

[Robaroni]

Originally Posted by Xringer

When I attempted to use a DC SSR, it was in series with the + line to the heating element.
http://i46.photobucket.com/albums/f1...psc5151ccc.jpg

It worked for a while then failed. Then I tried to use an AC SSR.
Also a failure..

Now, I'm testing with a real relay..

My hunch is that the DC SSR had the MOSFET on the high side or the inductive collapse was considerable. When you place a MOSFET (N channel) on the high side you need a charge pump to drive the device to saturation. Also the IRFP250N shows an RDSon of .075 ohms but you have to realize that manufacturers show the optimum statistics. RDSon varies at voltage. As voltage rises the RDSon could easily exceed 1 or 2 ohms. At 2 ohms the power the transistor waste in heat dissipation is I squared R (7x7x2) or 98 watts. That's considerable and the transistor will need massive heat sinking so use at least 4 to 6 transistors in your circuit with a good heat sink.
Since this is an inductive load make sure to use a substantial diode across the heating coil (even if you use a relay instead of a transistor) to keep the kick back from going through the device.

Rob

[Xringer]

I didn't think the resistance of the DC SSR would ever get up very high.
I never noticed any build up of heat when it was still working..
The 3-32v DC control voltage into DC SSR was 12V which put it into full saturation.
At least as far as I could tell with the voltmeter.

The series heater load of 13 ohms isn't very inductive.
If it was, the standard AC hotwater heater controls wouldn't last very long, with their small air-gap contacts.

Anyways, I've sworn off DC SSRs.. I will never buy another one..
They cost too much and die too easy. (Static discharge from my finger?).

Didn't have much luck when I tried the cheap way out, with an AC SSR..
Not sure why they shouldn't work. Seems like they should.
Maybe the first one was a dud, something intermittent in the control side.?.

So, when the Golf kart relay dies, I'll try the 200v 30a HEXFETs..
I am thinking of stacking them.

Thanks,
Rich

[Robaroni]

Originally Posted by Xringer

I didn't think the resistance of the DC SSR would ever get up very high.
I never noticed any build up of heat when it was still working..
The 3-32v DC control voltage into DC SSR was 12V which put it into full saturation.
At least as far as I could tell with the voltmeter.

The series heater load of 13 ohms isn't very inductive.
If it was, the standard AC hotwater heater controls wouldn't last very long, with their small air-gap contacts.

Anyways, I've sworn off DC SSRs.. I will never buy another one..
They cost too much and die too easy. (Static discharge from my finger?).

Didn't have much luck when I tried the cheap way out, with an AC SSR..
Not sure why they shouldn't work. Seems like they should.
Maybe the first one was a dud, something intermittent in the control side.?.

So, when the Golf kart relay dies, I'll try the 200v 30a HEXFETs..
I am thinking of stacking them.

Thanks,
Rich

Rich,
Here's a high side mosfet arrangement I've used. You want to keep the gate source voltage within specs, most are 20v max difference so if you have 200 volts on the source you want no less than 180 on the gate or it will blow. The 12v zener keeps the source/gate voltage in check.

With that said, it's much easier to run the mosfet on the ground leg. The source goes to ground and the gate never rises about 12 or so volts without zeners.

The thing that kills these guys often is the inductive spike you get from switching the mosfet on and off. That inductance has to go somewhere. You may not think it's much but engineers (including me!) with boxes of blown transistors will tell you they underestimated the inductive kick. If you have an inductance meter you might want to check your heater coil.

Rob

[Xringer]

Got a nice sunny day! Bad news for the golf cart relay..
The insulator (nylon?) on top started to melt..

http://ecorenovator.org/forum/attach...y-p1020909-jpg

So, it's on to plan 'F'..

[Robaroni]

Originally Posted by Xringer

Got a nice sunny day! Bad news for the golf cart relay..
The insulator (nylon?) on top started to melt..

http://ecorenovator.org/forum/attach...y-p1020909-jpg

So, it's on to plan 'F'..

What gauge is the wire and what current is it carrying?

[Xringer]

It's 12 awg running at 8A DC Max.. No problems with the wire.

The contact in the relay had very low resistance at first, but soon
I could measure about 1v across it. In the end, it was around 3 volts.
24w of heat on the contacts for a few hours is bad for low temp insulators..

[Robaroni]

Originally Posted by Xringer

It's 12 awg running at 8A DC Max.. No problems with the wire.

The contact in the relay had very low resistance at first, but soon
I could measure about 1v across it. In the end, it was around 3 volts.
24w of heat on the contacts for a few hours is bad for low temp insulators..

Two things:

First, relays have something called 'follow'. When the contacts touch as the relay operates they should very slightly rub, this keeps the contacts clean and prevents resistance build up. Normal relay contact resistance should be under 0.2 ohms.

Secondly, insulation melts because the wires get hot, regardless of the current. What those wires are telling you is that their diameter is too narrow. Measure the current through the wires and then look up a wire amperage chart on Google. Also the relay should have voltage and current specs on it. You might be able to read the numbers and get the specs off the net.

Rob

[Xringer]

It's 20 amp wire and I'm using 8 amps. There is almost zero loss in the wire.
The insulation that's melting isn't on the wire. It's on the relay terminals.

The contact points were heating up, that heat transferred into the 1/4|20 terminal bolts.
The bolts heated up and the insulation around the bolts got hot and started to melt..

[Robaroni]

Originally Posted by Xringer

It's 20 amp wire and I'm using 8 amps. There is almost zero loss in the wire.
The insulation that's melting isn't on the wire. It's on the relay terminals.

The contact points were heating up, that heat transferred into the 1/4|20 terminal bolts.
The bolts heated up and the insulation around the bolts got hot and started to melt..

The worst case power across the contacts is equal to I^2xR. or 8^2 x 0.2 = 13.6 watts. That might be enough to do damage but I just measured the contact resistance of two 12V relays rated at 30A using the 4 wire resistance method on a calibrated 6-1/2 digit Agilent meter. both relays came in under .002 ohms. Each, at less than 0.014 watts dissipation, would easily carry your current without problems or melted insulation.

You either have a bad relay, a relay that is underrated for the current or you have another problem. My guess is that since you blew an SSR and have a relay over heating you're missing something. Is it possible that the cold resistance of the element is increasing the current through it? How are you measuring the current? Did you watch the current from turn on?

Another thing, since the insulation melted that means that the current through the wires is exceeding their carrying ability. So I'd say you're not measuring the actual current. What's the element ratings? AC element?

Rob