Solid State Relay Puzzle

[Dockhead]

Well now here's something interesting!

I hooked it up and tried it -- the relay works.

But how about this -- the bloody Simrad AP seems to be backward from Raymarine -- there is power across the Drive Engage terminals when the autopilot is disengaged. Whereas, my Raymarine drive needs power to engage. Bloody hell. I can't believe I didn't read about this anywhere; and Simrad tech support didn't say anything about it.

So it looks like I need a normally closed relay, instead of normally open??!!

[s/v Jedi]

Originally Posted by Dockhead

Well now here's something interesting!

I hooked it up and tried it -- the relay works.

But how about this -- the bloody Simrad AP seems to be backward from Raymarine -- there is power across the Drive Engage terminals when the autopilot is disengaged. Whereas, my Raymarine drive needs power to engage. Bloody hell. I can't believe I didn't read about this anywhere; and Simrad tech support didn't say anything about it.

So it looks like I need a normally closed relay, instead of normally open??!!

huh?! No, I think you need to do the complete dock-side initialization routine gain. I think the Simrad AP can handle both ways, it tests that during the init routines.

Which Simrad AP do you have?

[Dockhead]

Originally Posted by s/v Jedi

huh?! No, I think you need to do the complete dock-side initialization routine gain. I think the Simrad AP can handle both ways, it tests that during the init routines.

Which Simrad AP do you have?

AC42.

It's now working -- I successfully completed the dockside initialization. Urra!

I had made a mistake in the wiring -- I was switching the negative rather than positive side.

So I'll add the diode and I think I'm set!!!

Thanks Nick, and everyone else who helped!!!!

[Dockhead]

Originally Posted by s/v Jedi

Yes, that's all you need. I'm pretty sure a 5A rectifier diode will be fine. Like you have in spare for the solar panel bypass?

For schematic: just put it parallel to the load (that AP solenoid) with the cathode (the pointy end marked with a band) pointed towards the positive side of the voltage source.

EDIT: but as near to the relay as possible, not at the location of the solenoid.

Something like this: http://www.farnell.com/datasheets/1026889.pdf ?

[s/v Jedi]

Originally Posted by Dockhead

Something like this: http://www.farnell.com/datasheets/1026889.pdf ?

Hmmm. I think it's still way overkill. Just get a cheap 1N4007 and that should last forever. You might not even need it but I don't know the solid state relay you use so why not. If you have another rectifier diode somewhere, use that. Take it out of an old radio

I love it how the Simrad pilots do their init routines.. finding out what kind of hydraulics you have, which way to turn etc.

[goboatingnow]

Sorry I missed this thread. Just a 1 to 2 amp diode is all you need forget the capacitor

On fact the MOSFET has a body diode protecting it any way. It's just a be sure to be sure

Solid state is more reliable then a conventional relay

In some very inductive circuits ( with significant parasitic and residual inductance ) with fast switching. You would usually use a snubber circuit , diode in series with a Parallel resistor capacitor, but not in this case.

Anyway it worked
Dave

[Paul Elliott]

I think you need a 4A diode, or at least one that can handle 2A transients. The coil current is 2A, so when the relay opens that 2A will be flowing through the clamp diode. The 1N4007 is a 1000V, 1A diode and I'm afraid that it wouldn't take the current.

OK, I just checked the 4007 specs and it is rated for 30A non-repetitive surge. It's probably going to survive, but I would prefer one with a higher current rating.
In any case, I think the transient clamp diode built into the relay will be sufficient.

[Dockhead]

Thanks, guys! I guess I'll put in the diode just to be sure. The relay wasn't cheap! Anyway it's very cool that it's all working now.

Now I move on to making a power supply for the radar and hooking up the heading sensor.

Have to pull cable to the binnacle and get the Triton and pilot control pad hooked up.

Then finish pulling cables in the mast and get the antennas all connected. Will have to learn to solder N type connectors between now and Monday evening. The cranes arrive on Tuesday morning to put the mast up!

[Reefmagnet]

I used to do 24vdc industrial control circuits for a living for a while. A 1N4004 freewheel diode will work fine. That's all I ever used and never had problems. Costs a few cents at your local electronics hobby store.

[s/v Jedi]

Originally Posted by Paul Elliott

I think you need a 4A diode, or at least one that can handle 2A transients. The coil current is 2A, so when the relay opens that 2A will be flowing through the clamp diode. The 1N4007 is a 1000V, 1A diode and I'm afraid that it wouldn't take the current.

OK, I just checked the 4007 specs and it is rated for 30A non-repetitive surge. It's probably going to survive, but I would prefer one with a higher current rating.
In any case, I think the transient clamp diode built into the relay will be sufficient.

Yes, it's more about how much energy is stored in that coil when the power gets switched off and can the diode dissipate that and survive. I think it won't even heat up, plus, how many times does the AP gets switched off per second? The duty cycle is very low, we only have to deal with a single event.

It's best to try it and test it with the multimeter. Schedule a maintenance test after a month of use.

[s/v Jedi]

Originally Posted by Reefmagnet

I used to do 24vdc industrial control circuits for a living for a while. A 1N4004 freewheel diode will work fine. That's all I ever used and never had problems. Costs a few cents at your local electronics hobby store.

Exactly, the 1N4007 is even tougher and costs are negliable. They are tough.

[Dockhead]

Originally Posted by s/v Jedi

I love it how the Simrad pilots do their init routines.. finding out what kind of hydraulics you have, which way to turn etc.

Me too -- very cool. Can't wait to see how the pilot works at sea.

Concerning the duty cycle of the relay -- indeed, it switches only when you engage or disengage the pilot. So what is that, 20 times max per day at sea? I presently do about 50 sea days a year, so 1000 cycles in a 10 year life? Bah, I think I've overthought and over engineered all this. I have a 10x overkill relay which cost 50 pounds. I'll throw on the diode and forget about it.

The main thing is the device is now fully set up and working!

[Paul Elliott]

Dockhead, I think you're doing the right thing, and I think that the 1N4007 will be just fine, and also the internal diode will work should the 4007 some how open up. Do be aware that when diodes do fail, they sometimes fail as a short circuit. Make sure your circuit breaker, etc, can handle this, because sometimes coils also short due to poor insulation or mechanical abrasion.

But a couple of points that have no real bearing on this particular installation:
1) In this application, the 1N4007 is *not* tougher than the 1N4004. The 4007 has a higher reverse voltage rating, but in this application there is no significant reverse voltage. The two diodes have identical current ratings, and this is the critical parameter.

2) Just because a particular diode works in some 24V applications doesn't mean that it's going to be appropriate for all. The diode requirements are going to depend on the coil specs, in particular the current, the resistance, the inductance, and to a lesser extent the parasitic capacitance of the coil. We aren't likely to know some of these parameters, so we mainly look at the coil current. This is current that the diode is there to clamp.

I know, I'm just being an annoying precisionist. That's what I do for fun these days.

[Dockhead]

Originally Posted by Paul Elliott

Dockhead, I think you're doing the right thing, and I think that the 1N4007 will be just fine, and also the internal diode will work should the 4007 some how open up. Do be aware that when diodes do fail, they sometimes fail as a short circuit. Make sure your circuit breaker, etc, can handle this, because sometimes coils also short due to poor insulation or mechanical abrasion.

But a couple of points that have no real bearing on this particular installation:
1) In this application, the 1N4007 is *not* tougher than the 1N4004. The 4007 has a higher reverse voltage rating, but in this application there is no significant reverse voltage. The two diodes have identical current ratings, and this is the critical parameter.

2) Just because a particular diode works in some 24V applications doesn't mean that it's going to be appropriate for all. The diode requirements are going to depend on the coil specs, in particular the current, the resistance, the inductance, and to a lesser extent the parasitic capacitance of the coil. We aren't likely to know some of these parameters, so we mainly look at the coil current. This is current that the diode is there to clamp.

I know, I'm just being an annoying precisionist. That's what I do for fun these days.

I used the autopilot for three passages since these discussions were started, without any diode on the bypass valve relay. The bypass valve did not always work. I will be talking to Raymarine tech support today -- I suspect either basic incompatibility with the Simrad pilot, or some kind of fault in the bypass valve, not anything related to the diode, but I will install the diode anyway tomorrow and see if that helps.

[goboatingnow]

Just to be picky, The diode is there not to clamp the current, but to prevent an inductive voltage spike as the magnetic field of the valve collapses. MOSFETS, particulary power MOSFETS have a relatively low trench breakdown voltage and hence can be damaged by such a spike. Anyone with a scope can easily see what is happening.

The resistance of the wire has a damping effect so the solid state internal proptection is actually probably sufficient, but good practice is to 'snub' the ringing at the coil. hence the diode. The actual current flowing when the diode is forward biased is very small, but the diodes forward voltage rating is a key factor.

dave