Author Topic: Inverter changeover relay issue  (Read 2363 times)

0 Members and 1 Guest are viewing this topic.

Offline MadScientist267

  • Impossible Condition Curator
  • Moderator
  • Hero Member
  • *****
  • Posts: 1514
  • Karma: +44/-4
  • Rules? What rules?
Inverter changeover relay issue
« on: January 28, 2016, 05:43:00 am »
5061-0

5063-1

5065-2

Typical everyday DPDT relay configuration:

 K      K
 C      C
NO   NO
NC   NC

Coil is connected to unit input, which is fed to the NO contacts.

Inverter output is connected to NC contacts.

Commons are unit output.

There is no other sense circuitry involved, the relay coil is it.

The issue is during fridge compressor start from genset, the voltage sags and causes the relay to chatter, which in turn kills any chance the compressor has of starting.

I can think of a couple solutions, but wonder what you guys think. None of mine are completely ideal.
Wanted: Schrödinger's cat, dead and alive.

Offline MadScientist267

  • Impossible Condition Curator
  • Moderator
  • Hero Member
  • *****
  • Posts: 1514
  • Karma: +44/-4
  • Rules? What rules?
Re: Inverter changeover relay issue
« Reply #1 on: May 16, 2016, 06:21:45 am »
This got left dangling... I realize it isn't really rocket science, but I should at least tag it on my way by...

The solution came in the form of converting the coil feed to DC and lightly capping it. It's only partially done as of the moment, and has been that way for a little while because I don't actually run the generator that often at all. I've logged about 42 hours since pulling the plug mid January, up to this point. With summer looming, it's much less, and actually the first of only 2 runs in anything resembling the recent past was a 2 hour "maintenance run", just to push something thru the carb and make juices flow. The second was after a week of crap sky straight (at which point a rather significant but unrelated problem with the generator itself surfaced).

The present fix is simply:

The tap point is at the selector switch to connect the load to either the grid feed or the generator. A power resistor is in series before a full wave bridge to limit current that would have destructive force if left to flow freely thru the relay coil, as impedance has been eliminated as a factor. On the other side of the bridge is 20uF/400V (2x10uF actually), across which the coil is connected. In the testing version I didn't use a bleeder, but of course that's highly recommended, and will be part of the final version.

The resistor was in lieu of an appropriate series cap that I couldn't lay my hands on at the moment I was tinkering with this, but I tend to have a wider variety of resistors on hand than caps, so that helped find the needed cap value. The voltage reaching the coil is about 55VDC. An "equivalent" cap (a relatively simple mathematical relationship based on the AC frequency) needs to be put in place of the resistor so as to not generate unnecessary heat and waste power. The former is the most important consideration here... it's only a few watts which isn't a big deal on generator, but even tho it's a beefy resistor (a 30W 2k "Dale"), it gets to pushing what I'd start to call "uncomfortable" to hold on to for any real length of time after it's been doing it's thing for a while. Not something you want trapped in a tiny closed up plastic box. A fan you say? Shirley, you jest.

Between the 20uF (which may need to be further adjusted after the resistor replacement) and the nature of an AC relay core, it sustains the contacts thru the dip of a normal compressor start.

It's rather noteworthy to add that this isn't entirely an ideal way to do this, as there can be magnetization issues that crop up when feeding an AC relay coil with DC, which once strong enough, can cause undesired and significant delays in the armature release. It varies in what I've seen the couple three times I've used this little hack for various reasons... anywhere from just barely noticeable all the way up to and including "can't really call that a delay". Jus' sayin'.

The better way to do this would be to replace the 120VAC relay with a DC coil version with the same contact rating. The remainder of the circuit (with values appropriately adjusted) can then be kept as is, without worry that the contacts will stick due to magnetization of the core.

My reasoning for not doing so is that I don't have a comparable DC relay on hand anywhere, and with increased contact ratings, relays indeed begin to put more than just dings in one's wallet. With infrequent operation, magnetization is less of a concern because it does tend to fade with time off. It's not permanent anyway... a quick blast with it's normal AC rated voltage or simply reversing the polarity will return it to normal (until it builds back up the other way in the case of the latter).

I'll close this out once the final version is built... The test rig is kinda scary looking anyway, and I really don't see a point in drawing such a simple schematic when it's only going to be modified. I'll post the final version with the update when that takes place.

Until then...
Wanted: Schrödinger's cat, dead and alive.