120VAC Time Delay Raley

[STB]

Hi all,

I'd like to set the wiring up so that my forward air conditioner can't run when either my stovetop or microwave/oven are in use.

In particular, I want to break the connection to the HVAC unit as soon as either device is turned on -- but delay by 5 or so minutes before turning it back on. This way, I don't short-cycle the HVAC unit as the thermostat turns the stove or oven on-and-off or as a human checks/stirs-and-resets the microwave.

I feel like others have solved this problem -- but the threads I was able to find didn't involve the time delay part.

Does anyone know of an appropriate relay to use?

Many thanks!

EDIT: I'd originally posted with one idea, but upon a more careful look at the spec sheet, I realized the time delay wouldn't work the way I want. I want the NC output to break immediately, but to delay coming back on. :-(

 

[KnotYet]

A medium current time delay relay from McMaster-Carr should do the trick.
They have a good selection and ship from various warehouses across the USA.

 

[rgano]

When I purchased my new dometic air conditioner a couple of years back, the dealer informed me that I should always turn it off via the air conditioner's control rather than the breaker because there are a number of subroutines which run in the electronics of the unit which need to shut down in an orderly fashion. The Micro-Air Easy Start in installed in this AC has a five minute delay built into the compressor start after a shutdown. I think a lot of AC units have compressor protection like that, but more expert opinion will soon chime in. IMHO, I would not attempt to install such a circuit as you discuss and would just handle the process manually.

 

[STB]

rgano,

Thanks so much for the feedback!

And, as you know, you aren't telling any lies. My pump starts 5 seconds before the compressor and stops 5 seconds after.

It is just that, for right or for wrong, the post-shutdown 5-second delay is something I'm not super worried about. I don't think a critical amount of cooling happens in that time, especially for the 1-3 times/day, max, I cook.

For right or for wrong, the reality is that the HVAC panel for this unit is at the front of the boat, and the breaker panel is at the helm near the galley. So, the relay won't be doing any worse than I've done -- I've been shutting it off at the breaker for years. I have the same bad habit w.r.t. switching loads when transitioning between shore power and the genset and back. I suspect keeping up with my strainer baskets (or not) will more than make up for it over the long haul.

I am more concerned that I get the restart time delay right so that I don't end up with rapid start-stop-restart cycles that cause it to kick back on before the refrigerant gets to settle and pressures equalize. That'll eat the compressor really fast.

I guess if a bunch of people beat me with sticks on the thread, I can adjust my goal slightly. I really care about the compressor /starting/ while cooking. I can handle the running load. And, I can handle, be it barely, one compressor starting while cooking. But, cooking and having two compressors kick in at the same time puts me way over the top.

So, what I could do is put soft-start kits into both units and see where that gets me. It could easily mitigate the need to prevent the HVAC from starting while cooking -- and do it without disabling the HVAC.

That is probably an $800 solution, which would be excellent if I knew it would work. But, I'm not sure I trust the "specs". I'd hate to spend the $800, look at the current draw on my ammeter, and not be happy with the load in reality.

The other thing I could do is to wire it with feedback so that, if the unit is already running, it stays running, but if not, it won't start. I fear that going that route would take me into the "I know something about electrical circuits so I'll make a contraption" space. And, when I've seen folks start down that slope, it has been a slippery one, and not often consistent with the KISS principle that I try to follow to keep things reliable, low maintenance, and easily debugged. Regardless, even if only for the thought experiment, I'd still need the time delay relay in this case.

Thanks again!

 

[STB]

KnotYet,

I don't think I'm seeing one that would work. I'm probably missing it. Any recommendations?

Many thanks!
-Greg

 

[ksanders]

This is not as easy as a time delay relay.

Remember that your stove top and microwave oven have power to them all the time.

You only want to turn off the HVAC when the stovetop or the microwave are operating.

In order to make this happen you are going to have to sense power flow and have that turn off your HVAC.

 

[STB]

Hey Kevin,

I totally get that. But, just as a for example, the neutral line doesn't normally have current flowing through it, but does when the device is operating.

I could also use a coil or hall effect triggered circuit. But....still need the relay.

 

[ksanders]

Hey Kevin,

I totally get that. But, just as a for example, the neutral line doesn't normally have current flowing through it, but does when the device is operating.

I could also use a coil or field effect triggered circuit. But....still need the relay.

Grainger sells a good selection of time delay relays with a variety of trigger voltages.

My opinion is that you need to work out your trigger circuit challenges first.

 

[STB]

Hey Kevin,

The trigger part isnt at all a concern to me. There are a lot of ways to do that. Like I said, the sensing can be based upon a coil sensor (or a hall effect sensor) or be directly switched off neutral. Everyone has familiar turf and less familiar space. That space is familiar to me.

There are already products out there that do most of that. A quick Google search turns up what might be an example:
-- https://www.dwyer-inst.com/Product/...tTransformers-Switches/SeriesMCS/Google#specs

I did see some relays at Grainger's that might work. But, I'm a lot less rehearsed at reading time delay relay specs than other specs.

I was just hoping someone had done this already or played with relays more often and could quickly make a recommendation.

But, thanks for the feedback!

 

[ksanders]

Hey Kevin,

The trigger part isnt at all a concern to me. There are a lot of ways to do that. Like I said, the sensing can be based upon a coil sensor (or a hall effect sensor) or be directly switched off neutral. Everyone has familiar turf and less familiar space. That space is familiar to me.

There are already products out there that do most of that. A quick Google search turns up what might be an example:
-- https://www.dwyer-inst.com/Product/...tTransformers-Switches/SeriesMCS/Google#specs

I did see some relays at Grainger's that might work. But, I'm a lot less rehearsed at reading time delay relay specs than other specs.

I was just hoping someone had done this already or played with relays more often and could quickly make a recommendation.

But, thanks for the feedback!

very cool, I had not seen those!

Use an “off delay” relay. Pick your poison.

The concept is that the relay will delay opening/closing the contacts until the trigger is removed plus the timer runs out.

you will probably have to use a interposing relay to get the current rating you need for the HVAC system.

 

[twistedtree]

Finder makes a line of time delay relays, and I think also current sense relays.


Good suppliers for stuff like this are:


Mouser - Top notch supplier. Usually the first place I try.



Factorymation - Generally better pricing on the things they carry.



AutomationDirect - Similar to FactoryMation



DigiKey - Good selection, but usually more expensive than other places.



Newark - Very wide selection, but I have had quite a few order issues with them and now avoid them unless they are the only supplier.



I'm not sure what you guys have found, but I find Grainger to be uniformly significantly more expensive than other sources. They are the West Marine of industrial supply. The good part is that they carry a lot of stuff and you can get it quickly, but otherwise they are last on my list, even after Newark.

 

[rgano]

rgano,

Thanks so much for the feedback!

.......

So, what I could do is put soft-start kits into both units and see where that gets me. It could easily mitigate the need to prevent the HVAC from starting while cooking -- and do it without disabling the HVAC.

That is probably an $800 solution, which would be excellent if I knew it would work. But, I'm not sure I trust the "specs". I'd hate to spend the $800, look at the current draw on my ammeter, and not be happy with the load in reality.


Thanks again!

OK. I can provide a bit more feedback about my real world experience. My little boat has two air conditioners, a rooftop (air-to-air) 13.5 kBtu unit and a marine 10 kBtu unit, both Dometic. The diesel gen is a mere 3.5 KW Nextgen. Prior to installation of the Easy Starts (by me with excellent support from Micro-Air), only one of these unit could be run at a time on the generator, and there was a hell of a thump whenever the compressor came on and noticeable change in rpm from the genny. Now they both run fine, and there is NO, repeat, NO indication anything has changed as the compressors come on. You have to really pay attention to tell. The Easy Starts are expensive, but the price comes down a lot if you just by the circuit board from them and get a plastic box to place them in. I got super lucky and got my two Easy Start CBs for a couple hundred dollars from a fellow boater who had bought but not installed them. These things DO work.

 

[KnotYet]

KnotYet,

I don't think I'm seeing one that would work. I'm probably missing it. Any recommendations?

Many thanks!
-Greg

As I read your desired outcome, you require 2 relays.
The first is a basic 'normally on' that turns off whenever the listed appliances are on.
The second is the time delay relay that incorporates the 5 minute start-up delay.
As long as you don't mind waiting 5 minutes every time you start the HVAC, that is.

There are many such programable relays that will do this from many suppliers.
I only mention McMaster-Carr because I have ordered these relays from them before.
Please correct me if I misunderstand your requirements.

 

[twistedtree]

I think you could also do it with a single time delay relay that has a NC contact. Wire the AC units to the NC contact, and have operation of the other appliances trigger the relay opening the AC circuit. When the appliance turns off, the delay will run out and the NC contact will close again.


For sensing the appliances, there are current sense relays that you could use. Run one of the wires from each of the appliances through the sense coil, and it will detect either of them running.

 

[STB]

Hi all,

Thanks for the help!

Kevin, KnotYet, TwistedTree:

I may just have bum luck paging through relays specs not really knowing how to search them, but what seems to be throwing me is that I only want the delay on the make, not the break. I've found all sorts of relays, but not exactly that. I can piece it together, but I just have a feeling that what I want is out there and that I just need to learn how to ask for it.

If I can find what I want, I shouldn't have to wait the 3 minutes each time, only after cooking. But, that may be a big if.

I may just give up on this approach and do the sense, state switch, and time-delay in solid-state and reserve the relay for carrying the load. Triacs get messy.


rgano:

I think I'm going to call them this week. It sounds like it'll be money well spent in any case. Thanks for writing about your experience.

If the MicroAir gives me the headroom to cook with both ACs on, that is the best of all worlds. Having said that, I might build a similar controller for the hot water heater, but that doesn't require a time delay and is a pretty low priority since i only use it for a bit in the morning, anyway.

Still looking for suggestions on a specific relay, though!

Cheers!
-Greg

 

[twistedtree]

This is the reason to use a SPDT relay, and use the NC side of it. Then the AC is normally on, and is triggered by load to turn off and stay off for a delay after the load goes away.

 

[STB]

Hey Tanglewood,

Thanks! Pretty please see the attached spec sheet. It is the one I originally thought would work, posted about, and then edited out.

I thought from the specs that it should work, but when I looked at the transition diagram at the bottom right of the spec sheet it seemed backward. It still looks like it should work with via the other terminal, but I dont understand why.

 

[twistedtree]

Hey Tanglewood,



Thanks! Pretty please see the attached spec sheet. It is the one I originally thought would work, posted about, and then edited out.



I thought from the specs that it should work, but when I looked at the transition diagram at the bottom right of the spec sheet it seemed backward. It still looks like it should work with via the other terminal, but I dont understand why.

That doesn’t look like it will do what you want. The timing diagrams tell the story.

I think you want something that:

1) activates on the leading edge of the loads turning on, and

2) stays on for a programmable time after the load turns off.

And I’m still not clear on how you plan to detect when the loads are on? It sounded like you had a plan, but I didn’t follow it.

 

[twistedtree]

I think something like those in this datasheet will work. I'm using the highlighted one for a timed bathroom vent fan, and I think it's pretty much what you need.

 

[STB]

Twistedtree,

Thanks so much!

The two basic "non contact" ways of detecting current flow are via some type of coil via inductance or via electron deflection in a hall effect sensor.

The idea behind the coil is that an alternating current moving through a wire produces an alternating magnetic field. The flux from this, in turn, induces a current flow in a nearby coil of wire. It is the same principle that lets the two coils of a transformer transmit energy, where in a 1:1 isolation transformer, or a step-up or step-down transformer, or to wirelessly charge a toothbrush or phone, or to enable those "light up AC voltage detectors".

So, basically what one does is to wrap wire into a coil and wrap that around the wire in which one wants to sense current, and then builds a circuit around it to let it drive the relay. Depending upon the details this likely involves an integrator and/or amplifier.

Either might need power to be able to drive a relay and/or the support circuit -- there is only so much one can drive off of a small coil:wire interface. If one did need power, one might want to take it from the output side of the circuit off to avoid disturbing the wire being monitored.

One could also use a hall-effect sensor, which basically takes advantage of the fact that (magnetic) opposites attract -- and sames repel. As current flows, electrons, which are negatively charged move, which can repel electrons in a tiny piece of metal, generating a tiny voltage differential across it (the potential energy of them moving back). This can then be amplified and used as above. The big advantage of hall effect sensors is that since they don't require current flow, they can be used for DC voltages and to detect static magnetic fields. In this application they'd be useful -- but, because of the tiny differential, require more support to drive a relay.

In terms of

One could get past this if one wanted to disturb the wire being sensed to add a voltage transformer. This replaces the inefficient wire with a more efficient coil. This is basically what many traditional current-sense relay do.

In terms of "contact" ways of doing things, one could add a transformer to the circuit being monitored and use it as a more powerful coil-based sensor. One could also add a "shunt", e.g. a very-precise low-resistance link. One could then measure the voltage drop across this link to deduce the current flow via Ohm's law and, in any case, amplify that voltage to drive the switch circuit.

Lastly, although the "hot" wire going to a device always has power, the "neutral" wire, which serves as the return path, is only a current carrying part of the circuit when the circuit is carrying a load. So one could come off of the neutral wire with a high impedance device to ground and sense voltage that way. Something like a MOSFET, JFET, etc can have doens to hundreds of ohms of impedance, so would have a negligible impact on the circuit. But, tapping anything is less clean than a non-contact sensor, and tappping the negative is particularly ugly as it is very unexpected.

At any rate, one could use any of the techniques to build a load sensor. And, it isn't hard to do. But, when ti comes to the passive sensors, which are likely the best approach -- plenty of people already have. They can be bought off the shelf.

Cheers!
-Greg

 

[twistedtree]

Got it on the non-contact methods. There are current sense relays available like the Finder devices I posted, maybe even from Finder. I just don't recall. But I have seen and used them with an external coil, and some have the coil built-in and you just thread one of the conductors through the hole as you would a stand-alone coil. I think if you google Current Sense Relay you will find a bunch.


You are sensing operation of two different devices, as I recall? Assuming they are on separate circuits, I think you could run the Line wire for each through the same sense coil. Then either of them will trigger, rather than having to separately sense each. The only issue I can think of is that they will need to be on the same leg, assuming you have split phase on your boat. It might still work with different legs, as long as the load current of both devices don't cancel each other out to a point below your detection current level. Actually, if they are separate legs, then I think if you run them through the coil in opposite directions, the detected current will be additive rather than subtractive.


Re sensing voltage on the neutral, I'm pretty sure that won't work, or at least not reliably. The neutral will always be at neutral potential whether the device is operational or not, so no voltage difference to measure. I think you would need to insert some sort of shunt then measure the voltage drop across it, which of course is a proxy for measuring the current

 

[STB]

See below. Somehow posted twice.

 

[STB]

Hey TwistedTree,

Think of it like placing a multimeter from hot to neutral. The mulitimeter acts like the high impedence parallel path with very low current low and very little disruption of the voltage drop across the original device. Once can use a semiconductor like a digital multimeter or a huge resistor and work from there like an old analog volt meter.


It is just another way of using current as a proxy for voltage. It is sort of like putting a shunt in series before or after the device, e.g. on hot or neutral, except since it is in parallel vs series, it is high impedence instead of low.

But, again, I think this is just plain ugly. No one would expect it.

 

[twistedtree]

Hey TwistedTree,

Think of it like placing a multimeter from hot to neutral. The mulitimeter acts like the high impedence parallel path with very low current low and very little disruption of the voltage drop across the original device. Once can use a semiconductor like a digital multimeter or a huge resistor and work from there like an old analog volt meter.

Whatever is being added across the load is acting like a high impedence parallel path of a voltage divider.

It is just another way of using voltage as a proxy for current. It is sort of like putting a shunt in series before or after the device, e.g. on hot or neutral, except since it is in parallel vs series, it is high impedence instead of low.

But, again, I think this is just plain ugly. No one would expect it.

I must not be understanding your description. Maybe a circuit diagram would help?

If you meter line to neutral, I think you will read line voltage whether the load is on or off. Now that’s not 100% true because there will some voltage drop due to the load. But other loads elsewhere will cause similar drops, so I don’t think it would reliably indicate whether your device is running or not.

And parallel resistors create a
current divider, not a voltage divider.

 

[CaptTom]

...the HVAC panel for this unit is at the front of the boat, and the breaker panel is at the helm near the galley...

So, if the goal is to avoid accidentally running both range and compressor, then why not a simple 1-2-off rotary switch? You flip the switch to "range" when cooking, and back to "Air Conditioner" when done.

I did something similar with my water heater circuit. I have two 30A legs, with two air conditioners on one side and the range on the other. Depending on what I want to run, I'd rather have the water heater powered from one leg or the other. So I added a rotary switch to choose which leg powers the water heater (bottom left in this picture):

You would do the opposite, with power from the switch being routed to either the range OR the air conditioner, never both.

I know, the "geek factor" is much lower. But it's cheap, simple and reliable.

 

[KnotYet]

I don't want to stop you from the fun of creating new solutions for doing this but, at
least in this situation, I see no need for non-contact current sensing to get it done.
You could use the existing wiring to provide the current to activate the relay(s).

 

[STB]

Hey TwistedTree,

Thanks! Indeed. I meant a current divider. I may have been spitting that out wrong the whole thread. I edited that one post to fix that. Thanks!

I don't think the problem is at your end

I think the problem is that this idea was never fully formed in my head because I just planned to go down the less intrusive and beaten path and use a coil or hall effect sensor rather than doing anything more intrusive, so it was just a bunch of ideas.

I just started to sketch it out and it would definitely need an accurate low-resistance shunt on the neutral side, which makes the idea the same as on the hot side.

I think I had it in my head that if the impedance of the sense circuit was high enough the resistance of the wire would do. But even roughly approximating numbers, that clearly isn't going to work out. Really roughly, a MOSFET is probably 40-50MOhms vs 150-200MOhms an inch for the wire. That's not going to play well in the real world.

One idea off the drawing board!

 

[STB]

Hi KnotYet,

Absolutely I could. A current sense relay can do it directly, if I can find one that also does the time delay right. But, I was having trouble doing the second part of that in isolation, so I figured I'd just do it with a coil.

 

[Ka_sea_ta]

If you are concerned about the delay a simple "delay on make controller" is pretty standard HVAC stuff... you will still need to use a contactor to handle the current for the AC unit

 

[twistedtree]

Hey TwistedTree,

Thanks! Indeed. I meant a current divider. I may have been spitting that out wrong the whole thread. I edited that one post to fix that. Thanks!

I don't think the problem is at your end

I think the problem is that this idea was never fully formed in my head because I just planned to go down the less intrusive and beaten path and use a coil or hall effect sensor rather than doing anything more intrusive, so it was just a bunch of ideas.

I just started to sketch it out and it would definitely need an accurate low-resistance shunt on the neutral side, which makes the idea the same as on the hot side.

I think I had it in my head that if the impedance of the sense circuit was high enough the resistance of the wire would do. But even roughly approximating numbers, that clearly isn't going to work out. Really roughly, a MOSFET is probably 40-50MOhms vs 150-200MOhms an inch for the wire. That's not going to play well in the real world.

One idea off the drawing board!

OK, thanks.


I think your choices for detecting when the devices are on are:


1) Some sort of current sense, or


2) Crack open the appliances to get access to it's internal on/off switch, and piggy back off that.


I know you could do this with two relays; one current sense and one time delay. I'm not remembering seeing a relay that does both, but that doesn't mean they aren't out there.