Fox T-Bird/Cougar Forums

I have about 40 sites where a very small portion of my scope is to give the customer the ability to see an hvac unit failure via dry contact closure.

I think I have a very generic layout below that "should work" on most any wall mount, split or packaged unit out there.

its based on a kind of digital approach,, if 1 and 1 is present then 0 output.
if 0 and 1 are present then 1 output.

I am planning to retrofit HVAC using the tstat call for cooling coupled with any internal circuitry that could drive the hvac contactor from either energizing or not.

the only part that is bothering me is that I don't know what the ohms are with a typical small 5 or 10ton unit so the added variable resistor I show may not be needed for what I call calibration.

thoughts?

Sorry, it won't work. The contactor and K2 are in parallel. The amount current or no current through the contactor will have no effect on the amount of current flow through K2.
A diagram of what is inside the "contactor" box would help.
Does the sine wave on the input to the contactor box indicate the AC power to run the compressor?

One problem here, I think the major problem, is that you're depending on the current flow to change appreciably to make the voltage at K2 dance around the threshold of activating K2, depending on whether or not that contactor activates. Another potential problem (depending on the answer to softtouch's last sentence) is that the control coil of the contactor might still be activating regardless of that input. If that input is high-power AC, that the compressor runs off, then the contactor still opens/closes. If this is a series of inputs that must be "satisfied" logically to activate the coil, okay, then no current flows through the control coil of the contactor.

However, back to how much energy is existing in the 24v line. I can understand your theory, but I don't believe you can control the input to K2 to the degree you think you can. I guess it depends on the current output of the tstat. I also wonder if those 3 total relays we net here, would either overload the tstat, or the tstat simply wouldn't produce enough current to drive them.

My experience with heavy duty contactors is over 30 years old. They used to make them with a "side switch" in them to be used to sense whether the contactor armature moved or not. A micro-switch was often used for this.

Yeah, that would be ideal, it would let you set up exactly the logic circuit you need.

This problem has been frustrating me increasingly over the last hour LOL. I have a double-E, and was doing this kind of buttstuffysis four or five years ago. The job I hold doesn't make me utilize this stuff...so I've lost it. Which infuriates me. What are you gonna do? /sidestory over.

I either forgot everything entirely...or this is simply a real-world circuit and I'm asking myself all of these real-world questions instead of simplifying the circuit. Not that those real-world conditions aren't relevant! - I think some specifics would be needed, resistances / impedances of all of the involved devices, "rated" current output of the tstat I suppose, good to know best case, what "is" our current.

The question is one, whether or not that resistor will cause too much of a voltage drop to activate K2. Two, do we actually affect how much current flows there based on how much current flows to the contactor from the 24v circuit (again assuming that failure trigger indeed "breaks" the circuit feeding the control of the contactor)

I downloaded PSPICE, which we used in my circuits class. Great free piece of software....aaaand I don't remember how to use it. I wonder if I still have my notes somewhere...this will be a fun dig through my closet. What the hell is wrong with me.


EDIT I found my stuff...thank you mom for not throwing that stuff out...I think we agreed ealrier this year to pitch my highschool stuff. I have a headache now but regardless of whether or not this gets solved here first, I am going to model this circuit this week. Thank you for inspiring me to re-learn this for curiosity's sake if nothing else.

what you do is subtract the relays and that is the oem circuit.  Tstat delivers low voltage +24v to the in series high and low pressure switch and then to the holding coil of the contactor.
once the holding coil energizes after a time delay, the large high voltage contacts close and 120 / 240Vac is sent out to the compressor.

all I am doing is intercepting and refeeding the tstat voltage to the contactor holding coil which I do not yet know its resistance value.
If the contactor activates, I assumed I do not want current to go out to my K2 coil so I add what am showing as a calibration or variable resistor in series with K2 coil.
I assume that current would take the path of least resistance so the variable resistor before k2 coil is a small insurance policy reinforcing that.

if the contactor holding coil does not energize, I assumed I could get k2 to change state and open my monitored dry contact to produce a "hvac fail" condition.

the pn of the Contactor is as follows:
allen Bradley (primary vendor)
pn: 400-NX98

hongmei electronics (secondary vendor)
pn: xmc0-403A.EBBd


softtouch- yes the sinewave indicates the Hi Line 120/240 to run the compressor
I am interested in what you said with the slide switch would be the silver bullet to turn this into a simple single relay application.
I was hopeful I could use these simple low cost solutions as a logic circuit so to speak.
My budget per site does not allow for wrap around rope CT's to monitor current flow through the contactor as a means to determine "unit fail" or not.  the CT's & itnterface  would be about 400$ per kit or per hvac unit,, and I could have several HVAC's per site.

I have a liebert 5 ton wall mount unit wiring diagram, I will see if this site lets me upload the pdf or not.

lets try

the above pdf shows low +24v on the right side and high 120/240 on the left side.

The Tstat would feed the "Y" terminal on the right side of the drawing with +24vac
as you follow the "Y" wire down, +24v goes through the "optional" LOR (lock out relay), the standard HP(high pres sw), the optional LP (low pres sw) , the time delay relay pin 3 for about five or so seconds, then out of the TDR pin 1 and into the C  (contactor holding coil C).  the white wire is ground up the right side.

now that the holding coil has gotten power, the compressor on the left can run to provide cool air.

I want to "monitor" the contactor holding coil in a simple way and only when the TStat has called for cooling and the holding coil does not get power, I want a dry contact relay to change state which will provide a "HVAC FAIL" alarm.

I have to assume in most cases (based on my many years experience dealing with things like this) that the "Optional circuits" were not chosen when the hvac units were deployed.,, therefore when I said "optional" above, and "standard" , I mean optional stuff probably doesn't exist where as the standard high pressure switch will always be in these units.

remember what a true "failure is".
loss of utility does not constitute an HVAC fail, a higher alarm step I am providing at the utility automatic transfer switch will provide that.
Also, a high temp does not mean an hvac failed, it means someone just might have the facility doors open too long or whatever.
A tripped AC feed breaker to the HVAC doesn't mean the hvac failed... the common denominator here is the tstat calls for cooling and therefore the contactor to the compressor either A- energizes or B- it does not.

based on those two situations is what I am trying to monitor.  once I get one to work, I can carefully put all the other hvac's in series so to speak to function the same way.
in my mind, if I can get K2 to be the overseer of the alarm , I can series up all my K2's in each hvac unit such that at any given moment if continuity is lost in any hvac unit K2 contact side, there will be an alarm back to the network operations center over SNMP indicating "site xyz" has an "HVAC Fail MAJOR" alarm.  The severity level "MJ" would trigger a trouble ticket and force a tech to dispatch.

and the more I look at this and how its applicable to most any small 5 to 10 ton unit in theory of basic pieces parts operation, on the +24vac side, the more I think I need to be focused on the time delay relay output.

In my diagram,, it is a must that K2 never ever ever ever change state unless there is a fail of the hvac. 
I do have a back up plan if K2 must temporary "toggle" when there is no failure.  I can time delay the SNMP Trap alarm out to the network operations monitoring center... through the software of the local alarm monitoring system made by Quest Controls.

I just emailed my HVAC tech support guy, I am asking if these contactors have slide switches or not built into them.. if so, that's going to be interestingly simple.... although it changes the diagram I have, its closer to solved.

you guys are gonna  when you hear how I am monitoring *any* diesel generator main circuit breaker handle without any moving parts!,, and all it takes is about 6$ in materials. (customer wants to know if their generator over current protection device is either tripped or has been tampered with or has been turned off.)

Wrote back to your PM...will read all of this when I get home...snack break, now back to work. Huh? LOL

I have to go back to the beginning to make sure we are all on the same page on some basics of how circuit stuff works.
When the Tstat energizes K1 and 24v are applied to the contactor and the resistor, it is not like a fork in the road where all the current goes to the path of least resistance.
The current goes through both paths at the same time. How much current in each path is determined by the resistance of each path.

Lets say the contactor has 1 ohm of resistance and the K2+resistor has 2 ohms of resistance.
24 volts applied to 1 ohm of resistance will have 24 amps of current flow.
24 volts applied to 2 ohms of resistance will have 12 amps of current flow.
This means there will be 36 amps flowing through the K1 contacts.
If the contactor breaks and has no current flow, the current through K2+resistor will still be 12 amps.
The current through K1 contacts would drop to 12 amps.

Why not add a current sensing inductor on the compressor feed? Then just add a normally open relay contact (K3) in the alarm loop that is triggered by the inductor (you'd have to step up the inductors output with a transistor to drive a relay with it). This way, if the 24v is on, K2 is on, but if the compressor is not running, there will be no inductive load on its leads, and K3 will not close, creating an alarm.

cougarXR, yes possible but the budget for that isn't there, those CT's or transducers are above the project budget.

softtouch~ yes I know what your pointing to... I know the illustration look ridiculous and I was hoping to use a difference of potential to trigger my K2.
However,, today I hatched another idea........
so my thoughts go back to "I want to know if the contactor has actually changed state,, and in reality,, I actually want to know if the compressor motor actually turned on.
I believe that a true fail would mean that the Tstat called for cooling and the compressor Motor does not spin
Its possible for the contactor to be fully operational but the motor not spin.

here is my other idea....
if I just focus on the holding coil of the contactor and ignore the motor for now,
when the contactor holding coil energizes, it should generate a magnetic field.  I can then use a magnetic reed switch which should change state.  this would provide the information I need... coupled with the TStat calling for cooling = no alarm condition.  example reed switches, like some people have for door or window sensors (wired) on their homes) >>>http://www.meder.com/reed-sensors-magnets-us.html

the issue I have is that im thinking the magnetic field will be present but since the low voltage to the contactor holding coil is +24vAC, I don't want the reed switch to clatter.
this would be a solve to softtouches earlier indication of a physical slider bar .. but using the contactors magnetic field instead.

If all you want to know is if the 24v is on, and the contactor is closed, why not just tap the output of the contactor? Is it 110v or 220v? Just stick a little 110v ice cube relay on there and add one of its N/O contacts to your continuity loop.

no,,
because the dry contact monitoring circuit is not allowed to toggle when everything is ok.  it can only toggle when things are broke.

if I do it the way you describe, the monitoring relay would toggle every time the contactor came on

Maybe I wrote it out wrong. But this is what I was thinking. The only problem I see, is the brief milliseconds of time between the Tstat kicking on and the contactor closing might cause a 'blip' on the alarm. You could add a 555 IC in there after K3's contacts if you wanted to make sure the alarm condition exists for so-many-seconds before breaking the alarm loop.

actually I can control a delay within the programming of the local alarm system... from milliseconds up to 18000 seconds

by the way,,, my shown K2 ,, when there is continuity from 87 to 30,, that EQUALS NO ALARM.  when the loop is broken, this is an alarm.  This is what we call "supervised dry contacts".
it helps cover the idea or concept that if the wire is cut *or* the alarm condition occurs, the circuit is open.

I understand the continuity from 87 to 30 being a no-alarm condition. The only way the alarm condition would exist in my revision is when the Tstat is on, and the contactor is not closed. That's what you wanted, right?

In a no-alarm condition, you would have 24v present, but the new K3's N/C contact (now in the open state because it's energized) would be preventing it from passing to K2 to break the loop. Does that make sense?

yep, it does.......

ok,, this should work,, using the 120v feeder to one relay helped a lot.

Yeah, that should work. It's cleaner using only the 2 relays instead of 3. I only suggested a 3rd because I thought this was some existing circuit that you were trying to add to. The vibration switch is also an interesting addition. I don't have any experience with those, so I'll be interested to know how that pans out.

That vibration switch is a new one on me too. I hope it works for you.

Looking at the Emerson schematic you posted a link to, I don't see a neutral conductor.
Your diagram shows the relay coils referenced to ground.
As I recall, the ground (green wire) connected to the metal boxes etc. is required to be a non-current carrying conductor for bonding purposes only.
You may have to use a 240vac(208vac) relay for K2 and hook to two of the phase legs to the compressor motor.
K1 coil will have be connected to terminal block Y (T-stat signal) and terminal block 1 (the other side of the 24vac transformer)

The field I worked in always had Delta transformers supplying the 3 phase power. 208vac between legs and no neutral.
Y (Wye) transformers have 240vac between legs and have a neutral conductor.

Be sure to use AC relays for K1 and K2. DC relays can vibrate and the contacts flutter if used in AC.

The pn of the switch is on it,, there is an online video of it in use as well here...........  http://www.adafruit.com/products/1766
I wont deploy the "very sensitive" shake switch till I test the sample I have coming tomorrow... then I speculate a starter order of 100.  I have a quality relay in mind as well, and will start with 200.  I will make up the mounting bkt and harness all by my lonesome because the installers may end up doing it wrong or just rush and things go wrong when that happens.  I have to "series" up all existing HVAC units on each site so I speculate a min of 4 per site and that is very skinny IMHO.

did you happen to notice how I am telling the installer to install it?  its attached right at the end of the zip tie, installer can  use a sleeve of heat shrink or whatever to attach.
If a good vibration point cant be gathered, I have another idea.... those door strike sensors which are a reed switch and magnet kit... well toss the magnet.  Mount the reed switch on the part of the AC motor that causes the reed switch to change state due to the magnetic field. I tested this in my garage on a few ac motors and it worked.... I tested on the back of my car alternator and it worked, the switch changed stated.  only problem is it doesn't tell you if the motor is spinning.  A shake switch helps with that.

The voltages at each site will vary and over time the delta element has been migrated away from greatly , mostly Wye service entrances for the most part.
When I see a small and big xfrmr on a near by pole though with aerial drops to the facility, it sure does make me want to verify a N leg "is" actually present.  Ive seen a few ways someone attempts to build a N leg but you'll know right away when you find the wild leg,, sticks out like a green fly in a sugar bowl.  I did not show a N/G conductor either to simplify the drawing.. but there should be N and G terminal strips within each unit if not each N / G conductor should at the least be on the same strip,, although that is becoming more and more of a No- No ,,, in order to isolate current carrying conductors from actual ground.  People don't understand N legs do carry current especially in an imbalanced panel board.

The Emerson drawing is very generic in nature to embody list options that "can be had".
You have noticed there are optional dry contact relay options "if" ordered.  Most of the time it is not so I have to at least be prepared. 
I see about 20$ per hvac unit in materials for this small chunk of my scope, and I'll give 2hr labor at 75$ billable per unit to integrate this "kit".

there are other options out there to "add on" made by other monitoring companies but they involve tstat probes, ct's, ect and could cost me in about 700$ mat per unit to monitor and that is out of my budget.  To beat it all , my design is actually gathering a better "fail" alarm condition than those other devices I just pointed to.  The CT's will detect current flow but nothing about if the motor is running. 

In replacement of the "shake switch" I actually want something else..... something like a "thermo-disc" like in your dryer.
It would be an open circuit for a while putting the hvac in alarm when K1 energizes but as soon as the thermo-disc cools off, the alarm would go away.
I have my alarm system programmed for a 15min delay prior to delivery of the actual alarm out to the network operations center.  This would give me the ability to .......say.....
Tstat triggers K1, K2 detects contactor coil moved Hi-Line voltage, thermo disc went from open to closed thus the hvac **IS REALLY COOLING**.

I just cant find a "non - buttstuffog" switch like that.  it need to be open or closed only,, nothing fancy.


gonna take your advice on the N & G legs, all the hvac's are typically bonded already and if they are not we will toss in a #6 Rhh/RHW chassis gnd back to the main ground bar.

Revised,
I found my "simple" switch to detect ductwork temp.
if anyone "knows" of a thermal couple sensor that goes "open" and "closed" which is cheaper per unit price of 10$, please!!!!!!! let me know.