Shop Products
Houzz Logo Print
mcguirev10

Voltage regulator for thermostat?

mcguirev10
12 years ago

I have three identical HVAC systems in the house. Recently I switched to those Nest thermostats which have a lot of nice features for gadget/control freaks. We've been very happy with them.

However, one system in the house blew an internal fuse (on the green fan power line) inside one of the Nests after just a few weeks. They sent a replacement overnight, and within a couple weeks more, it blew a different internal fuse (this time to the yellow heat/cool line). The Nest logs a ton of data and tech support requested permission to download that. They saw some voltage spikes and recommended that I install an inline voltage regulator.

They are sending a replacement and I have agreed to not install it until the overvoltage problem is corrected. I know I can pick up a little 3-leg TC-package 24V 1 amp regulator for about $2, but I'm not sure where to install it (or them).

One for each control line going into the thermostat? (My systems are all heat pumps using yellow, green, orange, aux-white, and red Rc... and blue common, of course.) Or just some of the lines? Or somewhere else in the system?

Comments (14)

  • brickeyee
    12 years ago

    The thermostat input is AC, not DC.

    A simple DC regulator is liable to cause problems without a bridge rectifier and some filtering capacitors.

    You might be better off just putting some film caps across the 24 VAC lines to try and filter out the surges.

    Thermostats have to live in an installation that has a lot of relays on 120 V and 24 V.
    It is a rather poor design that for the thermostat that did not provide adequate filtering of the input power.

    If it is a power 'stealing' thermostat (pulls power from the red and white lines instead of having a dedicated supply line to the thermostat) is is an even worse design and has inadequate filtering.
    Those have to deal with even more electrical noise.

  • mcguirev10
    Original Author
    12 years ago

    Thanks, I didn't realize it was AC, although I should have since I knew there was a transformer at the other end.

    Yeah I was a little surprised, too, that such an expensive piece of equipment couldn't deal with noise that a $25 thermostat ignores for years at a time. Although, on the other hand, the cheap thermostat didn't have much that could break in the first place.

    Same question if I go the cap route: would I want one per control line?

    I'm not sure how to test a transformer, but I'm wondering if I ought to be suspecting that instead, given that the other two identical systems in the house have been just fine this whole time. The systems are all about 7 years old. Tranes. (I know.)

  • brickeyee
    12 years ago

    Short of putting an oscilloscope on the line and trying to capture the transient it can be a real PITA to troubleshoot things like this.

    It may be something as simple as one set of wires being slightly longer or routed differently enough to cause more coupling to the lines from the noise source.

    The noise source can be as simple as the system relays opening and closing. Change in current flow change the magnetic field around the wire, and that can induce voltage in adjacent wires.

    You need to be aware that adding capacitors for noise can also cause larger currents to flow in the wires when they are switched.

    Tying the lines to common with a suitable rated (voltage mainly) film type capacitor can shunt noise, but without having some idea of the noise magnitude choosing a capacitor value is difficult (more like guessing and seeing if the problem goes away).

  • maryland_irisman
    12 years ago

    Brick, can the line voltage to the transformer spiking be the source of the transformer output spiking? I've seen that happen. Perhaps regulation at that point is needed.

  • mike_home
    12 years ago

    Do you have low humidity in your house? If yes, you may have a static electricity problem. Touching the thermostat may be creating an electro static discharge spike which is blowing out the internal electronic components. This may be the source of the voltage spikes.

    Whatever the reason it sounds like this is a poor design. Thermostats need to be designed to survive voltage spikes.

  • fsq4cw
    12 years ago

    It's possible that the assumptions here are all wrong. First of all, if it's a fuse that's blowing then the problem is over current and not necessarily over voltage. Noise on the line, if that's a problem, is likely introduced after the 24-volt transformer, as a power transformer by its very nature, being an inductor, will likely not pass transient high frequency noise, it's too slow to react unless it's part of a switching power supply design or an oscillator.

    The 3-legged packaged regulator (DC) referred to is likely a TO-220 (maybe a TO-92) package. A small value (capacitance) bypass cap may help if in fact there is some sort of RF transients being introduced (after the transformer secondary winding). Even though 24-volt AC feeds this device, it most certainly must convert this AC to DC (probably 12-volt) in order to function. There should be regulation and bypass built into the design.

    Another possible but expensive solution might be 'Whole House Surge Protection' installed right at the load center (electrical entrance). This should provide protection from voltage spikes originating from both sides of the load center (utility & internal). Besides, now that we have microprocessor controlled HVAC systems with ECM motors, microprocessor controlled 220-volt stoves, ovens, dryers etc., how else can we protect these sensitive devices from spikes? It's not just your large screen A/V system that needs this type of protection in today's world. You can't buy a Monster Power Bar for your HP!

    Another advantage of Whole House Surge Protection is that it ties all the separate grounds of the electrical utility, telephone lines and cable lines together to one common ground. Have you ever tried measuring the voltage (AC) between these grounds with a Fluke?

    The comment that is most interesting is from mike_home regarding humidity and static electricity, though I suspect the source of the problem lies elsewhere.

    Back to this Nest thing, what's directly on this fuse's circuit that's drawing so much current? This could be a design flaw and may not be up to you to solve.

    SR

  • brickeyee
    12 years ago

    "It's possible that the assumptions here are all wrong. First of all, if it's a fuse that's blowing then the problem is over current and not necessarily over voltage."

    Except that over voltage on semiconductors often causes them to fail just about instantly (faster then any fuse even if the current was rising) and then the fuse blows since the circuit current becomes too large.
    The result of the voltage spike appears as a blown fuse, and unless the equipment is expensive enough to be repaired, a new item is installed without seeing what else might have been damaged,

    "Noise on the line, if that's a problem, is likely introduced after the 24-volt transformer,"

    That is likely and what has been discussed.

    "as a power transformer by its very nature, being an inductor, will likely not pass transient high frequency noise, it's too slow to react unless it's part of a switching power supply design or an oscillator.."

    Power transformers often have significant capacitance between the windings, and that is how high frequency signals pass through them.
    The signals do not pass trough magnetic coupling like the 60 Hz power the transformer is operating on, but through the capacitance between the primary and secondary windings.

    Many components have much different than ideal performance outside their designed operating rage. Inductive devices and capacitors are notorious for this.
    At high frequencies many inductive devices allow signal through by capacitive coupling (even between the coils of a simple inductor) and the parasitic inductance of capacitors overwhelms their inherent capacitance. The capacitor then appears as an inductor to the higher frequency signal.

    Noise coupled from a relay closing and the fast change in current this produces in its operating circuits can couple to other circuits and can often go right through a power transformer almost as if it was not even present.

  • mike_home
    12 years ago

    I am a little skeptical about the information the Nest support people gave the OP. It was stated that internal fuses were blown. Are there actual fuses or is the output driver damaged by a voltage spike? What good is an internal fuse if you can't replace it?

    It is nice that the Nest support people can remotely log into the thermostat and read the data logs. Can they say how big were the voltage spikes? Were these 10 volt spikes or a 10KV spikes? The answer will determine the solution.

    Did the Nest support people recommend what voltage regulator should be used? You should implement the Nest recommended fix so if it happens again they will have no excuse not to send you another thermostat.

  • mcguirev10
    Original Author
    12 years ago

    The Nest is a two-part device. There is a base where the wires connect, and the control/display part plugs into that -- the "brains" of the device, so to speak. So it probably does make sense that the base contains surface-mount fuses (which are very tiny, as you probably know) to protect the much more expensive hardware in the other part, but since they aren't providing a replacement base plate only, it seems there is a disconnect between their design and their support policies!

    I spoke with a support guy who only said the engineers said they saw spikes. They did tell me the Nest can handle +/- 3V swings from the optimum 24V. I don't know if that's considered a large variation or not, but apparently it isn't large enough.

    I am actually on hold with them as I write this, since the information I was given during the first call was clearly inadequate...

  • mcguirev10
    Original Author
    12 years ago

    No luck there. The new support guy agreed the other guy's recommendation was misleading, but now they just want to refund my money. This is kind of annoying since my other two (identical) AC systems are also on Nests and working fine.

    This guy said there wasn't any mention in the notes about what they found from the logs, if anything.

    I suppose I'll call my local AC guy who installed my systems back when the house was built, just in case he has some ideas, but it sounds like I'm probably SOL.

    A real shame, too, I really liked these things.

  • mike_home
    12 years ago

    I am curious if you have a common connection (C terminal) on the problem thermostat and the other two thermostats.

    Is there anything unique about the problem thermostat set up compared with the other two thermostats?

    I find it interesting Nest has offered to refund your money after sending you two replacement thermostats. Perhaps they understand the failure mechanism and feel refunding your money is the easiest solution for them.

  • brickeyee
    12 years ago

    "What good is an internal fuse if you can't replace it?"

    Reduces the risk of fire.

    Not everything is about protecting the equipment.

    "They did tell me the Nest can handle +/- 3V swings from the optimum 24V. I don't know if that's considered a large variation or not."

    It is VERY small, and likely means there is no separate protection from the anti-static protection built into the components they used.

    If they saw anything it manes it was probably a relatively long "surge."

    The sampling rate is not going to be all that high, and anything shorter than the sampling rate is not going to be caught reliably.

    It sounds like another Mickey Mouse design.

    Like the one that clobbered the Neptune washers.
    The 'designer' failed to understand that heating elements pull more current when they are turned on cold and undersized drive transistors.
    The 'wax motor' used to lock the door closed then could not heat up to unlock.

  • mcguirev10
    Original Author
    12 years ago

    All three do use a common (C, blue) connection.

    Nothing unique about the setups. One of the working two is a smaller system, but still the same make and model line, and all were installed the same day back in 2006. If anything, the problem-unit has fewer running hours because it's our master bedroom area, so it is off more often than the others.

    The Nest guy claimed their equipment can be more fragile because they use solid-state switches. However, that sounded a bit like a cop-out to me, I'm 99% sure my old RiteTemp programmable thermostat (which has been running the "problem" unit for 6 years) is all solid-state, too.

  • mike_home
    12 years ago

    Attached is the link of the Nest thermostat tear down. It is an impressive device. It has an ARM processor, 64Mbytes of RAM, and WiFi. It even has a mini USB connector.

    I think Nest needs to add more robustness to the output transistors. It appears it is suspectible to electrical problems. I see on the Nest web site there are open positions for analog design engineers. Perhaps they are trying to hire the expertise for the next generation product.

    Here is a link that might be useful: Nest tear down

Sponsored
Pierre Jean-Baptiste Interiors
Average rating: 4.8 out of 5 stars76 Reviews
DC Area's Award-Winning Interior Designer | 12x Best of Houzz