STC-1000: A cautionary tale

[CJ92028]

I want to share my experience with the STC-1000 and why I don't use it anymore. I'm a wine guy, but I saw many posts here of people using this unit here and received a lot of help from this community's posts and felt I owed you all a long term reliability report on the STC-1000.

I built a large wine cooler and used a $20 refrigerator form Craigslist for my cooling unit. I chopped up the fridge and harvested the cooling system to create a compact cooling unit. The thermostat range was not adequate so I looked around for a remote thermocouple thermostat. I found many posts in this forum with great information and bought one. I built an aircraft style plate and front mounted it in a supporting wall under my cooler. It was perfect, inexpensive and looked great. The only negative was that it was centigrade but that was easy to deal with.

After 6 months of flawless operation I started seeing some oddities in it's function. I can't remember exactly because I suspected it was the way I was taking power from the refrigerator. It was staying on when not commanded and other oddities. Luckily I had a LaCrosse remote temperature monitor that would alert me via the internet when the temperatures went out of tolerance. But it never made much sense. And now that I look at the circuit, I see the secondary circuit though the relay is independent of the thermostat's power. And with my voltmeter I could see the relay was closed. Also it seems like the relays draw a lot of current. In fact, the whole unit seems to draw much power.

Then about a year after I got it, I got an alert that the temperature was much too low. I went to my cooler and smelled the smell no one wants to smell, electrical smoke. I pulled the unit out and the left rear was smoking and melted. The display was not indicating cool being commanded. The relay was not energized but appeared to be closed. But it was 'closed' (short-circuit) within the connector block. The plastic had melted enough on the connector block that an arc occurred and then things got really hot and the two contacts fused nuder a lump of melting green plastic.

I became the thermostat because it was a 100º+ weekend which meant I needed to find a quick fix. After it cooled I did a autopsy, cutting apart the case. I have worked in the electronics industry most of my life as a mechanical designer so I am familiar with failure modes. I noticed that the terminals are cheap knockoffs of the venerable Phoenix Contact screw-down type blocks. When I pulled it out I felt my wires up close to the unit and they were not hot. You see I'm not switching main power, my thermostat was controlling the fridge using the extant thermostat circuit which is much less current.

The heat buildup was do to high resistance in the connector block. I removed the cooling relay's connector block and cut it apart. The thickness of the metal conductors inside are woefully inadequate and IMHO not thick enough to carry the current of the relays. The connector should always be rated for 2 to 3 times the current of the relay circuit IMO. This was exacerbated by the fact that I had it enclosed in a wall in a garage with an ambient temperature of around 110º. Also when cutting the case away from the circuit board I could tell that is 'smeared' easily and has a low melting point. I also suspect it's not fire retardant.

I saved part of the case and swapped the terminal block form the heating circuit and got it working again. But it was doing strange things again. Then it magically started working again. So I left it for a time, hanging outside the wall. I got busy with other things and left it for about a month. I checked and I could feel the terminal block was hot. Too hot for my peace of mind. I removed that terminal and hard wired to the solder pads to eliminate the high resistance hot spot. But then when I powered it back up, it got into a cycle where as soon as it energized a relay, the unit would reboot. This left it completely useless. It's brain was damaged. The next day I went to the brewing supply company and bought a Penn (Johnson) A421 for $80. Everything in there was much higher quality and bigger. The real deal. I kicked myself for being cheap and not getting it in the first place. Sure I saved $60 but at the cost of dozens of hours and hassle. Not worth it.

I know how these things work, because it's happened to products I designed. Someone designs a unit like this and has the Chinese make it. Then magically cheap clones show up on the market. But they use cheap versions of all the components and materials. I believe something similar happened with this product.

I think the STC-1000 is good for simple projects, but don't ever entrust it with anything you value like a wine collection. Also touch the terminals, taking care not to electrocute yourself like I did (it only stings a little) , when your unit is running and make sure they're not getting hot. If they do, I recommend removing them and direct wiring to the ample solder pads under the relays. They are not solder masked. It's a trick to increase the current capacity of a trace on a PCB. The solder builds up when it goes though the wave soldering machine and makes it thicker. This makes it easy to solder a wire flat to it. I used my IR thermometer to check after I was done and there was only a couple degrees over ambient by this method. But, that's a lot of work to save $60 and still end up with a controller I can't trust.

I also recommend buying two so you have a spare if it goes south on the weekend. Sitting in a hot garage watching the temperature go up and down and flicking a switch on and off is no fun.

My other recommendation is not to use it in an enclosed space. It's not able to handle high temperatures. And the high resistance contacts make a good heater so if it's in a small box, it will heat up the air in that box quickly. And it will do it's imitation of Chernobyl.

 

[Hermit]

Mine has worked flawless with both heat and cool for years now.

 

[day_trippr]

As have the vast majority of STC-1000 installations. Indeed aside from the occasional stuck relay I've not read of anything as apocryphal as the OP's post...

Cheers!

 

[CJ92028]

I am sure I made conditions worse. I saw another post on a pot growers forum and the guy encased it in a Tupperware container, presumably to protect it from humidity. I've seen other photos, many on this forum, with melt downs where they were in a small box. Though over current and inadequate circuit protection was generally the problem with them. But heat makes things worse. And I was not operating past the current limit specified and had adequate gauge wire.

Another thought I had was that they might not all be the same. I looked online and found that there are indeed several knockoffs. In some posts I've seen here they mostly seem to be black Hongfa relays. Mine are the blue Goodsky. And my power transformer is a fraction of the size. Then I found this article that spells it all out. I believe there's at least one more knock-off also. See attached.

As an aside I notices another potential problem. A residential circuit breaker is at least 15 amps. These relays are only good to 10 and 12 amps. Therefore one needs to fuse these relays to protect from over-current.

 

[TechFanMD]

So the current going across the contacts of your relays was only what the thermostat circuit required (as opposed to actually running the compressor power through the relays)?

Not sure I understand your setup completely without a schematic or sketch, but there are conditions with too low of a current for the relay contacts, and you don't get the self cleaning action of a small arc. Over time resistance will build from natural surface oxidation (patina) which increases the resistance at the contact which in turn makes heat, which can then cause failure. We see this in our equipment in some systems where it was designed to run resistant heaters for diffusion vacuum pumps but was upgraded to turbomolecular pumps which draw a fraction of the current. The control system relays for the power to the pump suffer the issue I mentioned above and 'fail'. Could this be what happened?

 

[501irishred]

Unfortunately the same story is true with $200 OEM controllers as with the $20 STC. A case in point is a TRUE cooler I’m using had a relay weld closed in the OEM controller and ended up freezing everything inside, including a case of 12%ABV IIPA! I elected to go back with the STC instead of the $150 OEM (my wholesale cost BTW). After three years in an unconditioned garage, I couldn’t be happier.

 

[CJ92028]

"So the current going across the contacts of your relays was only what the thermostat circuit required": Yes. And no schematic necessary. Remove refrigerator thermostat and connect wires to the STC-1000.

"Patina" is oxidation which is high resistance. There was no oxidation or corrosion. I'm don't know how much current draw is, but it's significant. Enough for a good spark if one connects the wires. I also tin the ends of my wires to make better contact. It acts as a strain relief, prevents strands from breaking, and flattens out giving a large contact patch when tightened.

The other cause of heat rise is thin conductors. Take a thick piece of wire and cut out all the strands except one and put a current though it. The fully stranded portions will be near ambient temperature but the single strand will be hot. One strand is high resistance and you have just made an electric heater. These fake Phoenix connector blocks have thin segments internally, much smaller than the gauge. My Mark-1 eyeball wouldn't trust them for more than a couple amps.

There's a rule for temperature rise in conductors like wires. I always have to look it up. But I've learned over the years that I'm safe if it doesn't get hot. My highly calibrated digital (finger) thermal probe usually tells me all I need to know. Or my DeWalt infrared temperature gauge. Run max current expected and see if anything gets hot. If it does, the gauge is too small. I've spent about a decade of my life designing wiring harnesses, dealing with connectors, designing circuit boards. I start with the manufacturer specs, but in many cases, I'f found them to be optimistic so I always try to do practical tests. I don't trust, I verify.

My problem was that I was duped into a cheap rip-off. I hope this can help someone in the future avoid these and get a genuine STC-1000. But for a few bucks more, I recommend the A421. All the components are a n order of magnitude bigger. Look at the size of these terminal blocks. This is the kind of contacts one needs for running a compressor IMO. Also, the A421 has Fahrenheit.

 

[ScrewyBrewer]

Mine has worked flawless with both heat and cool for years now.

Same here using an STC-1000+ connected to a chest freezer and 40-watt Fermwrap heater.

 

[MaxStout]

Shouldn't be an enclosure problem with STC-1000. It's a mechanical switch and if functioning normally maybe a fraction of an ohm impedance, so only a tiny amount of heat is going to be generated, even at the full 10A. Methinks perhaps a bad unit--a ground fault or short somewhere, or bad contacts causing overheating.

AFAIK, all the STC clones are made in China. Thousands of people use them without incident. I've had mine running a 5 cu ft. freezer and 40w bulb off and on for 4 years now. One bad unit isn't cause for such alarm. Anecdotal evidence is anecdotal.