Why do my thermostatic electric fan switches keep failing?

The car in question is an 89 240 DL non turbo.

I replaced the pulley driven fan with an electric one and wired a thermostatic switch with a relay to control it. That was about four months ago, and I am about to install switch number three because they keep failing. They seem to trip gradually at a higher and higher temperature until the temp gauge is up into the red. I though maybe my gauge sender was acting up, but when I replace the aftermarket switch, it functions fine.

I have a 1" x 1/2" x 1" copper tee with a 3/8" npt adapter soldered together with a spade terminal soldered onto it to ground the tee. I have this in the lower rad hose, with the switch threaded into it. I have a relay and everything will work fine for a month or so and then it temp it switches at starts creeping up. I have a light so I know if there is power going to the fan or not. The light feeds off of the fan so the only thing the thermo switch is powering is the coil on the relay, which was supplied with the first switch I bought. The fan is from another model Volvo, I cant remember which one I pulled it from. It has a low speed and high wire, I am only using the low speed wire, which has provided ample cooling. If I bypass the switch and let it run whenever the ignition is on, I have no overheating problems.

I cut the last switch open to see if I could learn anything. I did not see anything that would cause the problem. It appears the contacts are pushed together by a shaft that is in turn pushed by a bi-metallic diaphragm. It is a 200 on 185 off switch, and I have a 195 degree thermostat. Battery and charging voltage reads normal. Alternator is a year old and I have a volt gauge on the dash, so I have a pretty good idea of what the alt is putting out at all times.

I have been getting my switches from the same retailer on ebay, they sent the first replacement one free. I thought maybe they were poor quality, bad batch, etc. but several reviewers on summitracing.com are having the same problem as me. Does anybody have a clue why these switches keep failing?

http://www.ebay.com/itm/131387544433?_trksid=p2060353.m2749.l2649&ssPageName=STRK:MEBIDX:IT

The temp switch made for the radiator fitting has continued to work for me for about 3 years now. I've had a couple of problems supplying the input side of the relay from time to time, but the temp switch has been, as they say, nails. It is, perhaps, measuring the wrong temperature, but it at least seems very consistent. My instrument cluster with the bypass of the compensation circuit never fully reaches the center dot without some problem other than the temp switch.

well I put a $20 probe and adjustable switch (cant recall brand) on a car a few weeks ago and it has less than 2 miles on it and the system has already failed. Fan ALWAYS ON.

I have a Mahle switch in my lower hose that is easily older than a year, thepornotaco on TB installed it when he had the car. Maybe it's time to try something not off ebay?

Just curious... when you pulled the switch apart and examined the contacts, was there any evidence of burning? Was there sufficient snap action to assure contact wipe? I would assume you are using a standard cube relay with 200mA coil and that such a switch should be more than adequate to survive any arcing from coil collapse (flyback) but if you were the experimenting type, you could tack a diode across the relay coil. If I were the vendor of that switch, I'd want to know. The failure could have some rather expensive and serious consequences to a driver not so vigilant as you are.

That is correct. Your problem is actually coming from the relay, not the switch. Choose a relay that has an internal diode across the coil. It is easy to add a diode, but why not get the correct relay if one is available. Each time the temp switch opens up, a large voltage is generated by the relay coil that causes an arc at the switch and burning of the contacts. A reverse connected diode prevents that problem. If adding a diode, the cathode end of the diode will connect to the battery side of the relay coil ( then, anode to ground). Put the diode right at the relay terminals.

cathode end has a white band ------|<------ anode end. Diode type 1N4004.

If possible, post a close-up of the internal contacts of a failed temp switch.

-Lazarus

Hasn't anyone thought of using a fan switch meant for VW/Audi? They come in 2 and 3 connector varieties.

Here is my relay, It also has markings "085" "963302" and "P2" This relay came with the same type of thermo switch, wiring harness and circuit breaker. It came as a kit, with a 185/170 degree switch, I didn't realize that it was too low a temp for my thermostat so I replaced the switch with a 200/185 one. I ordered it a few years before I actually did the fan conversion, so the retailer's info is no longer in my ebay account. Judging from the schematic on the side, it appears there is no diode in this relay.



And here is the inside of the switch itself. There is some filings scattered about from cutting it open and the notches in the disk are from my ever so careful hacksaw method.



Is this where the diode would go? Is this what is known as a reverse bias diode?



Where / how can I find the proper relay? And if one is not available will this diode work for me?:

https://www.digikey.com/product-detail/en/1N4004/1N4004FSCT-ND/1626118

Thanks for the help, you are not only gentlemen, but scholars as well.

I noticed a schematic on the side of a relay recently that had a diode included. That is all.

The 1N4004 will be fine. I have some 1N4007's on my A/C clutch and associated relay but only because I have a couple of hundred of them lying around.

Diode or resistor parallel to the windings in the relay. I had a similar issue with my old Galaxie, the fancy electronic fan control would cook itself in about two days thanks to the spikes from the relay opening. Wasn't a huge problem until it failed on in stalled traffic on I5, I went to restart once the highway reopened and didn't have enough left to kick it over.

4x4_Welder said:

Diode or resistor parallel to the windings in the relay.

Click to expand...

A resistor? It'd have to be a pretty low value to be enough to snub the inductive kick and you'd get quite a bit of parasitic heating while the relay was pulled in. On the other hand, diodes don't have a down side.

You would think so, but most genuine Bosch relays have had a resistor built in for years-

Unless they ran out of arrows to stick on the schematic.

Nope, you are absolutely correct. Turns out some relays use resistive damping because :

A) it's cheaper.
B) it's not polarity sensitive.
C) it's cheaper.

To get a resistor with a low enough value to provide the *same* amount of clamping a diode will provide is pretty much impossible, and means the relay is going to significantly increase power consumption and heat output, so they are generally used where a higher voltage inductive kick is likely to be tolerated. An electro-mechanical switch would qualify as a driver that was less sensitive to the inductive splash than perhaps a semiconductor driver. (data table found in TE Relay products http://www.te.com/commerce/DocumentDelivery/DDEController?Action=srchrtrv&DocNm=13C3311_AppNote&DocType=CS&DocLang=EN)

Bosch specific relay information found : http://bowery.com/maserati/home/files/bosch%20relays.pdf

Thanks, I've learned something valuable from this thread.

BradC said:

Thanks, I've learned something valuable from this thread.

Click to expand...

This is news to me too. Thanks indeed!

Xeroni said:

I have a Mahle switch in my lower hose that is easily older than a year, thepornotaco on TB installed it when he had the car. Maybe it's time to try something not off ebay?

Click to expand...

I think you called it. I installed a new relay with a diode in it and got a replacement switch from the same guy on ebay. The fan wouldn't turn on at all.

I connected it to a multi meter to it and used a piece of wire to build a frame that held the probe of the sensor and a meat thermometer in a pot of water. Well as it turns out the 200 degree switch didn't close until 215. I contacted the seller and he has stopped responding. I also notice now the listing says lifetime warranty (yea right)

So I bought a switch from Ron Francis wiring and my fan has worked flawlessly since. Now because I put in a suppression relay, I guess I can't know for sure that it's bad switches and not a voltage spike, but I think with the combination of 3 failures in a row, the last switch being 15 degrees off and shady seller is enough evidence for me.

ducttapeandzipties said:

So I bought a switch from Ron Francis wiring and my fan has worked flawlessly since. Now because I put in a suppression relay, I guess I can't know for sure that it's bad switches and not a voltage spike, but I think with the combination of 3 failures in a row, the last switch being 15 degrees off and shady seller is enough evidence for me.

Click to expand...

As an independent observer with an average level of common sense, I would lean toward the switches being defective. The idea of contact destruction by the relay coil flyback seemed like a stretch when I posted it, but now the gradual increase of make temperature dissuades me from that theory.

Thank you for following up and your thorough treatment of the subject!

Is the switch designed to have current passing through it? My Derale specifically states you can run power through the switch, otherwise I'd be trying to ground through it..

Use an OE 940 fan relay and decent brand switch in the rad or the SAAB T in the lower hose? I've been using ancient Wahler switches out of old dead rat SCAABS to repair steaming piles of various makes and junkyard 940/960/850 relays that are 20 years old with virtually no issues other than odd failure from a freak occurrence (flooding or wire rubbed through somewhere or fan drawing epic current due to drag/overheated/shorted windings). I don't think I have hardly any switches younger than about 10 years.

I've had the push-in regina switch fail a couple times, and seen a couple leaky ones in the junkyard (usually around a partially dissolved gasket from battery-acidic coolant/galvanic corrosion or something).

But this is all over a period of 15+ years and a bunch of really beat up POS cars I've seen. Granted, I didn't own these cars from new and do all the repairs either.

-Your friendly neighborhood cheapskate who virtually never buys new parts.