Transistorised Radiator Fan

Citroënbender · 31-10-2019, 09:03 PM

Anyone here been exposed to these on a smaller FWD Euro? Not a full CANBUS module, just tri-modal (off/slow/max) with a heatsinked transistor pack not unlike many cabin blowers.

Got what appears to be a faulty one on a Citroën (surprise, surprise). It won't return a signal to the PCM when actuated via diagnostics, and has a stored code reflecting the lack of this return signal in normal use. Runs constantly from a few seconds after startup and does not change speed.

**russellw** · 01-11-2019, 05:06 PM

I presume these are like the thermatic fans made popular a few years ago in as much as they use a sensor to detect the temperature and operate accordingly or do they rely on getting a command via the CANBUS?

Citroënbender · 01-11-2019, 05:55 PM

They’re a dumb fan that thinks it’s smart.

They have a relay for the high current feed, I think this is to limit a return spike when the motor stops.

The motor housing contains the pictured module encased in rubberised beige epoxy (mongrels!). Of the five pins, four are used - power and ground, control (modulation) signal and return signal. It’s like a 200A version of cabin fan controllers. I presume the choke visible is to restrict spikes, too. The fan when working, only operates low or high speed, it doesn’t pulse or vary beyond that.

Of course, a bit like BA steering columns, you have to buy everything in event of a failure - not just the motor.

Citroënbender · 03-11-2019, 04:12 PM

Here's the circuit. The high amperage supply is switched with KOEO, the commutator segments for circuits at 90°, not 180°, and the positive is straight from relay to the brushes - it's the earth which is modulated.

I'm hung up with no knowledge of what the control signal voltage or characteristics are, ditto on the return. Suspect it's plain old DC. Any suggestions on low-risk evaluation?

**X-AHH** · 03-11-2019, 07:42 PM

The control module seems to be temperature driven. The sig may just be to send a temperature dependant voltage to something else external to the fan unit?

Are you sure that the earth isn't straight to the fan body. (Difficult to tell from the photos)

Where's the choke you are referring to?

Citroënbender · 03-11-2019, 08:07 PM

The fan body is earthed, including heatsink and back of the PCB, but it doesn't carry directly to a brush pair. Choke as I called it, is visible in the pinky-beige rubberised view of the control module (second photo).

Fan only has two speeds, it may be capable of more by design but that's how it is operated by this PCM. Modes are: Off (cool weather, low load, no AC), Low (AC on, low to moderate engine load), High (all conditions exceeding these parameters or temp sensor failure). Module failure mode found by me is Low speed.

The engine only has one temp sensor; it informs the dash via the PCM and CANBUS and also does duty for the PCM itself.

**cs123** · 03-11-2019, 10:04 PM

How have you verified that the high speed isn't working. Is it being commanded by the PCM?

Kind of funky implementation of the fan isn't it. Is it PWM controlled? With the signal coming from the PCM and the fan controller driving it?

Citroënbender · 03-11-2019, 10:33 PM

I’ve got a crack of the factory diagnostic software, when I run either fan speed actuation (low or high) via the PCM it tells me the operation cannot be completed.

I presume it sends out the relevant command but is not getting the return signal expected. This is where I’m stuck without a scope.

Have been wondering how hard it might be to strip away the potting and see what’s under there. Don’t think it’s TO-3s as there’s nothing on the heatsink side that would reflect their riveting.

Edit to add, physically the motor is 100%. So I don’t feel it’s cooked its guts pushing a stiff motor.

**X-AHH** · 04-11-2019, 04:35 AM

Perhaps the speed high or low is dictated by sensed temperature?

The symbol on the circuit diagram seems to indicate that temperature plays a part in the module operation?

**cs123** · 04-11-2019, 05:56 AM

Quote:

Originally Posted by Citroënbender

I’ve got a crack of the factory diagnostic software, when I run either fan speed actuation (low or high) via the PCM it tells me the operation cannot be completed.

I presume it sends out the relevant command but is not getting the return signal expected. This is where I’m stuck without a scope.

Have been wondering how hard it might be to strip away the potting and see what’s under there. Don’t think it’s TO-3s as there’s nothing on the heatsink side that would reflect their riveting.

Edit to add, physically the motor is 100%. So I don’t feel it’s cooked its guts pushing a stiff motor.

Have a look on eBay. You’d be surprise how cheap scopes are. And you get to buy another tool

That coating will be problematic

Citroënbender · 04-11-2019, 07:38 AM

I’ve seen how cheap they are, and seen the kits to use with laptops, it’s a hangup to want one of the old school units in tidy, calibrated form. :-/

The icons on that circuit diagram indicate the unit function relates to (is influenced by / interplays with) water temperature and AC.

I still reckon it’s massive overkill by the maker to use this on a 1.4 litre “city commuter” car. My Alfa simply has a resistor on the low speed and toggles with a relay. It’s never overheated when the fan is working (once when the fan was literally jammed against the radiator post-collision, it did get warm until it selected high speed and self-clearanced).

Just as a side note, the fan is Faurecia brand (the successor to ECIA components) and the control module Sagem.

Citroënbender · 04-12-2019, 06:52 PM

I’m definitely no Jason ACT. :-/

Killed the unit dead trying to remove the coating.

Plan B came into action. Hotwired the fan after fully bypassing the control board. Set it up to run always on at least low speed, via a fan speed resistor (not unlike the BF). Used a Davies-Craig controller to bypass the resistor for high speed operation, set to 91° probe temperature.

One problem remains. Without a return signal from the fan motor, the AC clutch won’t stay pulled in. So I need to work out how it can be spoofed.

Citroënbender · 05-12-2019, 08:17 PM

Have shovel, will dig self a hole.

I'm now not sure if the AC clutching is a return signal issue or the solid state relay on its last legs (they are undersized). The clutch has a large air gap. A regular 40A relay with diode, might be clever.

Backprobed pins 1 & 2 of the three-way fan controller plug for DC voltage; 1 was V(bat) while 2 was 0.7V, seeming steady.