Technical Musings: DIY Fan Module (long)

[Martin Gutkowski <webmaster@xxxx>]

Hi all,

I thought I'd share with you my thoughts on the fan wiring system that 
I've been puzzling over for the past few days on and off. Here's what 
I've come up with (advanced apologies for those who already understand 
the system, but I thought it'd be an idea to explain things from first 
principles):

I caught myself wondering how the Fanzilla worked and how I could 
duplicate its function for my own car. I soon realised that the 
sequential engagement of the fans is really simple to achieve. So 
simple, in fact, that I reckon anyone with a bit of soldering experience 
should be able to build it, given the right components.

With the fan-fail bypass installed in your car, the fans are simply 
hard-wired to the relay that engages to turn them on. This relay is 
engaged under two conditions: when the engine reaches a certain 
temperature, and when the A/C compressor engages (which is does 
regularly when the A/C is on). It's the latter of these which cause the 
biggest headache. The reason for this is that the two fans draw a 
considerable amount of current on start-up (for maybe a second or so), 
and given the signal to the relay coincides with the A/C compressor 
engaging - let's just say I hope you're not drag-racing at the time!!!

There is a second problem with the system: The relay which powers the 
fans is a single 40A standard relay. The fans draw more than 20A each on 
startup. There are documented procedures for re-wiring the 
fan-fail-module's socket to let each fan have an independent relay. This 
is a good thing from an engineering point of view, but doesn't solve the 
"kick" problem, and doesn't give you your fan-fail light back.

The Fanzilla elegantly solves the "kick" problem by sequentially 
engaging the fans, with a couple of seconds pause between each, and 
initially a couple of seconds after the compressor engages. It is this 
functionality I wish to duplicate. It also replaces the fan-fail light 
function. Ironically, this function is the most difficult to achieve 
(I'm not trying to copy the Fanzilla here - I am merely 
reverse-engineering what it does to save myself some pennies!). I feel 
confident in providing a design for sequential fan engagement, but not 
to replace the fan-fail light function as yet (there are many many ways 
of doing this, but I'm not happy with any of them, because they are 
either not perfect from an engineering point of view, or too complex for 
a "hobbyist" to duplicate).

My design involves constructing two INDEPENDENT circuits - one for each 
fan, and each fan is driven by an independent relay (I believe the 
Fanzilla uses solid-state devices for driving the fans - this is a 
better system, if you can get it right, but I want to keep things simple 
and reliable for a "hobbyist" to build). Each one of these individual 
circuits uses only seven components - and only one has more than two 
legs! Add to this the relays themselves, and a protection diode, and voila!

Is this safe?

I work as a network engineer, and the words "single point of failure" 
exit my lips every day. I wanted to make sure that if something goes 
wrong with the board, you could only lose one fan (short of a fluke). 
You rely on the single wire to the original relay to turn the fans on, 
so from that point onwards, my design runs two seperate systems to 
control the fans.

My design will plug strait into the relay sockets for the fan relay and 
fan-fail module. You get two further relay sockets, and into these you 
plug a new relay for each fan (I found some neat relays in Halfords the 
other day which have a blade fuse socket as part of the case). The cost 
of the components (not including relays) is about $30 - and most of this 
is the cost of two new relay sockets! I got all my components from Farnell.

Would people be interested in this design? I want to build and test one 
myself before publishing the circuit, and I need to think further about 
how to achieve the fan-fail light. From a "pure" sense, the light must 
come on if one of the fan motors goes short or open circuit. The 
function of the light is not to check whether your relay has come on - 
which is all the "reverse" function of wiring the light to the fan 
output achieves.

Right - enough "blinding with science"! The circuit it simple, it's 
designed so you'd be hard pushed to bugger it up, it's easy to test, and 
should anything go wrong, you're far less likely to lose both fans than 
with the stock setup.

I'll have a crack at building mine next week.

Martin
#1458
DOC UK