[V8] Seat Heater switches

[Huw Powell]

I'm going to try to answer all the questions at once in one place.

Someone might want to copy this to the V8 and 20v lists if it doesn't 
make it to them.  I think I can post to them but I don't subscribe to them.

My answers and commentary are interspersed with Nick's two most recent 
emails.  My suggestions of solutions are after all that stuff, with an 
all-caps header: SOLUTIONS

As I understand it, your car(s) has heated seats, but only an on/off 
switch for them, and you want control over how much heat they generate 
so you don't have to turn them off manually.

Also, my reply is limited by knowledge only of the heated seat control 
circuits Audi used in the mid to late 1980s.

On 2/1/2015 12:33 PM, Nicholas Miller wrote:

> Not sure, obviously my electrical/electronics knowledge is fairly rudimentary.
> I've brushed with it for years at work, but mostly stick to mechanics or
> simple circuits..

I don't mean at all to be rude, but based on that, you might not want to 
try to do this. There are a lot of opportunities to burn your car to the 
ground with this project if you mess up.

Then again, if you build a stand-alone test setup first, it might be a 
great learning experience.

> This does look like a big endeavour.  I'd like to test at least one switch
> and see what the rotary wheel does, vary voltage or current.

It's neither, they "vary" resistance.  That is, they switch through a 
series of different fixed resistances.

A 1988 90Q switch reads 0-10-200-325-450-540 ohms, low to high (1 to 6)
A 1987 5000 switch reads 0-150-350-500-800-1000 ohms
I couldn't find my TT switches to check them.

(the lowest measurements are probably inaccurate; also I think the "0" 
position, which corresponds to "off", also read "0 ohms")

Now let us discuss the manner of operation of this system so you 
understand it clearly.

The seat heater elements (two in each seat, one butt and one back) are 
either on or off (and wired in series, not that that matters to you).

The butt element also incorporates a thermistor - a device that varies 
resistance in proportion to temperature.  This is the sensor.

The adjustment knob selects between a series of fixed resistances.

The control units, which look like relays, compare the resistance 
setting at the switch to the resistance of the sensor and turn the 
elements off and on periodically to keep the two values within a similar 
range.

> I checked my
> schematic for at least one set of seats, on my seats the temperature sensor
> is in the heater coil, all it gets is 2 wire signal from the switch, and
> everything else is sorted in the seat.

What seats are those?  The car you want to modify?  Be careful what you 
think are "signals" and which way current flows.  Also keep an eye for 
ground connections, since they also connect various devices.

> I think that may work, but in this
> case you would not get more heat, just variability between the current
> setting and off, maybe.  Best way to check would be to see what my current
> switch puts out for voltage, check inline current on the heater circuit,
> then check voltage output on the Audi switch and wire it inline to see if
> it varies current when inline.   I think it could work, even though not
> like it does in the original system.

More heat is not an option.

The Audi switch does not have a voltage output.

> Surely, the rotary switch does something with voltage/current even if it
> doesn't have the final say in the circuit, correct?  I understand in the
> end it seems to just signal a relay that then varies the power/heat level
> to the seat in the audi system, so my main question then would be how much
> current can the switch carry?  Probably not a ton, in this case.

Probably very little - perhaps 100 mA, if even that.  The Audi rotary 
switch does not pass current, it provides a specific resistance to the 
control unit.

> Thanks for the ideas.  I am wondering on the seat heaters, the current
> should be dictated by the heating element, right?  So that the heat
> adjustment should be accomplished by varying voltage, was my guess.

That is basically correct - the element draws a relatively fixed current.

But Audi adjusts the heat output by turning the element off at a given 
temperature, then turning it back on, and off again, etc.

SOLUTIONS

There are a few ways to achieve a lower heat setting than "full on".

Don't forget, whatever you might decide to implement, to install 
appropriate fuses!

One, the crudest, is to insert a series resistance in the circuit to the 
elements.  Your fresh air fan works this way - the switch (typically) 
selects between several taps on a high power resistive element which is 
usually housed in the fresh air plenum for cooling.

You could literally scavenge the parts from a fresh air fan control to 
do your project - the switch, wiring, resistor pack, if the resistors 
are of reasonably appropriate values.

This is crude for two reasons - one is the heat dissipation at the 
resistive device, and the other is that it wastes energy.  OK, those are 
really the same reason, and the second is that this method entails 
running high current wiring all over the place.

The second is the way the Audi seat heater systems work.  A 
thermostat-like system is used.  In the seat there is a thermistor - a 
semiconductor which increases resistance with temperature.  The switch 
encloses a set of low current resistors at set values.  A control unit 
(which looks like a relay, and does usually include one) compares the 
set resistance at the switch and the resistance of the thermistor and 
turns off the heat when the thermistor reaches the switch value.  It 
then turns it back on at some preset "swing" amount, like, say, 10% lower.

This turns the element on and off periodically, and has the advantage of 
running to a consistent temperature for a given setting no matter what 
the conditions.  It has the disadvantage of complexity, in that a 
retrofit would require obtaining and installing a suitable thermistor 
inside the seat.

The third way is probably best suited to a retrofit in a system that 
only provides on/off functionality (early to mid 80s small Audis were 
like this, by the way - just an on/off switch)

It allows you to control how much heat is generated via some sort of 
control, but there is no thermostatic feedback, just lower power settings.

There are two ways (at least) to do this - a high frequency duty cycle 
type system whereby the 12 volt supply is "chopped up" rapidly, 
effectively lowering the voltage supply in proportion to how long it is 
on versus off; and a (simpler?) timer system that turns the voltage on 
(and off) for varying fractions of a more macroscopic interval, such as 
10 to 50 seconds out of each minute in the "on" mode.

I can't tell you exactly how to do the former.  But, I am sure that 
there is some sort of packaged circuit out there you could buy that 
provides a "variable output" from 12 VDC, it's just a matter of it 
handling enough current.

You can measure the resistance of your seat elements and calculate this, 
with I = V/R.  Don't forget that V should be taken as about 15 to 
provide a safety margin.  If there are two separate connections, measure 
each one and add their current requirements.

The latter system can be achieved with one or two nifty little 
integrated circuits called "555 timers".  They'll run happily on your 
car's battery/ignition power, and for under $10 worth of parts can be 
set to turn on and off at intervals in the seconds to minutes range.

They can't handle much current, of course, so you'll probably have to 
cascade a pair of relays from their output.  The first will be a very 
small relay, with a miniscule activation current and the ability to 
switch up to 0.5-1 amp.  This will then control a typical automotive 
relay (the Bosch 30A relay coils draw about half an amp) which will 
switch the heavy current to the seat elements.

Once the components in this circuit are "dialed in", a variable or 
stepped resistance will be part of the circuit, and acts as your heat 
control by altering the "on" versus "off" timing.

If I were doing this I would also add a "full on" switch for the warm up 
phase - when it is  minus 10 dF out, you want those babies coming up to 
temperature as fast as possible.  That could even be on a timer, but 
it's easy enough to flip a switch when your butt is warm enough, letting 
the system go to the timed cycle which you have set to your general 
preference.

Advantages: easy and cheap to build.

Disadvantage: won't run to a specific temperature - it will put x watts 
of heat into the seats regardless of conditions.  A relative solution is 
where the Audi switches could come into play.  Since they offer a range 
of resistances, they could be used as the "variable resistance" in your 
555 circuit, and the numbers on the wheel would correspond to how damn 
cold it is out.

If you build a system like this, I would advise getting a 12 volt power 
supply and using your device to turn a bulb on and off in a place you 
will be for a while to make sure it does what you expect, and perhaps 
even to dial in your exact timing settings.

Best of luck,

Huw