12v pwm for two 10a fans?

I'm working on the heating system of an old Porsche. In each of the footwells, there is a 10A blower. The stock setup uses two large resistors to provide three distinct fan speeds. I'm hoping to bypass this and instead hook the two blowers up to a single pulse width modulator, thereby giving me infinitely variable speeds, as well as a reduced load on the (weak, expensive) alternator. Bonus points for a reversible fan. The closest thing I've found is this (which only outputs 0-5v): http://www.aliexpress.com/item/12V-30A-PWM-DC-Motor-Speed-Control-Controller-for-RC-model-LED-Bulb-Brightness/1157463020.html Ideally, I would want to output approximately 0-14 volts. Is my theory sound? Anyone have a better source for these things?

Google 12V 30A PWM controllers (or whatever current is appropriate) - there should be a few to choose from. Otherwise it's usually possible to add a MOSFET to a PWM output to boost the current rating, or relatively easy to construct a PWM circuit - I've provided a common 555 circuit various times on the12volt - eg: from dimmer on 65w halogen?. I'd set a minimum duty cycle of (say?) 10% for motors to prevent motor overheating. Note that a PWM won't have much effect on reducing the load - it's essentially still the same average current at a given speed.    As to the weak expensive alternator - especially if it has an external regulator - I highly recommend upgrading to an (Japanese) alternator with internal regulator (& internal fan!); in particular "2-wire" SL types (battery Sense and charge-Light rather than single wire (D+ or L = charge-Light) since they monitor the battery voltage rather than Ignition or alternator output voltage. For many vehicles this is usually not a problem since many share the same or similar mount types and dimensions, however Porsche may be one of those PITAs with unique design or confined mounting space. And by weak I assume "blows easily" whereby I suggest it's a Bosch alternator, and that is one alternator brand I avoid (they can't even handle being jump started!). You should find that Japanese alternators are far more robust. Alternator amperage should equal or exceed the OEM alternator rating. You can't have "too big an alternator" - the alternator only produces as much current as the loads require. (I've used 90A & 110A alternators in place of my OEM 23A generator. No way will I go back to externally fanned & externally regulated alternators!)

PWM motor controllers on ebay? Plenty to choose from, some with 40A switching capability.

oldspark wrote: Note that a PWM won't have much effect on reducing the load - it's essentially still the same average current at a given speed.

Wait, what? I assumed that a fan running at half-speed due to PWM would be using roughly half the juice. The controllers on eBay don't often specify what the output voltage is. Most are 0v to 5v.

A resistored fan at half speed uses about half the current. A PWM'd fan at half speed uses half the current. Both have inputs of 12V (or whatever the system voltage is). If motor speed is (also) proportional to voltage (and not solely current) then PWM will be more efficient. Alas I have long forgotten my motor theory so someone may wish to correct or qualify my statement. Maybe I should have stated "don't assume a lesser load". Some would argue a greater strain because of the chopping action, but many say bad grounds and thin wires cause a greater strain on alternators (ha ha). Unless otherwise stated, PWM outputs are the same voltage as the input.

PWM in simple terms is digital. In this application, the supply voltage is varied in time (pulse width). The output voltage is the supply voltage minus switching transistor losses...close enough to the supply voltage. The transistor is basically on or off, this can be seen in the heatsink sizing used. The control voltage at the potentiometer can be taken from a regulated (5 volt) source.

oldspark wrote: A resistored fan at half speed uses about half the current. A PWM'd fan at half speed uses half the current.

If I'm not mistaken, a resistored fan at half speed uses half the current, because the other half goes through the resistor? I don't see the case being the same with PWM.

Ok, if half speed means half the current, assume the 12V 10A blower (hence 12 x 10A = 120W & an effective resistance of 12V/10A = 1.2 Ohms) has 5A instead of its full 10A on a 12V supply. The PWM will chop the 10A current (on & off) at 50% duty cycle so the average current is 5A. Total power = 12V x 5A = 60W. The "halving" resistor will be 1.2 Ohms so at 5A it drops 5A x 1.2A = 6V (6V x 5A = 30W). The motor therefore has (12V minus 6V =) 6V @ 5A (30W). Total power is 30W + 30W = 60W (or 12V x 5A = 60W) - the same as the PWM. However you'll note the resistored blower power is half that of the PWM (30W versus 60W). If the blower acts like a resistor, then for the same blower power the blower would have 8.5V @ 7A (8.5V x 7A = 59.5W) - it is not 5A (ie, half of 10A) in which case the PWM uses 5A/7A = 70% of the total blower & resistor power. And that's where my motor theory is lacking. Is blower speed proportional to current and does it therefore have less torque (power) for the same current using resistive control? All I remember is that if you open the rotor field of a synchronous machine (motor or generator) it'll speed up to infinity... not that that is relevant. I do suspect I might laugh at these writings later on as my memories resurface...    But if your "half current, half speed" is correct, then PWM has no power saving. (It would in fact add the power used by the PWM control circuitry although that should be negligible, plus subject the alternator to a chopped current demand.)

Aha, that does help. Many thanks.

Keep in mind I don't know the answer - I've become like others that would rather spend hours & pages & delays debating rather than a 2-5 minute google search. (That's how we modern people keep ourselves too busy to have time for priorities.) But IMO my "half current = half speed" meaning no drop in power consumption is valid. However if half speed is a combination of voltage & current, then the PWM does save power - ie, it is not wasting the power in (heat from) the resistor. Certainly PWM is the most efficient method normally. It replaces other "dropping" methods that involve dissipating power. And for things like LEDs it is the only way to dim in proportion to a control signal (eg a pot or voltage). (Although a variable constant current feed could be used, but not for several parallel LED strings, or if mixed with incandescent bulbs etc.) I do now recall when building an AC motor speed controller way back last millennium (using phase controlled triacs or SCRs but essentially the same as PWM) that the source blurb referred to high or good torque at low RPM. In retrospect - namely now LOL - I reckon that must have been in comparison to alternate resistive methods. Torque versus power - now that's a tricky thing. But I know and can picture how my 400Nm 2 litre 45 year old car (135HP - I mean, 100kW) beat the crap out of 400HP V8s & rotaries & WRXs etc. (And I'm sick of telling people that an increase in engine power does not mean a heavier clutch and transmission - that is only needed if peak torque increases.) However I have little memory of electric motor torque (other than old steam cranes can get heavier things off the ground than modern electric etc cranes). Earlier than the drill/motor speed controller I fitted a dimmer to fluorescent lights. (That was when I was 15 years young but yes, I still strike people that claim it's impossible to dim fluoros!). That has to use phase control (not unlike PWM wrt current) to "merely" reduce the current but not the (peak) voltage - just like LEDs - tho AC fluoros require a filament transformer to maintain the high firing voltage. (The transformer replaced the fluoro ballast & starter. And we never replaced a fluoro after that except when they aged & lost brightness - in our case after 15 years. And kids had fun looking at the blue & white ripples traveling along the tube at lowest brightness.) I even used a PWM (the earlier 555 circuit) for a mate's stationwagon's 12V flouro tube. That makes me think PWM can be used to dim HIDs etc but that depends on their ballast circuit - modern day ballasts tend to be more electronic than the older ferro-magnetic (transformer) inverters and devices, and some will not tolerate a dimming signal. (hence why some AC LEDs and CFLs are dimmable whereas others aren't). And FYI, recently I reinforced that voltage cannot be used to dim LED arrays... I fitted a 3rd brake LED bar to my ute which I also wanted to act like a 3rd taillight. Instead of making a PWM I used a small dc-dc converter (thumbnail sized; ~$3 off eBay). I think it was ~8V (7.6V?) for the correct brightness. However I found that sometimes some LEDs would be dimmer of unlit - or occasionally brighter - in groups of 3 - the bar consisted of about 11 parallel strings of 3 series white LEDs.) At first I thought the darkies were blown but they proved ok. Temperature effected the LED characteristics and being like Zener diodes they'd take current from other strings. So yet again substituting voltage control for PWM was the solution. Incidentally, both solutions are equally efficient electrically. A dc-dc voltage regulator/converter (aka SMPS) is a (current) PWM with output voltage smoothing unlike "linear" converters that drop voltage effectively thru a resistor (even tho they use transistors, but they still waste the voltage drop times the current thru it/them (Watts) as heat). Oh well, that was yet more ramble - how unlike me! But a nostalgic ramble for me. It highlights how IMO so many things are the same (in principle), just different (in realising the circuitry etc). It also gives one of the 3 "impossible" things I achieved as a kid. I argued for year with those that said fluoro dimming could not be done DESPITE me saying I HAVE done it - it wasn't mere conjecture. (Another was the impossibility of making an electronic Halda - an accurate odometer for rally use. I don't recall the 3rd - it would not have been the 12V fluoro dimming as I consider that the same as AC fluoro dimming.) I recall 3 separate massive discussions - 2 with experienced electrical/electronic engineers & one with an expert/experienced electronic guru - but they could not be convinced. (Later I had demo videos so I could watch my opponents squirm, tho electronic Haldas have long been in existence, but so have dimmable fluoros.)    Not that anything changes... These days I'm mocked for suggesting CAS-less sequential ignition & injection (using only a distributor for timing - no cam position (cycle) sensing - and using traditional EFI; not ion sensing techniques). Plus a gamut of other things - often but not limited to automotive or electronic things. Just remember tho that PWMs are usually Open Collector or similar devices - they merely chop whatever supply (voltage) is applied to them.    Do not confuse voltage specs with their output - voltages usually refer to their control voltage. IE - since many PWM applications are computer controlled they may be bases on a 5V signal - meaning 1V to 5V or 0V to 5V etc. And 0V - 5V controlled PWMs (which may PWM 12V or 24V or 400V supply to the load) merely add a pot (variable resistor) for manual control - its ends to 0V & +5V respectively with its wiper going to the 0-5V control input.