Have 12v AC, converted to 12v DC, motor slow

[Mikel_NY]

Hello All,

I have a stater that puts out AC voltage, at max RPM it produces 40 volts. I have a voltage regulator installed parallel and it controls volts to max 14+v AC.

I need to run an 80 watt 12V DC cooling fan so I installed a little block that has 4 diodes inside. Possibly this is called a rectifier, I am not sure. I was told how to install with AC on two posts and DC load on the two posts that are labeled "+" and "-".

After install and test I have found that the fan is spinning slower than if I powered the fan from a 12V battery. ( I would like not to run a battery )


Would a DC motor run slower with a 12V AC converted to DC than just wiring it to a 12V DC car battery?




Please explain in simple terms as this is not my field of expertise.



This is a snowmobile motor with a 230 watt stater plate and I am cooling it with a radiator and fan similar to a car setup. the fan is 12v DC 80 watt.

Thank you in advance for any input to help with original fan speed to gain more airflow to cool motor.

 

[Baluncore]

Welcome to PF.
How much slower?

The assumed “bridge rectifier” you are using will have silicon diodes that will drop about 1 volt each when conducting. That reduces your 14V to 12V or less.

If you have wired it wrongly then you might be getting 12V for only half the time.

AC is measured as RMS. So the peak voltage of AC is √2 = 1.4142 times the equivalent DC.

Does your regulator set the peak at 14V for battery charging? You may need a small battery to maintain the 14V all the time.

 

[Kavik]

There's two main reasons I can think of that would affect your motor:

1. The output voltage of a rectified signal is not a flat DC voltage, it is actually just a sin wave except the negative part of the wave is reversed to be positive. The average voltage of this wave is lower than your RMS voltage. This picture represents that well (taken from isaac42's post on HarmonyCentral Forums). Just pretend the 24V RMS is your 14V and the ratio of 24V to 15.3V still stands.:
   
   
   
   
   Source: http://www.harmonycentral.com/forum/forum/Forums_General/acapella-94/293518-
   
   
   
2. There is a voltage drop across diodes, usually about 0.7 V. This will also reduce your effective DC voltage.

 

[Mikel_NY]

the outflow of air is noticeably less using the converted AC power than just using the batter power.

Knowing VERY little on this subject ... could a cap be installed to help in this problem?(I only know that a cap is like a very small battery. it stores power to be discharged later when primary voltage drops.)I would like to eliminate a battery for weight saving, can it be done in a different way? maybe I just need to run a battery.
What is the best way (and LIGHTEST) to run a 80 watt DC motor (fan) the fastest for cooling with a given 12-14 Volt AC supply rated for 230 watts? ( tomorrow I will meter the AC volt at idle and then after the gismo measure the "DC" voltage and see what drop there is.)

 

[Mikel_NY]

here is a picture of how my "gismo" was explained to me. It is the diamond shape in the middle.

would guess the left image is a transformer which I don't have because my stater is already outputting 12-14V AC.In the picture, on the right there is a straight line over a curve (frown) line. this I think is a cap. If this is true, homework would I size one if it would help the fan spin faster?

If one is installed ... is there a cycle/service life for a cap?

 

[Baluncore]

Does the stator produce 1, 2 or 3 phase output?
Is it wired as a 3 phase star with the centre connected to the chassis, or maybe a floating delta?

If your 14V regulator is for battery charging it will be regulating the peak AC, not the RMS

To use a capacitor instead of a battery;
An 80W fan on 12V will draw a current of 80W/12V = 6.67A
What is the frequency of the AC? If 100Hz then period T = 10msec.
C = I * T / dV; For a 1.0V droop in the capacitor during the cycle;
C needed will be C= 6.67A * 0.01sec / 1.0V = 0.0667F = 66.7mF, rated at 15V or better.

Maybe buy ten x 10,000 uF capacitors. US$2 each.
Wire them in parallel to give 0.1F that will handle the current at the stator output voltage peaks.

 

[jim hardy]

Are you sure that regulator gives 14volts AC not DC ?
I'm guessing that your snowmobile regulator is similar to the motorcycle regulators I've seen. Their operation verges on brutal.
What they do is first convert to DC then measure the voltage across the battery and figure out how much of the sinewave needs to be let through to keep about 14.5 volts there.
At some point in the sinewave the regulator throws a dead short across the alternator which removes the remainder of the alternator's ac sine wave.
In Kavik's sketch above, the regulator would chop off the wave as required to maintain battery voltage about 14.5 .

With no battery present to absorb current i'd be at a loss to predict how your regulator would work. It'd probably chop the wave early which would give you lower average voltage across the motor.
In Kavik's sketch, draw a line straight down from each sinewave peak and erase the right half of the sine wave.

That's how the regulators that I've seen work. That'd give you lower average voltage at the fan.

You might try a large capacitor across your fan.

80 watts at 12 volts is 6.67 amps.
Around 5000 microfarads of capacitor should make the regulator think there's at least a tiny battery out there. The more capacitance the better. You could try a few of these.
https://www.amazon.com/s/ref=nb_sb_...=6800&rh=n:306788011,k:6800&tag=pfamazon01-20

EDIT: Ahh i see Baluncore already answered. We're thinking pretty much alike.

 

[Mikel_NY]

I would say that the regulator is outputting AC because the fan would just twitch when turned on, it would not spin until I installed the square block and I have the meter set on AC to read the voltage output.

The regulator is wired parallel from the factory if that bit of information helps. I am clueless on Caps ... I like Jims post showing me what to buy, I like Baluncore's post telling me exactly what I need.

Just looking at the numbers, they are not the same. Am I wrong here? I like to ask ten questions and get this correct the first time than try once and fry some wiring.I think I can test for frequency, I do not know if the stator is single or 3 phase. If you describe a test method, I can report back. I know that the stator is a 6 pulse and there is 2 wires coming out, Positive and ground.

 

[jim hardy]

Balun and i assumed different frequencies, he used 100 hz and i guessed at 1000.

He used mf for millifarads, i used microfarads which also starts with "m".

We're not so far apart.

You just need plenty of capacitance. British motorcycle racers in the late fifties used a large capacitor to replace their battery. Manufacturers thoughtfully oriented the rotor magnets to provide a pulse of current just when the ignition coil needed it. If you ever kick-started a 500cc single you really appreciate that !

Anyhow - a few tens of thousands of microfarad should do the job. More won't hurt.

You might think of a capacitor as a tank that stores electric charge.
The unit of capacitance is the farad.
A farad is a lot of capacitance. We usually work in millionths of a farad, microfarads.
Think of a farad as a 55 gallon drum compared to a microfarad , which holds just a thimblefull.

Do you have a friend who'd loan you a car stereo capacitor? Walmart sells a 1.5 farad (1.5 million microfarads) for $25.
http://www.walmart.com/ip/Power-Aco...5F-1.5-Farad-Digital-Power-Capacitor/21607606

here's some 10,000 microfarad like Baluncore suggested

http://www.futurlec.com/Capacitors/C10000U25Epr.shtml

these are 25 volt rated . 16 volt would probably be okay but i like a margin of safety.
Ten of them in parallel would be 1/10th of a farad.

capacitor goes on DC side of your rectifier - right across fan wires. Observe polarity : + to +

 

[Mikel_NY]

I dialed meter to Hz and hooked probs to AC regulated wires. The voltage at idle (1500) is 12.3, the meter on Hz read about 150. When motor was reved to say 4000 the meter switched range to "K" and meter read ".5xx".

I than moved the probes to the "DC" output and it was 9.3V DC, no AC measurable.




DO the caps, when installed, boost voltage average like the average line in pic of post #3?

 

[Baluncore]

You have two terminals from the stator, that is single phase. If the chassis wire remains connected to the chassis, then the bridge rectifier will produce +ve and –ve to the capacitors and fan. Both the +ve and –ve will be swinging together in voltage relative to the chassis connection.

High RPM gives about 500 Hz so that will give a ripple frequency of 1kHz, T = 1 msec.
C = I * T / V = 6.67A * .001sec / 1.0V = 0.006670F = 6670 uF
Voltage will be lower at idle when cooling may be not needed so much.

For the same fan voltage at idle,
Idling, at 150Hz, will give a full wave rectified ripple frequency of 300 Hz. T = 3.3msec
C = I * T / V = 6.67A * .0033sec / 1.0V = 0.022011 F = 22000 uF.

Yes, the caps raise the average voltage. The caps are charged only during the voltage peak from the rectified AC, they must store enough energy to run the fan through the rest of the half cycle. If the current flows for half the time, the current into the cap will be twice the average current out. If the current flows for 1/5 of the time, the current into the cap will be five times the average current out. The peak current could be maybe 35 amp. That peak current must be handled by the bridge. The cap(s) will need to be low ESR. I prefer multiple capacitors to distribute the peak current amongst several smaller units.

 

[Baluncore]

Can the regulator handle a 35 amp charge current? Is it a permanent magnet alternator designed only to charge the battery?

If the alternator has no controlled field winding then you might consider removing the series regulator for battery charging, then use a single diode to the fan.

Also consider rectifying a chassis referenced AC, but half wave, with only two diodes into separate positive and negative rails, each with it's own capacitor(s). That will double the fan voltage.

 

[Mikel_NY]

the motor will max out in the 8500 RPM range, I only rev the motor on the stand to said 4000 rpm.

The stator is stationary, there is a flywheel that rotates around the stator with permanent magnets attached to flywheel.

The regulator is standard factory issue and installed as it comes from a stock factory install. that being said, it was not designed to run a DC motor. Just a few light bulbs ( about 150 watts total ) and heated hand grips that could be 45 watts.

They sell a 3 wire regulator that is designed to charge a battery if sled was ordered with electric start.

I would like to run the fan in my application as fast as possible ( w/o burning it out ) for max cooling without having the "heavy" battery. I'm trying to lighten the chassis as much as possible.
One of the two wires coming out of the stator goes to ground. I also have a ground wire from engine block to regulator body.



So with the new information of the motor max RPM in the 8300 rpm range and not testing it but measured at idle the Hz was 150, and mid 400 RPM Hz want to 500 ... DO I need to measure the Hz at top RPM to size the caps correctly?

the square block is a Motorola MDA 980-2, 8723 W Bridge Rectifier. I see a 2V drop on the DC side vs. the AC side.

 

[jim hardy]

MDA980 is a 1970's 12 amp rectifier.

Since the original alternator only needed to run lights and heaters there was no need to turn it into DC. So they might not have.

On motorcycles one disconnects the regulator and measures stator AC voltage with nothing but a meter connected to it. That's called "open circuit stator test".
That way you have nothing but magnets and wire active, no electronics to fool you.
Around twelve volts idle, forty or fifty at midrange rpm . Voltage should be linear with RPM.
Don't know what voltage to expect on yours. But if open circuit voltage doesn't go up with rpm it sounds like stator troubles.

If it voltage goes up nicely with rpm, doubling when rpm doubles, then stator and magnets are probably fine. Reconnect regulator and continue testing.

old jim

 

[jim hardy]

ps i see no need for voltage measurement at top rpm. I don't like to run engines that fast unloaded.

 

[Mikel_NY]

here is my wire schematic. the only part that is confusing is the stator plate is labeled engine ground and the regulator is chassis ground. should this be separate? I don't see a closed circuit if the two grounds are not connected.
I was asking If Hz measurments are needed at top RPM to size the caps correct. just guessing.

 

[Mikel_NY]

I have tested open voltage and it climbs with RPM to a may of 40V AC. the book says up to 45V AC is good.

 

[Mikel_NY]

I did all testing with book specs to stator as open voltage and wire to wire resistance. All very close to book. book also says 20% error is ok and testing should be at 68 degrees F.

 

[Mikel_NY]

MDA980 is a 1970's 12 amp rectifier.

Is this "big enough" for running a 12V 80 watt DC motor (rad fan) continuous for 20 minutes at a time in under hood heat of say 200 degrees F?

 

[jim hardy]

T to Brn must be the lighting coils? That's where you'd make the open circuit stator test.

I'd wager that W-WR is trigger coil for CDI and G-R-BR is power for CDI.

Surely engine is bolted to chassis so grounds are common - is there not a braided link like in modern cars?

 

[jim hardy]

sounds like your stator and magnets are fine

yes 12 amps is plenty for that 80 watt fan, about seven amps . Feel of it - if it's getting hot mount it on a piece of aluminum in the airflow.

good work you have eliminated a lot of potential unknowns.

With Balun's excellent tutelage you'll be going soon.

Have fun !

 

[Mikel_NY]

yes the Brown (ground) on the right and the Left wire that goes over the top of the stator is the Yellow wire (positive).

I have the engine bolted to the chassis and the case of the regulator has a 12g wire bonded to the engine case.that print is out of the Polaris book. when the motor is installed in a sled. the motor is rubber mounted, the carbs are rubber mounted and the exhaust is rubber mounted so I can't see where the original application's engine and chassis ground met. there was no bond wire I can think of.

With that being said ... I do not have this motor in a stock chassis. the motor is now hard mounted to chassis and grounded to chassis.

 

[Mikel_NY]

I thank both of you for your help. I just need a final answer as to sizing and quantity of the caps. then its off to the races so to speak.

 

[Baluncore]

It's not that easy. We need to identify the existing regulator capabilities. Is it a linear or switching regulator?

I suspect your existing regulator may not handle the 6.67A fan current when the rectified stator voltage is high.
I expect you will need a switching regulator for the 14V, 6.67A fan supply. Unfortunately, cheap buck regulators only work up to about 32V, like this example;
http://www.dx.com/p/10a-dc-cnoverte...ge-regulator-power-supply-214277#.VAPZRaOuDs4
The advantage is the high conversion efficiency. 6.67A out at 14V will only require 3.12A input at 30Vin.

 

[256bits]

It is a shunt regulator.

Most of these types use a zener, or amplified zener to chop off part of only one cycle of the waveform.
The reverse cycle should be the full voltage from the stator.
It is what is called a half wave regulator.
Simple and few part count, with the excess current being dumped to grd.

So what is going into the regulator is one half of a sine wave and a chopped sine wave on alternative parts of the cycle.

I doubt if this type of regulator is any different than what is used, or has been used for the last 50 years or more for running the always-on headlamp, taill lights and brake lights with no battery on snow mobiles. In fact, even snowmobiles with a battery and a DC charge circuit, the lights run on this AC waveform, but I have no idea of how the DC is made compatable with the AC so that cross currents do not flow.

Using a bridge rectifier complicates things somewhat as the GRD of the motor does not match the GRD of the stator, engine, or chassis.

 

[jim hardy]

Hmmmm well i learned something.

What i expected to find is in figure 22 here:
http://www.schematicsforfree.com/archive/file/Automotive/Other/Charging%20and%20ignition%20systems%20for%20atv%27s%20nbr18.pdf

but it seems you have something closer to figure 16, with your bridge tacked on behind.

I'll guess their battery equipped snowmobiles are closer to figure 19 which does what you said - clips one half of the wave and let's the other fall where it will.

The reason such simple regulators work with permanent magnet alternators is this:
Any current flowing in the stator makes a magnetic field that opposes the one from rotor's permanent magnets .
That makes an alternator inherently current limiting - when voltage rises so does stator current which weakens total magnetic field, pushing voltage back down.
That's why the resistor in figure 19, so voltage won't rise when headlight is switched off.
Whatever current is left over after lighting on positive half cycles can go into battery. Regulator makes sure it won't overcharge battery.

gotta go, more later.

see also this guy's page:

http://mastercircuits.blogspot.com/2012/07/half-wave-to-full-wave-conversion.html

 

[jim hardy]

from post #16: here is my wire schematic. the only part that is confusing is the stator plate is labeled engine ground and the regulator is chassis ground. should this be separate? I don't see a closed circuit if the two grounds are not connected.

I can't quite make out the symbols inside triangles.

Looks like regulator is across main winding , brown and yellow, as expected for a shunt regulator.

Brown looks like it's grounded to engine at triangle adjacent stator? Is that an "E" inside triangle?
It's also grounded at voltage regulator by triangle with a "C" inside?

Are those two grounds maybe one at engine and one on chassis?
If so the brown wire ties engine to chassis.

Now one wonders how alternator would behave if it powered the fans with no regulator. Would alternator's inherent limiting keep voltage manageable ?

Do your fans have electronics inside them or are they just plain old brush type motors?
Clearly the engine will run without a regulator else you couldn't have made that open circuit stator test.

Sorry, i wasn't going to meddle further. Just i sensed we were all missing something about the regulation scheme. It's halfway between my lawnmower(none) and a Suzuki 650(three phase full wave). Your term "half wave regulator" was new to me, i never encountered one before.
And living in South Florida nearly all my life i never encountered snowmobiles.

What other interesting tidbits have you for us?

Have at it you younger and quicker minds.

old jim

 

[jim hardy]

I thank both of you for your help. I just need a final answer as to sizing and quantity of the caps. then its off to the races so to speak.

Balun has sized them at 22,000 uf. I sized them at "a few tens of thousands of microfarads." We agree there.
Two or three 10,000's should do. More won't hurt.
Capacitors have a rating for AC current . It's called "Ripple Current".
Basically the larger the case size the more AC current they can handle - there's more surface area to get rid of the heat. That's why he suggested using several normal caps instead of one humungous one like i linked.
Feel them and if they're getting hot to touch we'll have to address capacitor heating.
Just for ballpark: this series of (very good) capacitors has ripple current rating of about four amps.
http://www.vishay.com/docs/28336/120atc.pdf
Without calculating anything, conservative common sense says :
Seven amps divided between just two of their 6800's would be fine.. four or five would be better to get more capacitance.
You'd want four or more of this slightly less expensive but more readily available capacitor with only about two amp ripple rating..
MAL213816682E3
http://www.digikey.com/product-search/en?vendor=0&keywords=+MAL213816682E3+&stock=1

That's one example of why i remarked about Balun's excellent tutelage.

Sorry if this seems like rambling. I try to explain "Why" as well as "what"; after all this is a Physics forum.. physics has a practical side though.

over and out...

old jim

 

[Mikel_NY]

Brown looks like it's grounded to engine at triangle adjacent stator? Is that an "E" inside triangle?
It's also grounded at voltage regulator by triangle with a "C" inside?

Yes, E inside triangle is engine ground and C inside triangle is chassis ground.

Now one wonders how alternator would behave if it powered the fans with no regulator. Would alternator's inherent limiting keep voltage manageable ?

The fan would not turn. I only saw the fan quiver when switched on/off when connected directly to Yellow Brown from stator.

Do your fans have electronics inside them or are they just plain old brush type motors?
Clearly the engine will run without a regulator else you couldn't have made that open circuit stator test.

I believe the fan is a plain cheap brush motor, nothing fancy.

The engine only spins the stator, I see two other circuits that power the CDI and sink the rotation timing.

The book has a test to unhook regulator and test fro AC up to 45V. mine tested to 41V AC. book says that is good.

the stock wire harness has the regulator wired in parallel w/ the Yellow Brown wires off the lighting side of the stator. the motor knows nothing if it is working or not.

I have the tach hooked to the Yellow brown wires to count the pulses and show RPM. If regulated or not that still reads pulses. I have another set of Yellow brown going to the bridge rectifier. the fan also hooked to the other two posts of BR.

With this setup, the motor runs (up to 5000 rpm tested), the light bulbs don't blow (new regulator) and the fan spins ( bridge rectifier).My issue now is fan speed ... I think the fan is spinning slower than when hooked directly to the battery.

 

[Mikel_NY]

Sorry if this seems like rambling. I try to explain "Why" as well as "what"; after all this is a Physics forum.. physics has a practical side though.


I find none of this "rambling"! I find this a type of school ... you know this, I don't. This is awesome to me that there are people that are good at something to share knowledge and be patient enough for me to learn another fraction of something I didn't know yesterday.

 

[Mikel_NY]

Two or three 10,000's should do. More won't hurt.
Capacitors have a rating for AC current . It's called "Ripple Current".

Do i understand that I will be installing the caps on the DC side, parallel with the fan?

Now I read that caps have a rating for AC ripple ... Do I install two different types before and after the rectifier? ... I'm only asking for clarification.

 

[jim hardy]

The fan would not turn. I only saw the fan quiver when switched on/off when connected directly to Yellow Brown from stator.

Correct. Fan is probably a permanent magnet motor. They won't run on AC like the "universal" motor in your Skilsaw or drill.
Yellow to Brown is AC . And it's regulated.

What i'd like to try is:
Regulator and lights disconnected.
Yellow - Brown connected to AC terminals on bridge
Fan connected to DC terminals on bridge, voltmeter same place to read fan voltage. Note fan won't be grounded.

Then start engine, observe fan voltage (and current if you have an ammeter?) at idle
then gingerly increase engine speed until you see almost 15 volts on fan, note RPM. (and fan current?). Don't overrev engine though.
Now we'd know the raw capability of the alternator .
And can figure out a test to determine whether that half wave regulator is appropriate.
Baluncore questioned that way back in the thread.

Do i understand that I will be installing the caps on the DC side, parallel with the fan?

Yes. Exactly.

Now I read that caps have a rating for AC ripple ... Do I install two different types before and after the rectifier? ... I'm only asking for clarification.

No. Not yet, anyhow...
Rectified AC makes a lumpy wave with peaks as in Kavik's post #3. The capacitor's job is to store current between the peaks , to fill in the valleys if you will pardon that informal phrase.
So current rushes into the capacitor on the peaks and dribbles back out between peaks. That's AC. And it's called "ripple". (see http://en.wikipedia.org/wiki/Ripple_(electrical) )
So your capacitors go on the DC side.

Try adding some capacitors across fan with the setup you have now, regulator in place. You might find some in a junk computer power supply or TV set from a trashpile just for an experiment.

I hope this project helps you understand the snowmobile alternator. It's an interesting device, closely related to the synchronous machine. We learn fastest by doing.

old jim

 

[Mikel_NY]

And can figure out a test to determine whether that half wave regulator is appropriate.
Baluncore questioned that way back in the thread.

I'm eager to see where this comment will be going, sounds like "2 ways to skin a cat". ( my interest is peaked). Are you thinking that the factory regulator is the cheapest production item that served the original application and there may be a better regulator for my current application depending on if the stator makes enough juice? (just a guess)

Edit: Just reread Baluncore's post 24, (that must be where I got the idea from)

Jim, I will do some more testing tomorrow and report results.

 

[jim hardy]

Edit: Just reread Baluncore's post 24, (that must be where I got the idea from)

Yep. I'm unabashed - will steal a good idea anywhere. Balun is a cornucopia of them.

Thanks Balun - as usual i was a day behind...

old jim

 

[Mikel_NY]

What i'd like to try is:
Regulator and lights disconnected.
Yellow - Brown connected to AC terminals on bridge
Fan connected to DC terminals on bridge, voltmeter same place to read fan voltage. Note fan won't be grounded.

base test:
Fan RPM of--1840 --(Directly hooked to a 12V battery at battery V of 12.4 sat overnight, not fresh charge)

Fan RPM is--1680 at 12.30-12.40V AC / 9.7VDC engine at 1500rpm idle regulator hooked up

Fan at 15V DC, unhooked reg/no bulbs tach 2800, fan speed 2250

Fan speed: 1720-1730 @ 10.3VDC engine idle 1500, reg plugged in, one 35V 10,000 mfd installed DC side in parallel
same engine speed, fan RPM increase of 40 RPM w/ cap.


Sorry, no amp readings at this time. Please ask or reask a question if more information is needed.

Mikel

 

[Mikel_NY]

Side bar question ... I have the on/off Fan switch on the Positive AC line going to the rectifier, then from the rectifier DC going to the fan and cap ( in parallel ).

Is this an accepted method of install? (my first w/ caps)

 

[Mikel_NY]

here is the best description of the one Cap that was installed: (about 1.5"dia, 5" tall)

Mallory Type CGS
10000MFD 35VDC
POS+85C
MAX surge 45VDC
CGS103U035R4L3PH
235-8406A

Bottom listing in pic.

 

[Mikel_NY]

while sitting back and thinking about this project I think I come to realize that the stator Hz is directly related to the RPM in that when I calibrate the Tach with a tone generator from my laptop that I use a tone of 500 Hz for a tach reading of 5000 and a tone of 820 for a tach reading of 8200.

In post #6:

To use a capacitor instead of a battery;
An 80W fan on 12V will draw a current of 80W/12V = 6.67A
What is the frequency of the AC? If 100Hz then period T = 10msec.
C = I * T / dV; For a 1.0V droop in the capacitor during the cycle;
C needed will be C= 6.67A * 0.01sec / 1.0V = 0.0667F = 66.7mF, rated at 15V or better.

the engine's AC Hz is RPM divided by 10.

Maybe buy ten x 10,000 uF capacitors. US$2 each.
Wire them in parallel to give 0.1F that will handle the current at the stator output voltage peaks.

Does this still apply now knowing that the motor will be run mostly at 820 Hz and sometimes at 150 Hz?

 

[Baluncore]

Does this still apply now knowing that the motor will be run mostly at 820 Hz and sometimes at 150 Hz?

Maybe. The time / frequency dependent values must be designed for the lowest frequency.

I do not know how your regulator works, or what load it is under now, even before the fan load is added. I am working here on the assumption that the alternator and regulator have sufficient capacity to drive the fan.

If you keep the regulator output chassis referenced, then only a single high current schottky diode is needed to rectify the regulator output into the capacitors. If you do use a single diode then you will need to design for the lowest RPM, giving 150 Hz, period T = 6.67 ms.
For 1V of ripple you would need C = 6.67 * 0.006.67 / 1.0 = 0.044 F = 44,000 uF.
I would then recommend ten x 4,700. uF capacitors in parallel.

Ideally you would take the alternator output before the regulator, then use a switching regulator. But if the existing regulator is doing things like shorting the alternator output to ground on negative half cycles, it would be impossible to use the alternator output while the original regulator was present. That would require a replacement switching regulator, and I see no cheap off the shelf solution for a 55V maximum input. I would build a switching regulator from available components, but I certainly don't recommend that for beginners.

An alternative solution could be to change the permanent magnet alternator for a more sophisticated three phase alternator, with a regulator to control it's field current.

 

[Mikel_NY]

To my knowledge, there is not an option of 3 phase replacement for my application.

I just completed a test of running the setup with:

no caps at idle 9.3V DC
1 cap at idle 9.8V DC
2 caps at idle 10.3V DC

Each cap rated at 35V 10,000MFD.

I'm thinking that this setup is very usable and if no one sees a safety issue, I'm going to try and run this setup at a track and see if it prevents overheating of the engine.

This is not to end this project ... I'm just itching to try the setup.

Is there anything wrong with using 35V caps? could 25V caps gain me a return?
I will look into sorcing 2 more like caps to gain the 44,000uf that you are stating I need.


When I go to the test, the 20,000uf that are installed now, what should I look for? maybe overheating?

 

[Baluncore]

Each cycle requires the current to flow into the capacitor and then out again to the fan, so ripple current is going to be very high. It is therefore better to use more capacitors of smaller value. Up to ten would be sensible. If possible use low ESR, high temperature capacitors designed for power supplies.

Capacitor weight, volume and cost are proportional to the voltage rating.
If the voltage will never rise above 16V then a 16V rating is OK. The same goes for 25V, 35V and 63V.
Capacitor dielectrics are optimised to operate between 50% and 100% of their voltage rating.

 

[Mikel_NY]

There seems to be many types of capacitors avalible. Would you please recommend the best type for my application.

I know a local source that has 16V , 1000 uF by the hand full. They are about 1/2" dua. And 1" tall with two wires at same end . The two wires are different lengths. I will look and see if they are "low ear" and high temp.

Thank you.
Mikel

 

[jim hardy]

Sorry to be gone so long from thread - some things piled up on me . ANd Balun's electronic expertise is better than mine, I'm kinda a fundamentalist able to help out with basics.

That CGS you picked looks like a very good capacitor with 5 amp ripple capability. I was hoping you'd find something like that but inexpensive on surplus market.

Does this still apply now knowing that the motor will be run mostly at 820 Hz and sometimes at 150 Hz?

You'll have more capacitance than you need at high engine RPM. If you don't spend much time at idle you might get away with two - that gives you around 10 amps of ripple capability ?

Is there anything wrong with using 35V caps? could 25V caps gain me a return?

Nothing wrong so long as your voltage doesn't go higher. I'd stay with the higher voltage caps. If you use capacitors at 70% voltage they last a lot longer than at 100%.

When I go to the test, the 20,000uf that are installed now, what should I look for? maybe overheating?

Yes. If they're getting warm it's almost surely that AC current heating them from inside. With only 8 amps of fan load i think you could be okay but not much margin left.
That's a little tricky to calculate because as Balun said current rushes into the capacitor then back out. So it's a long way from sinewave current, instead it flows in in big gulps on the voltage peaks. Those short bursts of high current have a higher heating value than a slow steady current and to try and put a number on that i'd be guessing ... there's something called "Crest Factor" we'd need to measure... meh - just feel of 'em and see if they're getting hot. One experiment is worth a thousand expert opinions.

I'm thinking that this setup is very usable and if no one sees a safety issue, I'm going to try and run this setup at a track and see if it prevents overheating of the engine.

You'll want to make a measurement of the voltage across fans with engine at track speed. That can be brief, just long enough for fans to get up to speed. I don't like to rev up an unloaded engine... What you want to know is your alternator doesn't apply 55 volts to a 35 volt cap and 12 volt fans.
Now - your fans being non-electronic just plain brush type motors will probably take some abuse, mild overvoltage. If you find your alternator making up around 25 or 30 volts perhaps we could rewire the fans in series ? Back in the days of DC railroad locomotives we had gigantic relays that switched the traction motors between parallel and series connection as needed ... that'd be a complication for your fans but not a major one.
I'm thinking that as your engine speed comes up and so does voltage, fan speed will increase giving more cooling at high power where the engine needs it . If you're real lucky you might just get away with no regulator !

Go ahead and give it a try .

Lastly, for now, i thought you'd be running lights too? They'll be less tolerant of overvoltage than the fans... perhaps we could regulate them...

Glad to hear of your success so far.
Golly i was going to butt out and let you guys have at it. Please forgive my "old fire-horse behavior". Old habits die hard.

old jim

 

[jim hardy]

Just an aside -- stumbled across an interesting hobbyist webpage describing a home-made voltage regulator .

http://members.iinet.net.au/~cool386/regulator/regulator.html

 

[Mikel_NY]

Hello all, an update.

I have ordered a few 4700uF,16v caps from one state away and they will be here Saturday at the earliest.

Jim, I will read the post you put up.

I have not " left the building" , I'm just waiting on parts.

What will changing my regulator type gain me? Maybe run a more powerful fan, an electric weight jacket. Lots of things I could convert to DC. ( The fan would be continuous, but a all else would be intermittent)

If there is good gains to be had, I have four or five snowmobiles that will be converted. It would be nice to keep phones charged while using them for GPS guidance.

 

[jim hardy]

What will changing my regulator type gain me? Maybe run a more powerful fan, an electric weight jacket. Lots of things I could convert to DC. ( The fan would be continuous, but a all else would be intermittent)

I try to worry ahead... was trying to figure out what to do if your voltage at track speed is too high for your fan motors.

A shunt regulator after the bridge might work but would need a diode between it and the capacitor&fan , just as in that article about Model T Ford regulator, so capacitors don't backfeed into it..

Just musing really. Let's see what happens when you try your setup.

Buck regulators hold promise too. I'm not familiar with these 'off the shelf' ones. Have wanted for a long time to learn about them for use on motorcycles.

old jim

 

[Mikel_NY]

I try to worry ahead... was trying to figure out what to do if your voltage at track speed is too high for your fan motors.

old jim

I just want to clarify that I my setup now is a stock stator that outputs up to 45V AC. It is rated at 230 Watts
-- the unit was designed to run a handful of light bulbs rated at 12V and totaling about 150 watts or a little more and can run 45 watt warmers w/o dimming the bulbs.
-- The stock voltage regulator is installed and regulates the AC voltage at top RPM all day long without blowing the light bulbs. That would say that at top RPM the AC voltage does not go over, say 15V or the bulbs would blow. ( I have not measured the voltage at top RPM to confirm.)

Now the modifications

-- I have removed about 120 watts of light bulbs and have installed an 80 watt brush 12V DC fan
-- I have installed a bridge rectifier to change some AC V to DC voltage to run one fan.

and the question

-- the motor will idle at 1500, this puts stator at 150Hz. To add capacitors in parallel on the DC side of the bridge rectifier without changing the regulator, is it still the 44,000uF that is required?


I respect all your expertise. I guess I'm missing something here. I would "assume" that wattage is wattage. Is there a difference in the load if I change from AC bulbs to simple DC brush motor?
If there is no difference than I have lightened the load on the regulator and would ask why I would need to change it? I ask only to understand.

 

[jim hardy]

I guess I'm missing something here. I would "assume" that wattage is wattage. Is there a difference in the load if I change from AC bulbs to simple DC brush motor?
If there is no difference than I have lightened the load on the regulator and would ask why I would need to change it? I ask only to understand.

By just inserting a bridge between the regulator and the load you haven't basically changed anything. When you add capacitors it does change something.With light bulbs before the bridge, that is right on alternator wires,
what does the alternator see?
It sees a load that draws current whenever any voltage is present.

With light bulbs after bridge but no capacitors, what does alternator see?
Alternator still sees same thing. Lights now see DC with lots of ripple instead of AC

Now add capacitors. What happens on DC side of the bridge?
The capacitors charge to some voltage on every peak then discharge a little bit waiting for next peak. Balun sized capacitors for 1 volt sag between peaks at lowest frequency.

So - what does the alternator see now?
It sees a load that draws no current until alternator voltage gets higher than whatever is leftover on capacitors from last peak. Then it draws a big gulp of current. That'll happen on every voltage peak.
So - the current demanded from the alternator has changed:
from a replica of the voltage wave (assume it resembles a sine as in Kavic's post #3)
to a series of short but very big gulps one at every peak.

Those big gulps of current will flatten out the voltage peaks. But I don't know how much.
That's why I was interested in voltage readings with capacitors installed and fans running and regulator disconnected. I think you posted them...
I'll go back and look what those numbers were after dinner - Fair Anne just set it out..

thanks

old jim

 

[Mikel_NY]

I received my order, can I use this style?

Lelon
16V 4700uF
RGA 105 degree C
H032(M)

I went to LELON's website and the only match I could find stated the cover is Black and White. I received green case with black print.

On the page that listed color, it stated if ESR is needed then use another style than RGA.EDIT: Jim, Thanks for clarification on what has changed. If these caps are usable, I will install 10-4700uF caps and disconnect the regulator and retest the voltage w/ fan on and report at what RPM the engine reaches at just before 15V DC at the fan.

ALL light bulbs are on AC side. there will be 2-7 watt bulbs on at all times, 2 may be continuous or intermittent. 28 watt total. on the DC side, the 80 watt fan will be on always and another DC 12V motor very intermittent. ( a drill motor to move wing )

 

[jim hardy]

from post # 35:

Fan RPM is--1680 at 12.30-12.40V AC / 9.7VDC engine at 1500rpm idle regulator hooked up

Fan at 15V DC, unhooked reg/no bulbs tach 2800, fan speed 2250

Fan speed: 1720-1730 @ 10.3VDC engine idle 1500, reg plugged in, one 35V 10,000 mfd installed DC side in parallel
same engine speed, fan RPM increase of 40 RPM w/ cap.

So at idle, if i understand:
with regulator hooked up the alternator made 9.7 volts at idle while driving the fan,
unhooking the regulator and adding a capacitor raised that voltage to 10.3 at idle.

The capacitor held the voltage up a little bit between peaks. That's probably why DC value came up.

At 2800 RPM your alternator makes 15 volts while driving the fan with no capacitor... doubtless the peaks are higher.
It'll be interesting to see what a few capacitors do for it above idle.

hang in there - call this R&D . Experimenting is one step at a time.

old jim