Taming the Vectrix Battery Cooling Impellers

Mr. Mik

1 kW
Joined
Sep 3, 2008
Messages
390
The discussion about this topic started a bit "off topic" in this thread:

http://www.endless-sphere.com/forums/viewtopic.php?f=2&t=8722#p133439

SInce then I have bought and tested one of these Potentiometers:
http://www.jaycar.com.au/productResults.asp?FORM=KEYWORD

" 5 Ohm, 15W Wirewound "

productLarge_7307.jpg


It produces smoke....but apparently not the magic kind of smoke, because it all still works.

The problem seems to be that the motors have a controller built in, and this controller stops the motor if the supply voltage drops too much. That is probably also the reason that the impellers often start running with a delay of a few seconds between them. One starts first, draws too much current and drops the voltage too low for the other. The, when it is fast enough and the current draw drops to about 2 to 2.5A, the other impeller motors starts to chime in.

I connected the 5 Ohm Pot in line with the negative cable returning from the impeller motors.

Starting with the pot turned to 0 Ohm, it runs fine. When the pot is gradually turned up, then the fans slow down. At some point around 0.5 to 1 Ohm the fans turn off, and then one starts running again, but stops when the other starts about 2s later; they alternate running for a few sec at a certain Pot setting.

Oh, and then there is that smoke coming out of the pot, the smelly kind. It did not start at first when I turned the resistance up, but once it had started it came back quickly each time the pot was turned up from zero and the impeller motors slowed down.

So far, it seems that only a small reduction in impeller speed and noise could be achieved with resistors, but this would cause a lot of heat and still use a similar amount of power. I guess that about 0.5 Ohm might work, maybe 4 x 2 Ohm (10W) in parallel.
 
I made up a resistor from 11p x 6.8Ω , 10W rated resistors. 10 of them in parallel were not enough to reliably start both impellers (because the already running one shuts down again when the second one starts to draw amps and drop voltage). Adding the 11th resistor seemed to just do the trick!


(How can I embed videos properly on ES?)

The results are not as good as I had hoped: The noise levels are very clearly reduced in the garage, but the lower hum of the reduced speed seems to transmit more easily into the house when the door is left open. With the glass door between garage and house closed, the noise-reduced version is also quieter inside the house.

The maths: ( 1/6.8Ω ) x 11 = 1.61/Ω
1/(1.61/Ω) = 0.61Ω

So this big thing is a 0.61Ω resistor.

More resistance and the fans will not start to run concurrently; less resistance and the speed reduction will be less.

The wattage rating for this resistor might be 11 x 10W = 110W. Not too sure of this, though, because when mounted against each other, the heat dissipation capacity reduces markedly. And I guess it's the heat that limits the wattage rating.

The 0.61Ω (110W) resistor heats up to 90°C (in 23°C ambient temp) whilst the impellers are running.
S4029096-1.jpg

Is this the best result that can be achieved using a resistor to tame the impellers?

The resistor-array (or what do you call this device more appropriately ???) could be mounted in the airstream which is created by the running impellers, that would keep it all cool.
 
Mr. Mik said:
The resistor-array (or what do you call this device more appropriately ???) could be mounted in the airstream which is created by the running impellers, that would keep it all cool.

But then it would be a heater. I suppose on a cold day that might be OK, or if you place it downstream of whatever the fans are supposed to be cooling it won't matter. That's a really big resistor...

You might try putting the two fans in series to see what happens. No resistor needed.

Did you check around the connector on the fan to see if there are any other wires or pins coming out? Some of those fans had a speed control built in.

With a higher resistance, you might be able to put a large capacitor across the resistor to get the fan started and still get a slow running speed.
 
Would there be any chance of using one resistor for each fan? Then at least the start up surge from one fan wouldn't effect the other fan. Sounds like what's really needed is a small DC-DC switching power supply that's powered by 12v and has an adjustable output in the range of 6-12v and can supply 3+ amps.

Lawson
 
fechter said:
Mr. Mik said:
The resistor-array (or what do you call this device more appropriately ???) could be mounted in the airstream which is created by the running impellers, that would keep it all cool.

But then it would be a heater. I suppose on a cold day that might be OK, or if you place it downstream of whatever the fans are supposed to be cooling it won't matter. That's a really big resistor...
They suck air through narrow slots between the 102 NiMH cells. Mounting the resistor downstream would be possible.
You might try putting the two fans in series to see what happens. No resistor needed.

I'll try that out one day when I have the battery housing apart for some other reason. It's a lot of work to get to the impellers. But I doubt if it will work, because the specs you found say that the minimum required voltage is 7V. And in series they would just get 6V each, right?
Did you check around the connector on the fan to see if there are any other wires or pins coming out? Some of those fans had a speed control built in.
I am fairly certain there were only two cables disappearing into the motor, nothing else. But I will check this too when the housing is off.

With a higher resistance, you might be able to put a large capacitor across the resistor to get the fan started and still get a slow running speed.

Nice idea! And thanks for your help!

Here is an updated schematic of how it is all wired so far:
ABCool26-1.jpg

 
lawsonuw said:
Would there be any chance of using one resistor for each fan? Then at least the start up surge from one fan wouldn't effect the other fan. Sounds like what's really needed is a small DC-DC switching power supply that's powered by 12v and has an adjustable output in the range of 6-12v and can supply 3+ amps.

Lawson

Maybe if I place one resistor each between the relay and the impeller motors?

Or better, between the Auxiliary power supply and the relay, so that the impellers always run at full speed when they are only powered by the stock 12V supply (That will usually be the case when riding on hot days and the full power is then needed).

I get a little confused about the "series" and "parallel" issues with this wiring diagram and the common negative cable connection.

Thanks!
 
Here are some pictures of the inside of these impellers: (click to enlarge)



More details at http://visforvoltage.org/forum/7593-can-vectrix-cooling-impellers-safely-run-lower-or-higher-speeds
 
At long last, it's done!

I can now run the impellers at low speed, but the stock system will override this with full power 12V if needed at any time.

The low speed power is provided by a 6s NiMH NHW10 battery stick (or several in parallel).

ABCool50-1.jpg

All the diodes in the diagram which have the Zener diode symbol are actually Schottki diodes. I could not find a proper Schottki symbol so far.
And the switch is really a SP/TT(Single pole/triple throw): It has an "OFF" position in the middle.
Click thumbnail for full size:

.

It works well in the garage, but there might be problems during riding. It remains to be seen. Throttling the relay with the trimpot might make it susceptible to vibrations, turning the impellers off when they are meant to be running at low speed, powered by the battery during riding. I hope the relay will function reliably for my 30min commute (when I need extra cooling in hot weather conditions).

But this design allows normal, or slow speed running of the impellers, either by a NHW10 battery alone, or supported by the 12V auxiliary power supply, at least whilst standing still.
It also allows for recharging (of the auxiliary battery) only, without impellers running.

Once the battery voltage drops to near the cutoff voltage at which the impellers would turn themselves off, the relay turns off instead. The trimpot is a multi-turn trimpot, which allows fine-tuning of the voltage at which the relays turns itself (and the power delivery to the impellers) off.

The battery gets charged (by auxiliary and / or stock 12V supply) at a rate which can easily be supplied by either of the power supplies (even if the battery is empty and the diodes are hot from recent use); but the charging will crawl to a halt (because of the diodes combined voltage drop) before the batteries get overcharged. The NiMH batteries end up getting charged at C/100 or less, with a temperature rise of 1-2degC. maximum, once they are fairly full.

.

When the impellers are being powered by the (initially full) battery and the auxiliary power supply, then they run for about 2 - 2.5 hrs before the relay turns off (when the battery voltage drops to about 7.45V).
.

I had to mix a few different diodes to get the balance right; the diodes I used initially had a higher voltage drop, causing thermal runaway and reduced forward voltage drop, etc. etc.
The current from the auxiliary power supply increased as the diodes were heating up, until the impellers were being powered and the batteries being charged by the auxiliary power supply. That is still an option, if more controlled heating / cooling of the diodes was implemented in the future.

The resistors are connected via screw-terminals; that will allow fine tuning. I changed them a few times to find the right values.

Heat-sinking the diodes is a bit of a tricky issue, but so far it seems like I have solved that problem. I'll have to check regularly if the double-sided-heat-sinking-sticky-tape continues to work!

I have propped it up with supporting hot-melt-glue; I am confident it will be OK.

See pictures for details. If you needs more details, ask away. I don't want to waste too much time describing a one-off device that no-one but me likes!
 
It would have been smart to actually attach the photos....Sorry!

Here they are:
 
Back
Top