WVO Centrifuge Design - thoughs ?

[veggie]

Hi AdeV

When I tested the gear pump on the PWM controller it was not pumping liquid.
Just free wheeling and probably pulling only 1 amp.
Not sure why it would cook a controller. (Yes, I double checked the polarities    )

Veggie

[BruceM]

It seems unlikely to be a brushless 12 motor with integrated controller; they are more expensive at present.

I suspect that instead, you have an EMC problem with the cheap PWM controller and a nasty brushed motor, emi-wise. EMI from the motor is glitching the MOFET gate or gate controlling circuitry, thus you see full speed until failure.  Capacitors at the motor, and a common mode choke might solve it fairly cheaply. A better designed PWM controller might be able to handle it. If your 12V supply is also full of EMI, that should also be addressed via passive filtering since the controller may not handle that well either.

MOSFETs have amazingly low "on" resistance so can handle a huge load in a single small package.  They do have some serious issues with EMI; their gates are voltage controlled and are much more susceptable to glitching than older bipolar transistors which are current controlled. 

The solution to avoid glitching is suppression at the source (motor), but also to use a very low resistance gate drive (1-10 ohm). Many cheap electronics will not in order to save cost.

Solving the problem at the source is a cost effective solution as it prevents the same EMI from damaging or cause intermittent faults of other connected electronics.  For a "junk box" filter for this motor, I'd start with 0.0-1 uF ceramic, and 10-100 uF electrolytic, and a 3+ mH common mode choke rated for your max current or better. 
 

[veggie]

BruceM

Thanks for that detailed explanation of the potential problem at hand.
The power source is a 12 volt deep cycle battery. Clean power.
I suspect poor Chinese electronics (again). In any case, it's dead. I won't be buy another anytime soon.
What do you think about using a Rheostat with an adequate amperage rating.?
Just vary the voltage instead of PWM.

Veggie

[veggie]

Hi Glort,

Ha...Ha. Yes Dr. Who was inquiring. It's a Tardis machine. It actually holds 3000 liters of WVO but it looks very tiny. 

I have one of those pumps that you mentioned (330 gallons per hour). I use it on my BioD processor for circulation, but for this application the feed to the centrifuge is only 10 gallons per hour. Just a trickle.
That's why I was hoping to slow down the gear pump a bit.

I will dig around more to see if I can come up with a better pump running off mains power.
The centrifuge and pump will have to run for 3 hours per batch so I need a small pump that can handle that duty time.

Perhaps for the time being I can run this gear pump at normal speed with a bypass valve in the discharge line which can slip some of the flow back to the tank. I could adjust the bypass valve position until I get the desired flow rate into the Fuge.
The WVO is open circulation from the tank to the Fuge (no pressure) so maybe the pump will last a bit longer without pressure/amperage stresses.

cheers

[BruceM]

If adding a filter is too much for you, then yes,  a power resistor to drop the voltage could be used.  Let ohm's law be your guide for value and watt rating. 

[veggie]

All good ideas Glort. Thanks.
Yes, there is an open steam port of the lid of the unit. I will report back on how effective that is.
The planned temperature for oil entering the Fuge is 80C.

BruceM, I will look into sizing a heavy duty resistor for dropping the voltage (speed) to something usable.
Maybe a light bulb or two as Glort suggested as an initial test.

Updates to follow...
Veggie

[BruceM]

I also use 12V light bulbs for power resistors, I've got 15, 25 and 50 watt bulbs in standard lamp bases. Walmart carries the 50 watt ones.  I also have a 30 year collection of power resistors left over from various projects, and where I guessed wrong on value.  Adjustable ceramic wire wound unit's in fairly high wattages are very handy...Digikey carries those.

I like induction motors, too, for something like this. A brushed DC motor on a load resistor will vary in speed with load.

[AdeV]

BruceM, I will look into sizing a heavy duty resistor for dropping the voltage (speed) to something usable.
Maybe a light bulb or two as Glort suggested as an initial test.

Why not see if you can get a blower fan speed controller from a scrap car? Certainly in older cars, these were simply resistor packs wired in series with the blower motor... If you can find a 3- or 4-speed one, then you've got a realtime speed control system AND a natty dashboard

[veggie]

** UPDATE **

I decided to go with mechanical flow control since I had most of the bits and pieces in my box of fittings.
The gear pump now has a discharge bypass which can relieve flow back to the suction side of the pump.
The RED valve lever controls flow. The gear pump can run at it's rated speed for the rest of it's life. (which according to Glort may not be for very long  )
The unit is almost ready to fire up.
Next is to mount the centrifuge motor variable frequency drive in a box to hide the 240 volt connections.
I may also add a switch on the front of the box to allow for quick power kill.

Here's the  latest...

[EdDee]

Hey Veggie,

A little trick I often pull with pumps is to use a regular water type stopcock type tap/valve in place of your ball valve to prevent overpressuring and stalling the pump....

You mod the stopcock by removing the rubber washer and then placing a fairly strong spring between the spindle and the seal holder... As you screw the tap closed, the spindle pushes the seal holder onto the seat and increases system pressure... this allows you to adjust flow later on in the line without stalling the pump... A full seal is not required in this application, hence the removal of the rubber washer

Just a thought...

Cheers
Ed

[BruceM]

Nice tip, Ed.  Veggie, does your gear pump already have an overpressure bypass relief internally built in?  If so, no worries on the ball valve.  Ball valves are great- but not so great for flow regulation near the low end, so Ed's solution solves both blowout and regulation adjust- assuming you have the right force spring.

[ajaffa1]

Hi Veggie, been watching your build with great interest. Nice work and great paintwork.
I run a couple of VFD  controllers in my joinery shop. Please be aware that they have internal cooling fans and create a lot of heat under heavy loads or long runs. You will need to cut some ventilation holes in your enclosure or things will go bang. Please don`t ask me how I know or how much it cost.
Here in Australia we have problems with mud wasps and other creepy crawlies that love to get into electrical enclosures so I now cover all ventilation slots/holes with fly screen mesh, have to blow the dust out of them regularly.

Looking forward to seeing it running,
Bob

[AdeV]

If it's anything like my cheap* Chinese gear pump - and it looks identical - it won't need an over-pressure or bypass system: the gears mesh so badly it'll just start spinning the oil if the back pressure's too much for it. Beware it's power consumption, though, which will soar above it's rated value and will cause overheating and damage in next to no time.

It's also hopeless if there's any pressure behind it: The oil/fluid will simply push past the gears and through the outlet pipe. A valve of some kind immediately before the pump would be a wise move, unless it's specifically lifting the working fluid against gravity, in which case you'll be fine (but it'll lose prime every single time you stop it; not a problem if you're not lifting more than about 6ft, as it will self-prime readily enough).

You may be getting the impression I'm a bit unimpressed with my gear pump. 

* Not that cheap either... about 3x the price of a diaphragm pump of slightly lower rating which - in practice - performed better with a lower electrical load.