Running Two Brushed 12V DC Motors

[Bryan Deily]

I am trying to figure out the best way to hook up two 12v DC peristaltic pumps to run as close to the same speed as possible. The manufacturing tests had the pumps running within 10ml per minute which is acceptable; when I hook the pumps up in series they are not even close. When hooked up in parallel they are in sync however they run intermittently 1 second on 1 second off. I am assuming I'd need more volts or amps due to internal resistance of the motors? The individual motors are 12v 3a max 1.5a continuous. The power supply I was testing was the same 36w. Do I need to get a 72w(12v 6a) to run both in parallel?

Also, what would be the best way to get the motors running dead in synchronized? I'm looking to use these pumps for a water exchange; pump old water out and new water in. Ideally the volumes are identical to keep parameters from creeping over time.

Thanks for any advice.

 

[billy_joule]

Do I need to get a 72w(12v 6a) to run both in parallel?

Yes.

Also, what would be the best way to get the motors running dead in synchronized? I'm looking to use these pumps for a water exchange; pump old water out and new water in. Ideally the volumes are identical to keep parameters from creeping over time.

Without feedback (measuring tank level or flow rates) you won't get them balanced.
A cheap float switch (>$10) would do.
Or a different design, why do you need to pump out? Why not an overflow?
Alternatively, you could arrange the pumps so the outflow pump flowrate will always be greater than the inflow while maintaining a set tank level.

 

[Bryan Deily]

I was thinking of wiring a float switch into power down both pumps in case there is not enough new water in the resivour. This would prevent pumping water out without replacing it.

If the numbers work out the pumps will be off by about 1/4 cup per day. My tank would never overflow due to evaporation however salinity would climb slowly. I may try an inline flow restrictor after each pump at 750 ml/min to see if that gets it any closer.

I was thinking of using a system of float switches and relays I just felt there were too many parts involved and was worried about failure.

 

[Bryan Deily]

Yes.

Thank you billy.

 

[Bryan Deily]

It seems most float switches couldn't handle 72w... so the plot thickens.

 

[Baluncore]

Peristaltic pumps are not calibrated if the grade or age of the flexible tube used is different. To match two pumps you should first fit them with tube from the same batch.

 

[Nidum]

(1) Fit feed back devices and use some very simple electronics to sync the motors .

Disc with a few slots or holes and a photo switch would be ok for the feedback devices .

(2) Use electronic variable speed drives and match the motor speeds manually .

 

[jim hardy]

Can you mechanically couple the shafts together ?

 

[Nidum]

Or even use one motor ?

 

[Baluncore]

If it is the mass transferred that must be the same then the pump revolution count is not important as errors will accumulate due to differences in tube characteristics and the temperature / density in the fluids. A counter-flow heat exchanger might improve things but the only reliable way is to regulate the level in the common tank with a float switch.

 

[jim hardy]

If the numbers work out the pumps will be off by about 1/4 cup per day.

If it is the mass transferred that must be the same then the pump revolution count is not important as errors will accumulate due to differences in tube characteristics and the temperature / density in the fluids. A counter-flow heat exchanger might improve things but the only reliable way is to regulate the level in the common tank with a float switch.

i think we all agree,, and our thought processes would benefit from an error budget. How much flow are we considering and how much volume does it take to reach the 'worry point' ?

Also, what would be the best way to get the motors running dead in synchronized? I'm looking to use these pumps for a water exchange; pump old water out and new water in. Ideally the volumes are identical to keep parameters from creeping over time.

I like clever mechanical gizmos. One can then build a circuit that does the same thing.
https://en.wikipedia.org/wiki/Differential_(mechanical_device)

A differential is a gear train with three shafts that has the property that the angular velocity of one shaft is the average of the angular velocities of the others, or a fixed multiple of that average.

A spur gear differential constructed by engaging the planet gears of two co-axial epicyclic gear trains. The casing is the carrier for this planetary gear train.

A differential gear connecting the peristaltic pumps would lock them in step when the averaging shaft is locked. Rotating the averaging shaft would raise speed of one and lower the other. Averaging shaft's speed would be set to a function of level error, the tank will provide integration.

RC cars have small differential gears, Ebay shows them in the $10 to $25 range..

Same thing could be done electronically by precisely measuring rotation of each shaft, perhaps with old mouse wheel optocouplers, controlling motor voltages individually. I like the gears, though, it'd be more fun to watch !

.......... That said, moving on to your other question...

when I hook the pumps up in series they are not even close. When hooked up in parallel they are in sync however they run intermittently 1 second on 1 second off. I am assuming I'd need more volts or amps due to internal resistance of the motors? The individual motors are 12v 3a max 1.5a continuous. The power supply I was testing was the same 36w. Do I need to get a 72w(12v 6a) to run both in parallel?

Yes you do need a larger supply.
Try your 6 amp car battery charger .

 

[Bryan Deily]

(1) Fit feed back devices and use some very simple electronics to sync the motors .

Disc with a few slots or holes and a photo switch would be ok for the feedback devices .

(2) Use electronic variable speed drives and match the motor speeds manually .

If I were to use variable speed drives and manually match speeds do you think it would need to be recalibrated down the road? Thanks for your input.

 

[Bryan Deily]

i think we all agree,, and our thought processes would benefit from an error budget. How much flow are we considering and how much volume does it take to reach the 'worry point' ?
I like clever mechanical gizmos. One can then build a circuit that does the same thing.
https://en.wikipedia.org/wiki/Differential_(mechanical_device)
A differential gear connecting the peristaltic pumps would lock them in step when the averaging shaft is locked. Rotating the averaging shaft would raise speed of one and lower the other. Averaging shaft's speed would be set to a function of level error, the tank will provide integration.

RC cars have small differential gears, Ebay shows them in the $10 to $25 range..

Same thing could be done electronically by precisely measuring rotation of each shaft, perhaps with old mouse wheel optocouplers, controlling motor voltages individually. I like the gears, though, it'd be more fun to watch !

.......... That said, moving on to your other question...Yes you do need a larger supply.
Try your 6 amp car battery charger .

I am trying to use the pumps for automated water changes in a saltwater fish tank. My tank is small relative to most which means the pumps would run a short time to exchange the desired amount of water. I'd like to see this scale to larger tanks as well which would require more precision. I will be running the pumps for about 6 minutes a day to exchange 7 gallons a week. The pumps are currently off by about 3.7% in volume. I'd prefer about a 1% difference or less.

I like the differential idea; I guess ideally I would run both pump heads off of one motor. Not sure what type of gearing this would require. The differential has three shafts? Two in with one out at the average speed between the two?

 

[Bryan Deily]

I spoke with the manufacturer of the pumps and they have stepper motors available. If I were to switch to stepper motors and have them run the same amount of steps would that be easier? I'd just have to figure out how to control them as I intended on running the brushed motors on a timer 6 minutes per day.

 

[Baluncore]

Stepper motors will not solve the problem. It would be easier to use one motor with the pumps on a common shaft. Water temperature and density will continue to vary as will the characteristics of the peristaltic tube and roller pressure.

If the waste pump drew water from the top of the aquarium and was run slightly faster, or with a bigger diameter peristaltic tube, it would always outperform the injection pump and could not lower the water in the tank below the waste inlet level. The waste pump would suck some air, but no float switch would be needed.

 

[Nidum]

Yes - stepper motors would be a very good solution . No problem with finding controllers for them .

With software control you can set up any timing schedule and still have the facility to make small adjustments as nescessary to balance the system .

It would be quite easy to add some feedback as well and get a self balancing system if you wanted to .

 

[Baluncore]

Can you put three smaller tubes in one peristaltic pump? Two would pump waste water while one injected clean.
With two tubes having slightly different diameters but the same wall thickness, the same could be done.

 

[Nidum]

If I were to use variable speed drives and manually match speeds do you think it would need to be recalibrated down the road? Thanks for your input.

This type of set up generally works quite well for long periods but certainly occassional adjustments would be needed .

Stepper motors would be my preference though .

 

[Bryan Deily]

Stepper motors will not solve the problem. It would be easier to use one motor with the pumps on a common shaft. Water temperature and density will continue to vary as will the characteristics of the peristaltic tube and roller pressure.

If the waste pump drew water from the top of the aquarium and was run slightly faster, or with a bigger diameter peristaltic tube, it would always outperform the injection pump and could not lower the water in the tank below the waste inlet level. The waste pump would suck some air, but no float switch would be needed.

The tubing is identical; water temperature and density should be close as the salinity and temperature are matched. I feel like if these were my concerns I'd be working in the 1% error range.

It is a great thought to have the waste pump at the surface of the return portion of the tank so it can not pump out more than is pumped in. I have an optical sensor that maintains evaporation by topping off with fresh water so if it were to pump more out than is coming in it would trigger fresh water to run and keep the level the same. I could run the pumps on opposite schedules to avoid that problem though.

I'd like to entertain running both pump heads off of one motor shaft. First, because I have the motors on hand. Second, because I think it is the more simple approach. I really don't know where I would start here however. Should I look at modifying the pump heads and existing shaft or would I use some type of gear box?

Thanks for all the great help.

 

[Baluncore]

I'd like to entertain running both pump heads off of one motor shaft. First, because I have the motors on hand. Second, because I think it is the more simple approach. I really don't know where I would start here however. Should I look at modifying the pump heads and existing shaft or would I use some type of gear box?

We cannot answer that question until we do not know what make and model peristaltic pump you are using.
Can you post a link to the technical data sheet.

 

[Bryan Deily]

 

[Bryan Deily]

I got these two peristaltic pumps from China. Manufacturered by Honlite Industries. They manufacture pumps for a high end aquarium controller company Neptune Systems. These are their brushed motors rated at 800 ml/min. They run on 12v 3a. Here are the specs to each individual pump.

 

[Nidum]

Do you actually need two pumps ? One pump and a spill weir will do the same job of maintaining the level if you can arrange for gravity drainage .

 

[Bryan Deily]

I do have flexibility with plumbing; I am not familiar with a spill weir. I assume I'd just pump water in and it would overflow out. I could make this work by adding a drain below my tank. The issue with this is when there is a power outage or the return pump is shut off the water level in the sump tank rises as the overflow and plumbing drain back there. All that water would be lost.

I'd like something that will work for more than just my application so I can share with the community. Two pumps would fit the bill better.

 

[jim hardy]

automated water changes in a saltwater fish tank. My tank is small relative to most which means the pumps would run a short time to exchange the desired amount of water. I'd like to see this scale to larger tanks as well which would require more precision. I will be running the pumps for about 6 minutes a day to exchange 7 gallons a week. The pumps are currently off by about 3.7% in volume. I'd prefer about a 1% difference or less.

Aha ! At last we are addressing the right question... "How do i keep desired level in my tank?"

Billy Joule had the answer... a weir at the desired levelPull the top of your bathroom toilet and look at the round tube in the middle of the tank. It keeps tank from overflowing by giving excess water a way out. Its bottom drains into the business end of the toilet..

A similar tube with its bottom closed , placed in your aquarium with your drain pump's suction line drawing from inside the tube should do the trick. Water will flow into your weir tube only when level is above is open end. Only then can it get pumped out because that's where the suction pump draws from. A piece of PVC pipe epoxy-glued to a brick would make a weir .

A simpler "virtual weir" would be to just fasten outflow pump's suction tube to aquarium wall at the desired level. That way it can only remove liquid that's above desired level , I don't know how to glue tygon tube to glass, though.

To protect against a siphon emptying the tank in case of tube break i'd locate the fill line's end slightly above desired level...

I'd also place a diode in series with the fill pump to make it run slightly slower than the drain pump, so that in case the system fails and turns them both on continuously you can't overfill..

Think simple, and draw from your everyday experience.
The humble toilet bowl tank can teach us a lot.
It's an analog flow integrator, , level = ∫(inflow-outflow) .
Its fill valve is a proportional controller with limits, inflow = k X (level error)

old jim