Peristaltic pumps MOSFETs and diodes for auto dosing system

I just ordered my Atlas Scientific EC and PH stamps and would like to add some peri pumps for an automatic hydroponic/aquatic dosing system. There are some cheap pumps on Ebay - http://www.ebay.com/itm/peristaltic-pump-BABYFISH-w-DC-motor-tube-dc-6v-For-Aquarium-Lab-Analytical-/400569497462?pt=LH_DefaultDomain_0&hash=item5d43cd7776

I'm not sure whether to get the 6v, 12v or 24v and the differences between them (some 12v have a lower flowrate than 6v :S). And then which combination of MOSFETs/diodes to use. I would like 3 pumps one for each PH up, down, and nutrients.

The system contains 135-180 Litres with a 15-20L resevoir. The PH pumps probably need a minimal flow rate.. as even a drop of adjuster makes a huge difference to a litre of solution. The nutrient pump will need to be big enough for a full dosing (180L), 5ml of nutrients per litre of water, but still able to make smaller frequent (hourley-daily) doses.

Any advice please?

Those ebay pumps may not be ideal for metering the addition of chemicals.

zoomkat: Those ebay pumps may not be ideal for metering the addition of chemicals.

Ok - why? They are for aquarium dosing which still uses PH up and down so whats the issue?

syphex:

zoomkat: Those ebay pumps may not be ideal for metering the addition of chemicals.

Ok - why? They are for aquarium dosing which still uses PH up and down so whats the issue?

They're driven by brushed DC motors with no encoder or feedback. You can set the voltage (or use PWM) to drive them at an approximate speed for a specific length of time for a pretty good guess at how much you've "dosed" but again you'll never know for sure how many revolutions the pump has made.

If you need to-the-drop precision then you should look for a pump driven by a stepper or one with some sort of feedback.

Chagrin:

syphex:

zoomkat: Those ebay pumps may not be ideal for metering the addition of chemicals.

Ok - why? They are for aquarium dosing which still uses PH up and down so whats the issue?

They're driven by brushed DC motors with no encoder or feedback. You can set the voltage (or use PWM) to drive them at an approximate speed for a specific length of time for a pretty good guess at how much you've "dosed" but again you'll never know for sure how many revolutions the pump has made.

If you need to-the-drop precision then you should look for a pump driven by a stepper or one with some sort of feedback.

I don't need drop-specific precision, that was a drop per LITRE, the total system is 180 litres. It will prettymuch constantly adjust itself so small differences in accuracy isn't really an issue - besides I can record the change in EC/PH per second the pump is on and use this data to estimate the amount of dosing required to reach a specific value. But it seems easier to just dose an arbitary (not too large to reduce shock) amount until the correct level is reached.

One drop I think is about 0.05mL. So a drop per litre for 180L is 9ml. The pump I posted is 100ml/minute so thats about 1.66ml per second. So if I run the pump for 3 seconds at a time it will dose about 5ml - that seems adequate. But how will the same pump fair at dosing to deliver 900 ml? That will take 9 minutes of continuous operation, doesn't seem too drastic really. But maybe I would be better off with a slightly larger flow rate pump? Thats what I am asking - also, what is the difference between 6v, 12v, and 24v pumps with the same flow rate? Maybe they are better at dosing in small amounts.

Quote from Ebay:

We then contacted two large aquatic companies and managed to get the information confirmed, that the life expectancy of a 12 volt DC peristaltic pump is approx 6 to 9 months in continual use but anywhere from 3 to 6 months when used intermittently. The reason the intermittent is shorter, is due to the arcing as the unit switches on which can cause pitting to the brushes wearing them away quicker than just running smoothly and constantly, just like the points in older cars.

Can this be solved with a flyback diode? It seems like marketing propeganda to me.

syphex: Can this be solved with a flyback diode? It seems like marketing propeganda to me.

I don't know the answers to your other questions, but a "soft start" in your software might help here. You could add a low-pass filter capacitor and ramp up PWM from 0% to 100% over the course of one second or so. You can also ramp down when shutting down.

A flyback diode protects relay contacts or transistors from inductive kickback when the motor is suddenly switched off, it doesn't do much at switch-on nor for the motor itself.

I have found that those cheap ebay peri pumps work well... for awhile. The silicone tubing that runs through the pump head had a tendency to wear out after 2-3 months. The wear on the brushes is exacerbated by inductive arcing caused by frequent starts and stops (the stops really). The flyback diode shunts the energy away from sensitive digital logic circuitry and back into the motor windings. It protects the circuitry not the motor. For a prototype, a cheap brushed motor works ok but wears out quickly. For a long term solution you should look into a stepper or higher quality motor. You get what you pay for basically. The voltage depends on what is most convenient or whatever power source you are most likely to use. In my opinion, 12volts switchers are pretty cheap and common. Just depends on your situation. 6 volts might be more suitable for a solar setup while 24 volts for industrial application.

Quote from Ebay:

We then contacted two large aquatic companies and managed to get the information confirmed, that the life expectancy of a 12 volt DC peristaltic pump is approx 6 to 9 months in continual use but anywhere from 3 to 6 months when used intermittently. The reason the intermittent is shorter, is due to the arcing as the unit switches on which can cause pitting to the brushes wearing them away quicker than just running smoothly and constantly, just like the points in older cars.

Can this be solved with a flyback diode? It seems like marketing propeganda to me.

When a DC motor turns, it generates a voltage that is the same as the voltage turning the motor. This limits current and speed of the motor. But when it isn't turning, it has only the resistance of the brushes and wires to limit current, so there is a starting surge of current.

You should be able to get the brushes to last longer with some kind of soft start.

Motors are complex things. Just coiling the wire makes them inductive, but the spinny part makes them generators. However, the voltage created by spinning will always be the same polarity as the driving voltage, and never higher than the driving voltage. Current reverses direction, but only as much as could be determined by Ohm's law from the generated voltage, and can even be much higher if you use a dynamic braking system (ie, short it with a relay).

For the inductive part, the -current- will always flow in the same direction as that which created it and is equal (at first) to the current originally flowing. Voltage reverses direction, and can rise to -much- higher than the original voltage.

A DC brushed motor is a complex conjugate of both of these (in addition to parasitic capacitance, oi!) and which one dominates depends not just on the construction of the motor, but on how fast it is turning, how much load is on it, is the load increasing or decreasing, was it accelerating or decelerating, are dynamic braking systems being used, is it being driven with PWM or a DC voltage, etc.

Worst case inductance is starting the motor with PWM with a long first pulse. Inductive effects will dominate, in which case a flyback diode is necessary.

The generator effect would predominate when there is a light load on the motor, it has spun up to full speed, and you are using dynamic braking without any resistance. You could get a very large current spike, depending on the size and speed of the motor, and the length will depend on the rotating mass, ie, the rotational momentum of the system.

The silicone tubing that runs through the pump head had a tendency to wear out after 2-3 months.

In the past I've used "tygon" pumps that apparently operate the same way by squishing the tubing flat and then expecting the tubing to rebound back to the approximate original size. Over time the tubing will lose some of its rebound capability and the pumping will decrease pre revolution of the pump. If the pump is controlled by some type of chemical concentration feedback, then the decrease in pumping efficiency may not be a big issue.

You should be able to find equivalent size Tygon tubing to replace the silicone in these pumps. Tygon is routinely used in small engines for fuel, oil and vacuum. A small engine parts counter should be able to sell you some.

But will tygon tubing spring back? If this is the stiff stuff I think you might be talking about, I can't see it working for a peristaltic pump.

The ebay pumps are good, but as with anything mechanical from ebay, you will need to replace it often. I would go with a 12v pump just because it's easier to find a power supply for it. For mosfets, get things that can handle 10a or so; that way you won't need to heatsink them.

As for the actual control, set a maximum pH and a minimum pH so it's not on 24/7. And be aware that the pH fluctuates over the course of the day, so before installing a dosing system have your Arduino log the pH at different times of day.

@syphex i have that 6v pumps to dose but i have problem with the tubing.

Tubing has a big resistance to motor. I had silicone oil but couldn't solve the problem completely.

You may think a stronger motor.

Will a motor driver/stepper motor help to reduce/solve the effects of startup on the brushes? Also I think some motors use a “shaded pole motor” to instead of brushes so they don’t wear out, is this the same as a “stepper” motor?

Also see some “dishwasher” detergent pumps - it seems like these might be made more for intermittent dosing.

These pump look like slightly better quality. But if these malfunction within 12 months I might as well buy twice as many cheap pumps:

http://www.ebay.com/itm/SET-3-DOSER-PUMPS-PERISTALTIC-FOR-BALLING-ABC-KALK-CALCIUM-OR-OTHER-NUTRIENTS-/181214590340?pt=UK_Pet_Supplies_Fish&hash=item2a313b3584

tylernt:

syphex:
Can this be solved with a flyback diode? It seems like marketing propeganda to me.

I don’t know the answers to your other questions, but a “soft start” in your software might help here. You could add a low-pass filter capacitor and ramp up PWM from 0% to 100% over the course of one second or so. You can also ramp down when shutting down.

A flyback diode protects relay contacts or transistors from inductive kickback when the motor is suddenly switched off, it doesn’t do much at switch-on nor for the motor itself.

Thoughts on this method? Please advise.

Please advise

Google "flyback diode" and do some reading. Its not a short answer, but the basic implementation is a diode connected across the motor terminals.

I was referring to "low pass filter capacitor" and ramp up the PWM until max and back down again

What size capacitor do I use? Attached in series or parallel to the pumps (12v)?

Do I need to find a DC adapter that outputs the exact current for the pumps or the motor will burn out?

Do I need a motor driver?

What is the optimum speed to drive the motor to increase the longevity?

I have the 30V / 60A N-power mosfets for switching are these too large?

A low pass filter suggests low frequency are allowed to pass thru while high frequencies are blocked. You want clean DC, no frequency, zero hertz which calls for a flyback diode. Decoupling capacitors close to your IC or microcontroller are also very important. Common practice is a .1uF with low ESC (ceramic). Using two capacitors of different orders of magnitude (.1 uF and .001uF) helps catch high and low frequency transients giving a little redundancy. Finally a fat electrolytic (100uF-1000uF) on your power rails helps when switching heavy loads. Without an oscope measure exact frequencies, you kinda have to follow general guidelines for decoupling.

The motor will only draw as much current as it needs as long as you dont greatly exceed the voltage rating. You will want to find a power source that can supply several times more current than your motor. Start up and stall current will be much higher then normal running speed (10-100x more).

Motor driver is probably unnecessary unless you want to reverse direction. Your mosfets should work fine as long as they are logic level.

I have these N-power mosfets to PWM my 12v pumps. http://www.adafruit.com/products/355#Learn

But I'm not sure how they work yet. The following tutorial has helped me understand them a bit more an even has a "Simple Power MOSFET Motor Controller" part if you scroll down. One question I have is the pins from left to right seem to be gate drain source, but in wiring diagrams it seems to be drain gate source.

http://www.electronics-tutorials.ws/transistor/tran_7.html

In this picture seems to depict drain gate source.. is this just a schematic convention? Or is that really the order of them?

How exactly does the PWM speed control work? The MOSFET seems to switch either on or off... there doesn't seem to be any variable resistance in the "saturation area". Is it the PWM varying the gate voltage which controls the voltage to the solenoid? Or is the MOSFET itself switching on and off in accordance with PWM to Vgs = 2.35V max?

Where is the - terminal for the DC motor?

syphex:
How exactly does the PWM speed control work? The MOSFET seems to switch either on or off… there doesn’t seem to be any variable resistance in the “saturation area”. Is it the PWM varying the gate voltage which controls the voltage to the solenoid? Or is the MOSFET itself switching on and off in accordance with PWM to Vgs = 2.35V max?

Read about the general principles here Pulse-width modulation - Wikipedia

Power Mosfets are operated only in saturation. PWM aims to control the average effective power by switching the power all the way on and all the way off, rapidly, varying the amount of time on vs off to achieve the desired overall power level.