# Accurate voltage control for motor

1. Jun 20, 2012

### lark

I have an air turbine, and it's critical that the motor speed be precisely regulated.
It's very voltage-sensitive, so the fluctuations in the line voltage between 120-126 A make the motor speed too variable.
How can it be regulated?
A UPS might work. But it would have to output voltage accurate within +/- 1%. THere's a rheostat on the motor, so the UPS could just regulate up, say, so it wouldn't be necessary to have both a step-up and step-down transformer. The motor uses 1-2 A at the normal speed. It uses max 9 A, but I never use it at that speed. I don't know, if you get a voltage regulator that's rated for a certain current, and more current goes through it, if it damages the voltage regulator, or it just can't regulate the voltage right.
It also needs battery backup for when the power fails.
Can anyone point me to such a UPS?
It would be great if the input voltage could be regulated by the actual airflow. The air turbine is powering an airline respirator, the airflow is 4-5 CFM through a hose maybe 3/4" inner diameter, with garden hose fittings. If anyone can help with that I'd appreciate it.
Laura

2. Jun 21, 2012

### Staff: Mentor

Hi Lark. You are going about it the wrong way, wanting to tightly fix the voltage. This means you are placing blind faith in the motor maintaining a constant angular velocity for a fixed supply voltage. What happens when the motor ages, or atmospheric pressure changes, or thermal effects show up?

Much better to use a feedback arrangement and try to maintain a constant airflow. But measuring airflow accurately may be difficult, easier is to measure speed of rotation and alter the voltage to the motor in an effort to maintain a fixed angular velocity. You can accurately sense the speed by sensing the passage of a pattern of dots around the shaft. Or maybe there is somewhere to measure pressure, and try to maintain that at a constant level?

Sorry, I don't have circuits in mind. Just pointing out what I reckon you should be looking out for.

If you're lucky, others may have some recommendations. Good luck!

3. Jun 21, 2012

### lark

Is it easier to tightly control DC voltage? It's a universal motor, it likely can run on DC just as well ...

The motor has a rheostat, so slow changes in the airflow/voltage ratio are OK. The variation I notice is from voltage changes.

It would be accurate enough to measure the airflow within about +/- 5%. A small change in voltage causes a big change in airflow, that's the problem.

I agree that it would be nicer to have a feedback arrangement, and I'm interested in how to do that. But it's likely more practical to control the voltage.

Laura

4. Jun 21, 2012

### sophiecentaur

Control the voltage by measuring the airflow - using an anemometer as the sensor - and use that as your feedback signal. An open loop system is almost bound to be less satisfactory and there is no need to avoid feedback just on the grounds of complexity. Feedback rules in most circumstances where you need control.

5. Jun 21, 2012

It would work well enough if I could tightly control the input voltage. However I don't know if there are voltage regulators that output voltage within +/- 1% that are reasonably priced. I asked about one, it costs ~$2700, which is too much. It likely has a lot of features I don't need, though. It would be enough to step up the voltage to a precise value, since there's a rheostat to decrease it again. About the air speed sensor, the air speed in the hose is about 10 m/sec. I would need a sensor that can fit in a 5/8" inner diameter hose and output a signal giving air speed +/- 5%. I don't know how to find such a thing, and once I had it, I wouldn't know how to use it to control the voltage input to the motor. I wonder if it could be done by software, send a signal giving airspeed to a computer, which then would send a signal to the motor. Laura 6. Jun 22, 2012 ### sophiecentaur I can see you have a problem. It seems you want a 'black box' approach to this as you are short of experience or facilities for building your own equipment. This usually means that you need to spend money to get what you want!! (A pain) The question with using open loop control is can you be sure that a particular voltage (however accurately controlled) will correspond to a particular air speed unless temperature and humidity are also controlled? If you are worried about air speed measurement then surely you will be measuring it anyway?? Else, why are you bothered? Would it be daft to suggest using the turbine speed as your feedback signal? If all you need is for it to be constant, irrespective of air flow, then a control loop from speed sensor, via a voltage regulator to the motor supply would do what you want. You may be between a rock and a hard place with this - either you have to put up with a system you aren't happy with or you need to learn about (or have local help with) control systems. You are right that a computer could be used. You could look into the Arduino system. It's a microprocessor board system which is aimed at home and student construction projects. It has many different interface modules for controlling virtually anything you could want. It is very easy to get started with loads of software and support available at reasonable cost but, of course, it represents another 'overhead' on your main project. (An excellent addition to the CV of a prospective job applicant, though). 7. Jun 22, 2012 ### lark It's an airline respirator, it pumps clean air to a facemask that I wear. If the air speed is too slow, the mask leaks. If the air speed is too fast, the turbine gets very loud. It's an industrial safety device that I'm using for a different purpose. Usually people when they use it, would just turn it up high, and let it be noisy. Not at all daft, regulating the turbine speed might well work fine, and it might be easier to find a motor speed sensor than an air speed sensor. I'm not sure if air temperature or pressure variations would make enough of a difference in the air flow/turbine speed ratio, to matter. Is there a motor speed sensor that I could attach to the metal housing of the turbine? I don't know if something could be attached to the shaft. 8. Jun 22, 2012 ### sophiecentaur From what you say, an air pressure sensing, rather than air flow sensing would suffice - because it'll be pressure that governs whether the air escapes from the mask. Did you even consider having a pressure relief valve on a T-off from the main air hose? I'd bet that would be the cheapest and most straightforward solution. As long as the volts supplied to the pump motor are always sufficient to give you 'enough' then, if they are too high, the valve would open and keep the pressure where you want it. Just another, alternative idea. 9. Jun 22, 2012 ### jim hardy is there a hands-on technician in your plant who might help you? Treadmills abound in junkshops now. Most have a 6000 RPM 90 volt permanent magnet DC motor capable of 2hp, and a power supply for it that gives steady speed control via a knob. I hesitate to advise a novice to mess with that kind of electrical power but it's a good source for a speed controller for like$10 instead of $2700. You'll need some practical help . old jim 10. Jun 25, 2012 ### lark Again, I'd probably do fine with a sensor for the line voltage, then something like a rheostat that would be adjusted based on the line voltage. That's what I'm doing right now by hand. I have a voltmeter, and for a given line voltage, I know how the rheostat should be adjusted. It seems fairly consistent. The turbine has a universal motor, it has brushes. Are speed sensors available for universal motors? Most people who use airline respirators use them at a high enough speed so that even with variations in voltage, they're always getting enough air. But I'm using it almost full time, I even sleep in it. So the noise it makes, which increases drastically with the air speed, matters. Laura 11. Jun 25, 2012 ### sophiecentaur That's worth knowing - a critical factor. You really do need to be running the motor as slow as possible and at a reasonably constant speed. It doesn't appear that the pressure / flow needs to be as well regulated as your original scenario suggested - you aren't using it for precise measurements, for instance. Is this a 'safety of life' issue or a way of keeping yourself isolated from ' room air'? The cost would relate to the answer to that question. Basically, I would say that there isn't an easy DIY solution to this. I see the solution in two parts: a regulated power supply (DC our would be cheaper) and a UPS. A whole range of UPS units exist from around £200 upwards and you would have to decide what down time it would need to cover. The regulated power supply is harder as it's not an everyday 'consumer' requirement. There is one scary solution and that would be to have ten 12V lead acid batteries in series, each one connected to its own charger. The chargers would need to be, say, 5A continuous rated and the batteries would need to have whatever capacity (in AHr) that you need to cover supply outages. The batteries wouldn't be cheap but the chargers could be ordinary auto battery chargers. You would need fuses all over the place and a well ventilated room to keep it all in but it would achieve a good, constant supply voltage for your motor and the UPS facility all in one go. 12. Jun 25, 2012 ### jim hardy Have you tried an incandescent light dimmer ? Like goes in a wall for ceiling lamp? I get a double outlet box, plastic from hardware store, and mount a dimmer in one side and an outlet in other side, nail box to a plank.. Wire dimmer to control one side of outlet. That'll control speed of a cheap old single speed electric drill which is a universal motor.. It'll give more constant voltage than does a rheostat. It'll run cooler too. old jim 13. Jun 26, 2012 ### sophiecentaur Great solution, Jim. Time was when you could get speed controllers with mains in and out connectors already fitted but now they're all inside the tools. She still has the problem of a UPS, though. No cheap solution there, I fear. 14. Jun 28, 2012 ### lark I get sick for several days when the turbine fails (allergies). But it's not life-threatening. Is there a good speed sensor for a universal motor? The utility voltage is pretty good - the engineer I talked to said it's usually between 120 and 124 V. Perhaps since the input voltage is already good, a less expensive voltage regulator would be needed to make it exact within 1%. There are slight symptoms of resistance in the neutral to ground connection in my house, and if it's reasonable to eliminate that, the line voltage would be more constant. When I turn on my 1 KW microwave, there's a 1.1 V neutral to ground voltage at the outlet that the microwave is plugged into. Also, the voltage to the turbine increases by about 1 V. And, the voltage at the side the turbine is on, is usually about 4 V higher than the voltage at the microwave side. The turbine draws max 9 A, and dimmer switches are 5 A max. I decided to use the rheostat that comes with the turbine, as a remote control, because of problems like this. Laura 15. Jun 28, 2012 ### sophiecentaur Imho the battery solution could be good for you because you could even buy a cheap lab bench supply unit for the lowest step in the 'Voltage ladder' and have, say, a continuously variable, stable 0-20V (say) range of control. The neutral volts would not be relevant in a system like that and many bench supplies have an external sense, which could easily be arranged to look at the 120V line. 16. Jun 28, 2012 ### jim hardy 17. Jun 29, 2012 ### sophiecentaur They are usually open loop control, aren't they so how would you use it in a regulator circuit, Jim? 18. Jun 29, 2012 ### jim hardy indeed they're primitive. My thought was since it controls firing angle it would make a stable voltage source . Poking around i found bigger 'lamp dimmers' selling to home shop enthusiasts for router speed controls. I think this 10 amp one is$39. A bit pricey yet just for an experiment.
http://www.wolfautomation.com/assets/15/ac03.pdf [Broken]

To close the loop is a significant jump in effort.

Of course nowadays there's IC's made for that, here's an analog one and a microcontroller based one.
http://www.mouser.com/catalog/specsheets/stevalill004v1.pdf

Gee- today's hobbyists have a cornucopia of gadgets !
Probably there's something on the hobbyist market based on these IC's just i haven't found it yet.

Closing loop on a lamp dimmer would be interesting challenge - but for this application one should stay with ready made parts , enclosed and safe.

old jim

Last edited by a moderator: May 6, 2017
19. Jul 1, 2012

### jim hardy

aha lamp dimmer with IR remote
so control is not connected to mains

20. Jul 2, 2012

### sophiecentaur

Good thinking Batman. Pass the soldering iron. ;-)

21. Jul 5, 2012

### lark

The line voltage is 120 V +/- about 5%, so it's already pretty good.
Could I get somewhere a voltage regulator that takes 120 V +/- 5% and outputs a voltage accurate to +/- 1%, either as 120 VAC or some amount that you specify?
Would the bench supply unit do that?
Laura

22. Jul 5, 2012

### jim hardy

There's a device called "Ferroresonant Transformer" that delivers nearly constant voltage, typically 3% Δout for ~15%Δin.
A company named Sola invented them in 1930's and they are still the industry workhorse. I see they now advertise 1% output regulation in their CVS series...

http://www.solahevidutysales.com/cvs_hardwired_series_power_conditioner.htm [Broken]

http://www.solahevidutysales.com/cvs_hardwired_series_power_conditioner.htm [Broken]

They are heavy, expensive, bulky and run hot but they last forever. Because of their longevity they are fairly common on the surplus market. Occasionally I see one in my scrap metal salvage yard. About all that ever goes wrong is the internal capacitor fails and it's an inexpensive part.

Their CVS series has harmonic suppression, but for just running a motor you wouldn't need that.

Is this a DIY project or can you get some corporate help?
Maintenance dep't of any good sized industrial outfit should have a couple Solas laying around.
EBAY has several but they're heavy so shipping can be deadly.. Perhaps you could find one small enough to fit in a priority mail flat rate box. I have forgotten how many kva you need.....

Last edited by a moderator: May 6, 2017
23. Jul 5, 2012

### sophiecentaur

I don't understand what you are saying here. You say that the regulation is "pretty good" but that you need better??. How does the motor speed change for +/- 5% change in volts? is that not acceptable? Is it the change in noise level that worries you or the pressure and 'air spillage'?

A low voltage (20V ish) regulator will give you more than accurate regulation for motor speed control but I don't think it should be necessary.
I think I'd already established that a DC supply is far far easier and cheaper to provide. If this will work your motor than that's what to go for. 120V power supplies are expensive and you could get by very easily using just series 12V batteries (individually top-up charged) They would give you the standby facility much more efficiently - just like the old telephone exchanges with a room full of massive accumulators plus a fairly good regulated voltage. All that you'd need would be some appropriate cheapo chargers and the appropriate Ahr worth of battery.
The only reason that people use UPSs with mains AC output is for flexibility and to suit any equipment (that equipment will have its internal regulators). You still need a battery backup of exactly the same energy storage capacity (even more, in fact). For a long standby time, the installation is very costly and you don't get voltage regulation. All the cheap ones will do is to give your computer system time to shut down safely after a brief period. You can't afford to have a shut down so your battery capacity needs to be greater than that. I actually don't think you would need more than 'just enough' series batteries to give you around 120V DC. When 'float charging' a 12V battery will give about 14.5V. Nine cells in series would give you 130V and eight would give you 116V. This should be enough to run the motor and some to spare. You could investigate what the volts are on the motor to give you your normal running speed and also the minimum volts that will give you 'acceptable' pressure.
Once you lose mains volts, the battery volts would drop fairly quickly to the 12V value and it could be necessary to switch in an extra battery to bring the motor speed up again but you could cope with that circumstance, I guess.
All this must seem a bit over the top and 'Heath Robinson' but it is an affordable and achievable solution. I could appreciate that it may not be attractive but the nice thing about car batteries is that they don't let you down and are 'consumer' priced items. The UPS (for your application) would be a lot more expensive and you couldn't rely on voltage regulation without even more cost.

24. Jul 6, 2012

### sophiecentaur

That constant voltage transformer idea is a good one, Jim. I had forgotten, about them, to be honest.

25. Jul 11, 2012

### lark

This motor is a universal motor, does a universal motor have inductance? I asked Sola about voltage regulators and they said with an inductive load, there are short-term current spikes that might overload the voltage regulator, causing it to go to 0 V. So they say, for a motor the voltage regulator has to be sized a lot higher than the nominal current draw. I don't know why a universal motor would have inductive characteristics though.
Would such spikes in current be something one could measure with an ammeter, or would they happen too quickly to see?

The motor speed changes drastically for +/- 5% change in voltage. There's some kind of feedback mechanism going on that causes this voltage sensitivity, I think. +/- 1% is the most I'd want. It's the noise mostly that's the problem.

Would this give voltage within +/- 1%?
I don't know how voltage-sensitive the motor is, using DC power. The guy who makes them says he's never run them on DC, he doesn't know how they work on DC. I don't know what DC voltage the motor would need. The rheostat is an AC rheostat, so I'd have to find a different rheostat.

There aren't power outages very often, so I'm not much concerned about that, just about the daily voltage variation.

thanks
Laura