120VAC digital potentiometer?

Thread Starter

JimG

Joined Dec 7, 2009
53
I would like to build a prototype circuit that is the equivalent of a digital potentiometer to control a 120VAC / 400mA load.

Purpose is to reduce the supply voltage for a 41W positive displacement pump so that the flow characteristics can be digitally controlled. The pump is essentially a spring-mass plunger driven by a solenoid at 60Hz. (I'm pretty sure there's an internal diode acting as a half wave rectifier).

The concept has been proven using a fixed resistor in series with the pump. Next, I would like to digitally control the value of the series resistor in order to vary the pump output.

I've already tried using a phase angle controller, but messing with the shape of the sine wave seems to confuse the pump once the delay approaches a half wave. This might yet turn out to be the best solution, but I'd like to try a digitally controlled resistor, too.

I've done some online searching, but haven't found anything that can handle the 120VAC and 400mA. Before trying to reinvent something, can someone point me in the right direction?

Thanks in advance.

Jim
 

R!f@@

Joined Apr 2, 2009
9,918
well, one way it to use a motor turn a pot, like in remote control volume in amplifiers, it has feed back to track it's position.
I think this a safe way, since only the pot will be live, the control will be low voltage.
What do you think?
 

Thread Starter

JimG

Joined Dec 7, 2009
53
I had not thought about using a small, remotely controlled motor to turn the pot. But it makes perfect sense -- thanks.

My preference would be for something simpler, and if possible no (more) moving parts.

I have been looking at solutions using an LED and a photo sensor to provide isolation. But I am not finding anything for the 120VAC side.

Jim
 

t06afre

Joined May 11, 2009
5,934
I do not which voltage range you want. But I can see a problem with heat dissipation. The energy not used by the motor will be converted to heat in your "resistor element" . The high voltage may also be a problem.
 

someonesdad

Joined Jul 7, 2009
1,583
The servomotor/pot approach is viable, although you'll probably need a hefty pot.

A solid-state approach would be to use some power FETs in parallel, as they can operate as variable resistances and the control side just needs a D to A converter. But you'll want good heatsinks to dissipate the wasted power.

If it was me, I'd look at a servo-driven Variac. Years ago I saw three big Variacs ganged together to compensate for voltage variations on a 3 phase system. A small 2 or 3 A Variac would work fine.

You may want to make sure the motor can reliably work at the lower voltages; some AC motor types will overheat when the voltage is lowered.
 

Thread Starter

JimG

Joined Dec 7, 2009
53
A solid-state approach would be to use some power FETs in parallel, as they can operate as variable resistances
Can you say a little more about this, e.g., maybe an FET product line to get me started? I think this might be what I'm looking for.

The D/A conversion should be fairly straightforward.

I will keep in mind your caution about overheating at lower voltages and run it through some tests. The pumps themselves are fairly inexpensive, so if I burn out the prototype it is not a great catastrophe.

Thanks.

Jim
 

Thread Starter

JimG

Joined Dec 7, 2009
53
I do not which voltage range you want. But I can see a problem with heat dissipation. The energy not used by the motor will be converted to heat in your "resistor element" . The high voltage may also be a problem.
Yes, the heat dissipation may become a challenge. The pump is designed to run on nominal 120VAC. My estimate is that I might want to drop the voltage by as much as 40V across the added resistor(s), so I could need to dissipate as much as 15W.

I suspect, however, that testing is going to show that much smaller voltage reductions will accomplish the goal, so the 15W estimate is likely to be very much on the high side.

Jim
 

SgtWookie

Joined Jul 17, 2007
22,230
I think I'd use power resistor(s) to drop much of the voltage, and use a MOSFET or power Darlington as a voltage follower in a narrow range for the fine adjustment. If you allow the voltage drop across the MOSFET/Darlington to become too large, you will burn it up. Good heat-sinking will be essential.

Marlin P. Jones & Associates has this 500VA 0V-130V Variac for $50:
http://www.mpja.com/prodinfo.asp?number=15162+TR
Not a computer-adjustable solution, but a Variac will at least have low power dissipation.

The least expensive solution might be to simply use relays to switch resistors in and out.
 

R!f@@

Joined Apr 2, 2009
9,918
Using power resistors in combination with relays can make a digitally controlled variable rheostat, but of course, resistor values have to be calculated to get the necessary load variation, and any other adjustment is not possible.
 

SgtWookie

Joined Jul 17, 2007
22,230
Using power resistors in combination with relays can make a digitally controlled variable rheostat, but of course, resistor values have to be calculated to get the necessary load variation, and any other adjustment is not possible.
True, but we do not know what resolution our OP really needs.

Infinite resolution is always nice, if you can get it. However, it is not often practical from a cost standpoint alone.
 
Top