Maximum Voltage on Voice Coil

Thread Starter

phogan22

Joined Nov 2, 2016
4
Hi all,

Be warned, mechanical engineer here. I'm working on a project with a DC non-commutated linear moving-magnet voice coil, controlled with a servo drive in current mode, and I can't seem to get a straight answer from the manufacturer on the maximum voltage the device is rated to.

The rated current is 2.4 A continuous/7.2 A peak, the back-emf coefficient is 83.6 V/m/s, the DC resistance is 12 ohm at 20 C, and the inductance is 6.5 mH at 20 C. The magnet wire is AWG 22, Type-E rated to 600 V. As a worst-case operating point at peak current and max velocity could be as high as 1.9 m/s. Accounting for a 50% higher resistance due to increased temperature, the highest voltage I could see is ~160 V_bemf and ~130 V_R (ignoring the L*di/dt term) for a total of 290 VDC. I only really plan on leaving it at 2.4 A, so this voltage is likely higher than I'd see.

I've sized a servo drive and power supply that are able to deliver this voltage, but my question here is: is the high voltage itself bad for the voice coil? What issues might there be even if I use the servo drive to keep the continuous current below 2.4 A and the voltage stays below a reasonable safety factor on the wire rating?

Thanks,
Paul
 

#12

Joined Nov 30, 2010
18,223
Speaker coils don't die from voltage or current, they die from heat. That's voltage times current, or watts.
Another way to say it is current squared times resistance.
2.4 amps through 12 ohms is 69.12 watts continuous.
That's how much energy you are allowed to apply to the coil in the long run.

Obviously you can get away with 7.2 amps, peak, for a few seconds...very few seconds.
This is based on the principle that audio has amplitude peaks 10X the average sound level.
So, yeah, you can hit it with 622 watts, for a few miliseconds, but don't plan on that being your goal.
 

Thread Starter

phogan22

Joined Nov 2, 2016
4
Thank you for the responses.

#12,

Specifically you're talking about resistive voltage drop across the coil, correct? Which is why P =V_coil*i = R*i^2? With a back-emf your voltage supply could be much higher, but at the same current and resistance the power dissipated by the coils would be the same regardless of the velocity. If I never let the current rise above the continuous rating then we wouldn't exceed that 69.12 W.

Paul

Edit: This is the specific voice coil I'm using: https://www.h2wtech.com/product/voice-coil-actuators/NCM08-35-450-3LB
 

#12

Joined Nov 30, 2010
18,223
Back EMF opposes the applied energy, by definition. In all my years designing audio systems, I have never seen an occasion where the back EMF increased the power dissipated in the voice coil because the back EMF reduces the current, and therefore, the applied power. I can't quite wrap my head around why you want to defeat the back EMF, but that's irrelevant. You want to guarantee a rate of acceleration? Fine. Just keep your average energy below 68 watts (per the specs) and you should be OK. There is no spec for heating time constant of the coil, but I would guess 1/2 second is the maximum time for peak energy input. (That would seem to demand that you started at Tambient (25C) and will demand a zero energy input for the next 4.5 seconds.) See where I'm going here?

The ultimate goal is to keep the temperature below the melting temperature of the insulation. The wire will heat up a lot faster than the total mass will heat up. I see you're moving 4.6 pounds in a total mass of 12.56 pounds. That's quite a lot of thermal mass, so external temperature measurements are just about useless for protecting the insulation on the wires. You could measure the DC resistance to find the temperature of the wires, but the practical way to avoid damage is to obey the specs for total energy: 68 watts continuous or 620 watts at 10% duty cycle.

Then there is the idea of cooling the total mass below 25C, but the thermal time constant of the total mass is way too slow compared to the thermal time constant of the coil for that to affect the coil heating by more than a few percent of total power. You can smoke a 68 watt coil in a very few seconds at 620 watts, even if you start at 0 degrees C. Just control your input power and you will be OK.
 

Thread Starter

phogan22

Joined Nov 2, 2016
4
I guess that's one of the sources of my confusion: if the maximum power rating is 68 W, does that refer to i^2*R power dissipated as heat in the coil? Or the total power inputted to the system? I think it's the former, because when I plug in (2.4A)^2*12Ohm I get almost exactly 68 W.

With a servo-drive operated in current-mode, don't you get constant output current, where the supply voltage will rise/fall with the back-emf? All the servo drive sizing guidelines I've been following recommend sizing everything to supply the maximum expected back-emf plus the I*R drop in the coil.
 

#12

Joined Nov 30, 2010
18,223
I get the feeling you're chasing a non-problem. What are you trying to do that requires such precise tracking and running on the frightening edge of over-heating?
 

Thread Starter

phogan22

Joined Nov 2, 2016
4
I'm trying to set up the voice coil with return springs and pistons and oscillate at a resonant frequency to pump fluid. I wouldn't say it requires very precise position tracking, and the exact force output isn't absolutely critical, but it'd obviously be nice to run as close to the continuous rating as possible (isn't that what continuous is supposed to mean anyways?)

With a 200 N continuous rating for the voice coil and operating at resonance, I didn't think this would be so complicated.. :/
 

#12

Joined Nov 30, 2010
18,223
I didn't think this would be so complicated..
I don't think it is complicated. Set up your frequency, connect the wires, and increase power until 68 watts happens.
I would do this with what I have on hand: A frequency generator and a 200 watt audio amplifier.
At resonance, the impedance and efficiency of the coil increases. That's where the back EMF opposes the applied energy the most.
When you start pumping, that load will pretty much squash the resonance effect. Your attempts to compensate for an impedance peak at resonance are mostly imaginary. The system won't act like a resonance under load. It will act like a resistance, as in "resistor", not like an inductor.
 
Top