# Voltmeter protection

I want to drive a couple of analog voltmeters 0-10V using a pwm controlled mosfet to supply 12v to the voltmeter. I'm not using it as a real voltmeter, i just needed an easy to control realistic gauge.

The highest number i will use to analogWrite the pwm will be 212, which is equivalent to 10v average. As far as i know, pwm doesn't really lower the voltage, so peak voltage will still be 12v.

Any chance to hurt the voltmeters by doing this?

Yes, they are designed for DC. You need to design a low pass filter and place it between the PWM and the meter. You are also correct PWM does not lower the voltage but the integration will take care of that and give you 10V at 212 if your calculations and design are correct.

gilshultz:
Yes, they are designed for DC. You need to design a low pass filter and place it between the PWM and the meter. You are also correct PWM does not lower the voltage but the integration will take care of that and give you 10V at 212 if your calculations and design are correct.

Will this filter add any delay to the response time of the voltmeter?
As i see, people are designing these with a particular duty cycle in mind. I need this to control the whole 0-83% range. Would that be a problem?

If that's a regular-old electro-magnetic meter movement it's actually current-operated and the coil inductance will smooth/filter the current so it will be electrically safe, and the inertia will smooth/filter the movement so the needle shouldn't visibly "vibrate" with the PWM.

Will this filter add any delay to the response time of the voltmeter?

I don't think you'll need a filter but yes, depending on the R/C time constant. And, that's the whole idea... You want the meter to smooth/average the PWM rather than responding to the on/off nature of PWM.

DVDdoug:
If that's a regular-old electro-magnetic meter movement it's actually current-operated and the coil inductance will smooth/filter the current so it will be electrically safe, and the inertia will smooth/filter the movement so the needle shouldn't visibly "vibrate" with the PWM.
I don't think you'll need a filter but yes, depending on the R/C time constant. And, that's the whole idea... You want the meter to smooth/average the PWM rather than responding to the on/off nature of PWM.

Oh yeah, the meter works as expected, indicating the average voltage. I was only worried of using pulsed 12v as it's higher than the maximum of 10v displayed on the meter.

Look at the meter scale or the meter
specs: Does it say something like
20,000 ohms per volt? That means
that the meter movement has a
sensitivity of 1/20,000 or 50
microamps full scale. So, if it
reads "on scale", it is not

Think of an analog meter movement as a very specialized form of electric motor - motors are PWM driven
all the time without issue.

gilshultz:
You need to design a low pass filter and place it between the PWM and the meter.

No. The inertia of the voltmeter IS the low pass filter, you will absolutely not need any other filter.
You will not need any special protection either as long as you dont brutally exceed the RMS value.

you ought to look closer a the meter.
Usually they are mA meters and display volts through a resistance in series.
Your 10v meter may be e.g a 1mA meter with a 10KΩ resistance in series or a 0,2mA meter with a 50KΩ resitance in series. In that case you replace the resistance with a half small resistance and you can drive it directly from the PWM output without any extra FET transistor.

As stated before an analog meter is simply a coil in a magnetic field working against a spring. I see no detriment to providing a 10% overvoltage especially since it will not force the needle over its 100% of range position.

Thought, How stable is the 12V? Any change in the 12V value will directly effect your meter reading.

RIN67630:
you ought to look closer a the meter.
Usually they are mA meters and display volts through a resistance in series.
Your 10v meter may be e.g a 1mA meter with a 10KΩ resistance in series or a 0,2mA meter with a 50KΩ resitance in series. In that case you replace the resistance with a half small resistance and you can drive it directly from the PWM output without any extra FET transistor.

Won't be doing that at this point since i already have the pcb with the mosfet installed but it's very good to know. I don't get what's the formula though. Can i just take a 25v voltmeter and replace the resistance with a fifth of the original?

JohnRob:
As stated before an analog meter is simply a coil in a magnetic field working against a spring. I see no detriment to providing a 10% overvoltage especially since it will not force the needle over its 100% of range position.

Thought, How stable is the 12V? Any change in the 12V value will directly effect your meter reading.

Should be pretty stable. I'm using a MeanWell supply rated for 2.5A, and i won't even use half of that.

petronel:
Can i just take a 25v voltmeter and replace the resistance with a fifth of the original?

Roughly, yes if the value is high enough (> 50kΩ and the internal resistance of the coil can be neglected.

petronel:
Won't be doing that at this point since i already have the pcb with the mosfet installed but it's very good to know.

Do you mean to say between yesterday and today, you have constructed and assembled a PCB based on not having actually worked out how to implement your project? That is simply shoddy planning.

While you do need to take some care in modifying the extremely expensive meter pictured, you do not need a FET or a PCB to connect it directly to an Arduino PWM pin using one resistor! It makes no sense to add a ridiculous amount of additional circuitry.

There should be no harm in driving the meter a little over the maximum scale reading and you can calibrate the readings in code (and you do not need floating point to do that).

Meter movements experience only very tiny forces, 20 times full scale might be an issue, twice is no problem, they typically have little padded end-stops for the needle.

Paul__B:
Do you mean to say between yesterday and today, you have constructed and assembled a PCB based on not having actually worked out how to implement your project? That is simply shoddy planning.

While you do need to take some care in modifying the extremely expensive meter pictured, you do not need a FET or a PCB to connect it directly to an Arduino PWM pin using one resistor! It makes no sense to add a ridiculous amount of additional circuitry.

[/quote]

I had the PCBs printed and installed for a while now as i planned to use a different type of indicator that was controlled by a FET but was not satisfied with the result so i opted for these voltmeters.
There are various inputs and outputs connected to the arduinos, not just the voltmeter, and i need a robust build, i don't want want to mess with duponts.