Go Down

Topic: Driving proportional solenoid valve using PWM (Read 741 times) previous topic - next topic

TomGeorge

#15
Oct 18, 2018, 11:29 am Last Edit: Oct 18, 2018, 11:30 am by TomGeorge
Hi,
You will need a PWM frequency of 1.2kHz.


Tom.... :)
Everything runs on smoke, let the smoke out, it stops running....

zeus2kx

#16
Nov 07, 2018, 07:54 pm Last Edit: Nov 07, 2018, 09:34 pm by zeus2kx
I would modify this circuit to deliver about 1.2A max. To do that, the voltage divider has to deliver 1.2V to the op amp pin 3, when 5V is input. R1=3.3K, R2=1K should produce about 1.16V.

The valve drive and op amp power supply voltage must be greater than about 8V. 10-12V should be OK but you will need a heat sink on the power transistor.

Sorry for the long delay. I was away for a while and also took me some time to arrange all the required parts.
Good news is that it is working! Thanks a lot.
I used a pot for PWM. A small issue is that there is a high pitch sound, though not loud, coming out of valve. Is there a way to suppress this sound?

I have a different valve (PVQ31 0-165mA 24V).
Resistance of the coil (without exciting) is about 100 ohms.
What changes do you recommend in the circuit?

Thanks.

Hi,
You will need a PWM frequency of 1.2kHz.
I used digital pin 5 which has PWM frequency of about 980Hz.

TomGeorge

Hi,
Good to see you have a result.

What are you using as a power supply for the solenoid?
Can you post a copy of your circuit diagram please?

Thanks.. Tom... :)
Everything runs on smoke, let the smoke out, it stops running....

outsider

Datasheet don't say anything about suitability for PWM or whether dither is needed. Good luck.  :smiley-confuse:

jremington

#19
Nov 07, 2018, 10:48 pm Last Edit: Nov 07, 2018, 10:49 pm by jremington
Quote
Is there a way to suppress this sound?
No, but you could try different PWM frequencies.

For the new coil, you have to change the circuit constants to result in 165 mA max = 16.5V across a 100 Ohm coil. Study the design hints in reply #14.

That cannot be done without a higher voltage power supply, say 24V.


TomGeorge

Datasheet don't say anything about suitability for PWM or whether dither is needed. Good luck.  :smiley-confuse:
I agree, no notes about PWM, you need to confirm with supplier/manufacturer that the valve is happy with PWM or a DC current.
If PWM is okay, they should have quoted an operating frequency.
Tom... :)
Everything runs on smoke, let the smoke out, it stops running....

zeus2kx

#21
Nov 08, 2018, 07:34 am Last Edit: Nov 08, 2018, 07:46 am by zeus2kx
Hi,
Good to see you have a result.

What are you using as a power supply for the solenoid?
Can you post a copy of your circuit diagram please?

Thanks.
I am using a wall wart to power the circuit.


I have tried contacting IQValves but they have no information about this "retired" valve.
Also, I couldn't reach SMC (the other valve company) for technical issues. I will retry.

Thanks.

zeus2kx

#22
Nov 08, 2018, 07:44 am Last Edit: Nov 08, 2018, 07:47 am by zeus2kx
No, but you could try different PWM frequencies.

For the new coil, you have to change the circuit constants to result in 165 mA max = 16.5V across a 100 Ohm coil. Study the design hints in reply #14.

That cannot be done without a higher voltage power supply, say 24V.


How do you recommend to change the frequency tp about 1.2KHz using arduino?

If my calculations are correct, R1 = 10K, R2 = 340 Ohms to deliver Vout = 0.165V. Please correct me.

I don't understand why the current value corresponds to voltage (0.165A to 0.165V).
Also in case of Parker, voltage divider calculation yields the same result, i.e. 0.172V or 0.085V corresponds to 0.172A or 0.085A.


jremington

#23
Nov 08, 2018, 10:10 pm Last Edit: Nov 08, 2018, 10:11 pm by jremington
Quote
I don't understand why the current value corresponds to voltage (0.165A to 0.165V).
Ohm's Law. V=IR

Quote
R1 = 10K, R2 = 340 Ohms to deliver Vout = 0.165V
Correct.

You could also increase the value of the current sense resistor, for example to 10 Ohms, leading to 1.65V for 0.165A through the coil, and then use the original voltage divider.

zeus2kx

Ohm's Law. V=IR
In this particular case when coil resistance is 100 ohms, yes V=IR.
In#22, both coils can't have same resistance but still their Vouts are similar to output current. Doesn't make sense to me.

You could also increase the value of the current sense resistor, for example to 10 Ohms, leading to 1.65V for 0.165A through the coil, and then use the original voltage divider.
Don't have 10 Ohms resistor with higher wattage.
I will test R1 = 10K, R2 = 340 Ohms combination.

Thanks.

TomGeorge

Hi,
That valve is not designed for PWM control, the circuit and the info sheet you posted the link to shows it is a DC current type proportional control.
The circuit you are using is for DC current control of the valve, not PWM.
The circuits input control variable is DC Volts, not PWM.
The values and equations are for DC current, NOT PWM.

If you replace the PWM from the Arduino with a variable 5V supply, I think you will get better results.

With PWM your meter is measuring AVERAGE Volt drop and hence AVERAGE current of the PWM system.

The circuit is, via the 1Ohm resistor, current limiting the Valve Coil current, ONLY on the PWM ON time, so the peak current is limited, BUT the average is lower, dependent on the duty cycle.

I hope I explained that clearly enough.
ALL the parameter refer to DC current control, NOT PWM.

Tom.... :)
Everything runs on smoke, let the smoke out, it stops running....

jremington

#26
Nov 09, 2018, 03:27 pm Last Edit: Nov 09, 2018, 08:05 pm by jremington
Quote
In#22, both coils can't have same resistance but still their Vouts are similar to output current. Doesn't make sense to me.
In both cases you have a 1 Ohm sense resistor, producing a voltage proportional to the coil current, according to Ohm's law.

zeus2kx

#27
Nov 11, 2018, 03:38 pm Last Edit: Nov 11, 2018, 03:39 pm by zeus2kx
I hope I explained that clearly enough.

Yes thank you.
I will test using a regular 10K pot with 5V as input. I can then vary the voltage.
I have also ordered digital pot X9C103S.
I hope everything would work out as expected.


zeus2kx

In both cases you have a 1 Ohm sense resistor, producing a voltage proportional to the coil current, according to Ohm's law.
Oops, I missed that out. Thanks.

MarkT

I used FQP30N06L instead.
Its a logic level N-Channel MOSFET with Vgs lower than 3V.
No, the FQP30N06L requires a minimum of 5V gate drive, this is very clear from the specs and
graphs.
[ I will NOT respond to personal messages, I WILL delete them, use the forum please ]

Go Up