Hello,

I'm doing a project on driving a solenoid valve in BOSCH. The input is a 0-10V sine wave (Which is actually an output from a D/A converter) and it is to be converted to a PWM signal of 300Hz with 0-100% range dutycycle. This duty cycle is to be used to inject current to drive a solenoid valve. For 100% dutycycle, the current should be ideally 1.4A to close the valve completely.
Would using a micro controller give a more accurate PWM signal? Please help me out.

Thank you.

 

Almost everything you've described is a fairly straightforward application of PWM control and could be accomplished without a micro controller. But where does the sine wave come in? What aspect of the sine wave will control the duty cycle?

 

Do you have a link to the BOSCH(?) solenoid valve spec's?

Ken

 

The sine wave is the feedback signal from a pressure transducer which decides the dutycycle of PWM. I have solved the PWM generation part. My main trouble lies with how to convert the dutycycle of PWM signal to corresponding current to drive the Solenoid Valve

 

Re: post #3

Ken

 

The sine wave is the feedback signal from a pressure transducer which decides the dutycycle of PWM. I have solved the PWM generation part. My main trouble lies with how to convert the dutycycle of PWM signal to corresponding current to drive the Solenoid Valve

Firstly you choose a supply voltage to apply to your solenoid, considering your driver drop and the solenoids rating. Then your going to place a current limiting resistor on the solenoids return line to control max current flow, at 100% duty.

Would using a micro controller give a more accurate PWM signal?

A micro may give you the option to implement a variable control algorithm.

 

To be more specific, i'm designing a Solenoid Valve Controller to control the pressure in the common rail system which is used for testing fuel pumps. It is tested at various pressures. So the pressure keeps changing continuously in the rail which is taken care of by their custom made software. The sine wave is the output from this software.
This means that the sinewave amplitude keeps changing every 30sec or so and hence the corresponding dutycycle of PWM and the current to the solenoid valve.

The only detail about the solenoid valve I know is that the max current needed to close it completely is 1.405 Amps and its insulation resistance is 1MΩ.

 

So, you are saying that the proportioning solenoid valve only has a minimum-current-to-close spec' and no resistance or voltage spec'. I don't see how you can design a driver without at least one more detail...voltage or resistance. Insulation resistance doesn't count. Can you ask someone?

Ken

 

The problem is, the solenoid valve is already fixed in their common rail unit. They didn't manage to get much detail before installation i guess. This was all the specs they had when we questioned them .

 

Kind of like telling you to put a tire on their vehicle without telling you what make and model it is. I think you are going to have to push them for more information. Even if they have to walk up to it, and read the model # on it.

Ken

 

The problem is, the solenoid valve is already fixed in their common rail unit. They didn't manage to get much detail before installation i guess. This was all the specs they had when we questioned them .

If they have the valve, don't they have the amplifier to drive it? The very first bit of information required is the full model number of the valve.

 

They've given us the maximum current value and there are no more data available . We can set up a circuit for a particular voltage output and use a suitable resistor to limit the current to 1.4A. But my problem lies with designing a circuit which will swing the voltage based on the PWM signal.

 

How would you select a resistor to limit the current if you don't know what the supply voltage or load is? With no more information than that maximum current number, you're just spinning your wheels. Just buy a PWM controller kit, assemble it.
http://store.qkits.com/moreinfo.cfm/MX033
http://store.qkits.com/moreinfo.cfm/K8004

Ken

 

I checked the valve today and after much convincing, the operator allowed us to measure the voltage at the coil . At 100% dutycycle, the current through the coil is 1.565A and the voltage across the coil was 10.8V.

 

OK, this might work as an off-the-shelf solution: http://www.axiomatic.com/pwm-controller.html

Ken

 

Just one doubt!!! What must i use to switch my voltage from positive to negative? The voltage must be positive during T-ON and negative during T-OFF.

 

Just one doubt!!! What must i use to switch my voltage from positive to negative?

Please post your circuit.

The voltage must be positive during T-ON and negative during T-OFF.

Are you referring to the device in the link?

Ken

 

Hi adithya.rp,

I’m going to provide my mind-reading services to the group and comment on what I “think” you are after. Refer to attached schematic.

The PWM signal is 300Hz, period is therefore 3.3mS. However, the actual ON time could vary from; say 1% to 99% where the solenoid would be expected to switch fully on or fully off in 33uS. Is this what it does?

On the other hand, if it is a control valve where flow is proportional to steady DC current then an averaging filter is required unless the solenoid coil inductance (unknown) is designed to do the averaging for you. Is this what it does?

The current sense feedback is not required, but it would be a better design if the actual current reading is used to calibrate the 100% value. The solenoid is calibrated for current versus flow rate because the copper wire used in the coil winding is very dependent on ambient temperature by 0.4%/C. This means that a fixed voltage may not translate accurately into the maximum flow value over a range of temperatures. Is high accuracy required?

If high accuracy is not required then use a limiting resistor in series with the coil. Calibrate when the coil temperature is at a normal operating value it is most often at.

It sounds like you need a high-side coil driver arrangement where “positive during T-ON and negative during T-OFF”. Does this mean that one side of the solenoid is grounded and must stay that way?

Use an off-the-shelf controller if you can.

The devil is in the details. The quality of the advise offered on this forum is proportional the the quality of the request for advice.

Regards,
Ifixit

 

I've attached a simple circuit that i tried. But the problem is, i'm getting 1.4A at the load irrespective of my dutycycle. Its 1.4A for 100% dutycycle (Which it has to be) and the same for 0% dutycycle too ...

 

Hi,

At see at least two problems...

1. The 2N3796 is a low power, depletion mode, audio amp. You need a P-Channel, enhancement mode, MOSFET switch that can handle at least 1.4 Amps (Id).
2. The source pin of the P-Ch MOSFET should go to +12V. The gate voltage is applied with respect to the source pin. I.E. Gate = +12V = MOSFET off, and gate = +2V = MOSFET on.

P.S. Since the 2N3796 is reversed biased, it may have shorted out and is no longer any good.

Good Luck,
Ifixit