H-bridge to drive latching solenoid

[mn1247]

I need to drive a latching solenoid. After doing some research, I am realizing that I need an H-bridge, and would like to use an IC for this.

However, I've never used such a device before and I am hoping to find an example of how one is operated. Also, I'm confused as to whether I should be using a "full-H" or two "half-H" circuits (I've seen chips with each). Moreover, I'm hoping the chip might output a pulse in response to a continuous input (since the solenoid only needs a brief pulse to switch).

The solenoid is 9V, and needs 1.5A for 10msec to switch.

Any advice or examples? Particular devices I should look at? Thank you.

Eric

 

[Mikebits]

I can see no reason why you need an H bridge to drive a solenoid. An H bridge is usually used to drive a motor with the feature of reversing the motor direction.

I think you can get by with a single transistor to drive your coil.

 

[mn1247]

Because it's a latching solenoid, it needs one current pulse in the forward direction to latch it, and one current pulse in the reverse direction to unlatch it.

Hence, I think, the H-bridge.

Eric

 

[Mikebits]

Can you post the link to the data sheet or part number of the relay?

 

[ronv]

H Bridge

You should be able to use something like a 754410 (in parellel). You need to tell us a bit more about how you are switching it before we can help develope the pulse to drive it.

 

[Boncuk]

Here is what you are looking for.

Use an MC34151, a high speed dual MosFet driver (Motorola) with totem pole outputs (1.5A).

Here are schematic and PCB suggestion with a relay connected.

Boncuk

 

[Mikebits]

It would first be helpful to know what relay the OP is using before suggesting a solution.

 

[mn1247]

Thanks for the ideas. I'm afraid the specs on the solenoid are limited... here's a link describing what I'm trying to drive (the best "spec" I have):

https://www.digcorp.com/assets/0000/0205/S305DC_ValveAdapter_catalogPG33.pdf

My understanding is that a forward pulse (10msec) will latch the device (and thus open the water valve); a second 10 msec pulse then unlatches it (and closes the water valve). Thus, operation requires the ability to reverse polarity on the output of the driver.

I am hoping to keep the pulse as narrow as possible as I want to conserve power - I will be running the whole circuit off of a 9V battery.

I had not planned on using a "relay". Should I?

Eric

 

[Boncuk]

It would first be helpful to know what relay the OP is using before suggesting a solution.

It really doesn't matter if the OP is using a latching solenoid or a latching relay. If both of them have one coil only it needs to be connected reversely to unlatch it.

That is what the MC34151 does, dependent on the logic input.

Latching solenoids or relays are manufactured to preserve electric power. A short pulse in either direction will latch and unlatch them.

Boncuk

 

[Boncuk]

Thanks for the ideas. I'm afraid the specs on the solenoid are limited... here's a link describing what I'm trying to drive (the best "spec" I have):

My understanding is that a forward pulse (10msec) will latch the device (and thus open the water valve); a second 10 msec pulse then unlatches it (and closes the water valve). Thus, operation requires the ability to reverse polarity on the output of the driver.

10ms is the minimum pulse width specified. You can't further reduce that for proper function.

I am hoping to keep the pulse as narrow as possible as I want to conserve power - I will be running the whole circuit off of a 9V battery.

I had not planned on using a "relay". Should I?


Eric

You don't have to use a relay at all. The suggestion I made is universal. Just treat the relay coil being the coil of the valve.

To have the solenoid latch logic input A should be positive (for at least 10ms) while logic input B stays low. Output A will then have VCC-level and Output B will have ground level - latching the solenoid.

To unlatch logic input B should be positive (again for at least 10ms) while logic input A stays low. Output B will have VCC-level and output A will have ground level unlatching the solenoid.

Boncuk

 

[mn1247]

Thank you - you've been very helpful. Two follow-up questions if I may...

Do I have to worry about transients if both A and B are low or high simultaneously? Does it "short"?

Can the device limit the output pulse width? For example, if the input to A is +9V for 1 second in duration, can the chip output a briefer pulse (say, 20 msec)? Or is there another device that will do both?

To clarify a bit: My input signal is driven by an LED-photodiode pair, wherein the beam is either unbroken or broken depending upon whether an object enters the path. The solenoid should activate when the beam is broken, and should deactivate when it is unbroken. The durations involved are seconds to minutes, and I don't want to consume power during those entire intervals. So, I need to have pulses of 20 msec or so just at the rising and falling edges of the photodetector output.

Again, your help is much appreciated.

Eric

 

[mn1247]

(duplicate)

 

[mn1247]

One more quick question if I may...

I was reading the datasheet for the MC34151 and it says (in bold type) that I should not try to breadboard with this device. That's a problem for me - I will need to breadboard before I try to make a PCB. Is there a device that's a bit less "fussy"?

Thanks
Eric

 

[ronv]

754410

Did you rule out this chip? There are 4 half bridges in an ic so they are easily paralled to get 2 amps. $1.65.

https://www.electro-tech-online.com/custompdfs/2010/07/sn754410.pdf

The 34151 is made to drive fets so it is very fast (nice chip) but also only good for .4 amps continuious. Never know when that might happen. Can you post your detector circuit?

 

[mn1247]

The 754410 looks good - it's my understanding that two of these can be physically stacked one atop the other to allow a greater current load? Is that right? (sounds a bit bizarre). Do you just solder them together pin-for-pin?

The LED circuitry is based upon a project kit....

https://www.quasarelectronics.com/kit-files/electronic-kit/3130.pdf

I plan to remove the relay and tap that voltage to operate the pulse-generation circuitry (yet to be designed). The voltage at that point is (I believe) high if the beam is unblocked; low if it's blocked.

I will also be looking at reducing the power of this circuit - 45+mA is too much. Most of that is in the LED, but I don't need 15 meters distance (maybe 0.3m at most). So, I'm hoping a lower-power part will do.

Thanks
Eric

 

[ronv]

Yes, you can stack them but there are 2 full bridges per ic so you can just parallel them in the one ic. You just have to be careful you get a to a and b to b if you know what i mean.
I used about the same circuit your going to use for my son's jump-o-meter for his vollyball team. Had 24 transmitters and receivers spaced at 1 inch intervals to measure how high the girls could jump.
I think the output is ground when in the light if you repace the relay with a resistor.
What's the next step?

 

[Boncuk]

Hi Eric,

I've also read the bold type remark not to used bread boards for experimenting. It may refer to loads with high capacitance.

The miniature solenoid valve you're going to be using won't draw more than 100mA. The MC34151 can easily provide that current (also using a breadboard). I'd risk one fried IC to find out the proper function.

I guess using the SN754410 will be a lot overkill - requiring also more space.

To shorten activation pulses you might think about using non retriggerable monoflops connected to the IR-barrier delivering a pulse length of e.g. 12ms - connected to the logical inputs of the MC34151.

The two outputs follow exactly the two inputs:

InA-H -> OutA-H, InA-L -> OutA-L.

The same applies to InB and OutB.

If both inputs are H both outputs will be H as well, meaning no current flow through the solenoid - hence not a short circuit. Vice versa will be the same.

Regards

Boncuk

 

[ronv]

Specs.

The reason they don't like the MC34151 on a breadboard is because the added inductance along with the fast switching speed means a lot of inducitive voltages flying around. I don't sell chips but I don't think it makes sense to use a 40 ns chip to drive a 10 ms solenoid. The solenoid spec puts the resistance at 5 ohms which at 9 volts comes out to 1.8 amps. Probably won't be that high because the chip will drop a couple of volts at that current but still above the spec. Anyway if you decide to use the 34151 make sure to add the diodes shown in figure 21 of the data sheet and add some good decoupling -- both close to the driver ic.