# Driving electromagnetic coil "out of spec"

I'm designing automatic watering system for my garden. I've bought these bistable electromagnetic valves.

But specs are somewhat confusing: voltage range is 3.6-6 V, current consumption is 200-300 mA, coil resistance is approx 9 Ohms. I've measured the latter parameter for all valves bought and got mean resistance of 9.08 Ohms with SD = 0.13 Ohms so coils have pretty much the same resistances.

But I can't figure out numbers for volts/amps - seems like "datasheet" and description are simply contradictory. Given that coil resistance is valid I've got the following current values for boundary voltages (assuming R = 9 Ohm): 400 mA @ 3.6 V and 667 mA @ 6.0 V; while the specs are 200-300 mA.

As I remember from my previous practice with relays the device itself (relay switch) is just rated for some voltage (at which I guess it's safe to use it without insulation breakage, arcing etc) but the coil is characterized with resistance and working current. As soon as bistable valve is just a coil it's performance should depend only on current value and not the voltage directly.

I've conducted small test: feeding SN754410 with 10.9 V and powering my coil in series with 2 parallel 47 Ohm 1 W resistors I've got that for 9.8 V drop on the whole circuit (coil + 2 par. resistors; -1.1 V due to drop on SN's outputs) I'm getting 265 mA and valve works well (I've tried to change pulse polarity too). At the same moment voltage drop across the coil is just 2.4 V. Seems like it is the current which specified right in valve datasheet and not the voltage!

And the question is for what reason the seller has specified 3.6-6 V range? Will I get any troubles/glitches with valves using them undervolted? Or it will be ok while current is enough to stay in specified 200-300 mA range?

No one here can tell you what's in the mind of a seller. But you've done a wonderful job characterizing the devices. So you have all you need. Just ignore the seller's writing and use what you've learned about them. (I'm actually quite impressed with the detail work you've done in preparing yourself.)

Most relays (and yes, I'm aware that strictly speaking these aren't "most relays") are designed to operate (this is an "always operate" spec) at about 70% of their voltage rating. This is a $$\3.6\:\text{V}\$$ device. If I applied that "rule" here, I'd guess that these would always operate at as little as $$\2.52\:\text{V}\$$. Just as an educated prediction, anyway.

That seems to be what you are finding, as well.

I don't find it any problem that you are sourcing these the way you are, if you are willing to go to the lengths you have already done in validating what you are getting. Again, I'm really impressed with your practice, here.

In any case, you know everything you need to know to drive these. Believe what you have observed and use that information.

If on the other hand you are making something for others where you need guaranteed specifications to limit your own legal liability, then all of the above advice goes out the window. Then you need something from the supplier or manufacturer that places them in the middle between you and someone who may feel harmed. That's an entirely different topic.

• Thanks for detailed answer. But regarding undervoltage: I always thought that this is the current who determines the strength of electromagnet, not the voltage. So as soon as I'm feeding enough current through coil it should work fine, right? Commented Jul 1, 2021 at 18:35
• @DrobotViktor As an educated guess, anything between about 2.5 V and 3.6 V should be sufficient. I'd select a convenient voltage rail power supply regulator IC in that range -- for example, 3.3 V -- and use that with a transistor switch. A transistor will drop negligible voltage from the 3.3 V and the valve will have enough to work with. You just don't worry about it. If you already have a 5 V source, then I think I'd design a constant current circuit for 350 mA instead, because of the excess overhead voltage. What is your situation? What rails are convenient to you and what drives the ON/OFF?
– jonk
Commented Jul 1, 2021 at 18:40
• Because these valves are bistable I must use some kind of H-bridge to generate pulses (approx 30 ms) of opposite polarities. I use SN754410 for this project but its outputs are dropping up to 2.9 volts max in total at 500 mA (and this is why I made all of the tests). My plan is to use +12V from battery directly to feed SN754410 (and then use 2 parallel 47 Ohms/1 W) or use LM7808 for that purpose (then I need 2 parallel 27 Ohms/1 W). The whole LM7808 will be driven from +12V rail and will be switchable with high side driver, which in turn is controlled by ATmega64A Commented Jul 1, 2021 at 18:47
• Also I wonder if it is safe to drive 1 W resistor in pulsed mode (I need only 1 pulse of approx 30 ms to switch the valve) but with momentarily power dissipation of 2 (and may be even 4) watts. At the worst case my valves will be switching every minute but I think that short 30 ms pulse per 60000 milliseconds is negligible even with so huge heat dissipation Commented Jul 1, 2021 at 18:50
• @DrobotViktor Geez. Sorry. I had completely forgotten about the SN754410! I would use an inline current limiting circuit with BJTs. You just place that in series with your valve and between the two half-bridge outputs. This would self-limit the current. You would then face asking yourself, "What current, exactly?" For this, I'd go with 300 mA and see how that works. That's about 2.7 V for the valve, which is enough above what I think it can use to make me feel somewhat safe. More conservative, I'd use 350 mA. Inline current limiting is really easy as this application doesn't need precision.
– jonk
Commented Jul 1, 2021 at 18:53

The picture of the datasheet is below the text, and it says that working voltage is 3.6V. There is no mention of 6V. The datasheet never even mentions about current consumption, so who knows where the specs you are reading from the text are from.

Trust the datasheet and specs of the device, not the text. If in doubt, don't buy from places who don't know what they are selling, but from reputable distributors.