Diode In Series for Solenoid

Harry Stoner

I have a BS2 relay controller that's been running great on a breadboard for many months. I am now putting it on a PCB. One thing I thought about was putting a 1N4004 diode forward-biased in series before the solenoid coil. The 1N4004 would protect the transistor or MOSFET from receiving back-EMF voltage spikes if the reverse-biased diode physically attached to the solenoid fails. In other words, add a 10 cent diode to protect a $1.50 transistor and possibly the BS2 behind it.

Coils operate on 25VDC and draw maybe 1-3A and are normally pulsed on for 25-100ms. By putting the diode in series I am dropping the voltage to the solenoid coil to maybe 24V but I don't think that's a problem.

I am worried if the diode will fail in normal operation from normal forward-biased currents. Will say continual 3A bursts of 50ms hurt the 1N4004 (rated for 1A continuous and 30A spikes forward)? Just a question of reliability. I don't want to have to replace these diodes on the board because they are not up to the task. Maybe I should just live with the possibility that the transistor might take a hit (but then, will the BS2 I/O pin be next in line?).

Any thoughts?

Thanks.

Harry

Guenther Daubach

Harry,

usually, protection diodes are not placed in series with the solenoid but in parallel instead (see the attached picture).
This way, the diode shorts the back-EMF, this protecting the transistor.

I have used 1N4004 or 1N4007 quite often for that purpose, even with bigger relays, and never had problems, although there are various fast switching diodes available who would do an even better job. In a PWM-controlled motor driver application (24V up to 5A), I use BYW98 diodes for that purpose.

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Greetings from Germany,

G

pjv

Hello Harry;

A forward biased diode in series with a solenoid coil will do nothing to protect anything.

If indeed you have a solenoid with 3 amps flowing through it, and you switch the current off; at that very instant the inductance of the coil will cause the current to continue to flow at 3 amps and the voltage will rise to whatever level needed to push the 3 amps through whetever resistance it finds. Generally this means that the voltage will spike high enough to break-down whatever silicon there is (the transistor collector junction), until the current can flow. This break-down spike is what takes out the transistor. One cannot predict in what manner the transistor will fail; it might short collector to emitter, but also collector to base. In the latter case, the spike may then also appear on the circuit driving the transistor, and depending on the value of the drive resistor, may also take out the drive (BS2) circuitry.

As Guenther said, put a sufficiently large diode in parallel with the coil, this provides the path for the 3 amps in the coil to circulate and hence dissipate. The current will die down relatively slowly compared to the switch-on build-up time. In fact, 24V (of the supply) divided by 0.8 V (of the parallel diode) = 30 times slower. Depending on the value of the coil's inductance and its resistance, the decay time could be a millisecond or more. So you need a sufficiently sustantial diode to take a 3 amp decaying pulse for whatever period of time.

With close observation on an oscilloscope you can observe the voltage (0.8 V) accross the diode, and the duration of the current flow. It is an interesting experiment to add a 1 ohm resistor in series with the diode and observe the oscilloscope trace giving a 3 Volt peak for the 3 Amp pulse. Note that in this case the decay time will be shorter than the "diode only" case by a factor of about 3 V/ 0.8 V = 4. Then if you increase the value of the resistor to 2 ohms, the 3 amp pulse will give a spike of 6 volts, and the decay time is halved again, to a factor of 6 V/ 0.8 V = 8 approximately.

So the decay (dissipation) time is a function of the resistance of the current circulating loop; the higher the resistance, the faster the decay, BUT the higher the voltage (and power) the circulating circuit must tolerate. Its all a matter of balancing one thing off against another. Generally though, with relatively small inductance values and small currents, the standard parallel diode is a good way to deal with the current induced spike. This is not neccessarily the case with larger current coils, or those with large inductances. Automotive airconditioner style clutch coils are an example of what to be wary of.

Hope this insight helps you understand it all a little better.

Cheers,

Peter (pjv)

Harry Stoner

Thanks Guenther and Peter for the very good information. Thanks for the great detailed math Peter on the different characteristics. - this is great stuff to know.

Please understand that there is already a back-emf diode in parallel - I'm just seeing if I can add a failsafe protection mechanism in case THAT diode fails (e.g. due to a lead breaking off because of mechanical vibration or whatever). I guess it's overkill, but now is the time to think about it before I make up the PCB!

Thanks again.

Harry

pjv

Hi Again Harry;

As before, the series diode does nothing for you. If you are concerned about reliability, then put that second diode in parallel with the first. They will somewhat share current (not likely equally), but if vibration could cause one to break, then you still have number two.

Cheers,

Peter (pjv)

Crystalionx

It looks like you've gotten all the help you need. I am interested, frankly, in applying the relay system to opening a door. What type of Mosfet and solenoid did you utilize - and for what purpose, if you don't mind.

KenM

A no brainer approach to energizing "most" loads, and a good option for a solenoid that may pull well over an ampere.

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Ken

pjv

But if the LOAD is inductive Ken, don't forget the parallel back-emf diode to prevent all that smoke from escaping!

Cheers,

Peter (pjv)

Harry Stoner

Crystalionx said...
It looks like you've gotten all the help you need. I am interested, frankly, in applying the relay system to opening a door. What type of Mosfet and solenoid did you utilize - and for what purpose, if you don't mind.

Well there are lots of parts out there. The coils I am using are not standard off-the-shelf ones. The MOSFET I am using is ST 22NE10L, which is in a TO220 package and is rated at 22amps/100V max (not at the same time though). Maybe you need something smaller or larger or with different characteristics.

Thanks.

Harry