Fast Snubber for Solenoid Valve

SgtWookie

Joined Jul 17, 2007
22,230
Mind if I crash into this thread? :D
Please do! :D

The company I work for happens to manufacturer snubber networks for several major relay manufacturers. We do it for them as we can handle bare die and produce a simple string of parts with special wire on the ends. "Special" here means solid cupron bus wire which can both be soldered and welded.
Cool. :cool:

I've personally designed relay timers that mount inside the relay housing and they all use this same snubbing scheme.

These snubbers consist of a 1,000 volt diode in series with a 36V zener. It's not a fast diode either so think of it as a 1N4007. We once spent some time trying to figure out why there is a zener in there and a associate of mine nailed it: it gives a fairly large voltage for the coil to "discharge" into. (I use "discharge" here as I don't know a better word to say when the energy stored in the coil is being released.)
Yep - that's exactly what the simulations I posted in reply #3 show. The 1st image compares using just a diode with a diode & resistor in series across the coil. The voltage scales on the left show the higher voltage on the lower plot, and the current scales on the right show more rapid decay of current with the resistor.

The 2nd and 3rd images are using Zeners (progressively higher in voltage) in series with diodes instead of resistors; the plots illustrate how much more quickly the current stops flowing with a higher voltage Zener.

I've never been much of a spice user, I typically use excel as my spice as it allows me to write equations for predictions and play with values.
Well, Spice is only as good as the data it's given, and the models are based on varying levels of accuracy. It's really easy to leave out some critical component(s), and then real-world results are very different from the model. As I mentioned earlier, I've had to simply make guesses at some of the parameters I fed to it; so I could be off by quite a bit.

The coil of a relay may be modeled as an inductor and resistor in series. You can measure the resistance with most ohmmeters but the inductance takes a little more work (more later).
Yep - well, the inductor model I used has fields for inductance as well as resistance and parasitic capacitance; so it has provisions for that. Since our OP said 500mA @ 12v, I gave the inductor a resistance of 24 Ohms. I assumed an inductance of 0.2H; that would be with the solenoid core centered in the winding. That's simply a guess, of course; the inductance could be a fair bit higher or lower. I used a parasitic capacitance of 100pF ; that was basically a random shot in the dark - I really have nothing to base that on, but it was better than using zero. If the capacitance were considerably higher, it would increase the rise time of the spike.

Consider this little circuit that I would recommend using:


We have a MOSFET driven directly off an Arduino's output pin. There are some very good MOSFETs that are called "logic level" as they reach very good performance with just 5V on the gate, allowing such a direct drive scheme. D1 and D2 are the subber I am suggesting. The GREEN current is when the FET is on, and the RED current is during the turn off of the FET when the inductor current must be dumped.


That's just about the circuit I had in the lower schematics of the 2nd and 3rd simulation; just the diode and Zener were swapped - I was using a TIP121 Darlington instead to approximate the ULN2xxx drivers that were being discussed.

As the coil is truly an inductor with inductance L, the current thru it builds up as an exponential governed by R and L. As these are fixed inside the relay itself you are stuck with whatever the relay has. If you want a faster turn on you need to pick another relay.
Actually, increasing the initial voltage will cause the current to build up proportionately faster; but current limiting would need to be used to prevent the winding from burning up.

As for the turn off, remember we have an inductor with a current flowing thru it, and the current thru an inductor cannot change instantaneously, it must flow from one value to another. That's why you get an inductive kick: the voltage has no constraint, so if you interrupt the current path very fast the voltage will grow to the point where that same current can flow somewhere. Typically it breaks down the transistor and destroys it. I have seen over 1,000 volts on an unsnubbed relay.
We've been talking about that as well. I was able to get ~7kv out of the modeled inductor, but that's with an instantaneous turn-off (ideal switch), and probably too little parasitic capacitance in the inductor itself.

The voltage across and the current thru an inductor is generally given by:

V = L ΔI / Δt where ΔI and Δt are the change in current and change in time.

Now when V is fixed there is a nice linear relationship (meaning we can ignore all the calculus and use this simple equation) and we get a time to discharge the inductor as:

Δt = L I / V where I = ΔI = current flowing just before we open the switch

Obviously, the larger the V the faster we discharge the inductor.

A resistor may be substituted for the zener but will not be as fast as the voltage will decrease as the inductor current also decreases so it takes 5 L/R time constants.
Yep, that's basically what was shown in the simulations.

The MOSFET drain will see a rise in voltage as it turns off, as the Zener and diode voltages add to the supply voltage. Here we get about a 36V + 1V rise over the 12V supply for a 49V spike on the MOSFET. I like to derate things at least 50%, so I would pick a 100V MOSFET for this. Also, as the typical current is 500mA I would pick a device capable of at least 1A.
I agree with all of that, too. Have to also keep in mind tolerances for the solenoid windings, resistors, etc. - so the de-rating keeps our OP out of trouble (meaning smoking parts). The snub could occur more quickly if the diode/resistor or Zener/diode went to ground instead of +12v (because you wouldn't have to subtract the supply voltage to the Zener rating or resistor calculation), but there would be an efficiency hit.

The IRLD110PBF is one such device, you may want to search out others. I found it using DigiKeys search engine for MOSFET, then single, then Logic Level gate, then 100V, then 1A, then finally Thru Hole.
I really like those IRLD's being in a 4-pin DIP, but the cost is going to be a big hit on this project. Our OP wants ~210 solenoids being controlled by his uC, that would cost him ~$260 just for the MOSFETs alone! :eek:
Mouser carries the IRLD120 (100v, 1.3A) for less:
http://www.mouser.com/ProductDetail...LD120PBF/?qs=sGAEpiMZZMvMXbh32ZmHAO5BiyVBSKlA
but that would still be ~$207 for 250 of 'em.

If he went with the ULN2065B from Avnet Express, that would be $76.45+shipping for 55 of them (enough for 220 solenoids, but more spares would be a good idea), and the parts count would decrease significantly (no base/gate resistor, one driver IC per 4 solenoids). The ULN2065B has built-in diodes which may or may not be fast enough; the performance of the ULN2803A's or ULN2003A's that he already has could give an indication on how fast they are, or if they could work.

If those were used, a single Zener from the COM terminal to +12v might take care of it, if the Zener were rated for enough power and fast enough. Since multiple channels dump through the COM terminal, a simple resistor probably won't work well, as you'd have to plan for multiple solenoids (up to 4) turning off simultaneously.

I'm trying to find some kind of a reasonable trade-off where he gets the performance needed, with decent reliability, without increasing parts count too much, at a cost that's still within reason.

Now given all that here's a simple way to measure the inductance of the coil if you wish: Given such a circuit as above, measure the width and height of the voltage spike (voltage is the portion above the 12V) at the FET drain. Then you can solve for L by:

L = V Δt / ΔI

You would need L if you wish to use a resistor only as a snubber. You pick the R by first choosing the maximum spike you can stand. If we stay with the 37V we already have:

R = E / I = 37V / 500mA = 74 Ohms.

The time constant for an inductive circuit is L/R and needs 5 times to decay, so the turn off time is then:

5 * L / R where L is the inductance found just above.
That's all good - the rise time is also something of interest, as that will help determine what the parasitic capacitance is.
 
Last edited:

SgtWookie

Joined Jul 17, 2007
22,230
SgtWookie:
You are right! Nice detail you have realized about the IRF MOSFETS being controlled from a microcontroller. I don’t know how but a guy made the IRF540´s work directly form a 74HC595 microcontroller. (See http://www.electro-tech-online.com/microcontrollers/114468-waterfall-printer-simulator-2.html post #15. Here you can see some videos of the cascade “working” http://www.electro-tech-online.com/...deas-reviews/114242-waterfall-printer-11.html ).
The IRF540 is not specified for logic-level operation. The Vgs(th) (the threshold) where Id=250uA is anywhere from 2v to 4v; and 250uA is "just barely" conducting. I suggest that you will have a great deal of grief if you attempt to use components outside of their ratings. Now there IS a plot in the IRF datasheet where typical values with Vgs=4.5v and Id=10 are shown - but these are not guarantees. You want to make certain that things are going to work.

It is sad that the OP in that thread gave up on their project. I firmly believe that their main problem is caused by the lengths of tubing on the output of the valves. You must not make that mistake; the shorter the distance from the physical valve to the water exit, the more satisfactory the flow control will be. The shape and alignment of the water exit aperture is also critical. If there is even a minor burr, imperfection, etc. on the water exit aperture, the effect will be spoiled.

I am not a member over there; you might inform the original poster of that thread my suggestion of removing the solenoid valve exit tubes and trying it again. It may be messy the way they have it rigged, but it should work much better. They may also be having problems with their snubbing of the solenoids.

Tell me if my solution/idea is correct: I already have plenty of ULN2803 (logic). I can connect the ULN inputs to the 74HC595 microcontroller and the outputs to the IRF710. This will help to convert the “logic” to an analog output that the IRF MOSFETS needs (and also 12V instead of the 5V that the uC provides). This will also fix the maximum I/O limits of the uC when all solenoids are ON.
Just to keep things straight, the 74HC595 is a serial-in parallel-out 8-bit shift register w/latch & clear outputs, not a uC. :)

I suppose something like that could be done. However, your parts count is climbing! The more complex you make this project, the more likely that things are going to break.

Keep in mind that the ULN2xxx drivers can only SINK current, they cannot SOURCE current. So, you would need to add pull-up resistors to source current to all of the MOSFET gates. It will also invert all of the logic, and make the solenoids turn ON by default. In order to turn them all off, you would have to clock a stream of 1 bits down the '595's, and latch the outputs to pull all of the gates low.

I don´t have an oscilloscope but I can have access to one in the labs of the company I work. The same day I measure the R, L and C of the solenoid, I will analyze the peak voltage of the solenoid when disconnected with no snubber. This will help a lot to choose a correct snubber.
Yes, it will. You test 5% to 10% of the solenoids that you have to get a decent statistical sample. Test them with the actuator present within the coil; as the inductance will be higher that way.

I'm very inclined to trust Roman and Ernie's judgements/statements - but this is your project, and if errors are made it could be quite costly to you. Gathering empirical data via testing actual components that are going to be used will minimize the chances for mistakes, and also help a great deal in building more accurate models to simulate with.
 

Thread Starter

Tomas_2

Joined Oct 8, 2011
15
If you think that 74HC595 + ULN2xxx + IRFXXX are not the best choice, feel free to suggest ANY alternative possible. Some post ago you mentioned to use 74HC595 + ULN2065, this will reduce the chain in one element. I must buy many new things so now is the best moment to hear suggestions.

I would like to buy something reliable, the less expensive possible and easy to find. I liked MOSFETs because they can handle high voltages without damaging, so the solenoid won’t need a big snubber, ergo closing faster.

Tomas
 

ErnieM

Joined Apr 24, 2011
8,377
When all else fails, go back and re-read the original post.

I just did so, and followed the link to the Osaka Station clock. Very cool project!

I ran some numbers guessing that a kid's head is 6" tall so the tiles on the back wall are 12". From some static pictures I've seen it looks like the smallest burst is 6" long as it emerges from the top. It seems that gravity is the sole source of velocity as I'm guessing a 7' height of fall.

Since for an object starting at rest:

d = 1/2 a t^2
where:
d == distance
a == acceleration of gravity
t == time

solve for t:

t = sqr( 2 d / a) )

for a 7' fall:

t = sqr ( 2 * 7' / 32') = sqr(.4375) = .66 seconds

And that is about all the squirts take to travel from top to bottom. This implies the squirts are not pushing the liquid out with some initial speed, the water is simply "falling."

OK, so if we assume the water is simply falling, to make a 6" droplet we need to turn this thing off when the bottom of the drop has fallen 6" ( = .5'), so:

t = sqr ( 2 * .5' / 32') = sqr(.03125) = .176 seconds

If I was doing this project I would start with the solenoids and jjst drive one of them to see how fast I could turn it on and off to see if I could get the very short squirts this thing needs.

Note the smaller you make the unit the shorter the time the squirts need to be on for a given shape. If you instead want a 1" droplet (1/12') the time is only:

t = sqr ( 2 * 1/12' / 32') = sqr(.0052) = .072 seconds

Now a given "slow" solenoid may well still be useable if you drive it in a special way: You first overdrive it to get the current to build up fast, then drop back to a lower sustaining drive, then turn off thru a large voltage to get the quickest turn off.

Please keep us updated on this very cool project!
 

SgtWookie

Joined Jul 17, 2007
22,230
Ernie,
The "overdrive" you're suggesting is used all the time with stepper motor drivers; if a motor is rated for 10v @ 1A, they might use 40v to 100v to get the current moving quickly, and a "chopper driver" to limit the current to what it's rated for.

I've seen solenoid/relay drivers that do something similar, but those are usually designed for maximum economy without sacrificing the initial actuation time.
 

Thread Starter

Tomas_2

Joined Oct 8, 2011
15
Hello!

When I have a chance, I will measure with an oscilloscope the peak voltage. If it is below 350-370 (400 supports the transistor), will I need a snubber?

ErnieM:
d = 1/2 a t^2 only behaves like that when no friction is in game (ideal conditions). In real life there is something called “terminal velocity”, when gravity force is equal to the friction the sums of the forces is cero, ergo acceleration cero, V=cte. Actually the real speed of a drop is between 4.3-4.7 m/s (in the experiments I’ve done, depends on many factors) BUT at the moment that falls, V is quite slow because pressure of water is slightly higher than atmospheric pressure. Lots and lots of tests remain so for the moment I haven’t found the optimum pressure, diameter and type of tip to use and many others I have work around.

Now that I read about overdrive voltage, I might be a really good idea! What crosses my mind at this moment is that the length of the peak must be short enough for not to interfere with the closing of the solenoid/valve. Just to say a value, the peak must not exeed 4-5 ms. Just another idea is to use a lower value than 12V and excite the solenoid with a peak and maintain it with 8-10 volts…

Is there an easy way of creating an overdrive short enough no to interfere the closing?


Thank you people! You can’t imagine how much you are helping me.

Best regards,
Tomas
 
Last edited:

SgtWookie

Joined Jul 17, 2007
22,230
Tomas,
Do the solenoids have an AC rating as well as DC?
This whole snubbing situation could possibly be avoided if the solenoids were driven using AC and TRIACs; the TRIACs will automatically stop conducting during the zero-volt crossover when they are not receiving current input. In a 60Hz system, there are 120 zero-volt crossings/second, with 50Hz, 100 crossings/second.
 

THE_RB

Joined Feb 11, 2008
5,438
Hello!

Thanks THE_RB for the 2SC2335 data, I’ll have it in consideration (BTW, already ordered 20x of them for testing purposes!). When I have a chance, I will measure with an oscilloscope the peak voltage. If it is below 350-370 (400 supports the transistor), will I need a snubber?
...
I would be very surprised if you see 300v from a solenoid run on 12v. But yes, that would indicate a snubber or damping resistor or some system is needed to limit the peak, even with a proven rugged 400v Vce transistor I would not expose it to continual 300v peaks if there is no need.

With a 12v supply and a simple resistor across the solenoid you should get it well under 100v, which would be a nice operating peak voltage for that transistor.

You originally said 150 solenoids, so you need a large safety margin built in. the last thing you want is failing diodes or transistors on a big expensive display, where breakdown costs will be high. That was part of the reason for my suggestion to snub using just a resistor, not only does it keep costs low but a resistor is the most reliable component you can use to damp that back-EMF spike.
 

ErnieM

Joined Apr 24, 2011
8,377
Tomas_2: Those are traditional "back of the envelope" calculations only intended to try to define the limits of the problem: turning a value on for a second is lots easier then turning one on for .01 seconds. The second case you need very fast on and off times.

One simple way to "kick" these on is to have a double drive from a higher supply voltage:



I've left the values out for now just to show the concept, but say you have a 24V supply and a 12V zener on Q2: the solenoid is initially turned on with 24V by turning on BOTH Q1 and Q2. The current builds up faster then with 12V, and after some time (you would need to measure this) when it is built up you turn Q1 off and allow Q2 to supply it with 12V to sustain the current.

To turn it off you then just turn off Q2.

For a faster turn off, bigger is better. Every time you double the snubber voltage you half the turn off time for the current, though there is probably still a fixed mechanical time limit. If you have a 400V switching device I'd advise using half that maximum so you make a 200V snubber with a 200V zener.

A zener, due to it's fixed voltage characteristic will always be faster then just a resistor for a given discharge voltage.

I would be very surprised if you see 300v from a solenoid run on 12v.
I wouldn't, I already have seen that and more. Try using a mechanical switch, a low capacitance probe and a fast scope.
 

Thread Starter

Tomas_2

Joined Oct 8, 2011
15
Tomas,
Do the solenoids have an AC rating as well as DC?
This whole snubbing situation could possibly be avoided if the solenoids were driven using AC and TRIACs; the TRIACs will automatically stop conducting during the zero-volt crossover when they are not receiving current input. In a 60Hz system, there are 120 zero-volt crossings/second, with 50Hz, 100 crossings/second.
The type of solenoids I bought are DC. My reasons:
-DC has a constant time to open/close (e.g. 12 ms) while AC might be faster in some cases but it has some variability (e.g. 9-13 ms). I need all valves to open/close at the same speed, if not the figures won’t have good geometry. Repetitive actions are very important in this project.
-DC valves are safer (generally DC is rated 12-36V while AC is 110/220V).
-DC solenoids last longer. If the mechanism blocks open-positioned, in a DC it will just get hot. The AC will burn!

The pros of AC are known: Snubber isn´t needed and lot cheaper! No need to buy DC power source, caps, etc, etc.

I would be very surprised if you see 300v from a solenoid run on 12v. But yes, that would indicate a snubber or damping resistor or some system is needed to limit the peak, even with a proven rugged 400v Vce transistor I would not expose it to continual 300v peaks if there is no need.

With a 12v supply and a simple resistor across the solenoid you should get it well under 100v, which would be a nice operating peak voltage for that transistor.

You originally said 150 solenoids, so you need a large safety margin built in. the last thing you want is failing diodes or transistors on a big expensive display, where breakdown costs will be high. That was part of the reason for my suggestion to snub using just a resistor, not only does it keep costs low but a resistor is the most reliable component you can use to damp that back-EMF spike.
OK, I´ll have this into consideration. My idea is to have a reliable electronics and to leave the weak points to the mechanical parts. When testing, I will try some resistors and see how they perform in the retard of the solenoid. From what I have read, a diode + zener is the fastester but a resistor is a very good alternative also. The idea is to use a safe peak voltage as long as it doesn´t affects much the retard of the valve. In a beginning, the idea is to control 200 LEDS instead of valves. Then 2 valves and if it works then 20. Finally 200 valves. If it works very good, then analyse if 400/500 is viable. Remember that each valve is one line and the open + close time is the pixel length.

Tomas_2: Those are traditional "back of the envelope" calculations only intended to try to define the limits of the problem: turning a value on for a second is lots easier then turning one on for .01 seconds. The second case you need very fast on and off times.

One simple way to "kick" these on is to have a double drive from a higher supply voltage:

I've left the values out for now just to show the concept, but say you have a 24V supply and a 12V zener on Q2: the solenoid is initially turned on with 24V by turning on BOTH Q1 and Q2. The current builds up faster then with 12V, and after some time (you would need to measure this) when it is built up you turn Q1 off and allow Q2 to supply it with 12V to sustain the current.

To turn it off you then just turn off Q2.
I really liked this idea but it has a cons: I must double the outputs of my “PLC”. Also, create very narrow relation between them as 2 outputs control the same output VERY dynamically. I see it viable but not that simple couse I will need to double everything!
One question, if I understood correctly, the zener diode works as voltage regulator (in your example from 24 to 12V). ¿Does this zenner supports 0.5A@12V from 24V of continuous dutty? As I said, my electronics are quite basics!

Once again, I´m very pleased with your collaborations.
See you,
Tomas
 

ifixit

Joined Nov 20, 2008
652
Hi Tomas_2,

The attached circuit idea show how a resistor and cap in parallel can be used to apply more current to the solenoid for 10mS or so in order to speed up activation time. How effective this is depends on the design of the solenoid. You would have to test it.

IMO a simple cheap snubber does as well as zener/diode combos to limit the EMF spike. Use a driver device with a max Vce of twice what the spike voltage is limited to. Margin for error.

Solenoid specs would be nice.

Regards,
Ifixit
 

Attachments

All you need to do is use an "avalanche rated" power FET, such as the IRF520. The person on the other thread was using an IRF 540, but that's bigger than you need.

Have a look at the first page of the data sheet: http://www.irf.com/product-info/datasheets/data/irf520n.pdf

Notice the symbol for the body diode; it is the symbol of a zener diode. That family of FETs can be operated with the body diode in breakdown without damage.

You see on the data sheet that the avalanche current can be 5.7 amps, which is much more than you need. The avalanche energy is also specified. If you are only pulsing your solenoids a maximum of 10 times a second, but typically less than that, you can probably use the single pulse avalanche energy rating if the FET is cold at the start of each instance of turning off the solenoid current. If you determine the inductance of your solenoids, you can calculate the avalanche energy dumped in the FET each time you switch off your load.

You can expect that with a 100 volt rated FET of this family, the breakdown voltage of the body diode will be about 120 volts.

I have used this family of FETs with no problems in a design that used the body diode to clamp the voltage when turning off an inductive load.

You will have to get more than 5 volts of gate drive to make sure the FET is fully turned on, but you will avoid the expense and hassle of separate components for a snubber.
 
Last edited:
Something that may surprise you is that the turn-on and turn-off times of 1N400x and 1N540x series diodes is not specified (!)
At least one manufacturer has specified turn-on time for a 1N400x series diode; see about 1/5 of the way down the page here: http://www.cliftonlaboratories.com/diode_turn-on_time.htm

That page is an investigation of whether a 1N400X series diode has a short enough turn-on time to be safely used as a snubber for inductive loads.

The specified turn-on time for a 1N4004 is about 220 nS. What I have found in practice is that the distributed capacitance of typical electro-mechanical inductive loads such as relays, solenoids, etc., is sufficient to slow the rise time at turnoff so much that a 220 nS turn-on time is plenty fast enough.

Finally, as an alternative, consider the 1N493x family of fast recovery rectifiers; they are 1 amp diodes, similar in ratings to the 1N400x family, similar in cost and availability. This family has the same 30 amp half-cycle surge rating which is much better than a 1N4148 signal diode. The1N493x diodes have a maximum reverse recovery time of 200 nS; the forward recovery time is not specified, but it is surely no worse than 200 nS.
 

Thread Starter

Tomas_2

Joined Oct 8, 2011
15
Thanks for the answers!

For the time being, I won’t be adding an “overvolt” system. If necessary, I will implement it in a future. It will help a lot the open/close time but it is not so simple and lots of components must be incorporated (remember its 200x of each thing).

The Electrician:
As SgtWookie said, IRFXXX are not designed for logic operation (they are controlled by a 74HC595). It might work (already bought some of them for testing) but I’m not shure if they will work in “nominal” values (the 5 volts of gate drive to make the FET fully turned on). THE_RB recommended me to buy some 2SC2335 as they should work OK with logics. Do you have experience with IRFXXX and 74HC595? If yes, I would be great to hear your experiences.

I bought some diodes for testing:
UF4007 1000V 1A Ultra-Fast Recovery diode
1N4937 1A 600V FAST RECOVERY DIODE
1N4007 DO-41 IN4007 1A 1000V Rectifie Diodes
1N4004 DO-41 IN4004 1A 400V Rectifie Diodes
1N5383B 150V 5W Zener Diodes
1N4752A 33V 1W Zener Diode
1N4757A 51V 1W Zener Diode

Hope the 1N4004/4007 works well as the other fast diodes, they are much cheaper and easy to find! Will tests them and compare with others.

Best regards,
Tomas
 

John P

Joined Oct 14, 2008
2,025
One way to make the "double drive" work without extra outputs from the PLC would be to have one transistor that just goes on when the output is on, and another (controlling the higher voltage) where the gate is driven via a capacitor with a pulldown resistor, so it has a predictable decay time and will give a pulse of high voltage for every turn-on. It adds to the component count though, obviously.

If you have two voltages, you can run the freewheeling current back to the higher supply, for a simple rapid turn-off.
 

John P

Joined Oct 14, 2008
2,025
Let me say before anyone else does--having tried to sketch it out, I don't think the idea I had a few minutes ago is workable. It seems as if the powering and snubbing parts of the circuit aren't compatible, and there would have to be too many power transistors. Back to doodling on the margin of the Sunday paper. How do the existing digital fountains do this, anyway?
 
The Electrician:
As SgtWookie said, IRFXXX are not designed for logic operation (they are controlled by a 74HC595). It might work (already bought some of them for testing) but I’m not shure if they will work in “nominal” values (the 5 volts of gate drive to make the FET fully turned on).
The spec sheet for the 74HC595 I found on the web show the input and output parameters for operation on 6 volts (maximum is 7 volts). If you were to operate your 74HC595s on 6 volts, that would give you an extra volt of drive to the IRF520. The FET wouldn't be fully saturated but since you only need 300 mA for the solenoids, with 6 volts of gate drive you would have more than enough. And, you won't need any snubbers at all.
 

Thread Starter

Tomas_2

Joined Oct 8, 2011
15
Hey folks!

I have been working in the prototype of the "Water Printer". The electronics works like a charm with 0.1 ms precision. Valves open/close time is 6ms, fast enough for me.

Finally I use common 4004 diodes, I tested a lot of zenner diodes and they are pretty much the same (5 ms open/close time), as 4004 are cheaper, I used these.

Here you can watch some videos of the first beta, a lot of work remains....

Any suggestions are welcomed!

http://www.youtube.com/watch?v=OoJ6TRb85JQ

http://www.youtube.com/watch?v=O-kkP-_2Kdo


Best regards,
Tomas
 

crutschow

Joined Mar 14, 2008
34,281
At least one manufacturer has specified turn-on time for a 1N400x series diode; see about 1/5 of the way down the page here: http://www.cliftonlaboratories.com/diode_turn-on_time.htm

That page is an investigation of whether a 1N400X series diode has a short enough turn-on time to be safely used as a snubber for inductive loads.

....................
The turn-on time of most silicon diodes is very fast and not that different between normal and fast recovery types since diodes don't have significant forward recovery time, only reverse recovery (the time for the carriers to dissipate). If you look at the waveform for the 1N4005, you will see that the turn-on is much faster than 220ns. The 200ns he mentions is for the slight overshoot after the turn-on which likely is fixture related since diodes don't have overshoot by themselves.

So there's no reason to buy a "fast" diode if you are only concerned about turn-on. A standard recovery diode is just as fast for that.
 
Top