HIgh Current Mosfet Toggle noise issues?

Hello, I built a circuit from this link:

formatting link
entitled "High Current Mosfet toggle switch with debounced push button. It worked perfectly on the breadboard. When I installed it on a pcb board with an external switch (inside an enclosure) the circuit now powers on active. In other words, the IR510 Mosfet has power on the gate pin, which is not how I want it to work. Can anybody suggest what values I might change the10k/10uF on left side of circuit to, to resolve this?

The other issue I have is with the load. I connected a relay (k1,

24vdc coil)to the load terminals and had what I will call debounce, where the relay doesn't always close, but bounces. It may take several times to activate it. I put a diode across the coil and that resolved that problem. But my final load is 120vac 11amps so I am driving a much larger relay which we will call K2. K2 has 120vac coil and contacts. Now when K2 is in the circuit, I get the same issue. Can I put a diode across the ac portion of the coil to control this spike or bounce? If I remove K2, my problem is gone.

Thanks, Stan

Reply to
stan.portis
Loading thread data ...

This works:

formatting link

John

Reply to
John Larkin

I'd add, say, 1M from gate to Vss, so it comes up off. :-)

Cheers! Rich

Reply to
Rich Grise

But then it wouldn't toggle. The gate has to float.

John

Reply to
John Larkin

Do you have an RC snubber?

formatting link

Reply to
Hammy

This is just a guess.

The L time constant on the relay coil is causing a Hi Z long enough to allow the 10k RC via the diode to charge up enough for the power fet to obtain enough bias to turn on before the coil in the relay saturates.

Maybe a low value non inductive resistor across the coil will help. and also reduce the value of the .1 uf cap on the 3904 tranny.

--
"I\'m never wrong, once i thought i was, but was mistaken"
Real Programmers Do things like this.
http://webpages.charter.net/jamie_5
Reply to
Jamie

Try modifying the original circuit by adding a diode and a 470 uF cap, like this:

Original: +Vcc | +----------+ | | [10K] [Load] | | etc etc

Modified:

+Vcc | +-----+---|
Reply to
ehsjr

(Repost cause first didn't show up)

A very simple debouncer might be something like

formatting link

Its not the best and is asymmetric. R2 and R4 form a voltage divider when the switch is closed and because of the way it is set up you'll need R4 about < R2/10. It depends on the gate threshold voltage though and you can set up optimal components for different times. The time constants are about R2*C1 and R2*C1/10 so you can choose C1 from the second(max switch settling time < R2*C1/10)... noting that it forces the second time constant to be 10 times longer. This means if your switch needs more than 100ms then it will require 1s.

This circuit isn't for a momentary switch though but could be modified by adding a flip flip. Not sure if thats what your using. You can add a pulldown to the gate which also will slowly discharge the cap over time if no power is supplied.

The circuit could probably be improved but would require you to mess around with the components to get the optimal result. (I don't know your supply voltage or gate threshold voltage so I don't know what to choose for the resistors). You could use something like a zener or other things to do a better job but I tried to keep it a simple as possible.

You can offset the asymmetry somewhat and it will mainly just require a zener or a few diodes. I can go into the theory more if your interested.

Basically though tC = R2*C1 and, say, tD = R4*C1 where tC = charge time, tD = discharge time.

if tC = tD then R2*C1 = R4*C1 ==> R2 = R4. Problem here is that because R2 and R4 form a voltage divider you'll never discharge C1 enough in most cases except if you limit the max voltage on the gate to be about 1 voltage above its max. But then it might be somewhat unstable. (depending on how critical the application one can improve on the circuit to make it symmetric so there is no practical stability issues)

About your specific problem, I do not know what you mean by power to gate. Also, a diode should not stop bounding on a DC relay. Seems like something else might be going on there.

To main issues usually associated with relays are

formatting link

formatting link

Reply to
Jon Slaughter

I added two relays. Each time I added one it caused the circuit to not always latch the output closed, appearing to bounce the relay. When I added the diode across the coils it took that away. But now that I added the third and final relay which is 120vac coil I can't add a diode across the coil (or can I since it's ac).

Also Vcc is about 18volts dc.

When I mentioned power to the gate, by that I mean when there is voltage at the gate of my mosfet, the mosfet is turned on. When the gate is close to zero, the mosfet is turned off. Turned off is the condition I am looking for when I power this circuit up. I preferably would like to stick with this circuit since it's built and already in the box installed.

Thanks---

Reply to
stan.portis

No diode on AC because then you supply DC current to the load through the diode regardless of the state of the relay. You'll need a snubber over your load and relay.

I'm not sure how complex your circuit is but you might need to troubleshoot some and figure out where the problem is. Try using some led's and resistors for your loads to see if the load is the problem or if its the switching.

Theres several problems that could be happening. Your switches could still be bouncing because you didn't choose the time constant properly. The relays could be bouncing to due to some type of oscillation that the switching circuit is having. There could be some inductive kickback working its way to the DC relays causing them problems(and hence why the diodes work).

It sounds like the problem is probably inductive kickback though but it could be a combination. Try putting diodes on the DC relay's and an RC snubber on the AC relay. (Although why do you need an AC relay to switch something on when you can use a dc relay since you already have the DC setup)?

You realize that AC relays take AC current to switch it on and off but both AC and DC can switch AC current.

It sounds as if you have a DC relay controlling a DC relay which controls an AC relay which controls a motor.... but why not just control the motor with a DC relay directly?

Also, make sure you are using the right type of relay for the job,

formatting link

Also, from wiki,

"Alternatively a contact protection network, consisting of a capacitor and resistor in series, may absorb the surge. If the coil is designed to be energized with AC, a small copper ring can be crimped to the end of the solenoid. This "shading ring" creates a small out-of-phase current, which increases the minimum pull on the armature during the AC cycle."

Easiest way here is to get a switch that has the opposite normal state than the one your using.

You could exchange the BJT and 10k resistor positions and that should work as it inverts the voltage states to the mosfet but if you already put everything to gether it might be more work.

Jon

Reply to
Jon Slaughter

I did some testing. But first let me clarify what I am trying to do. I am trying to control a woodworking dust collection system that draws

120vac/11amps. Initially I didn't realize that it pulled more than 10amps and therefore I wound up with three relays. The first relay is a very small pc board dc relay (mounted on the pc board). The second is an external relay mounted in the case. It is a 10amp dc relay. The final relay is a 35amp ac controlled relay. (120vac coil). I found that if I disconnect the third relay everything works as planned with no issues. When the third relays coil is connected to my 10amp relay it causes the circuit to switch on and not switch off, or not switch on unless I try it 4-5 times. That third relay is screwing everything up. Funny enough about the rc snubber circuit, I don't even have the load connected to the third relay yet. I am still bench testing. But, that 35amp relay does make quite a thump when it pulls in. That is why I figured something needed to be done to the coil control voltage. Should I still connect an rc circuit? If so, what size values for the resistor and capacitor? I calculated 12uF and 1ohm at 11amps (22.5uF/.7ohms at 15amps).

The other idea that I think Jon was suggesting is to eliminate the second and third relay by purchasing an 18-24vdc controlled relay capable of switching approx 15amps ac on it's contacts. Is that correct?

Reply to
stan.portis

So your first DC relay is controlling the second DC relay which controls the AC relay? The second DC relay is actually letting in AC(most likely the mains?) into the AC relay?

What are you putting on the coil of the AC Relay and whats its coil ratings? I hope you are not putting DC on the AC Relay's coil?

This is the easier solution.

A quick search, for example, gives

formatting link
$file/G5CA_0607.pdf

(not recommending this one but just showing and example)

In this case you can control it with just one relay.

Reply to
Jon Slaughter

So your first DC relay is controlling the second DC relay which controls the

Yes. I was really trying to keep ac off of the pcb board.

What are you putting on the coil of the AC Relay and whats its coil ratings?

I'm putting 120vac on the coil, switched in and out by the second relay. The ac relay is a GE CR7CE 600vac max, 35amp relay. The Coil only says 120vac/60hz. It is a big one! I salvaged it off of something years ago. I looked for the spec sheet online but haven't run across it. It is definetely the one that causes my troubles. Without it things are perfect. I like your idea of the omron electronic relay. I will price that out.

Reply to
stan.portis

Ok, if you want to keep it off the pcb then use a secondary smaller dc relay or mosfet to control the coil of the main one.

Although your AC relay should work I suppose(although the fact that its AC triggered complicates things unnecessarily but if thats all you got it should work).

You might want to first test the relay if you can. Put the mains across coil and some load across the switch like a lightbulb or two.

Mains ---SW--+ +---+--- light --- Mains | | Coil Switch | | Mains -------+ +---+------------- Mains

where SW is some type of switch that you can cut the mains to the relay which should turn the light off. Of course it goes without saying that you need to be careful when working with the mains.

Jon

Reply to
Jon Slaughter

I ordered an electronic relay, 15amp 120vc contacts with 12vdc control voltage. Cost $2.83. May have it by tommorrow. Thanks for your help. I was also trying to use that large relay in case I ever upgrade to a larger 220vac dust collector. This particular contactor has three poles on it. Must just be too noisey?? Seems like there must be some way to isolate. (Opto-isolaters ?) But, I also like the idea of simplicity. I will be able to elimate all three relays with this new one. I might need to put a resistor on the Vss since its 18vdc and the new relay is 12dc. I posted a picture of the circuit

formatting link

If you know of a better way to post schematics let me know. Its kind of small.

Reply to
stan.portis

Yeah, it should simplify things a great deal. If you need to upgrade and the relay cannot handle it then just get a new one and its pretty much just plug and play.

Remember to put a snubber on that relay though or it will wear out.

If you put 18vdc on the new relay then it will draw an excess amount of current and probably over heat and burn up unless you have good ventalation. (it will probably work fine but shorten the life) A resistor would probably work fine for your purpose but its not ideal. If the coil draws I amps at V volts then that means it has an equivalent resistance of I/V ohms and increasing the voltage effectively decreases the equivalent resistance so you can make that up with a series resistor(make sure it can handle the power). (a real relay doesn't have a constant effective resistance since it depends on the current but it should be fine)

Also note that it would have probably been better to use a latching relay to simulate a switch so that excess power isn't wasted in holding the relay closed... I think that what your doing but its probably not much power(few watts maybe) so no big deal. (using a latching relay would probably require a new circuit though so...)

I use

formatting link

when I need it but there are many others too(search google for free image hosting)

Jon

Reply to
Jon Slaughter

The solid-state relay worked perfect! I received it Friday in the mail and installed it right away. It eliminated all of the relays and solved the noise issue and the current capacity issue. The coil is driven by 12volts dc and the contacts are rated at 120v/15amps. My circuit pulls 11amps. I connected it to my dust collector and it works just as planned. Thanks to everyone's contribution ! That solid state relay costs $2.83.

Stan

Reply to
stan.portis

ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.