Printer Friendly Forum Posts - pulse relay common problem?

Hello,

I spent several days looking for a timer/pulse relay and came here where I found I could make one. Thrilled, I gladly bought up some stuff from an electronic store and assembled two different relay circuits. My question deals with only one of them, the Constant > Momentary. Before I install it, I want to test it to make sure I'm not going to blow anything up within my car, however I'm having issues.

Tyco 20/30A 12V Relay, 3A Diode, 10K Resister 1/2w, 1.5uF Cap.

When I give the coil 12v, while monitoring 87 and 30, I don't get any activity between them, nor does the relay click. I realize the idea is a 1/2-3/4sec contact pulse but I don't even get that. If I check 85 & 86 while hot, 86 shows .1v, while checking both sides of 85 (cap/res) it shows 12v. Tells me the relay should have triped but I don't get anything.

Any thoughts/comments welcome. Small photo below to show how its wired.

Thanks!

posted_image

Nobody huh? Not even to call me an idiot for using the wrong cap? Or suggest searching (which i did btw with mixed results)?

i'd of rather been called an idiot then left to ponder on this one for a few days.. oh well...

Okay, you're an idiot for using the wrong size cap - posted_image

But not really, think you just missed the "K" if you were using this diagram as your reference.
Instead of 1.5uF, you probably want to try 1500uF, which is a much larger size cap than you're trying to use.

I normally would suggest adding a transistor driver stage to control the timing, but it seems that most people usually like to avoid that stuff, not being comfortable with working with electronics. That being the case, you can just try getting a larger cap to wire in, and it should work.
One caution if you try the brute force method; when your switched 12V signal turns off, the cap may be shorted to ground by your signal, so make sure it can handle the current surge.
You should check the driving circuit to see if it tries to drive it to ground when the signal is turned off, and if the spike in current could damage anything.

Finally, someone calls me what I am... An idiot.. posted_image

I was trying to use that diagram for this application. I also made up the Radio on until Door trigger one as well, but no fancy anything on that one. But thats another topic.. ;-P

My intent with this pulse is to trip the traction control switch in my car. It defaults to being on with a momentary switch on the dash to turn it off. I want it Auto-Off on start and figured this would work. The switch shows 0v off, 10v AC On, 12v Run. Wires are tiny tho, maybe 22ga. What I thought was tapping into the +V line for 86 & 87 since it only shows 12v on Run and using 30 to the Negative wire of the switch, with 85 to Ground somewhere. Tho after hearing your note with brute force (which that circuit by default is i believe), I'm wondering if I shouldn't rethink how it's wired. One thing I do not want to do is blow out any part of the switch or the tcs system.

So now I'm thinking about the diagram I saw which is the same Con > Mon circuit, only in reverse polarity with a Neg pulse. Only problem there may be that the switch/line is wired for +12v short so a bridging the pos line wouldn't do anything.

I'd be interested to hear your transistor driver stage... =) I have no problems with electronics, soldering, etc.. Been in the computer industry for 18+yrs.. I'm no slouch, but apparently have trouble reading here and there.. lol al-la 1.5 vs 1.5K posted_image

Thanks!

Adding a transistor stage to control the timing allows you to use a much smaller capacitor to get longer pulse times, and is easier on the driving circuit, because it won't suddenly discharge a supersize cap.

Something like this can be used, where a transistor is used to switch the relay coil on and off, and the delay timing components won't be affected by the impedance of the relay coil, as it would be when using a direct drive kind of setup.
The cap value here can be made smaller to decrease the pulse output time. If a smaller resistor is used, it'll also decrease the pulse time. Just as a ballpark number, multiply the resistance x capacitance to get the approximate pulse time. The values here would give in the range of 5 sec.

The BS170 MOSFET shown is sensitive to ESD, so it has to be handled more carefully than regular transistors when installing them. A regular NPN transistor can also be used, but I just like these because they're voltage controlled, and effectively don't need any current to drive them.

Reading your description, I'm wondering about the "10V AC On" you mentioned. That's not AC as in alternating current, is it?
The constant to momentary circuit is meant for a switched DC signal, not an alternating current (AC) signal; the thing would just switch on and stay on if AC was being supplied as an input- not what you'd want.

Maybe if you could sketch out what you're working with it'd make it clearer what's needed, whether a positive trigger or negative one.
Not really sure what's involved from the description-

Is the timing of that +V signal you were planning to use the same as the ignition switch (comes on as you turn the ignition to run?) If this relay were to simulate the switch closure at the same time as the voltage came up, would it still work to switch your TCS off, or is some time delay needed? Some of those details need to be worked out before making up something to do it.

Thanks Dualsport!

I dunno how you make your schematics, but damn man! [8D]

When I said 10v AC, I meant Accessory possition of the ign switch. It's a car, so everything is DC, thankfully.. lol

Unless there's a program out there that would aid in making a legible diagram, as I'm about as much of a freehand artist as a 8yr old.. :/ Wanna know how to use a computer w/o a keyboard or how to program, I'm down for, drawing, nu-uh...

I'll try to describe again, hopefully better. well, i'll try anyway...

TCS Switch has 4 pins, 2 for a dash light, 2 for the switch. Tossing the light circuit away, I have 2 wires. 1 Pos, 1 Neg. Ign Swtich: Off = 0v Pos, Accessory = 10v Pos, Run = 12.5v Pos.

All I want to do is automatically connect those two wires momentarily on ignition start. I don't need any voltage or ground to either wire if I can just connect them for 0.5sec or so. If there's a way to energize the relay for that 0.5sec and then shutting down or halting contact until power is removed, I think that would work. If I could connect 30 and 87 to + and - side of the dash switch, and use some form of pulse on the coil side I should be fine. I just don't want any forward voltage going from the coil to term 30 or 87.

Maybe I'm asking too much, or maybe I'm not just understanding relays correctly. I've read confliting info on relays so maybe thats part of the problem.

I hope I've described with more detail or info that will help. I don't think what I want should be that complex, but maybe it is..

Thanks again dual!

Sounds like you should be okay with the above circuit then-
The relay contacts are going to just momentarily connect between 87 and 30, so wiring them across the two contacts of your switch should simulate your pushing the button and releasing, until the next time you turn power off and on again.
You should be able to use your 1.5uF cap now, to get a shorter pulse time. The 10k resistor you have is probably too small, so you'll have to get another with a higher value, to lengthen the pulse. You can experiment with different values to get the pulse time you want, though it doesn't sound like it's too critical in this situation.
Bigger value of C and R will give a longer pulse, and vice versa.

BS170's can be had cheaply from mouser.com; if you don't want to stock up, I bought a bunch at the quantity prices-
posted_image

BTW did you check if it does what you want when you hold the switch closed at the same time you power on? If there has to be some delay before you push the switch, you should account for that-

Sweet! So tonights project is to make the sob.. =D

Thanks again for all your help! I may order a few of those mosfets from mouser to see how they work, or from you if yer willing. =D It would be sweet to not have a huge cap to hide/shove somewhere but that'll make a nice secondary project in the comming weeks.. ;-)

Out of curriosity, how could I make the relay/circuit for the tcs like we've been talking about, however, use the +12v side of the switch for my power source as well, with the catch being zero feedback thru the 12v line? Would adding a second diode on the +12v line before the relay serve the purpose correctly?

Honestly I want to tap the least amount of wires as I can get away with which is why I ask.

Again, Thank You!

The only power being drawn will be from the relay, so if you're able to confirm that the +12V switch input can supply the current, you could use that.

Adding a diode couldn't hurt, to block possible spikes from the relay going back to the supply line. If that line is just an ignition switched input, it probably wouldn't be a problem either way.

The BS170's are pretty neat to play with; effectively they're tiny solid state relays that switch to ground just from a low trigger voltage, about 3V. You can switch them on and off just touching the gate input with your finger, because they work on voltage, not current. I've got a bunch stocked up, so if you don't want to buy a hundred of them (though even that is so inexpensive you might want to get a bunch to use for future projects), you can get a few from my stash- They're under a dime a piece in quantity, so postage costs more than the part - ;-)

Thanks Dual. So over the last couple days, I've soldered, spent parts, and come up empty.. My last itteration had a 2200uF cap with a 15K resisitor. It wires up fine and in therory should work fine, in reality it doesn't. I've come to realize the tcs system must disable user input while it self-tests which I didn't think it did. So what I need to figure out now is something that will delay the pulse output for xx seconds on power-on. Something tells me I'm gonna end up with two relays and alot more wires.. =(

So what hell am I in for now...

I assume you're using the original direct drive setup with the big cap then?

I mentioned earlier that you need to check if the switch can be turned on at the same time power comes up, or if a delay was needed; is that what you're coming up against now?

Try to find out exactly what kind of timing requirements you need, and you may be able to put together something to do it. Using some solid state components will make it easier to do, so you may have to look into doing some electronics work.

Yep, I'm using the direct drive huge cap brute force method.. =P

Well, my original thought on the switch was the tcs doesn't lockout anything or self test since its part of the abs system, but appearently I'm wrong. Another problem is the relay will energize at 10v instead of 12v. I didn't count on that one happening either. So what I have now is something that will work, but only after the system self-tests itself. I'm gonna say a 10-15sec initial delay would be about right. How to get that is where I'm at now... To say I'm a bit baffled would be accurate.

It's easy enough to add the delay if you use the transistor drive- I figured you'd need the delay after turn on, that's why I asked if you confirmed it works without the delay first-

I can draw it up if you want to try-

Please!! I'd like to get my dash back together enough to drive it.. posted_image

Also, sorry about me not checking first on the delay on.. But I honestly didn't think it would be a problem while I was testing the plug with a dmm. =(

Well I did ask twice if it would work without a delay- posted_image

Shouldn't be hard to add a delay to the trigger; pretty sure it's been posted previously, but I'll draw it out if I can't find the post. If you don't feel comfortable with electronic stuff, there are commercial products that will do it for you too. Probably overkill for the application though, cause they're chock full of features you won't be using.

yea, i'm knockin my head on the wall for that one..

i have no problem working with electronics and such.. while i wouldn't want to build a motherboard, i could.. but why when they're 50 bucks these days? posted_image

just need to know now what parts to buy and where to connect/solder everything. i have yet to hack apart my latest relay setup in hopes that i just have to add to it.. =)

thanks again dual!

Here's something you can try:

The added inverter stage is used to control the pulse delay timing, and the transistor circuit is to set the length of the pulse and drive the relay.
You can adjust the resistor and cap values to change the delay and pulse width as necessary.
The two diodes on the input of the inverter is to protect the input in case your input signal has any spikes or transients- might be able to do without it if your signal is clean, but there's really no reason not to add it in.

posted_image