cheap 12V Delay on Break Relay for fog-light "turn assist" lights.

[(*steve*)]

One last question, If i wanted to use a different relay with a higher resistance coil, the load on the transistor should go down, right?

That is correct. If the relay can handle the power demanded, then it is perhaps a good idea. The only other consideration is that automotive relays are made for harsh conditions, vibration, and high temperatures that another relay may not be.

I am currently still bench testing the board, but i just left it on 100% for about over 60sec and the transistor didn't get much above room temp.

When it's 100% turned on, the transistor is saturated and is dissipating very little power. The dissipation increases as the capacitor discharges and the reduced base current means a higher voltage across the transistor.

The worst case is probably if the relay only just stays pulled in in the interval between flashes.

Ok, After testing i think i may swap the 10K pot for a simple 10K resistor. would is hold a bit longer if I replaced the 470Ohm that is in series with the pot with a larger one? I think it hold's enough as is, but I wouldn't mind it holding a tad longer.

Increasing the resistor will increase the delay up to the point at which you can't supply enough base current and the relay either doesn't pull in, or drops out soon afterward.

If 10k + 470 ohms is not enough, I would consider a slightly larger capacitor (what are you using now?)

Once you have determined a suitable delay, measure the resistance of the pot and replace it with an appropriate resistor.

 

[Resqueline]

Yes, using the higher resistance relay you mentioned would lower the dissipation in the transistor.
And be aware that some auto relays are as low as 67 ohms and those would make the transistor exceed the allowed 625mW, but the short time could make it survive.
Yes, you can replace the 470 ohm with a larger one. Up to 4.7k with Steve's circuit, or you can remove it & the pot completely if you went with my circuit.
Very low temperatures might make the circuit change the timing slightly so I guess it's best to have a margin on the hold time.

 

[TbirdMan]

Yes, I was going for plenty of margin, I haven't tested it at the duty cycle of the flasher exactly, but I have tried to simulate it, and it seems like it might be okay, but a couple extra tenths would be nice. I think I will stick with the relays i have, for the reasons mentioned as well as the biggest reason, they're free.

I am currently using a 470uf cap. with the pot maxed out it seems to hold about right, that was why i was considering bumping the resistor a tad, Perhaps i could remove the 10k pot and 470 resistor and throw a 1k pot in series with the 2.2k?

 

[Resqueline]

Perhaps i could remove the 10k pot and 470 resistor and throw a 1k pot in series with the 2.2k?

Ok, so you were talking about "my" version. Yes you can, and that'll effectively make it "Steve's" version. You could also have a 470 ohm resistor + a 5k pot.
: Minimum resistance around 470 ohms (to "contain the magic smoke"), maximum around 5k (to ensure sufficient base drive current).

 

[TbirdMan]

Sounds good, if that doesn't do it Ill look into a larger cap. would a 1000uf be reasonable?

 

[Resqueline]

Yes, that should be plenty. If you want to save space you could always add a transistor in a Darlington configuration and thus get away with a 10uF capacitor.

 

[TbirdMan]

Space isn't really an issue, the entire circuit minus relay takes up about a postage stamp. with the largish realy the circuit is plenty small, and simple.

 

[TbirdMan]

ok, Tell me if this circuit is workable? since Radio Shack doesn't carry 5k trimmer pots I put a 10k pot parallel with a 10k resistor, which should serve as a 5k pot, no? I will likely prototype it off the board to determine what resistance i want and replace it with a simple resistor. I wasn't sure if i should leave the 470Ohm between base and ground, as I understand this is to make sure the transistor turns off?

EDIT: lamp has separate ground.

 

[TbirdMan]

found some info on my relay: RELAY,A/C CMPR(4 TERM)(280 MICRO)(HI-PWR)(W/RESISTOR)(UNSHROUDED)(GRAY)(GRAY)(35AMP)(4-TERMINAL 280 MICRO "HI-PWR" W/RESISTOR UNSHROUDED)(3.670)(ACDelco #15-50688)
That should be more than enough for a 50W bulb, correct?

 

[(*steve*)]

I put a 10k pot parallel with a 10k resistor, which should serve as a 5k pot, no?

Well, the minimum resistance will be 0 and the maximum will be 5k. However the response won't be linear.

The first 25% of the movement of the pot will give you 40% of the change in resistance, while the last 25% will give you only 14%. This is fine if you expect most of the adjustment to be in the upper part of the resistance, less so if in the lower.

Did they have 4.7k or 6.8k? One of them may have been suitable.

 

[(*steve*)]

35A should be ample. (It allows for a load over 400W)

 

[TbirdMan]

and that circuit should work?

in this circuit I am thinking the adjustment may be in the upper end. I want plenty of hold, i would rather it stay on a little after the signal cancels than turn off too early. if I end up getting too much resistance the next step will be to switch to a 1000uf cap.

 

[(*steve*)]

It should. I'm not absolutely certain of what you were using before, or the changes involved here, but it seems fine. placing that extra resistor in parallel with the pot will reduce the time delay. And that's what you're expecting, right?

 

[TbirdMan]

The reasoning for that was that earlier it was stated that the most resistance i would want there would be about 5k, "to ensure sufficient base drive current"

 

[TbirdMan]

Ok, after some trial and error, I think this is the final circuit i am going with, it is simple, and holds for almost a full second, perfect.

Thank you to everyone who helped me, your advice was invaluable. all this actually has me turning over in the back of my mind a switch from Mech Eng to EE for my Major.

 

[Resqueline]

Simple is often best, & I'm confident it will serve the purpose well. Good luck with the installation, and the education/graduation.
I see the Shack has only 1k, 10k, & 100k trimmers, although they do have a bigger 5k pot btw..

 

[(*steve*)]

I see the Shack has only 1k, 10k, & 100k trimmers, although they do have a bigger 5k pot btw..

Let me guess, they come beautifully packaged with data on the back of the card they're on, and cost about ten times more than anywhere else?

I think that's how I remember them in Australia.