Hi Tim (2)
Combination of everything. That would work but still need a diode between turn lamp and Mosfet / capacitor / discharge resistor. That is getting more components but I don’t think there is any choice. The discharge resistor can be high so there is some advantage there, cap charging current would be smaller.
Cheers Bob
Hi Tims (1&2)
Just thought of another possible thing against a relay without Mosfet. What does changing a bulb size do to the turn signal repetition rate. This would depend on how the signal is generated.
Without the Mosfet driving a relay the turn lamp will have the relay coil resistance across it at all times. This resistance could be about 85Ω and to get about a 2Sec time constant would require about 23500µF. This might have an effect on pulse rate and thus the required time constant, so chasing tails could be the result.
So I think Mosfet switching a relay would be the go. A 2 Sec time constant would then require 100kΩ / 20µF combination.
Pulse timing is still a bit up in the air as Tim (1) has not let anybody in on this little secret to date. If he does not let us in on this little gem soon I personally will call it a day and bow out as I think we have come up with most of the suggestions with the available info.
I just did a bit of revision. I somehow got the impression that the relay was switching the 21W lamp. It isn’t, it is switching a 5W lamp so only about 0.5A. So one of the 10A small relays would do nicely.
Cheers Bob
Hi there Bob,
I personally will call it a day and bow out as I think we have come up with most of the suggestions with the available info.
Let me start off but clarifying that I am nothing more than a guy with a problem he wants to fix, but is unsure of how to fix it. The reason I started a thread here is to gain knowledge from people with much more expertise and know-how. I’m happy to see you and Tim (2) tag in and offer some solutions and explaining your thought processes and I do appreciate them.
Please keep in mind that as I read your inputs, I have to process them and figure out what is actually being said. Sometimes I read them again to make sure I understand, as this knowledge does not come naturally to me.
That said,
Pulse timing is still a bit up in the air as Tim (1) has not let anybody in on this little secret to date.
When chaning from incandescent bulbs to LED, I also switched the turn signal relay. The old one worked off resistance from the original bulbs and the new relay is electronically timed. So regardless of the bulb, the frequency remains the same.
What I can tell you about the frequency is that I counted 67 flashes in 60 seconds, so let’s call that 1.1Hz. What that means for the new components, I don’t know yet.
If you feel like tapping out, then that’s fair. I’m not holding you against your will, I hope.
Like I said, I’m just here to gain more insight in electronics and I consider forum uses like yourself to be experts compared to myself. So thank you once again for your contributions so far. They do help!
Greets
Hi Tim (1)
OK, that is what we need.
Swapping the relay out is another useful bit of information. We now don’t have to worry about what affect the bulb changing operating conditions will or will not do.
Approximating your count and without any means to measure it would seem that your pulses are a little better than 1 sec so we need to hold a relay up for at least 1 and probably a bit seconds. I will have a look at the data sheet for a suitable commonly available Mosfet and come up with what I think will be a suitable time constant. At a guess I would probably say about 2 seconds. We have to have enough voltage to hold the Mosfet and relay on until the next turn pulse.
Don’t worry about me being old and cranky these days. Happens all the time. But getting information is like pulling teeth with some.
Cheers Bob
Tim(1),
I agree with Bob that a relay alone without a MOSFET is a problematic.
This is my current view of how I would approach it.
Tim291322v0.1.pdf (14.0 KB)
The values of R1 and C1 depends on the length of your flasher “on” pulse (tOFF), the length of its “off” pulse (tOFF) and the characteristics chosen MOSFET . Very roughly, C1*R1 = tOFF + tON / 2.
If I was to do a proof of concept on the bench, I would substitute each of the two LED arrays with a single (“jelly bean”) LED in series with a 470R resistor. (The resistor limits the current, otherwise 12V would cook a standard LED, which wants only 2V approx.)
Regards,
-Tim(2).
Hi Tims (1 & 2)
That is the sort of circuit that you can’t go wrong with.
I don’t quite agree with your choice of diodes though. The 1N400X series are really power rectifiers and once you get into the kHz area they exhibit much the same characteristics as a piece of wire. Also only 1A max. C1 charge current could approach this. I would prefer to see something like MBR735 7A Schottky. As the magnetic field in this case will decay fairly slowly you might get away with 1N400X for the relay but why not be sure and use adequate Schottky here also. Up to the end user.
Timing:
In the absence of actual measurement the turn pulse time appears to be about 1 Sec. I think it would be safe to assume that the on and off times are about equal so a time constant of 0.75Sec would be OK. We will assume that this timing is chosen.
Now the relay has to be ON for about 0.5Sec. Fortunately the voltage across the cap and thus gate voltage is predictable and at 0.5Sec will be 6.2V (assuming 12V charge voltage on C1) at which the Mosfet will still be very much on.
Unfortunately the actual Mosfet OFF point is not so predictable. The threshold voltage is listed as 2 to 4V meaning it could be off anywhere between. Also the actual relay release point is a bit more up in the air and the actual would depend on the individual relay.
But the gate would be at 4.4V @ 0.75Sec which is border line gate threshold and 3.16V at 1Sec which should be close to off. So unfortunately the exact relay release time is subject to a few variables. It could be a subject for experimentation but I would suggest 0.75Sec time constant would be a good option. The release time after the last pulse should not be much more that 1Sec which should be acceptable.
C! = 10µF. Tanntalum 35V type if possible.
R1 = 75kΩ
All = 0.75Sec time constant.
The 12V relay supply could come from the parking light. There is no need for relay operation without the parking light.
100Ω resistor in series with Diode 1 would be nice to tame down the inrush current a bit and will not affect C1 charging noticeably.
Mosfet type is a good choice.
Cheers Bob
Note: 100Ω resistor minimum 1W. 1.5W required but only for a very short time so 1W should be OK.
Or 2 X 220Ω 1W in parallel might be a safer option.
1 Like
