Syvecs Forums

Long story short I stripped the guts from a shift light because it was inaccurate and connected it directly to the ecu, obviously the internals of the ecu allow some current to flow in the off state, enough to make the led dim constantly.
Shift light works as expected at correct rpm.
What can I do to have the led completely off?
Thanks
Graeme

The leakage current of these ECUs does seem quite high - can't be used for a DCCD drive, for example, as the leakage means the DCCD never goes fully open.

A typical injector driver device only has a leakage current of 100uA or thereabouts, so the output stage in these ECUs must be a little unusual. Without knowing the exact output stage schematic, your only option is probably to use an external relay to drive the LED.

Thanks for the reply.
I thought about putting it through a small solid state relay actually, probably the only way to fix it!
Thanks
Graeme

Would a transistor work? Using the ecu ground to trigger the gate? If so what transistor would be best?
Graeme

You could no doubt find a suitable mosfet, which would need some kind of pullup resistor on the gate. I would still be tempted to use a through-hole solid state relay (which would still need a series resistor to limit the current flowing through the trigger LED) - they're just a few £'s - simply because choosing them is a little easier than trawling through loads of parameter tables for mosfets; I would have thought.

Thanks Tim
I have some cheap Chinese 20a solid state relays, bit large physically but I guess it would save me buying anything else!
Graeme

Why not just wire in a little bulb in parallel somewhere to sink the current when off. Should force the LED off.

Or maybe a resistor could do the same job ?

Actually, a pull-up resistor might do it. Without knowing the leakage current we're dealing with, nor the characteristics of the shift light input, its difficult to be sure, but if it was, say, 10mA, then a 220R 1W resistor from the output to battery volts would allow the shift light input to sit near to battery volts.

This might keep it off, but, equally, I don't know if the input stage of the shift light would be happy with this.

He say's he's gutted the shift light, so it is just a simple LED now, triggered from the ecu

So there are no controls actually in the light anymore ?

Oh, yes,sorry. So a pull-up to battery should do the job. Value needs to ensure the shiftlight connection remains just above battery volts minus the forward voltage of the led(s), with the leakage current flowing, and wattage chosen to ensure it is OK with full battery volts across it.