Odd Electrical Problem

[Drewmotty]

I spotted an oddity on my way to the Devon TR meet last week. 
As the overdrive is disengaged the indicator and ignition warning lamps flash. They are both LED and have been for some time but I’ve not noticed them flash before. 
I’ve looked at the wiring diagram but can’t work out what’s happening. 
I  haven’t got round to any intrusive diagnostics yet. 
Any ideas?

Edited November 10 by Drewmotty

[RobH]

I rather think this may be a 'back-emf' effect from the overdrive relay or solenoid. 

When a relay coil is de-energised the collapsing magnetic field produces a pulse of voltage called a back-emf, which propagates on the power wiring.   Normally it is of short duration and power, and does not cause any problems in ordinary electrical circuits other than to cause arcing on switch contacts.  

Ordinary bulbs are not affected because of their low resistance and the time lag in heating the filament,  LEDs though are capable of responding very quickly to low currents so may well flash. 

If it annoys you it can be stopped by putting a reverse-biassed diode across the offending coil, which will 'short out' the pulse locally.

[john.r.davies]

RobH,

Lighten my darkness, please!   A 'back-emf' will be of the opposite polarity, won't it?   And diodes only light up if the polarity is correct?

John

[RobH]

True for the light-emitting-diode elements themselves but not necessarily for the bulbs they are used in.  Many of the LED bulbs sold now are dual-polarity - particularly the ones for dash tell-tales.  It's possible that is what is going on here.  

 I can't think of any other rational explanation for the flash. 

(Of course the coupling may not be direct on the power lines. It may be through capacitive or inductive effects between wires in a cable-form in which case polarity at the bulb may depend on the relative cable orientations.)

 

 

Edited November 10 by RobH

[Drewmotty]

13 minutes ago, RobH said:

(Of course the coupling may not be direct on the power lines. It may be through capacitive or inductive effects between wires in a cable-form in which case polarity may depend on the cable orientation.)

Thanks Rob, you’ve answered my next question. I guess that those wires all sit alongside each other under the dash. 
I’ll see if I can separate out the overdrive wires and if it all gets too much trouble I’ll put standard bulbs in those two tell tales. 

[RobH]

As I said an easy cure is a diode across the relay like this (reverse the diode if positive earth)

Any 1 amp diode with reverse voltage of 50v or more will do e.g 1N4001 through 4007 :

 

Edited November 10 by RobH

[harlequin]

Now that's interesting, I wondered why my indicator repeater light flashed when I turned the ignition off ever since I changed to LED indicators 

George 

[keith1948]

Our house wiring for the lighting is 2 wire with no earth or at least it was. I fitted the compact fluorescent energy saving light bulbs in bedrooms with a light switch by the door and another by the bed. With the 2 wire connection between the switches, one is always live. This induced a voltage in the disconnected wire (lights in off position). This voltage charged up the bulb capacitor which would then discharge with a flash at regular intervals. The solution was to replace the 2 wire strap between the 2 switches with 2 wire plus earth. The earth wire between the 2 power wires acted as a screen which stopped the live wire inducing a voltage in the other wire. When we had tungsten light bulbs this was not an issue but with the fluorescent style bulbs it was. Not noticed any issues with the LED bulbs we have now but then we have the 3 wires now with the earth screen effect. Modern wiring I believe is twin plus earth for lighting.

As an experiment I connected one wire of a length of 2 wire cable to mains +. I measured the voltage between the connected wire and earth and got 240v. I then measured the voltage between the unconnected wire of the pair and earth and got about 112v just by induced voltage. Doing the same with a cable with a central earth wire connected to earth gave a voltage of about 30v.

So I guess it is entirely possible that induced emf might be an issue if the cables are in a loom or adjacent to each other as Rob has suggested. Tungsten bulbs would not have this problem.

Keith

[stuart]

26 minutes ago, keith1948 said:

Our house wiring for the lighting is 2 wire with no earth or at least it was. I fitted the compact fluorescent energy saving light bulbs in bedrooms with a light switch by the door and another by the bed. With the 2 wire connection between the switches, one is always live. This induced a voltage in the disconnected wire (lights in off position). This voltage charged up the bulb capacitor which would then discharge with a flash at regular intervals. The solution was to replace the 2 wire strap between the 2 switches with 2 wire plus earth. The earth wire between the 2 power wires acted as a screen which stopped the live wire inducing a voltage in the other wire. When we had tungsten light bulbs this was not an issue but with the fluorescent style bulbs it was. Not noticed any issues with the LED bulbs we have now but then we have the 3 wires now with the earth screen effect. Modern wiring I believe is twin plus earth for lighting.

As an experiment I connected one wire of a length of 2 wire cable to mains +. I measured the voltage between the connected wire and earth and got 240v. I then measured the voltage between the unconnected wire of the pair and earth and got about 112v just by induced voltage. Doing the same with a cable with a central earth wire connected to earth gave a voltage of about 30v.

So I guess it is entirely possible that induced emf might be an issue if the cables are in a loom or adjacent to each other as Rob has suggested. Tungsten bulbs would not have this problem.

Keith

In the world of sound engineering the problem of small induced voltages between cabling is well known and in multicores there is two wires and a foil covering for each pair to prevent whats known as "Cross talk" There can also be a problem when microphone cables run next to heavy duty lighting cables inducing an annoying buzz!

Stuart.

[keith1948]

36 minutes ago, stuart said:

In the world of sound engineering the problem of small induced voltages between cabling is well known and in multicores there is two wires and a foil covering for each pair to prevent whats known as "Cross talk" There can also be a problem when microphone cables run next to heavy duty lighting cables inducing an annoying buzz!

Stuart.

Hi Stuart

I have a 1960's Broadway Plectric guitar that suffered from mains hum until I lined the pickup cavity with earthed copper foil.

[Spit_2.5PI]

What do you call your Broadway Plectric guitar? Humphrey?

Cheers, Richard

[mkmick]

Very good,  I'll buzz off now

[TR Rob]

I installed LED bulbs in two outside lights that are both controlled by individual electronic timers.

Result both LED bulbs did a continuous flashing. I then installed a Screwfix Elkay 4uf power capacitor across the terminals and the flashing stopped on one but not the other !

Had to replace LED bulb with normal bulb to stop flashing.

On both LED bulbs the capacitor was wired at the light end of the circuit .

Anyone any ideas why this should happen ?

[RobH]

It might have been the electronic timers if they had enough leakage current to power the bulbs slightly even when 'off', Rob. It probably would not have happened with old-fashioned mechanical timers, which break the connection completely.  

There is a fair bit of electronics inside mains-powered LED bulbs, to rectify and regulate the power to them. That must include capacitors and something may have been charging up from leakage until there was enough to power the LED, which would then discharge it to allow the cycle to start again - as with Keith's CFL above .

One of the perils of modern complication - odd unintended interactions can happen. 

[TR Rob]

Hi Rob, funny thing is the addition of the capacitor stopped the flashing on one LED / timer circuit on front door.

But the other LED/timer circuit on garage kept flashing even with the addition of a similar capacitor, this is the circuit that now has an incandescent bulb in it ?

Can’t fathom out the rationale behind it !!

[keith1948]

18 hours ago, RobH said:

It might have been the electronic timers if they had enough leakage current to power the bulbs slightly even when 'off', Rob. It probably would not have happened with old-fashioned mechanical timers, which break the connection completely.  

There is a fair bit of electronics inside mains-powered LED bulbs, to rectify and regulate the power to them. That must include capacitors and something may have been charging up from leakage until there was enough to power the LED, which would then discharge it to allow the cycle to start again - as with Keith's CFL above .

One of the perils of modern complication - odd unintended interactions can happen. 

Hi Rob

With your knowledge of all things electrical you might know the answer to this one. If you fit a capacitor across the terminals or the LED has capacitors inside then presumably these will charge and discharge power depending on the wiring. Is this a tiny hidden constant drain of energy we are not told about? Fitting a capacitor across the terminals seems to be hiding this power drain because the LED or CFL no longer flashes to give a visible sign of what is really happening.

Meanwhile just mended the horn push on the TR so it does not fly apart when I press the horn!

Keith

p.s. I hope the original post has now been resolved or at least explained. Looking at the wiring diagram I would clean the connectors on the control box and the ignition switch. Temporarily removing the LG/G fuse in the fuse box should stop the indicator warning light flashing and that would suggest the issue is towards the ignition switch I think. My first port of call on electrical issues is to check earth connections and not just on the obvious electrical component. 

Edited November 12 by keith1948

[RobH]

(apologies for thread-drift)

Without knowing the details of the circuitry it is difficult to make any meaningful guesses Keith.  As you know, a capacitor will allow an AC  current to flow but the amplitude depends on the capacitor value  (Xc = 1/2Pifc) where Xc is capacitive reactance in ohms, c is farads and f is frequency.  Lots of electronic units have small suppression capacitors fitted to protect the circuitry from voltage spikes on the supply so it's quite common for some extremely low current to be drawn all the time.  Most of the 'wall-wart' power units do it and I have known instances where lots were in use on the same mains circuit and the cumulative leakage was enough to cause nuisance tripping of a 30mA RCD. 

Why Rob's external capacitor stops the flashing is anyones guess but it is a large value at 4uf, so It may just have been forming a capacitive potential divider which prevented the stray voltage at the lamp rising far enough to make it flash.   Again-why that didn't work with the second lamp is open to speculation but there must be differences in the circuits.