Coping with voltage spikes

[Tim StevensParticipant @timstevens64731]

PS: the requirement for not changing the electrics does not apply to all cars. Otherwise I would be required not to have a brake light, and to drive with one rear lamp only, no flashers, etc. Please ensure that claims about what the law says are complete – specifying which cars (etc) are exempt. Otherwise you are likely to be seen as part of the problem.

Tim

[duncan webster 1Participant @duncanwebster1]

I'd just increase the series resistance, OK it will be a little dimmer most of the time, but you don't want to be dazzled by the oil pressure gauge. I'd guess it has about 1k at present, add about 400 ohms and see how it goes.

[Maurice TaylorParticipant @mauricetaylor82093]

I would try connecting the led across the battery at the battery terminals ,the led will light.Then start the car if the led stays lit ,there is more than likely a problem with the circuit where it it is usually connected.It will prove whether the led is suitable or not.

[Michael GilliganParticipant @michaelgilligan61133]

Posted by Tim Stevens on 13/08/2020 23:10:07:

.

There is no part number – the failing LEDs were purchased on e-bay to a brief specification – 12v narrow beam 3mm white.

 

.

Tim,

I tried your ebay search phrase and found a seller whose ‘description’ reads thus:

”Lens Type : Water Clear Lead Length : 20cm Optimum Supply Voltage : 9v – 12v for highest brightness Minimum Supply Voltage : 5v with much reduced brightness Maximum Supply Voltage : 14v Current : 20mA @ 12v Viewing Angle : 20~25 Deg”

IF those are the same LEDs then 14v looks a risky Maximum

MichaelG.

Edited By Michael Gilligan on 14/08/2020 00:01:52

[HopperParticipant @hopper]

If you are still running the old style voltage regulator with coils and points inside it (Lucas, Bosch, Miller, Delco etc)you might try replacing it with a modern electronic unit. Those old mechanical regs put out massive spikes and dips all over the place. You can see them with a cheap digital volt meter across the battery. The numbers just keep up a manic dance. I use an old analog meter to set those regs for this reason.

But your glowing indicator warning light might also be a sign of a partial short in the wiring somewhere too. I can't visualise how voltage spikes could jump across the switch contacts and make it glow?

[Danny M2ZParticipant @dannym2z]

A ferrite ring with the power line looped through and around it a few times is useful to surpress high frequency spikes.

* Danny M *

[Robert Atkinson 2Participant @robertatkinson2]

Posted by Tim Stevens on 13/08/2020 23:10:07:

Two more answers:

There is no part number – the failing LEDs were purchased on e-bay to a brief specification – 12v narrow beam 3mm white.

The vehicle is a 1928 Lea Francis model P 12/40 coupe-cabriolet. The dynamo is CAV driven at 1.5 wengine speed as original, and the voltage regulator came from a British supplier whose details I have, but not handy at 11.pm – sorry. It has been converted to a two-brush system in place of the original 3-brush which was controlled by the driver when he remembered and turned the dynamo on, or off (ie too much or not enough). If this information really helps to diagnose the problem do please let me know how.

Tim

OK,

So the lighting regulations don't apply

Sounds like cheap LEDS running at maximum current coupled with spikes (negative as well as positive) from the generator / igintion system. Put a 180 ohm 1/4 watt resistor in series with each LED and a 12V 0.5W zener diode across each LED (on LED side of resistor) Anode to negative side of LED.

If all LEDs come on at once a single resistor and diode will work. Let me know how many LEDs you have on the circuit and I'll calculate values. I've probably got suitable parts in "stock" If so I'll send you them for postage cost.

Robert G8RPI.

Edit,  Shoulld have said, the zener is cheap, simple and protects aginst both positive spikes above 12 V and negative spikes above 0.8 V

Edited By Robert Atkinson 2 on 14/08/2020 13:07:35

Edited By Robert Atkinson 2 on 14/08/2020 13:08:06

Edited By Robert Atkinson 2 on 14/08/2020 13:08:49

[Tim StevensParticipant @timstevens64731]

Some further info which might help:

In my (limited, practical) experience, most 12v LEDs intended for vehicles are fitted with a connector fitting – bayonet is favourite for side tail & brake lights, and for internal lights, some arrive with a selection of connections so you can match the original bulb type. In the metal parts of the assembly, there is usually a fairly complex circuitry feeding the LED with interrupted voltage derived from an oscillator and rectifier. This is contrived to allow the LED to work with anything from 8v to 30v, so suit trucks as well as cars and motor bicycles. Some even still work down below 6v – ideal for owners of Austin 7s and Model A Fords. The frequency is above the flicker-limit for most folks – but shows if you turn your head quickly near a bright LED, which will be seen as a series of blobs of light. The same circuitry is used in household bulbs, and for eg traffic signs. I saw a TV scene this week where a green traffic light was visible going on and off regularly – I interpret this as the lack of synchronicity between the camera 'shutter' and the oscillator in the LEDs used in the signal. ( … discuss … ?)

These larger, more complex devices seem not to be bothered by ignition spikes, and so most of my system works well 'out of the box'. These lamps all need to go on and off at various driving (etc) circumstances, turning, braking, dip etc, so a 'universal' supply would perhaps struggle.

Thanks to your suggestions here, I will try these options:

A: using a plain LED with a separate voltage reducing device
B: and a capacitor across the supply to the LED, or
C: the same but with a ferrite toroid with the 'live' feed wound in and out a few turns;
D: a Zener diode with any other combination of these ideas;
E: demolish a known-ok small 12v LED fitting and use the internal oscillator etc circuit to feed my oil-gauge LED;
F: A separate circuit comprising a 3v battery or cell, to feed the 3mm oil-gauge LED, perhaps even with a charging system which is only live when the panel lamps are not on.

So, lots to play with – and thanks again for all your comments.

Regards

Tim

PS I wonder if our editor would like a few paragraphs on the ins and outs of LEDs for model makers?

[duncan webster 1Participant @duncanwebster1]

In the period of limbo between SWMBO announcing she is ready to go out and her actually being ready I connected a 3mm 20mA LED via a 820 ohm resistor to my power supply. On 15.3 v I got 15.2 mA, but no noticeable reduction in brightness until the voltage was down to around 9v, 7.5mA. This suggests that a 1k series resistor would be OK which would stand 23v input (20mA). If reverse voltage spikes are an issue, a 1n4001 diode would block. All a lot simpler than voltage regulators etc

Edited By duncan webster on 14/08/2020 21:22:02

[Tim StevensParticipant @timstevens64731]

Duncan – the problem is not 'how can I reduce 12v to 3v' – that is done before I get them if they are 12v LEDs. The problem is that diodes do not cope well with spikes – of an unknown voltage* and very short duration, caused by the collapse of the field in the ignition coil as it it switched off every half a revolution. Even if I knew the size of the pulses, I could not expect to mock up a similar spiky supply. The only way to test a possible cure is on the car itself.

* I don't know how you are measuring your applied voltage – but anything other than an oscilloscope will not register the voltage of a spike. In my opinion, based on not very much training – none in LEDs – back in about 1968.

Best wishes – Tim

[Anonymous]

Posted by Tim Stevens on 15/08/2020 14:58:49:

The problem is that diodes do not cope well with spikes – of an unknown voltage* and very short duration, caused by the collapse of the field in the ignition coil as it it switched off every half a revolution.

I've read this thread through twice and I can't see any evidence that points to the problem being ignition pick up. The di/dt and dv/dt of an ignition system can be quite high, but the energy is low. The only time I've had a problem with ignition pick up was when fitting thermocouples to a racing (F1) car. In fact they were a darn sight more reliable at measuring rpm than the "proper" rpm circuit.

The LED circuit is fairly low impedance so it is unlikely that enough current will be induced to cause the problem. There's an art to solving these sort of problems and the first step is to gather data without making any guesses at what is causing the problem. That's not as easy as it sounds. If one thinks one knows what the problem is then data collection is often biased towards verifying the expected cause. There are some simple questions we can ask:

If I've understood correctly the LED fails when the car is started? Is this repeatable, ie, every time the car is started a new LED fails.

Can the LED be powered without the car actually being started, and if so does it work?

If I've got it correct some other LEDs flicker when the car is started/running which is suspected to be due to variations on the DC supply. It would be useful to quantify those variations, although I'd agree with Tim that a 'scope is the only real way of doing so. An analogue meter will give some indications but will tell one very little about the nature of any variations.

One more general question; does the wiring for the failed LED run significantly closer to the ignition system than any of the other units fitted?

As an aside you've only got to look at the startups that have come and gone over the years, along with investors money, having had the wonderful idea of wireless charging ones 'phone/tablet/camera only to find that it's pretty darn tricky to reliably transfer significant power between randomly placed coils.

Andrew

[Tim StevensParticipant @timstevens64731]

Andrew:

The LED in question works as required when fitted. If I then drive without the sidelights on, there is no perceptible problem. But when I turn on the sidelights this particular LED (out of 7 altogether) does not work.

I suspect (but as you surmise, I don't KNOW) that the cause of the failure is related to the running of the engine, and the fact that LEDs are sensitive to brief too-high voltage pulses leads me to seek a way to reduce or eliminate the pulses. If that does not work – ie if the LED still fails, then I have eliminated one possible cause. Groping in the dark it may be, but I have not been able to come up with an alternative method with the limited kit I have.

If it were an easy option, I would try to isolate the ignition wiring etc from the rest of the lighting loom. Perhaps I will have to if my first guess turns out to be totally wrong.

PS I have not yet tested the action of starting the car – it may be that the nasty spikes arise from the starting solenoid etc, rather than the effect of ignition voltages. Next time I need to start the car I will try this.

Cheers, Tim

Edited By Tim Stevens on 15/08/2020 17:02:01

[SillyOldDufferModerator @sillyoldduffer]

Posted by Andrew Johnston on 15/08/2020 15:45:12:

Posted by Tim Stevens on 15/08/2020 14:58:49:

The problem is that diodes do not cope well with spikes – of an unknown voltage* and very short duration, caused by the collapse of the field in the ignition coil as it it switched off every half a revolution.

I've read this thread through twice and I can't see any evidence that points to the problem being ignition pick up. …

There's an art to solving these sort of problems and the first step is to gather data without making any guesses at what is causing the problem. …

If I've got it correct some other LEDs flicker when the car is started/running which is suspected to be due to variations on the DC supply. It would be useful to quantify those variations…
…

Andrew

Andrew is right about the need for more evidence, and his point about ignition pick up being unlikely is well made!

In that case, the other suspect is the car's electrical system.

Tim explained: 'The vehicle is a 1928 Lea Francis model P 12/40 coupe-cabriolet. The dynamo is CAV driven at 1.5 wengine speed as original, and the voltage regulator came from a British supplier whose details I have, but not handy at 11.pm – sorry. It has been converted to a two-brush system in place of the original 3-brush which was controlled by the driver when he remembered and turned the dynamo on, or off (ie too much or not enough). '

Tim describes a pre-war dynamo system with a relay based regulator. Though clever these crude regulators are a long way from the smooth volts supplied by a modern alternator system. The regulators relays control voltage and current by switching a resistor as necessary to increase or decrease current flowing in dynamo's field coil. A third relay cut-outs the battery to prevent overcharging. Correct on average, like the rpm of a hit and miss engine.

Could be three unsuppressed relays putting inductive spikes on the output. My chief suspect though is the rough voltage regulator itself. It's notion of '12v' is undoubtedly crude, plus a relay has slow reactions, meaning volts could rise quite high before the field current drops. I guess the relay kicks in and out more or less continuously too, putting a series of spikes on the output. Filament bulbs and the battery are unlikely to mind, but LEDs are voltage sensitive.

Suggests a cure as per Robert's Zener, just in front of the Tim's 12v LED, eg.

The circuit is a simple voltage stabiliser, with the Zener shorting out anything over 12v. Self criticism, the zener and 180 ohm resistor powering the oil-pressure gauge lamp will soaking up whatever excess the dynamo happens to be chucking out. Might need to be generously rated. An oscilloscope would reveal all.

A better solution would be to put a modern stabilised power supply across the dynamo and use it to provide clean power to anything remotely modern on the car. I think the box I recommended earlier has a good chance of fixing Tim's problem.

Might I have seen Tim's car on TV? Just the job for Downton Abbey!

Dave

PS,

I see Tim posted just before I hit send, and maybe he's added an new clue.   I'd assumed the LED was being destroyed, which definitely points to spikes.  Now I'm wondering if Tim means it goes out temporarily.  Does the LED come back on when the sidelights are turned off?   Going out temporarily isn't spikes.

Edited By SillyOldDuffer on 15/08/2020 17:24:30

[Tim StevensParticipant @timstevens64731]

SoD: No the light does not come back on, nothing I do to the dead LED will light it again. (Long rambling thoughts about a defunct Norwegian Blue …)

And the voltage regulator is not a pair of relays under brass lid – it is a bunch of electronic gismos sealed in an aluminium slab – from: Dynamo Regulators in Fareham, Hants UK. I will enquire of them about the likelihood that their box of tricks will cause my problem. It is 100% better than the original system, charging the just-started battery at about 8 amps for about a mile, and dropping back to about 14V at anything above a slow tickover. But remember that the dynamo is charging all the time, and I can use the car without turning the LED on, and turned off back in the garage the LED still works OK. Looking at the failed LED, there is no difference, no melted plastic, no gaps where wires used to be. I do not have a microscope to peer at the actual source of the light. Neither have I tested the bulge in the LED wiring which I guess is a resistor – another job for next time in the garage.

The problem I anticipate with your 12v Zener is that it is connected across the battery and dynamo. The battery sits at 12.7v when 'off' but on charge is at least 13.8V, so perhaps I need something more like a 16V Zener ?

And the car is not ideal for posh TV shows – it is a two-seater with mother-in-law seat in the back. And the hood, when up, only covers the front seats.

Cheers, Tim

[Robert Atkinson 2Participant @robertatkinson2]

Posted by Tim Stevens on 15/08/2020 17:52:24:

SoD: No the light does not come back on, nothing I do to the dead LED will light it again. (Long rambling thoughts about a defunct Norwegian Blue …)

And the voltage regulator is not a pair of relays under brass lid – it is a bunch of electronic gismos sealed in an aluminium slab – from: Dynamo Regulators in Fareham, Hants UK. I will enquire of them about the likelihood that their box of tricks will cause my problem. It is 100% better than the original system, charging the just-started battery at about 8 amps for about a mile, and dropping back to about 14V at anything above a slow tickover. But remember that the dynamo is charging all the time, and I can use the car without turning the LED on, and turned off back in the garage the LED still works OK. Looking at the failed LED, there is no difference, no melted plastic, no gaps where wires used to be. I do not have a microscope to peer at the actual source of the light. Neither have I tested the bulge in the LED wiring which I guess is a resistor – another job for next time in the garage.

The problem I anticipate with your 12v Zener is that it is connected across the battery and dynamo. The battery sits at 12.7v when 'off' but on charge is at least 13.8V, so perhaps I need something more like a 16V Zener ?

And the car is not ideal for posh TV shows – it is a two-seater with mother-in-law seat in the back. And the hood, when up, only covers the front seats.

Cheers, Tim

Hi Tim,

No you don't need a higher voltage zener. the additional 2 or 3 volts is dropped across the 180R resistor This "wastes" about 0.05 W (50 milliwatts) of power.
I think your problem is the LEDs are running at maximum at 12 V and the 13.8-14.2 V of a modern regulator plus any spikes from other loads is the last straw. The 180 ohm / 12V zenner circuit ensures the LED and resistor nver sees more than 12V (you could even go down to 10 or 11 V which will further protect the LED and reduce any flicker from voltage variation. One advantage of this simple ciricut is it protects itself. If you use a voltage regulator IC you really have to add diodes and capacitors to protect the IC.

SOD,

Thanks for doing the circuit (and credit). I looked again at the unit you linked to. It seems to be specfically for German cars and reversing cameras powered from the reversing light. It appears some models use some kind of PWM to the light. I'm guessing here, but it's probably a series diode and a electrolytic capacitior to give smooth DC at low current.

Robert G8RPI.

[Maurice TaylorParticipant @mauricetaylor82093]

Hi Tim,

I would do the following,

1 Make a test lead ,with a 12v led ,length of black wire ,black crocodile clip,length of red wire,red crocodile clip.

2 Measure battery voltage with engine running.

3 Turn engine off

4 Connect test lead , led should light.

5 Leave test lead connected,start engine ,hopefully led should stay lit.If it stays lit there must be a problem with wiring to pressure guage.

6 If it blows ,fit new led to test leads,then disable generator either by disconnecting it or removing fan belt.

7 Connect test lead and start engine,if led stays lit ,this indicates a problem with the charging system,if it blows it suggests a wiring problem.

8 Make another test lead 12volt 5 watt bulb ,put this in place of led at pressure guage to see if it works or not.

Hope this helps,I know it seems a bit long winded but it takes longer to read it than do it.

Regarding voltage spikes ,put a supresser capacitor on th e HT coil , like when fitting a car radio.

Maurice

[Tim StevensParticipant @timstevens64731]

Even more info:

Sorry, but there is no fan belt, The dynamo is driven by a helical gear on the camshaft. I will try as you say but removing the dynamo connections.

Fan belts as we know them today were not invented until about 1930.

Cheers, Tim

[Anonymous]

Posted by SillyOldDuffer on 15/08/2020 17:15:31:

Whatever you do; don't give up the day job!

The LED is the wrong way round. Even with the LED the correct way round the Zener won't do anything against positive spikes. If there's 12V across the Zener there will also be 12V across the LED and since current in a diode is an exponential function with voltage that's a lot of current. In practise the resistor or LED will most likely fail long before the Zener breaks down.

Against a negative spike the Zener will protect the LED as it acts as a normal diode with a forward voltage characteristic much lower that the reverse breakdown voltage of the LED.

Andrew

[Dave DanielsParticipant @davedaniels93256]

Wow, this one's gone on a bit

I think Malcolm was on the right lines. Connect a LM317T as a current limiter with a 120R resistor will give circa 10mA.

That's where I would start. Cost about 50p. If the problem still persists then it may be you need someone with a scope to see what's really going on. Meters, analogue or digital are often of little help under such circumstances.

D.

[Anonymous]

Posted by Dave Daniels on 15/08/2020 22:30:47:

……you need someone with a scope to see what's really going on,

That's where I'd start. Identify the problem and then one can create a solution.

Andrew

[Anonymous]

Posted by Tim Stevens on 15/08/2020 16:57:47:

The LED in question works as required when fitted. If I then drive without the sidelights on, there is no perceptible problem. But when I turn on the sidelights this particular LED (out of 7 altogether) does not work.

Is it safe to assume that the LED in question only comes on when the sidelights are turned on? If yes all we can say is that the LED fails to come on when the sidelights are turned on. We can't say whether it failed when the car was started or when the sidelights were turned on, or somewhere between the two events.

Andrew

[Rod RenshawParticipant @rodrenshaw28584]

I really, really, hesitate to second-guess Andrew but I think SOD's circuit, though flawed, could indicate a simple solution.

The LED symbol in SOD's circuit is the wrong way round, as Andrew says, but it's also a simplified version of reality. The LED that Tim is using has a series resistor or other circutry built in, as supplied, to render it suitable to run off a 12 v supply. The zener circuit (once corrected) supplies 12 v whatever the input voltage, as long as it's > 12v and within reason ,Then the LED should light happily?

Rod

[Speedy Builder5Participant @speedybuilder5]

As Tim wants to reduce power consumption, he could run a fiber optic cable from a lamp in his system back to the instrument that needs illuminating – almost as simple as some of the electronics being discussed ??

[Anonymous]

Posted by Rod Renshaw on 16/08/2020 11:18:58:

The LED that Tim is using has a series resistor or other circutry built in, as supplied, to render it suitable to run off a 12 v supply.

If that's the case then I agree that the Zener might do something, albeit not necessarily that useful. The problem with Zeners is that they're not overly precise, the knee is quite soft at low breakdown voltages and they have a temperature coefficient.

An interesting test which might help solve the problem is to put an example of the LED on a bench power supply and see what happens as the voltage is increased. That should tell us if there is a maximum input voltage before the LED goes phut.

Ignore this bit if you're not an electronics nerd!

Most Zener diodes are not in fact Zener diodes at all. The Zener effect (quantum tunnelling made possible by very high electric fields in the p-n junction) is only dominant below about 5.6V. Above that avalanche breakdown dominates. The Zener effect has a marked negative tempco while avalanche breadown has a positive tempco. Look at the tempco of the breakdown voltages and you'll find a near zero value at about 5.6V. In old school electronics (before sooper-dooper IC voltages references were available) a 5.6V Zener was often used as a reference. Or even better in conjunction with a forward biased silicon diode where the tempcos of the two components tended to cancel. That's why old school voltage references were often at 6.2V.

Andrew

[SillyOldDufferModerator @sillyoldduffer]

Posted by Andrew Johnston on 15/08/2020 21:45:33:

Posted by SillyOldDuffer on 15/08/2020 17:15:31:

Whatever you do; don't give up the day job!

The LED is the wrong way round. Even with the LED the correct way round the Zener won't do anything against positive spikes. If there's 12V across the Zener there will also be 12V across the LED and since current in a diode is an exponential function with voltage that's a lot of current. In practise the resistor or LED will most likely fail long before the Zener breaks down.

…

Too late for the day job, I'm retired. That gets me Bert Ashton's tee-shirt: I'm entitled to strong opinions, even if they are wrong…

The LED being the wrong way round is a fair cop, I forgot to rotate it.

However, the circuit works because although I showed Tim's LED as a single device, he's using a 12v LED, ie with a built in dropper resistor.

The full circuit, polarity corrected is

I don't see any way the boxed LED can get more than 12V because a conducting Zener will pull the LED-side voltage down through the 180R dropper. A half watt Zener and resistor would safely drop a continuous 20V overload to 12V, so should deal OK with spikes. I agree an oscilloscope is really needed to confirm what's going on, but the Zener circuit has a good chance of fixing it.

Interesting point about Zener Diodes really being Avalanche Diodes, but it's in the same class as the Tomato being a fruit, not a vegetable. Was an OCP71 really an OC71 with the paint scraped off? How I miss the good old days!

Dave

[Author]

Posts