Welcome to our resource for MG Car Information.
|
MG MGB Technical - LED blinks with fuel pump
Sorry to monopolize the threads recently. All of you have great input and troubleshooting. I converted to LED's for safety and less stess on the old loom. I got all of them from Moss however they didn't have them for the backup lights and I got the best quality I could find on Amazon. https://www.amazon.com/gp/product/B008E94ZVE/ref=oh_aui_detailpage_o00_s00?ie=UTF8&psc=1 However, when the fuel pump ticks the bulbs flicker in harmony. None of the other LEDs do this. The vendor is at a loss. Its not the bulb as they both do it. Vendor claims a voltage drop but battery is fully charged and I measure 12VDC+. Later I'll put a meter on it to see what happens with the fuel pump to see if there is a voltage drop. I have the Moss festoon ones in the cabin and the boot and those do not flicker (Moss doesn not recommended for backup). Curious. When lit they are great and the incandesant are yellow by contrast. |
Max71 |
Just a follow up. I am seeing a voltage drop for the backups. -1VDC = 11VDC. All others including those festoons measure 12VDC+ Cleaned all connections. Not sure where the drop is coming from... |
Max71 |
I had the led stop/tail bulbs flicker with the fuel pump. Short answer I removed them. Fuller story here. http://www.mgb-stuff.org.uk/electricstext2.htm#ledpark But how often are you using back up (?reversing) lights |
Michael Beswick |
I have LED DRLs and they flicker with everything, it's a 'feature' of them. Whilst LEDs take very little current compared to incandescent, so are less susceptible to dimming with relatively small increases in series resistance from old connections, they are more susceptible to voltage changes. This is because whilst incandescent have to heat up over a finite period to full brightness, they also dim over a finite period when the voltage is interrupted. When the interruption is only brief or slight you probably won't notice anything on an incandescent. Bt because LEDs light up 'instantly' they also extinguish instantly, so any fluctuation in voltage becomes visible. How OEMs get round it I don't know, but obviously our systems were never designed with them in mind. Back-up lights have a wired earth that is shared with the fuel pump, which takes about 6 amps for each pulse, so I suspect your vendor is correct albeit nothing to do with the battery, but the shared earth. It could also be the shared white/circuit fed through the ignition switch, but as others don't flicker perhaps not. As far as the volt-drop goes it depends on what two points you are comparing, it could be occurring anywhere in the circuit from the battery, brown circuit, white circuit, green circuit, reverse light switch, etc. You would have to take series of readings at each of those points to see where it (or the bulk of it as there are likely to be more than one) is occurring. But as Michael implies, does it matter with reversing lights? |
paulh4 |
Max, you could try fitting ballast resistors to each of the reversing light bulbs, which will often fix led bulbs flickering. See this link for more info, http://www.classiccarleds.co.uk/collections/ballast-resistors/products/2-x-50w-6ohm-led-indicator-ballast-resistors-fix-hyper-flash-turn-signal They also supply reversing light LED's, which are a bit different from the ones you have fitted. http://www.classiccarleds.co.uk/collections/festoon-bulbs/products/2-x-classic-car-led-reversing-light-bulbs-glb273-5630-smd-6-led Andy |
Andy Robinson |
Fitting 50 w ballast resistors into a circuit that originally took 47 w seems to rather negate the purpose of fitting leds. |
Michael Beswick |
Hmmm. All quite interesting comments. None of the other LED's flicker. When measuring across the bulbs they all show 12VDC+ Talked to several MG parts dealers and none of them stock LEDs until they can work it out. However, the Moss ones work properly. Maybe that's why they are so expensive. I have them in the rear lights for stop, running and turn signal. No flickering. Solid. I also have ones from SuperBright for the side running lights and no flicker. LED tech is something I'm learning now as far as voltage drop. Did consider ballast resistors but Michael is right about current draw. Thanks for all the links I'll check them out! At the end of the day, correct, I don't use the backup lights all that much. Just nice to have everything working correctly if they can work correctly. Interesting Moss doesn't see LEDs for the backup and maybe this is why. There is an English gentleman selling on eBay claiming LEDs working for backup. Mine look identical. Perhaps not. |
Max71 |
Those ballast resistors are for use with LED indicators, to make the 'normal' flasher units flash at the correct rate. Without them the original flasher wouldn't flash at all, and the later electronic flasher units would flash very rapidly. However you then lose the warning feature that tells you when a corner isn't working. If you have LED indicators, what did you do about the flasher unit, out of interest? They will emphasise the effects of any bad connections, and probably not reduce flicker. As said they will negate the lower-current feature of LEDs, as well as making them dimmer than otherwise they would be. What meter are you using? On an analogue you won't see the voltage variation that is causing the flicker, and may not on a digital. Try a temporary earth direct from the battery to the reversing lights to see if that stops it. If it doesn't, then it could still be bad connections on the 12v side. The running lights are fed off the brown i.e. not through the ignition switch or fusebox, and whilst the indicators are it's debatable whether you would see a flicker from the fuel pump when they are flashing on and off anyway, unless you temporarily bypass the flasher unit. |
paulh4 |
When I got the LEDs I had to buy a digital flasher unit so the LED turn signals flash normally. Although I tried to cheap out on the front and get Sylvania LEDs for the turn and they wouldn't work properly. I'm using a DMM with true, fast rate RMS, but as that article accurately states you can't see the flicker on a meter, you need a scope. I can just measure the voltage. From the schematic I see its the green wire and I went ahead and cleaned all connections but I'm sure there's one at the top of the tranny I can't access, as that switch engages the reverse lights. All this was quite fascinating and gave me a more in depth understanding of LEDs for cars. I have a friend with a 1971 GTV that I was going to tell him to get all LED's. Now I'm not sure. |
Max71 |
Max, Attached is the circuit showing the common power between the fuel pump and the reverse lights, within the greyed area. Any or all of the connections in the common circut can contribute to that voltage drop. I will also post a circuit for a possible fix. Herb |
H J Adler |
Herbs fix is what I suggested to a pal who is in the process of fitting a third brake light, using red versions of the 'eagle eye' LEDs I'm using for DRL, which flicker. However it won't do any harm, but 3 amps is way over the top for the diode, the LEDs only take a few milli-amps, a 1 amp would be fine. |
paulh4 |
Having tried it with my DRLs - eight LEDS so I used 220uF - it reduced the flicker but certainly didn't eliminate it. |
paulh4 |
Thanks Herb. I posted that schematic in a different post I believe and did my best to clean all the connections associated with it. I never thought of putting diodes / cap in there. Agreed Paul, 1A should be enough. If it includes a cap I'm kind of surprised it didn't fix the flicker. The cap should hold enough juice during flicker moments. 1K uF should be big enough. If we use a fast recovery diodes and ultra low impedance caps that wouldn't make a difference? BTW, what is the symbol stand for with that circle, two small circuits/line. Is that switch icon? |
Max71 |
I was rather surprised as well. It's probably because the LEDs are still using system voltage. If they could be run at a lower voltage it would almost certainly have worked. "what is the symbol stand for with that circle, two small circuits/line" Not with you, whereabouts is it? |
paulh4 |
The symbol is just before the Backup light symbols. Looks like a circle with eyes and line. I spoke to a friend who's been an electronics technician for over 30 years. Here's his response: You could increase the value of the capacitor to start with. They store voltage, and as in power supplies for amps, they filter out the ripple on the supply voltage. If that's not enough, you can look up "voltage regulator circuits" on google (images), and experiment with that approach. The value of the capacitors can be almost anything, the bigger the better generally, especially on the output side. The output voltage can be chosen (or adjusted) depending on the number of LED's and the desired brightness level. Don't forget to isolate and heatsink the device properly when mounting. |
Max71 |
Do you mean the thing labelled 'Brake Switch'? With a green wire one side and a green/purple the other? A voltage regulator relies on the supply voltage being higher than the circuit voltage, the difference between the two is the amount of ripple or flicker that a capacitor can compensate for. I used that on an electronic brake fluid level circuit, as the audible warning on that used to cheep and squawk depending on what other circuits were operated. That dropped the 12v supply voltage down to 6v for the circuit. As part of looking into how the fuel pump affected other LEDs on my pals car, I photographed the attached scope trace. 'A' represents the zero volt line, and the upper line the 12v supply. 'B' is the instant the solenoid is energised, and 'C' when it is de-energised. It's very brief, but you can see the supply voltage dipping by almost half at 'B'. Then for the rest of the time it is energised, with its six amp load, it is about half a volt down. Whilst in a theoretically perfect world these volt-drops wouldn't happen, the Lucas Fault Diagnosis manual states that volt-drops of 10% i.e. 1v to 1.5v on most circuits are acceptable, the exception being the starter circuit where 0.5v is the maximum it should be. |
paulh4 |
Yes, I thought it might be a switch icon and as you embarrassingly pointed out if I learned to read it was self evident. :) I've been going back and forth with the tech about these comments. His solutions are ones already talked about and require a different LED design. I think a larger cap should work. I'll try to play around with it and see. |
Max71 |
These 12v LED packages contain an internal resistor, as each element only requires around 2v (varies with power and design). LED arrays may wired in series, or parallel, which will determine the value of the resistor required. If you can get into the package non-destructively you have the option of providing a regulator to drop the voltage from 12v to some lower voltage, and then a capacitor between the regulator and the LEDs should do the job. A lot of faffing around though. |
paulh4 |
The trouble with any series component is that they produce a voltage drop, hence less brightness. Another solution is to use a relay. Feed the contacts with clean power, straight from the main brown wire, and the coil via the normal path. Herb |
H J Adler |
"The trouble with any series component is that they produce a voltage drop" Indeed, a diode such as would be added in an attempt to eliminate flicker can result in a volt or more being lost. Being a semi-conductor that is fixed, not dependant on current like a resistor. But it's not going to make a lot of difference to illumination of an LED. If you meant the manufacturers internal resistor in these packages then that is essential precisely because it does drop the voltage, from the 12v supply to the typically 2v max of an elements. |
paulh4 |
Curious if that is a standard points SU pump? If so what type of diode etc is across the points? |
D mckellar |
Cross-purposes, I think. Later SU pumps were diode quenched (now replaced with a bi-directional device), but that's nothing to do with why LEDs flicker with the pump. The diode referred to earlier is in series with the supply to the LED, which together with a capacitor in parallel with the LED, was an attempt to cure the flicker. |
paulh4 |
This thread was discussed between 14/02/2017 and 03/03/2017
MG MGB Technical index
This thread is from the archive. The Live MG MGB Technical BBS is active now.