MG-Cars.info

Welcome to our Site for MG, Triumph and Austin-Healey Car Information.

Parts

MG parts spares and accessories are available for MG T Series (TA, MG TB, MG TC, MG TD, MG TF), Magnette, MGA, Twin cam, MGB, MGBGT, MGC, MGC GT, MG Midget, Sprite and other MG models from British car spares company LBCarCo.

MG MGA - Question Regarding Electronic MGA flasher Relay

Hi folks. My MGA 1500 roadster is having problems with one rear brake light and turn signal. No brake light or turn signal on one side. Not a ground issue. Anyway, I have to do a bit more testing to pinpoint the problem, but more and more, it is looking like the 8 terminal relay box on the firewall may be the problem. I checked the Moss Motors website, and they no longer sell the old style mechanical relay box. They now sell an "electronic" relay box, that presumably uses some sort of computer chip instead of mechanical relays. Has anyone installed one of these electronic relay boxes? Are they reliable? Anyone know what kind of warranty they carry? I understand that if one of these electronic relay boxes is installed, that the usual troubleshooting procedures used with the old relays are no longer valid. Does anyone have an updated troubleshooting routine available? Are these new relay boxes repairable if/when they crap out? (Probably not) Lastly does anyone know of a source for the old style Lucas mechanical relay boxes, or anyone that can rebuild a used one at reasonable cost? Frankly, I think I would prefer to use the old style of mechanical relay if possible. I have managed to clean/re adjust my mechanical relay several times over the years when it stopped working, and like having that capability. Thanks, Glenn
G Hedrich

Glenn-
Have you actually checked for an output from the 8 legger on the appropriate terminal?
Have you studied the diagram and looked inside?
Have you looked at Barney?

No clue what is in the new electrogizmos, but I don';t like it! There may be discrete sub miniature mechanical or electronic relays inside. Sometimes these are in plug sockets, more often soldered to the board - but you can remove them and install sockets for plug in relays.
Electronics people do not understand real world field service; I had some mighty discussions about putting the damned failed relays on sockets in some industrial equipment. Finally just told the suppliers of the $8000 machine to put the new bigger relays in sockets, or else the next time I had to send it back for a 6 week service repair, it would arrive attached to my person, along with a large club. Our entire operation at the time was dependent on that device, and besides the 6 weeks, it cost about $100 to ship it back and forth. Of course, the bigger relays I specified cost a whole $.03 more than the originals, and lasted for about 18 years now with no problem, so I never got to use the $.10plugs!

There are diagrams floating around about replacing the guts of the 8 legged box with several standard relays, and I have read of people who will do this. I have not kept this, since it is rarely necessary, and I could figure it out if I had to, easier than finding the instructions!

FRM
FR Millmore

See http://mgaguru.com/mgtech/electric/ts202.htm
Andy Bounsall

I have a electronic relay box in my TF; have had it for about three years now with no problem.
David Werblow

Actually there are two pages, and the second one is just as important as the first:
http://mgaguru.com/mgtech/electric/ts202.htm
http://mgaguru.com/mgtech/electric/ts202a.htm

The Moss "electronic" relay unit is not really electronic, just two miniature relays on a circuit board, and I think they use the same relays I did.

There is a little irony here. When I first converted mine to miniature relays I expressed some concern for the temperature rating to the relays, being only 55dC (131dF). We should all know that the engine bay runs higher temperatures in warm weather. Later I reported that my hard wired conversion was working fine at 4-1/2 years and 30,000 miles. About six months later Moss Motors was selling the circuit board version of this unit. Immediately thereafter mine suffered a contact failure, and I had to replace the relays.

So my experience, based on a single test unit and a single failure, says that Mean Time Between Failures may be about five years. The original relay unit lasted more than 40 years before it began to have problems. At this rate the "electronic" relay unit may have to be replaced several times to make a similar service life. Only time will tell. If temperature related failures do turn out to be a regular problem, the proper solution is to source high temperature relays, non plug-in sockets and a stash of spares.
Barney Gaylord

Barney-
"non plug-in sockets " Why izzat?

If the problem is burnt up contacts, it is not a temperature problem but current through contacts, with zillions of make/breaks. You could easily have a hundred or two of cycles in rapid succession at a long light, and the contacts are not rated for that abuse; that info is found deep in the technical descriptions for relays and contacts. And the problem is much greater when you use a relay primarily designed for AC in a DC circuit, which is not reflected in the ratings printed on your relays - DC switching ratings are typically 1/10 of AC. (There is a reason arc welders run on DC!)

The cure is to use arc suppressed relays, quite common. Relays usually have schematics printed on, which show these elements. Or, add suppression yourself. Best is varistors, but diodes and/or resistors are also used. The electrical life should then equal the mechanical life. If diodes are used, polarity is important. The advantage of self added suppression is that it works with easier to find and cheaper non suppressed relays, should you ever actually need to replace them.

Another easier approach approach is to use standard automotive relays rated far above actual load, IE 20-40A, which can be had for as little as $2 ea. They are still only 1" cubes, at most, but there are smaller ones with these ratings. But, bigger relays usually indicates greater and therefore faster contact separation, which quenches arcs better, and bigger also gives more heat sink/disperse capacity inside the relay where it matters to the contacts, as well as bigger = more mass of contacts.

FRM
FR Millmore

When I first bought the car, the original LUCAS box was defective. I had brake lights and turn signals on one side only. The small coil on the defective side was found to have shorted its winding wire. After removing this coil, I had it rewound with new coil wire of the same gauge and the same number of turns. I put the unit back together, and hey presto the little gadget worked perfectly. I now have b/lights and turn signals on both sides. Always refer to Barney's site before attempting this operation, I did, and you can't go wrong.
F. Camilleri

The relay does not click on-off with each blink of the turn signals. It clicks on just once and stays on while the flasher unit does the on-off cycle switching. When the turn signal switch is released, then the relay will switch off. Therefore there is only one relay cycle each time you use the turn signal.

There are also separate relays for left and right turn, so they are only used half the time, depending on which way you turn. They are rated for 10,000,000 mechanical cycles, or 100,000 cycles under rated full load and full rated voltage. They are not running under full rated load or voltage, so they may last longer (except for the temperature issue). If you use the turn signals once left and once right every 5 miles, the relays might be expected to last half a million miles minimum.

They are rated for 5 amps per contact at 240VAC or 24VDC, so they are running at half voltage and way less than the rated current. The way I have mine connected the brake light contacts carry less than 1-amp per contact, continuous while you hold the brake pedal down. Front signal light contacts carry less than 2-amps with 50% duty cycle while the signal is flashing. The highest load contacts are those connected to the flasher unit, carrying 3-1/2 amps with 50% duty cycle while flashing. Everything about these relays is fine and dandy, EXCEPT FOR THE MAXIMUM OPERATING TEMPERATURE.

If you plug in a trailer with 21-watt brake and turn signal lamps, then the flasher contacts could see 5-1/4 amps at 50% duty cycle. Done that, and it doesn't seem to cook the relays for short term use. Then again, there was so much resistance in my trailer wiring the it probably wasn't seeing full current. I have since converted my trailer to LED lamps where the additional current is negligible, and the trailer lights are now very bright.
Barney Gaylord

Barney-
You are of course correct re the contacts not operating with the flasher; reckon I have a stuck mental relay!
But, you have a high resistance problem to match, both mental and possibly in the circuit.

It is manifestly impossible to run a 21W lamp on less than 1 A current in a 12V system; the current must be around 1.75 at nominal 12V. But system V is typically above 12V, and lamps are commonly rated at 12.8V for the 21W. Above 12.8 they give more light, so actual load will be 2A plus. If you have less, measured, then you have high resistance in the circuit - and no light.

Further, there are huge inrush currents on a cold filament, up around 10X, so your initial contact load is around 20A.
See: http://cache.freescale.com/files/analog/doc/app_note/AN4049.pdf
There may also be some unknown inductive effects from the coil or other items on the power feed circuit, a big reason for suppressed relays in the first place.
Data sheet for relay:
http://datasheetz.com/data/Relays/Power,%20Over%202%20Amps%20@%2030%20VDC/OMI-SS-212D,500-datasheetz.html

It is for these reasons that I always use 40A relays for headlamp loads, which are nominally 10A for two H4 lamps. And I note that many relays are double rated, at say 20A make/40A break, for these reasons.

I will note that your analysis is precisely that of the Branson Ultrasonics engineers I mentioned in my story. It was a listed option on the equipment, but an uncommon one which I needed; they had little or no actual experience with it, a shortcoming I soon rectified. My contention was that without trying to account for all the possible factors, we had a failed relay due to obvious overloading, and the simple fix as to use a better rated one, and make it easily changed.

I said to them and I say to you my opening line: "If the problem is burnt up contacts, it is not a temperature problem but current through contacts". The fact that you had an intermittent failure, circumvented by repeated operations, indicates that you almost certainly had bad contacts, not a temperature related electrical failure. Most such devices have welded internal connections, so the sole electrical failure would be coil insulation breakdown, likely to be permanent fatal after the first fail.

FRM
FR Millmore

I think I missed an earlier question from FRM. -- Why no plug-in relay sockets? In this case with two relays it would add 16 more push-together electrical connections to the electrical system in a automotive environment, meaning 16 more things that could possibly go wrong in the future. It is similar to having too may snap connectors in the wiring harness, and I suspect everyone on this party line can relate to that. If you have good relays, it is likely that the plug-in socket would fail before the relay would fail, meaning a reduction in the overall reliability of the system.

For the newer message, I didn't say less than 1-amp per bulb. I said the way I have mine wired I have less than 1-amp per relay contact for the brake lamps. Reason is, there are two relay contacts in parallel sharing the load for each bulb.

For a 21-W bulb to draw 2-A requires almost 14-V supply. With the size of wiring in the MG it is very unlikely to ever have 14-V at the bulb socket, even when my alternator is putting out 14.5 volts.

I didn't need a 7-page tech article on MOSFETs and Xenon lamps. Do you have a little chart like that showing inrush current for incandescent bulbs? If the inrush current lasts 10 to 20-ms it may be just about enough time to warm the relay contacts up to normal operating temperature. This is not like inductive motor start up.

The data sheet link shows "404 Page Not Found". Not sure what you had in mind there.

I didn't dissect the failed relay, so I'm not sure it was a "burned" contact. A relay contact might fail for other reasons related to operation above the rated temperature limit, like warping of the contact carrier mounting, or shorting of some windings in the coil that may reduce the pull-in force without actually breaking the coil circuit.

Since I am not manufacturing a million of these per year, it seems prudent to slap one together in home garage style and see how it works. Now I have 8-years and 52,000 miles of real world testing and so far only one relay failure (which is statistically entirely inconclusive, except to prove that it is possible to have a failure).

Frankly, with 131dF max operating temperature rating I was rather surprised it didn't have a meltdown on the first hot spring day. I say give it another five years and see if it fails again. I believe the Moss Motors "electronic" units use the same relays, so perhaps everyone who buys one of those will be a Moss Motors field tester.
Barney Gaylord

"I think I missed an earlier question from FRM. -- Why no plug-in relay sockets? In this case with two relays it would add 16 more push-together electrical connections to the electrical system in a automotive environment, meaning 16 more things that could possibly go wrong in the future. It is similar to having too may snap connectors in the wiring harness, and I suspect everyone on this party line can relate to that. If you have good relays, it is likely that the plug-in socket would fail before the relay would fail, meaning a reduction in the overall reliability of the system."
>>A spurious argument, one I know you are fond of. These are static connections, and if correctly assembled with appropriate grease, eg OxGard, they do not degrade with time, by my measurement.

"For the newer message, I didn't say less than 1-amp per bulb. I said the way I have mine wired I have less than 1-amp per relay contact for the brake lamps. Reason is, there are two relay contacts in parallel sharing the load for each bulb."
>> OK, I missed the double contact; however, it is likely that a single contact makes the initial connection, and that one is the problem. Until it arc burns and then the other one goes to work.

"For a 21-W bulb to draw 2-A requires almost 14-V supply. With the size of wiring in the MG it is very unlikely to ever have 14-V at the bulb socket, even when my alternator is putting out 14.5 volts."
>> I routinely achieve and expect that measured voltage anywhere in the system is within 0.5V of alt+ to earth, or Batt+ to earth, depending on whether or not it is running. The only significant alteration required is relays for the headlamp load - in plug-in sockets, connected to the harness by standard Lucas bullets.

"I didn't need a 7-page tech article on MOSFETs and Xenon lamps. Do you have a little chart like that showing inrush current for incandescent bulbs? If the inrush current lasts 10 to 20-ms it may be just about enough time to warm the relay contacts up to normal operating temperature. This is not like inductive motor start up."
>> ALL the charts are for halogen bulbs H1-55W. Halogen bulbs ARE incandescent. There does appear to be a misleading fact in that the expected draw for one such bulb is c5A, but the charts show 10A, so it appears that whatever system they are discussing has two bulbs in parallel, as you might expect in a headlamp system.
I repeat, this is not a constant current heating failure, but a make/break arcing failure. That is usually what degrades contact performance over time.

"The data sheet link shows "404 Page Not Found". Not sure what you had in mind there."
>> Right you are, and I have no idea why, as that was a copy/paste from the URL bar when I had that document up. The Datasheet is NLA from the manufacturer (Tyco), as the relay is obsolete by them; there are many listings for the data sheet by search on the part number.

"I didn't dissect the failed relay, so I'm not sure it was a "burned" contact. A relay contact might fail for other reasons related to operation above the rated temperature limit, like warping of the contact carrier mounting, or shorting of some windings in the coil that may reduce the pull-in force without actually breaking the coil circuit."
>> Maybe. All the modern relays I've taken apart have steel frames and all welded construction; case issues are usually not much.

"Since I am not manufacturing a million of these per year, it seems prudent to slap one together in home garage style and see how it works. Now I have 8-years and 52,000 miles of real world testing and so far only one relay failure (which is statistically entirely inconclusive, except to prove that it is possible to have a failure)."
>> Entirely agree. My approach says use the biggest relay I can, if the cost is minimal and space is not an issue. I hate fixing things twice.

"Frankly, with 131dF max operating temperature rating I was rather surprised it didn't have a meltdown on the first hot spring day. I say give it another five years and see if it fails again. I believe the Moss Motors "electronic" units use the same relays, so perhaps everyone who buys one of those will be a Moss Motors field tester."
>> I think the max op temp is derived from coil heating at spec temp in continuous load, which would be far greater than what you have here. And in the other direction, that the temps are based on free air, not an enclosed case. I have some experience asking these questions of manufacturers, and that is a normal sort of case. You can commonly utilize components at double the ratings once you know the limiting factors; sometimes that results in a reduction in expected life, say from 10 million to half that - who cares?. In one case, a varistor was rated at just below what I needed, but the special order hi temp one would maybe have netted a 10% life increase at max temp/max conceivable load, and I was not near max load even though my temps were possibly/occasionally over the rating.

FRM
FR Millmore

Here's a simple question. The relay is rated for 100,000 cycles with 5-amps at 24-VDC. Does running it at 5-amps with only 12-volt load (half the power throughput) increase contact life? If so, how much?

Some day I should put a temperature probe on the relay cover to see how hot it really gets while driving. I'm still surprised it lasted five years.
Barney Gaylord

I would expect the contact life to be essentially unaffected by pure power load, assuming the contacts were "as new". I would expect the arcing deterioration of contacts to be slower for lower voltage, all else being equal. Once the contacts start downhill, a direct reflection of arcing however faint, constant current losses and heating would be proportional to contact resistance, and arcing on make/break would generally increase as the contact surface becomes less smooth. That's why you "clean" = file points.

The actual power dissipation by Joule's law is P = I[sqd)R, so for constant current it would actually be directly proportional to R. Since greater R translates to less I at constant applied voltage, the actual power dissipation could decrease sharply as R increases, a self limiting condition that probably saves our butts more often than we know - but it gives crappy light or heater blowers. A complication comes in the fact that we are taking about one resistance in a chain though, so E (V) across the resistance of interest varies too, by the rules of series resistances.

Other forms of Joule's are: P=IE or P=E{sqd}/R, but note that E is applied voltage across a single resistance, not system voltage.

It is all complex and sometimes not intuitive. I have seen relays of say, "30A", with contacts as small as a pinhead or nearly 1/4" diameter, and they are made of various materials or material pairs, which all affects this. But bigger usually means bigger contacts and mountings to act as heat sinks, and bigger separation to quench arcs, and heavier spring loads for better contact pressure = lower resistance over time. Hence my use of "over rated" relays, which has empirically been successful - so far. For aircraft or spacecraft or F1 cars, might be a different story.

Mouser show the relay in a 10A form as special order 1640 minimum. $1.53 ea.
datasheet (actually for the relay family but not showing this exact one! Why? But it does show contact life for loads at 24 & 30VDC to give you an idea. The ratings are for pure resistive loads, not filament loads, and do show 100,000 ops at rated contact load for those conditions. Maybe this one will work.
http://www.mouser.com/ProductDetail/TE-Connectivity-OEG/OMI-SS-212D500/?qs=8wHch9UpSvYWE6g259Ib/rJfWm6Gh9Iv

Just remembered (and found), I have a splendid .pdf Bosch relay application design catalog I'd be glad to send you, don't have the URL where I found it. Typically, lamp loads are 1/2 to 1/3 of resistive loads for rated life, or the inverse. But, some relays show no difference, while others show lamp loads as Not! Good little charts of current characteristics at switching. And the rated life is much greater than the Radio Shack ones.
Unfortunately, Bosch sold their relay operation to Tyco, which now owns the entire world relay business, much as Federal Mogul owns all the engine parts. But I think the part numbers are still good.

FRM.
FR Millmore

Would it be possible to add a capacitor with a resistor across the contacts to try to combat the arcing, thusly?

http://relays.te.com/appnotes/app_pdfs/13c3236.pdf

http://www.industrologic.com/mechrela.htm
Del Rawlins

Del-
Yes, but you really need to do the calcs. When I was doing this, I came to the conclusion that the most versatile method was MOV or varistors, mentioned near the end of your first exhibit. These were then made exclusively by Harris. Since sold to Littlefuse, and also produced by a zillion third world companies.
Harris had a superb book, free for the asking, which may still be available online from Littlefuse. I studied it a lot, as we were having voltage spike problems that were causing random mayhem, on one occasion life threatening - my life at that! I am the last customer on an REA line, the worst possible position to be in, and of course have machine tools and compressors in addition to the computer operated equipment we were building. Surge suppressors for computers were expensive and hard to get, unlike now.

So, I have varistors all over the place - in my breaker box, outlets with motors or computers, etc.
Among other items, you can have transients exceeding 1000V in automotive electrical systems, which really raise hell with electronics. Lucas alternators blew up regularly at first, then Lucas added suppression, but the suppressors are NLS and I've yet to find an alternator shop who knows what "them funny lookin' things" are. Usually when you get a rebuilt, the suppressors are gone or the leads are clipped. Since they only cost pennies, I have a lot of them, though I am now a bit rusty on application data.

An interesting point is that surge suppressors are most effective very near the source of the disturbance, but do protect the entire circuit to some degree, so more is/are better.

FRM
FR Millmore

This thread was discussed between 22/07/2012 and 24/07/2012

MG MGA index

This thread is from the archives. Join the live MG MGA BBS now