Automotive Air Conditioning Information Forum (Archives)

Provided by www.ACkits.com

We've updated our forums!
Click here to visit the new forum

Archive Home

Search Auto AC Forum Archives

Relay Coil Resistance

Gary Harrison on Sat December 23, 2006 12:51 PM User is offline

Year: 1994
Make: Honda
Model: Accord EX
Engine Size: 2.2L

I have had some recent problems w/ my son's 94 Accord EX Main Fuel Relay (MFR). The MFR is a constant source of problems for Hondas and Acuras. I replaced the relay w/ a spare relay that I thought was OK, but it soon exhibited the same problem; failure to start after sitting in warm sun for several hours.

As a temporary fix, I hard wired the relay to bypass the relay function allowing the car to be used. I now have a new relay and intend to install.

My first question is about what is failing on these relays. Many attribute the Honda MFR problem to solder joint fatigue and have successfully repaired the relays by resoldering the solder joints. However, I have cleaned and resoldered every joint w/ no change in the problem.

I measured the current demand for the relay and it seems OK to me, about 330 mA. However relay gets very warm ( ~ 150 F) on metal part of the relay. This power calculates to about 4 watts at 12V. Is this enough heat output to warm a relay part to these temps? This amount of self-generated heating seems high to me. However, no-start problems usually occur after sitting in a parking lot warming for hours and occur at first start, not after fully warming the part.

I measured resistance of the coils and got some puzzling (to me) results; different resistance on different resistance scales of VOM. I believe this may have something to do w/ the diode in series w/ the relay coil (constant voltage drop across diode?), but my electrical eng. knowledge is a bit fuzzy here. Any help? I got 640K ohms on 2K resistance range and 5.3K ohms on 20K resistance range of my digital mutlimeter. Multimeter manual says a constant current is directed across a reference resistance to measure voltage proportional to resistance.

Another puzzling thing is I got same resistances across non-working and working MFR's. This makes me think the problem is something other than the coils, but there's not much to go wrong except the coils, the series diodes, or solder joints.

Other question relates to design of these relays. Why not use a solid state relay? It would generate much less heat and should be infinitely more reliable. Honda/Acura MFR's have left untold number of motorists stranded in sunny parking lots. This seems to be an ongoing problem for Hondas and Acuras through the 90's and may still be in newer products, but I don't know for sure.

This is more a matter of curiosity that necessity, but I know you guys love a puzzle and hoped you might offer some possible solutions.

Regards

-------------------------
See you down the ROW.

NickD on Sat December 23, 2006 1:22 PM User is offline

Most control automotive relays I have tested including those on 80's Honda had a coil that drew 100 ma at 12 volts for a coil resistance of 120 ohms. Measure the resistance directly across the coil, not with a diode in series, must be an isolating diode of some type. I find a coil drawing 300 ma to be a bit high, the relays with 100 ma coils have contact rating of 30 amps and seems like if Honda is having problems with this relay, look elsewhere. Can't imagine the fuel pump drawing much over 4 amperes, some way over pushed GM fuel pumps pull as high as 8 amperes, but still far less than 30 for a contact rating.

If you are measuring your coil resistance with a diode in series, you will get crazy readings on different ohm scales, nothing unusual about this. I expect a 12 volt relay to pull in at 6-7 volts as a safety margin, when cranking an engine the starter on a cold day can pull the battery voltage down to less than 9 volts and if you need that relay to run your pump when cranking, then your car won't start if the pull in voltage is higher. Relays have been around now for 150 years, shouldn't be a problem, only problem I have had is dirty contacts.

I am dumping my motorhome AC clutch relay in favor of all solid state, easier to wire, clutch coil draws 4 amperes, easy to do.

Gary Harrison on Sun December 24, 2006 10:34 AM User is offline

thanks Nick

Yes, I was measuring resistance w/ diode in series of necessity. I can't see the component side of the circuit and don't know exactly where to place probes to measure coil only resistance. I might find the correct points by trial and error.

Regards

-------------------------
See you down the ROW.

NickD on Mon December 25, 2006 8:02 AM User is offline

Depends on how many contacts on the relay in question, a fuel pump relay only needs one set of Form A contacts, the movable armature contact, and a fixed normally open contact. Coil contacts are generally parallel to each other and located on opposite sides on the relay.

Most covers can be removed to see the contacts and easy enough to trace where the coil leads are, but able to remove even a crimped on metal cover, clean the contacts, check the wiring, and recrimp the cover back on. Some plastic covers are lightly glued on that can be snapped apart with a utility knife, snapped back on with a touch of RTV on the seam.

Hardly ever replace a relay unless it is fried to nothing, in the 50's and early 60's, my company was making computers with all relays, preferred over the vacuum tube for more reliability and literally thousand of relays were used. Telephone exchanges also had thousand of relays and stepping relays that would follow your rotary dial, even back then, these cost 150 bucks a piece, and if you dialed a seven digit number needed seven of these, each one stopped at its number location to give a unique voice path through the closed contacts. Work with Bell Labs to replace those troublesome costly things with touch tone that was much more reliable and a lot cheaper, but they still increased the price for touch tone if that makes any sense. Ha, just like plasma and LCD TV's today, cheaper to make than a vacuum tube, but selling for five times or more as much, marketing.

Electronic components are made in huge production runs, just sounds like your relay has a shorted coil fault in which case, all the relays have that same fault, use a different brand.

Gary Harrison on Mon December 25, 2006 9:15 AM User is offline

All the relays used by Honda/Acura are made by Mitsuba and have been for 15 yrs or more. Given the regular problems, it's hard to understand why Honda hasn't changed manufacturers. Even aftermarket parts suppliers buy the same OEM relay and resell. Only way to change brands would be to "roll your own" from components using the Mitsuba base connector. The problem is not disastrous, sometimes a relay will work for years,... and then leave you stranded in some parking lot when you least expect.

I can't see the component side of relay assy w/o desoldering the connector base from the component circuit board. I don't want to do this on my new relay. I'll try to puzzle out the positions for determining true coil resistance.

regards

-------------------------
See you down the ROW.

NickD on Mon December 25, 2006 2:46 PM User is offline

You means this relay doesn't have a socket and is hard soldered in?

FrankD. on Mon December 25, 2006 3:44 PM User is offline

Gary can you post a picture of the relay?



-------------------------
FrankD.

Gary Harrison on Mon December 25, 2006 8:35 PM User is offline

Mitsuba Main Fuel Relay for Hondas & Acuras

Here's the relay. The connector base is soldered to the trace side of the circuit board, and a shroud encloses the relays and components are on the opposite component side. It' impossible to see the components, except for small parts of the relays.

This is same relay used from 1990-1998 to my knowledge, and beyond. See similar relay from 2000 Accord 2.4L engine.



Regards

-------------------------
See you down the ROW.

NickD on Tue December 26, 2006 12:00 AM User is offline

That has got to be a socketed relay with a keyed connector, sure just the socket isn't soldered to the PCB, can't you just unplug the relay?

Have a very easy to pop off dust cover so you can examine the contacts, only four terminals, two for the coil and two for the normally opened contacts, just about the most simplest relay made.

Gary Harrison on Tue January 02, 2007 7:45 PM User is offline

Sorry, I'm not getting notices of posts.

The cover comes away easily, but the component side of the circuit board is between the circuit board and the base (connector). You have to desolder the base from the circuit board to view the component side. I didn't want to do this on my new relay and I don't have an old one w/ me just now. I will try again later. I have two relays that won't work.

regards

-------------------------
See you down the ROW.

NickD on Wed January 03, 2007 8:17 AM User is offline

I am not getting the picture, Gary, are you saying you cannot unplug the relay, and are you sure the relay is the problem? Is it getting the signal to energize the coil from the PCM?

And when you state the relays are bad, what is bad about them? Do they click with 12V applied to the coil? That is the worse case, an fried or open coil. Had one of those, a rare can't find anywhere, including the FAA stock of a current sensitive relay that was keeping a very expensive to replace visual approach system from working. Had to no choice but to disassemble that relay and rewind the coil to precise specifications. The armature can be cleaned and lubricated and even the contacts replaced.

Had to order 30 new relays for another old system, ha, reburnished the contacts on these relays just to many times, not much left, time for new ones. Ironically, the contacts are rated at 30 amperes just like your little relay, except these contacts are 3/8" diameter and yours less than a third of that size. Automotive has their own rules and no wonder why we have problems.

So what is wrong with your relays, contacts, coil, armature stuck?

Gary Harrison on Wed January 03, 2007 8:36 PM User is offline

The internal ends of the connector male contacts that you can see in the picture penetrate the circuit board at 8 locations and are soldered to the circuit board. I attempted to remove the contacts from one of the non-working relay circuit boards earlier and even w/ a solder sucker it was impossible to clean well enough to release the circuit board. I finally resoldered all the connections hoping a bad solder joint was the problem. Same result when installed, cold starts in morning are no problem, but no start in hot sun.

The relay that drives the fuel pump fails to close. I found what I thought might be a high draw for the coil and the coil definitely gets very hot (~ 150F), even though relay is apparently closed. However the problem usually originates when a start is first attempted after sitting in the warm sun, so not clear if high current draw is the problem or not.

Regards

-------------------------
See you down the ROW.

NickD on Thu January 04, 2007 5:47 AM User is offline

I only see four terminals on that relay that seems right, two for the coil and two for the contacts.

Are you sure you have a Honda and not a Ford? That module mounting the fuel pump relay sounds like the Ford ICM all riveted together with the fuel pump, fan control, and AC clutch relays all soldered to a PCB so if one stupid relay goes bad have to get a bank loan to replace the entire module.

I have a Weller temperature controlled desoldering station with a real vacuum pump, works like a champ, got that after fooling around with those cheap solder suckers that suck in more ways than one, good for lifting the lands off a PCB, most are not temperature controlled and will burn themselves out if left plugged in too long.

Gary Harrison on Thu January 04, 2007 8:48 PM User is offline

Click on the link to get the 1st picture. This is the correct picture. The 2nd picture is not the relay I've been describing. Actually, there are only 7 pins and not 8 as I earlier described. One is missing from one row.

regards

-------------------------
See you down the ROW.

Dougflas on Thu January 04, 2007 10:12 PM User is offline

Is it possible the heat from the relay welded itselfto the mating plug on the board? If you have another relay, why not take a dremel tool and grind the relay away. Then you can unsolder each lug one by one. Relay contacts get hot by either poor connections or by heavy loads.

NickD on Thu January 04, 2007 11:05 PM User is offline



Relay on the left, it's mating socket that should be soldered into the PCB on the right that is keyed so the relay will only plug in one way.

No reason to remove the socket that would require desoldering from the PCB unless it is badly burnt, since it shows male spade lugs, the relay end should be female and recrimped for a tight fit and the soldering of the socket pins to the PCB should be good. Removing the dust cover from the relay will reveal how each pin is wired by a visual trace.

Gary Harrison on Fri January 05, 2007 11:38 PM User is offline

I will try to get a picture to post. Sorry for the confusion.

regards

-------------------------
See you down the ROW.

NickD on Sun January 07, 2007 6:50 AM User is offline

In the 80's and to model year 1990, all my family members and me had Honda's until the Japanese devaluated the buck, so I haven't kept up. Seems like what made the Honda's back then special slowly disappeared and they are making later models like everyone else with new problems. Perhaps why I purchased a new 2004 Cavalier, brand new price was cheaper than a 97 Civic sitting on the dealers lot with a zillion miles on it. In my opinion, the Honda is not the same car as it was in the 80's, but seems like the buying public hasn't caught on yet.

Didn't shed too much of a tear when I drove my 86 to the wrecking yard three years ago, unibody was so badly rusted, couldn't jack it up anymore, but that 300 foot of vacuum hose under the carbed hood was a nightmare to work on. Have to say my 92 DeVille is far superior to anything Honda ever made, easier to work on and even gets better fuel economy for a much larger car than was my 85 Accord SEI got. But GM couldn't leave the Caddy alone and went downhill from there. Ha, recall spending a full day just trying to change the thermostat on that Honda, with a zillion things in the way, took less than an hour with the DeVille.

Yeah, post a picture, got me curious as to what you have, seems like something like a relay shouldn't be a problem.

Back to Off Topic Chat

We've updated our forums!
Click here to visit the new forum

Archive Home

Copyright © 2016 Arizona Mobile Air Inc.