Welcome to our resource for MG Car Information.
MG Midget and Sprite Technical - Electrical problem – need expert help.
|I apologize in advance for the long and complex posting that follows:|
Ive put about 500 miles on the Ankle-biter since getting it back on the road after 38 years. Woo-hoo! It runs and handles surprisingly well.
Ive tried to add a 12v power outlet and radio, but all is not well with the electrical system.
For background, Ive replaced the dynamo with a Mitsubishi alternator, negative ground conversion and installed a new wiring harness. The +12v for the add-ons is taken from the brown wire (added fuse) under the instrument panel that is designated for the radio/cigar lighter in the wiring diagram.
A) I purchased a cheap 12v battery and alternator tester (http://www.harborfreight.com/12-volt-battery-alternator-tester-66842.html). This plugs into the power outlet and has three LEDs that show the battery charge and three to show the alternator function. When I plugged in the tester all three lights lit showing a full battery charge. When I started the engine the lights for the alternator lit for a second and then went out, as did the lights for the battery, leaving only one LED that supposedly shows the battery is discharged.
Thinking the tester was defective I brought it back and got another and tested it in my everyday car on the way home. It worked fine. After putting it into the midget I watched more closely. After starting the engine, the alternator lights came on indicating normal voltage with a slight blinking on the over-charge indicator. As I closed the choke and the engine rpm lowered, the over charge indicator blinked brightly a few times and all of the LEDs went out except the one that showed a discharged battery. I cooked the second little voltage tester.
B) The radio works great when the engine is off, but as soon as I start the car the radio begins to blink on and off.
Because the testers were destroyed and the radio was acting in a very funny manner I thought perhaps the alternator was to blame, more specifically the voltage regulator (the original control box is completely out of the circuit). I plugged a cheap digital multi-meter into the power port. The display was flashing showing different numbers anywhere from 0 to over 100 volts. I then got out my older analog meter and the needle sat very steady at about 14.5 volts with the engine running. My assumption is that the cheap digital volt meter was showing instantaneous readings and the analog meter was showing the average voltage. I dont know if that is true, but I dont have any other explanation.
C) I took the alternator out of the car and brought it to an automotive electrical service facility that repairs alternators. They bench tested the alternator with an oscilloscope and said that everything was working as it should. The only explanation they could give for the odd behavior was that the case of the alternator was perhaps improperly grounded and the voltage was spiking because regulator was not receiving the correct input voltage.
Back home again, I cleaned the brackets and reinstalled the alternator. I also put a screw into the back of the case and wired a jumper directly to the negative terminal of the battery making sure there was a good ground to the alternator. I started the engine and the radio and voltages were the same as before.
Next I disconnected the +12 volt and ground connections for the radio and wired them directly to a spare battery that was not installed in the car. The speaker wires and antenna hookups were not changed. The radio worked perfectly with and without the engine running.
Lastly, I jumped the +12 volt and ground wires for the radio directly to the car battery. Engine off - everything was fine. Engine running - the radio was again blinking on and off.
Next, I got my hands on a PowerProbeIII (http://www.powerprobe.com/pp3.php) and measured average voltage, positive peak to peak voltage, and peak to peak voltage.
Average voltage with the engine running measured 14.6 to 14.7 volts. Interestingly, the higher voltage was at the slowest idle.
The positive peak to peak mode captures the highest detected voltage. While probing the +12v connector (radio disconnected) I would find between 15 to 15.4 volts. However on a couple of occasions (I tested the circuit over and over for about 5 minutes) the voltage peaked above 16 volts. Also, with the probe tip floating (not contacting a circuit) the display should read 0.0 but it registered between .2 and .6 volts.
The peak to peak mode measures the difference between the positive and negative peak voltage levels over a 1 second period. With the engine running the voltages vary from .2 to .6, but then on occasion spiked to 5.3, 8.8 and once to 11.3 volts, again over a several minute time period. A normal peak to peak reading while testing a charging circuit is usually under one volt according to the Power Probe manual.
One last thing, I have an electric fan. When the coolant rises to the fan-on temperature, I can hear the rely switch on and off rapidly, perhaps two or three times a second. That is not normal either, I assume the relay should switch on once and then not operate again until the coolant temperature drops.
If the alternator is operating properly, as I am told is, then Im stymied.
The next thing I did was disconnect the three electrical connections to the alternator, thereby removing it completely from the circuit. I thought if the alternator was the problem then everything would run okay. It did not. The radio still cut in and out and the peak voltage spiked to 15.7 volts. The voltage from the battery was 12.7 while the engine was running with the alternator disconnected.
Could the electric fan or fan controller be the problem? I disconnected the +12 to the controller, removed the fan relay from the base, and removed the fuse to the fan motor. After starting the engine again the problem persists, radio cutting in and out and peak voltages showing near 16v. I dont see how that is possible. The guy at the alternator shop suspects a bad battery. I have a new WestCo AGM battery installed. Everything works fine when the car is not running, no voltage spikes, radio is perfect. Ive tried tapping the battery with a rubber mallet to simulate vibration but no spikes. Only when the engine is running.
The only other thing that I can think of would be the coil or the condenser. The alternator shop says neither would cause the problem. Other than buying a new battery I dont know what to try.
|Paul, one thing that you don't mention is the rev counter. I assume your 1098 uses an electrical rev counter that is in series with the LV ignition wire and therefore also the alternator input wire. I don't have a proper explanation, but just wonder if that could be the source of your problem?|
Its a mystery!
|I would certainly consider disconnecting the green feed to the rev counter and see if it makes a difference.|
I also have a vague recollection of someone else having a problem with one of these plug-in testers, although it performed normally in a 'modern'.
|Dave O'Neill 2|
|wow this is an interesting one!|
It does sound like the ignition circuit is inducing voltages somewhere along the way. Try bypassing all the cabling related to the +ve side of the ignition by running a wire directly from the coil to the dizzy (bypassing the tacho). If that "fixes" the problem you are one step closer to finding a fix :-D
If it doesn't, you're on your own ;-D
|oops, ^^^ that should read -ve side of the ignition.|
|Are you using unsupressed ht leads - ie wire cores?|
I've had the most awful issues with that recently, in my case affecting electronic ignition and cpu unit. As they are at the centre of where all the wires pass I just wondered, as a long shot, whether this might be an issue.
|I'm not familiar, with the Mitsubishi alternator. Are you saying that the o/p is unrectified unregulated AC?|
What exactly did the oscilloscope trace look like at the auto electricians?
Modern alternators usually contain both the voltage regulation and rectification circuits. Hence you won't read peak to peak voltages. You should be reading recified DC with zero to peak voltage peaks.
Do you have access to a Lucas alternator to install a a test?
PS, you should be careful about running the alternator without any connections to it. You might damage it.
|The mystery continues. I took the green lead off the tacho and no change. I removed the white lead from the coil and jumped the coil back to the battery, thereby removing the wiring and the switch from the circuit. No change. I have been running the tests with the alternator disconnected and my battery voltage has dropped to 12.2, but the highest peak voltages are still reading in the upper 15 volt range and I am getting 4-5 volts difference on the peak to peak reading. |
I have an AGS battery from a golf cart that is the same size, amp-hour, etc. as the Westco battery in the car. But it does not have the round posts so I cannot switch the batteries. I tried removing the cables at the battery terminals and using jumper cables to connect to the spare battery but I could not get enough amperage through the connections to engage the starter so I have not been able to test the theory that the battery is defective. Other than buying a new battery, I am not sure how to test that. The battery from my other car simply wont fit.
Lawrence, yes, I should be reading rectified DC with less than a volt difference in peak to peak. I am reading .4 to .5 volts and then it jumps to several volts. The alternator guy would be perfectly happy to replace the voltage regulator and diode rectifier if I wanted him to but he says both are working perfectly fine. Since the alternator is currently out of the circuit, I dont see that as the cause.
Last time the engine was running I noticed that the spikes seemed to occur about the same time the fuel pump operates. I have a new SU negative ground pump and it makes a kind of thump, thump, (rather that the original tick, tick) when it operates. Could it be there is a fault in the wiring or the pump that would cause a significant voltage draw when it starts? When that voltage draw is released could that cause the battery voltage to spike? I am not sure how to test that. If I disconnect the pump the engine wont run for very long. Perhaps I could rig a small container and suspend it to feed the carbs by gravity. I will give that a go and see what happens.
|Very odd symptoms,|
With the alternator disconnected there is no way you'll ever get more than battery voltage in the circuit anywhere.
Sure you're meter is working correctly?
Mine gives really odd readings when it's battery is low.
|SR Smith 1|
|Okay, so it is not the fuel pump. I disconnected the wire at the fuel pump after filling the float bowls and started the engine. The engine ran long enough for the spikes to occur and the radio to cut in and out. |
At this point I have removed from the electrical system the alternator, the tachometer, the fuel pump, and wired direct from the battery to the coil and the problem persists.
Things I have not done include replacing the battery, the coil or the condenser. I am running solid core wires with suppressor clips and standard champion plugs. I have a side-entry distributor cap.
I'll install the old coil that was removed prior to the rehab and see if that has any effect.
|SR, yes I am confident that the meter is working properly. Additionally, the radio is behaving oddly and I have destroyed three of the little plug-in voltage testers. There is definitely something wrong. |
I have been told that I cannot possibly be getting voltage spikes higher than the battery voltage, yet I am. ??
If you think it's the battery, either fit another or link in another battery via jump leads. You don't need to start the engine, if the alternator wires are detached anyway it'll be no different.
I've been an auto electrician for 30 years and have heard of this sort of thing. There must be a simple explanation.
|SR Smith 1|
|You can get more than the battery voltage as spikes, that's the way the ignition works and all you need is an inductor and some way of making and breaking the supply to it.|
You have already discounted the pump. Other inductors are the wiper motor, electric cooling fan, heater fan and of course the ignition coil. Most of these can be taken out of circuit to see if they are the cause. The ignition coil can't be taken out of circuit. I'm just wondering if the condensor is dying causing spikes? It's also worth putting the old coil and leads back just in case. New coils seem to be notoriously un-reliable.
Modern voltage meters can react very easily to spikes and if asked to measure peak volts will do so with spikes that have a surprisingly fast rise time.
Also, the voltage regulator feeding the instrument supply does make and break as part of the way it works. I'm just wondering if a strange earthing fault behind the dash could be implicated some way?
|Rob aka MG Moneypit|
|Rob, the voltage regulator making and breaking is not a likely cause. I have the alternator and the old control box completely disconnected as well as the cooling fan and controller. |
I just put the old coil back in and that is not the cause either. I do have a new condenser in hand and I am going to replace that next. Regarding the heater fan and wiper motor, both are turned off. Could they possibly cause a spike when the circuit to them is open?
Should the condenser prove not to be the problem, then I think I am down to the spark plugs, spark plug wires and battery. All of which are new.
Before replacing the new battery, I will go for the less expensive option of getting new spark plugs and plug wires. What is the best option here? I have solid core wires with suppressor clips which I guess I should replace with suppressor wires. Can I use the suppressor clips with the suppressor wires? Even though this a rehab and not a restoration, I want to keep the original look as much as possible.
A recommendation on the type of plugs and wires would be helpful to me.
I believe that Rob was referring to the voltage stabiliser for the gauges, not the regulator for the dynamo
|Dave O'Neill 2|
|Dave, I never knew such a thing existed. Where is it located? I have spent many hours going over the wiring diagram and don't recall ever seeing it. |
|Ah, as you have a 1098, you probably have the early fuel gauge which doesn't use the stabiliser. |
I would have to check when it was introduced, but I'm away from home for a couple of days.
|Dave O'Neill 2|
|Dave, in the Horler book on page 105 he says, "The fuel gauge was also wired through a bi-metal voltage stabiliser to prevent rapid oscillation of the needle." This is in the chapter on the Sprite MkIV and Midget MkIII. |
I have a MkII so I guess I don't have one. I don't know how you guys remember all of these details. I never cease to be amazed.
|If you did have a gauge voltage regulator, it would be here ( white circle, UK car) and look like this.
|SR, I was reading over the posts again and you said, "You don't need to start the engine, if the alternator wires are detached anyway it'll be no different." |
Yes, it does make a difference. If the engine is not running everything works perfectly. Only when the engine runs does the problem appear.
I detached the cables from the installed battery and used jumper cables to connect them to a spare golf-cart battery of the same spec. The starter would not engage. The voltage was there, but apparently the cable clamps did not provide an adequate connection to provide the needed amperage to engage the starter. I need to make a set of cables with the proper connections for both ends so they can be tightened down properly. I don't want to link to the Midget battery, I want to remove it from the circuit altogether for a reliable test.
I don't have one. Thanks for the photo. If I ever do see one at least I will know what it is.
|Yup, I know you don't, you said earlier. I just thought you would like to know what and where if you did. :).|
|I found the cause of the voltage spikes. After removing or bypassing everything possible in the electrical system I just started replacing things. |
After swapping the standard Champion N5C plugs with resistors (NGK G Power 7082) there was no change. Next up were suppressor wires to replace my solid core wires and suppressor clips. Not wanting to buy things I didn't need I searched my junk box and found the plug wires and distributor cap that was on the car when it last ran 38 years ago.
As luck would have it, they were suppression wires (Packard TV R 2Q-71)with standard clips. I swapped out the new dizzy cap and solid-core wires with the old dirty, greasy suppression set. The engine started right up, the radio was no longer cutting in and out, and the voltage spikes were gone.
I've read in the archives that the side-entry caps should not be used with suppression wires because the wires won't hold up. I don't know how long those wires were in use, but I drove the car from '72 to '76 and they were on it when I bought it.
I found a set of Bosch wires that are claimed to work with both screw-in and push-in caps (http://shop.advanceautoparts.com/p/bosch-ignition-wire-set-f102037181bos/10809348-P?searchTerm=plug+wires) and I will give those a try. If they work as well as the old Packard wires I'll be very happy.
|Just a heads up tip for checking the battry in the future|
The usual suspects, auto zone, oriellys, advance auto and even the smaller chains like sears and battry plus... will all check your battery for free.... takes about 10-15 minutes....and will even charge it over night for free if its to low to get a reading....there equipment will tell you if any thing is wrong and the general life of our battey
Plus there is no condition you have to buy a battry from them, they just want the oppertunity to sell you a battrey
They can also test starters and alts free
|Prop and the Blackhole Midget|
|Prop, thanks for the info. I never did take the battery out of the circuit. Despite the comment made by the alternator technician, I could not fathom how the battery could be the cause of the spikes. But then I don't really understand how the solid core wires were creating the excess voltage, nor do I have any clue as to where it was being picked up in the electrical system. I just know it was there and the solid core wires are where it came from.|
I just received the new spark plug wires. They have boots and clips on both ends. The product description said that they can be used on both top and side entry caps. Since there are no instructions regarding the side-entry caps I assume I would just cut the ends off, stick them in the cap and screw the spikes down. I think the boots on the dizzy end will help keep moisture out, and the fact that this is a rehab rather than a restoration, I am going to get a top entry cap.
|Be very careful with top entry cap! The leads tend to foul the steering column and can wear, as well as having a nice hefty bit of earthing material in probable contact.|
I have never liked the idea of silicon leads with screw fittings, hence my short term dabbling with copper centred wire. I don't have a radio and thought suppression didn't matter, Big mistake - took a month on and off to realise the error of my ways.
|maybe it was a simple case of high voltages being induced in the plug leads?? When all else fails use Occam's razor, the simplest answer/solution is usually the correct one|
|R W Bowers|
|Hi Graeme, I am in the US (left hand drive) so the top entry cap is okay, plenty of room. I like the vintage look of the side entry cap and I could always switch back, but who will notice the difference when I am zipping down the expressway? ;)|
R W, I am sure that is the case. But how does it get from there to the rest of my electrical system? Always a mystery (although I do understand the principle behind induction). I had never heard of Occam's razor before. Thanks, I always learn new things here.
|Of course! Forget the idea that the steering layout is "mirrored" but the engine isn't. If your leads have push-in terminals that must be better in the correct cap than using the screws in the side entry cap.|
As I said above, don't understimate the induced voltages (called EMI I believe) surrounding unsupressed ignition leads. In my own experience it caused major malfunction in both electronic ignition and a digital ignition management system.
|" --- I don't really understand how the solid core wires were creating the excess voltage, nor do I have any clue as to where it was being picked up in the electrical system. I just know it was there and the solid core wires are where it came from."|
Well with hindsight it's pretty simple.
When you pass currrent down a conductor, you generate a radiating magnetic field. When that magnetic field cuts across another conductor, it induces a voltage in it. In your case it was inducing volatages large enough to cause the power supply in your radio to act up.
Unsupressed HT leads often cause audible radio frequency interference, but I've never had them cause the radio to switch on and off. I've got unsupressed copper (stranded) HT leads on my Sprite, I've always used them, but they don't cause my CD/Radio to cut in/out.
I guess that your particular radio must be sensitive to EMI, in the way that some modern electronic ignition systems are.
I wonder if a wound suppresor fuse in the live feed to the radio, would have solved the problem too?
This thread was discussed between 12/11/2014 and 22/11/2014
This thread is from the archive. The Live MG Midget and Sprite Technical BBS is active now.