Coil Failure(s)

[Sinitro]

1 More clarification point about coils...
Since they use an oil bath to cool down its internal windings, if possible its best to mount the coil vertically. Also as mentioned previously, today there are many sub-par, low quality ignition parts made in China. Note that China has the know-how to build reliable components but unfortunately today everyone is pushing for low prices and here China responds. Since we have been sourcing electronic components from China for >20 years we know this category very well..

Just my $0.02...

 

[Ranger16]

I know this could be a dumb question, but does it matter what side of the resistor the wire to the coil comes off of?

Yes it does. One side is full 12v (for starting), the other is I think about 9v (for running). Put the coil on the wrong side and you run full 12v all the time and probably reduce the coil life.

The reason I ask, the shop manual for my '67 appears to show it coming off the left side, while manual for my '64 shows the opposite.

Forget left or right side. Turn it upside down and the right side becomes the left side.

 

[Dave6T4]

What I meant to ask here, does it matter if the plug-ins get switched side for side on the spades on the ballast resistor? I am thinking, no.

 

[AR67GTX]

No, it wouldn’t matter. There’s no forward or backward on a resister.

 

[Dragon Slayer]

It is interesting coil failure. The distributor would not matter it just triggers the ECU. The ballast resistor protects the coil by limiting current to the coil other than starting voltage.

So, if some sort of ignition switch flaw caused full voltage on output side of ballast resistor, your coil has full voltage all the time and would overheat fail. IF somehow there is a short on the ballast resistor (unlikely) unless some metal shorting the winding.

So the only other item could be the ECU somehow generating a longer dwell to the coil by some bleed of current flow to ground. Not sure if that is possible and still trigger a secondary voltage firing event.

I guess lastly the secondary side can have some problem via bad spark plug wires. With ignition wires removed from coil you could read the primary coil resistance and the secondary on the coil, see if one side or the other is deteriorating.

If available you could move the bad coil to another vehicle and see if it remains the problem. If not, may not be coil, but system effect.

I have seen these orange box deteriorates and produce a weak spark. But you're saying with a new coil it runs fine for a while.

 

[RJRENTON]

It is interesting coil failure. The distributor would not matter it just triggers the ECU. The ballast resistor protects the coil by limiting current to the coil other than starting voltage.

So, if some sort of ignition switch flaw caused full voltage on output side of ballast resistor, your coil has full voltage all the time and would overheat fail. IF somehow there is a short on the ballast resistor (unlikely) unless some metal shorting the winding.

So the only other item could be the ECU somehow generating a longer dwell to the coil by some bleed of current flow to ground. Not sure if that is possible and still trigger a secondary voltage firing event.

I guess lastly the secondary side can have some problem via bad spark plug wires. With ignition wires removed from coil you could read the primary coil resistance and the secondary on the coil, see if one side or the other is deteriorating.

If available you could move the bad coil to another vehicle and see if it remains the problem. If not, may not be coil, but system effect.

I have seen these orange box deteriorates and produce a weak spark. But you're saying with a new coil it runs fine for a while.

Having dissected a few orange box ECUs, the circuitry turns on the case mounted transistor, charging the coil's primary winding, thru the ballast resistor (which is bypassed during starting) to ground. The distributor's pick-up coil, in conjunction with the reluctor, generate the pulse, that triggers the internal dwell timing network, comprised of a programmable uni-junction transistor (UJT), which looks at the time interval between in coming trigger pulses (RPM increases) to allow sufficient charge time for the coil (rate of change of the primary current flow), which in turn, turns OFF the external case mounted transistor, causing the coil to produce the spark. The external case mounted transistor is a Power Darlington design, capable of withstanding the collapsing primary winding's induced voltage, yet allow for primary current flow. It usually fails because of a mis-matches coil and or incorrect ballast resistor, allowing excessive current flow and induced voltage conditions reflected back to the ECU. That's how the origional ECU worked.....the current Chinese stuff will likely have different components of possibly different designs....I've not had the opportunity to see. It is a basically coordinated components selection....change of one component will result in a failure......just my opinion of course,............
BOB RENTON

 

[Geoff 2]

There have been many design improvements in semi conductors & electrical components in general since the Chry ECU started life in 1971. A lot of people criticise the current Chinese made ECUs & that the exposed transistor is a dummy. One might find that the transistor in these units that switches the coil current is actually more reliable & has better reserve capacity, not switching current near it's maximum limit.

 

[RJRENTON]

There have been many design improvements in semi conductors & electrical components in general since the Chry ECU started life in 1971. A lot of people criticise the current Chinese made ECUs & that the exposed transistor is a dummy. One might find that the transistor in these units that switches the coil current is actually more reliable & has better reserve capacity, not switching current near it's maximum limit.

Please provide specific info on the referenced transistor....there are literally several dozen types that could be used for this application it's a NPN junction with a internal MOV transient suppressor, with a Shockty quick recovery junction. Since many of the Chinese units use fake TO3 transistor on the case for appearance, its difficult to accept what you DO NOT see. Show us your data to compare......
BOB RENTON

 

[Geoff 2]

If you do not believe that semiconductor [ as well as other electrical components ] have improved in design & reliability since the early 70s, then you must have been hiding under a rock....
Not to mention a hugely expanded range giving more options.
I built a 3 transistor radio in 1961 that used a Germanium OC72 Phillips output transistor. It cost me $7, which was about 3/4 of the weekly apprentice wage I was getting at the time. Also, the $7 was cheap because my Dad worked in the electrical industry. Today, a similar & better silicon transistor cost 50c...or less.
It's called progress...

 

[451Mopar]

With the Mopar ECU the hardest (most heat) on the coil and ECU is having the ignition on, and the engine not running.
The ECU power transistor is "ON", and only turns "OFF" when it sees the signal from the magnetic pickup in the distributor.
With the engine not running, current through the Ballast/Coil/ECU transistor is basically only limited by the resistance of the coil windings and ballast resistor.
When the engine is running the coil has inductive reactance, a resistance that increases with frequency. This reduces the current through system as RPM increases.

For cranking, the ballast is by-passed just while cranking the engine.

What I'm saying is use the ACCessory key position for power when the engine is not running.

 

[Dragon Slayer]

I have taken a lot of OEM and after market boxes apart. While chip tech has improved, quality still varies and construction of unit matters too. There is a reason automotive and mil spec cost more. On left OEM on right a cheapo box. Notice the board difference wrt heatsink and noise supression and quality of the mounting of the components. I have seen better circuit boards in toys than the cheapo ECU. The power transistor was only heat sunk by friction with the ground tab. Only one ground joint to the ECU body. Mopar OEM unit much higher quality.

Back to the orig problem, if the coils are failing, it does have to be the over current as I mentioned or some secondary effect.

ECU see an awful lot of voltage spikes. Regardless of the TVS diodes and suppression protection, over time those components can fail and the power transistor/FETs can become damaged. Diodes and those power transistors than leak or fail (internal diode).

 

[Geoff 2]

Disagree with some of the comments:
- What exactly is wrong, or what is worse, about the mounting of the components on the aftermarket unit?
- you cannot gauge the 'quality' of components [ resistor, diode, capacitor etc ] by looking at them.
- the aftermarket unit uses a TO-220 or similar power transistor. These are smaller than the original big TO-3 design used by Chrysler & are used in modern day electronics. They require less heat sinking. A similar, or same, design is used in the GM HEI module, which switches more current than the Chr transistor & is no less reliable.
- the coils cannot 'overcurrent' because current is limited by the 0.5 ohm bal res.

 

[Dragon Slayer]

Ok, but you are wrong. Look at the quality of the PCB. It is built with EMI shielding embedded and thick trace. The ground is via 2 tabs. Look at the solder joints. The device leads are bent over to lock in place with a solid solder joint on the trace. This prevents vibration cracking. I know the TO-220 device is current packaging, but it should not have been floated for heat sink purposes. Exposes it to much higher temperature which is a leading killer of electronic parts besides voltage spikes. You can also visibly from the paint marks determine the tolerance of the resistor and also the quality and wattage rating. If you purchase electronic parts, and I do for a hobby business, there is a big price difference between common parts and milspec parts of the same type. That board probably has $1 worth of parts at the overseas manufacturer. And less in labor.

The circuit board in that aftermarket is a POS lowest cost assembly. It is obvious if you looked at it.

Yes you can overheat and saturate a coil even with a ballast resistor. In operation the dwell determines how much power has to be dissipated as heat, and heat can damage the internal insulation.

Ballast resistor are normally higher too not .5 unless it is using a higher performance coil and ECU meant to use a low ohm ballast. For racing as example. High RPM and less time between firing events.

Even with a 1.2ohm resistor you can saturate the coil if it is charged too long.

 

[Sixpactogo]

Bob, I made that statement out of frustration, and I apologize for it. I returned the failed coil to my parts guy, and he replaced it. He said sometimes a bad part can get out there. We both agreed if my car eats another coil, something else is going on. I am going to do resistance test on my ballast, and test voltage of my charging system. As with a lot of owners of vintage cars, the least understood automotive system, for me, is the electrical system. I am getting better at it , but still have a ways to go.

While you are doing all those resistance checks, see how many ohms you get between the neg post on your battery to the case of that orange ECU. If it is more than .5 ohms. it is not a good ground and that will also overtax your coil. It also does no favors to the plate coil in the distributor and the coil in the ECU itself. All three of those coils can get overheated and fail just from one piss poor ground.

 

[451Mopar]

I have taken a lot of OEM and after market boxes apart. While chip tech has improved, quality still varies and construction of unit matters too. There is a reason automotive and mil spec cost more. On left OEM on right a cheapo box. Notice the board difference wrt heatsink and noise supression and quality of the mounting of the components. I have seen better circuit boards in toys than the cheapo ECU. The power transistor was only heat sunk by friction with the ground tab. Only one ground joint to the ECU body. Mopar OEM unit much higher quality.

Back to the orig problem, if the coils are failing, it does have to be the over current as I mentioned or some secondary effect.

ECU see an awful lot of voltage spikes. Regardless of the TVS diodes and suppression protection, over time those components can fail and the power transistor/FETs can become damaged. Diodes and those power transistors than leak or fail (internal diode).

View attachment 1373060

View attachment 1373061

That is a really clean Mopar ECU board, how did you get all the potting material off it? Looks like the one from the old Orange MP box?
I think counterfeit parts and poor quality control is creating issues with replacement parts, not to mention they are trying to get the cost down to make the most profit.
Most of my old Mopars have been converted to MSD 6-AL (from around 1990) or Crane HI-6 (2000's) ignition systems. They still use the original magnetic pickup distributor to trigger the boxes. On the Charger with the MSD, I use the Crane PS-91 E-Core ignition coil mounted on the inner fender. On the Coronet with the HI-6, I use the MSD Chrome cannister ignition coil because it is mounted in the stock coil bracket.
The '73 Jensen still has the factory setup.

 

[Geoff 2]

Post #33. No, I am not wrong. I will deal with the BS:
- On the aftermarket cct bd, I can clearly see the gold tolerance band on some of the resistors. This is the 5% tolerance range for tighter resistance control; 10% is more common, & 5% cost more. Far from scrimp on quality, I see a 1watt res on the aftermarket unit & only 1/4 & 1/2 watt on the factory unit. The 4th resistor from the left at the top of the stock unit looks burned; one reason for that is underrated in wattage. Maybe that is upgraded in the new unit with a 1w resistor.
- The EMI shield. It has one, it is called the metal case. The most sensitive part of the cct to spurious emissions is the p/up leads, where a few induced millivolts can cause false triggering. I use Co-ax cable for the p/up leads on some installations. Chry chose not to use any form of shielding for the p/up leads. I see no problem with the cct board & is representative of types used today.
- the power trans could well be the SOT-93 package, rated at 15 amps. Some TO-218 types look very similar & are rated for 25 amp. Since the trans is only switching about 4 amps, it is well below it's rated capacity & doesn't need a heat sink. Look at the so-called heat sinking of a GM module on a Chev dist. Dist at rear of engine, copping all the engine heat from convection. Then, the dist is being heated by conduction being bolted to the engine. This hot surface is supposedly a heat sink for the power trans inside the module. The so-called heat sink in this instance may well be hotter than the device trying to shed heat...but it survives!
- overheat & overcurrent are NOT the same. I said overcurrent. Because of the design, it would be near impossible for 0.5 ohm the bal res, which is limiting the current, to reduce in value & cause an overcurrent condition. They mostly go open cct, cannot see how they would short cct. - the coil has copper windings; these increase in resistance [ less current ] as the coil heats up.
- if the coil overheats with the correct bal res, then the combination was badly designed. This is not the case with the Chry system; the engineers allowed for prolonged periods of idling/slow running.
- if a coil failed from 'overheating', such as insulation failure, then the coil design was faulty or something external to the coil caused the failure. Like anything else, if used as designed, it will not fail unless the part itself is faulty.

 

[Bee1971]

And the original poster still hasn’t responded to my first question from weeks ago of how many volts at the positive side of the coil while cranking and with engine running in regards to the Ballast , also mentioned the ground on ECU

So many easy things to check first

And now YIKES

 

[Dave6T4]

Sorry! Car is in winter storage for now. I will look at it again in the spring. I want to get back to finishing my '64 Polara. I truly appreciate everyone's help and suggestions. Dave

 

[Dragon Slayer]

What is the source of over heat? Current. I am sorry your wrong and inserting partial truths as red herrings. You seem to want to argue for argue sake. You can't see the quality of construction and soldering method?

You might want to do some reading on ignition systems and also electronics. Even on circuit board there needs to be emi protection. The metal case does not prevent interferrence on the circuit board itself. The signals on the board have to be conditioned and protected. Even trace routing especially on a multilayer board can effect how circuits accuracy and function. That cheapo one watt should be compared to the wire wound resistor on the OEM board. Also the circuits are different. A good ECU will do Voltage Regulation at input. Condition trigger signal, protect from stray signals and suppress voltage spikes to protect components, control dwell. The OEM has 4 transistors vice 3, more components for conditioning and protection.

It is hard for me to believe you think the cheapo board is a better board?

I used a heat gun and acetone to remove the potting. Which is another difference between the two units. Proper potting also has heat dissipation qualities. The OEM board was harder to clean because the potting surrounded the board. The cheapo was barely in contact and just a sealant.

I am not arguing that newer electronic can't work, I am saying some of these ECU made overseas (that have a reputation of failing) are not made well. Contruction and quality of components matter.

There are some good articles on coils on the net. Worthy of a read. Besides resistance, actual inductance, ratio and overall construction and magnetic coupling internally matter. Part of the reason simply going to a GM coil can give a much-improved spark.

 

[RJRENTON]

What is the source of over heat? Current. I am sorry your wrong and inserting partial truths as red herrings. You seem to want to argue for argue sake. You can't see the quality of construction and soldering method?

You might want to do some reading on ignition systems and also electronics. Even on circuit board there needs to be emi protection. The metal case does not prevent interferrence on the circuit board itself. The signals on the board have to be conditioned and protected. Even trace routing especially on a multilayer board can effect how circuits accuracy and function. That cheapo one watt should be compared to the wire wound resistor on the OEM board. Also the circuits are different. A good ECU will do Voltage Regulation at input. Condition trigger signal, protect from stray signals and suppress voltage spikes to protect components, control dwell. The OEM has 4 transistors vice 3, more components for conditioning and protection.

It is hard for me to believe you think the cheapo board is a better board?

I used a heat gun and acetone to remove the potting. Which is another difference between the two units. Proper potting also has heat dissipation qualities. The OEM board was harder to clean because the potting surrounded the board. The cheapo was barely in contact and just a sealant.

I am not arguing that newer electronic can't work, I am saying some of these ECU made overseas (that have a reputation of failing) are not made well. Contruction and quality of components matter.

There are some good articles on coils on the net. Worthy of a read. Besides resistance, actual inductance, ratio and overall construction and magnetic coupling internally matter. Part of the reason simply going to a GM coil can give a much-improved spark.

Your first paragraph regarding "red herrings" without any substantiation is spot on. #36's experiences at fixing things is fine, but without any fundamental knowledge of the how's and why's things operate other than fix by substitution, are just an opinion, which he is entitled to offer.....but....Supposition on component values is speculation; example...This is the 5% tolerance range for tighter resistance control; 10% is more common, & 5% cost more. Far from scrimp on quality, I see a 1watt res on the aftermarket unit & only 1/4 & 1/2 watt on the factory unit. The 4th resistor from the left at the top of the stock unit looks burned; one reason for that is underrated in wattage. Maybe that is upgraded in the new unit with a 1w resistor. Based on what ? .....without knowing circuit design parameters....voltage drop or current flow or duration thru the component. One cannot determine a component's power dissipation capability visually or by its physical size. Now, the word wars will begin.....or ....."I know more than you do because I'm older and have bern fixing cars longer than you".....just doesn't cut it.....without supporting documentation.....
In the meantime, I support Dragon Slayer's premise and comments and information wholeheartedly. Just my opinion of course.
BOB RENTON