The condenser (aka capacitor) is located internal to the coil. The one in the photo was added by someone to repair a failed (open) internal capacitor instead of digging the old one out and replacing it.
The condenser is connected in parallel with the coil points. Its purpose is to prevent or suppress arcing across the point contacts to prevent contact pitting and wear. The condenser does that by allowing the coil current to continue flowing for a brief time after the points open; allowing the contacts to physically separate from one another before current flow stops abruptly, producing a very high voltage across the coil points that is responsible for inducing the high voltage in the coil secondary winding which is connected to the spark plug electrode and produces spark. The high voltage developed across the coil points will cause current to jump across the contact gap if the contacts are not sufficiently separated from one another. The capacitor (condenser) also improves the spark quality and duration as energy is transferred from primary coil winding to the high voltage secondary coil winding.
It would appear the condenser, or a connection thereto, inside the coil has gone open circuit. I've seen this kind of repair done a few times.
A potential problem (apart from an incorrect substitute) is that if the internal condenser comes good, then the capacitance value increases.
So does that mean if your coils internal capacitor is in proper condition your points will pit less ... or not at all?
The flow of electricity is like the flow of water. If you turn a water pipe off suddenly the flow of water does not want to stop instantly and a tremendous pressure will build up. That is why most homes have a tank or pipe built into the water system so to take up this pressure.
When you turn off the flow of electricity by opening a set of points, a pressure builds up and can cause a spark across the open points. A capacitor charges up like a battery absorbing the extra current flow. Then when the points close that charge is released helping to start the electric flow.
I hope this was understandable.
Mark, Yes, that is the goal of the capacitor.
Note that proper point adjustment is critical in meeting that goal. Early points used a fixed, upper contact and a movable lower contact attached to a steel element that is pulled down towards the coil core when the primary coil current reached sufficient level to generate spark with enough energy to ensure proper combustion. The problem with those early points was electrical contact was broken almost immediately when the lower contact pulled away from the upper fixed contact before the lower contact got moving very fast. This is similar to a welder striking an arc; moving the electrode away from the work slowly intending to draw an electrical arc.
The solution used on KW points was to change the upper fixed point contact with a movable point contact that follows the lower point contact for a fixed distance when the coil is energized to produce spark. Both point contacts remain in electrical contact while they travel towards the coil core with increasing speed in response to the building magnetic field. The upper point contact eventually reaches a stop provided by the limit rivet head; allowing the lower point contact to break electrical contact abruptly and move away from the upper point contact very fast. The capacitor can now do its job of continuing coil current flow for a brief interval while the lower point contact moves away from the upper point contact to greatly reduce the ability of the coil current to jump the point gap and produce an arc that pits the point surface.
The upper point contact is attached to the cushion spring which provides a downward force to keep the upper point contact physically connected to the lower point contact as both contacts travel towards the coil core when the coil is energized. Insufficient cushion spring tension will result in the lower point contact pulling away from the upper point contact almost immediately causing an early, weak spark; allowing the point contacts to close again and repeat the process but even weaker and later (retarded). This is known as a "Double Spark" illustrated below; which is undesirable as the spark occurs too soon (1.5ms, Advanced) and of poor spark energy (4.8A, cold spark). The photo below shows the coil current (vertical axis) versus time (horizontal axis) when a double spark occurs. This condition is easily observed on the Hand Cranked Coil Tester (HCCT).
Increasing the cushion spring tension can correct the double spark maladjustment but that alone does not guarantee the coil points are properly adjusted. The cushion spring tension may still lack sufficient tension to ensure the upper point contact remains in physical contact with the lower point contact as they travel towards the coil core. If that happens, the lower contact pulls away from the upper point contact slowly drawing an arc across the contacts. This permits coil current to continuing flowing with time (retarding spark timing, 2.630ms) and weakening the spark energy (4.5A cold spark) degrading engine performance and causing damage to the point contacts (pitting). Here is a photo of coil current that arcs across the slowly opening point contacts, delaying spark. The capacitor simply can not keep the current flowing long enough for the point contacts to separate enough to prevent arcing because the lower contact pulls away from the upper point contact too slow in this situation.
This maladjustment is NOT readily detected on an HCCT. Detecting this coil point maladjustment requires the ability to actually measure the point dwell time to fire spark using specialized electronic equipment or an Electronically Cranked Coil Tester (ECCT) designed specifically for his purpose to properly adjust KW coil points to help ensure optimal ignition timing and let the capacitor do its job to extend point contact life. Here is what coil current looks like with properly adjusted points: Fires a hot spark at 6A coil current on time after 2.150ms dwell time without point arcing.
Testing was done on 12V battery operation.
(Message edited by mkossor on April 27, 2016)
Mike - WOW! I wonder how many other Model T guys are like me and have about "zero" understanding of things electrical!
Here's an anology that couldn't be more opposite of your post Mike,.....I was told that the condenser compares to the short vertical pipe commonly inserted in a household water pressure line (one for hot and one for cold) that prevents the water pipe(s) from "banging" when you shut off the water faucet abruptly.
How's that for a Model T guys "understanding"? FWIW,.....harold
Harold, You got it!
My motorized hand cranked coil tester easily detects upper bridge points not working. It displays a double spark when the upper bridge points do not follow the lower point set. All hand cranked coil testers that I have ever seen do this. It is the main purpose of the device. The other main purpose of a hand cranked coil tester is to make all coils fire at the same time. This is achieved by adjusting all coils to the same current.
When operating the coil properly on AC voltage on a Hand Cranked Coil Tester you will see excessive sparking at the points when the capacitor is bad, or the wrong value, or both.
What was this business about the condenser sending voltage back through the coil windings in the opposite direction of current flow to collapse the magnetic field in the coil and generate high voltage?
The anti-water hammer analogy is very good; to a point. The capacitor does absorb some of the energy as it slows down (delays) the collapsing magnetic field; allowing the point contacts to move physically apart and limit how high the voltage (pressure) gets; both these actions suppress the tendency to arc. However, we do NOT want the capacitor to absorb all of the energy as in the case of the anti-water hammer or the coil would not produce a spark at the plug.
We want the magnetic field to collapse eventually, once the point contacts are far enough apart to resist arcing. We want the voltage (pressure) across the point contacts to get large (200V-250V) just not too large, too quick. We want to transfer the energy from primary coil winding into the secondary coil winding to produce spark. The induced secondary voltage will by about 80 times larger than the primary voltage; 16000V to 20000V. Too little capacitor value will promote arcing, point contact wear and weak spark energy. Too large capacitor value will slow the magnetic field collapse too much, limiting the voltage (pressure) value too low producing a weak, delayed/retarded spark.
Clarification; the HCCT is a time proven, viable method of coil point adjustment capable of achieving excellent engine performance but it does have limitations. Adjusting coil points for the exact same average RMS current will absolutely guarantee different firing times if the coils are not identical in every way (same inductance, resistance, etc.) or coil points arc randomly during opening. That's because the HCCT uses coil current as an indirect approximation of coil firing time instead of directly measuring firing time. Since the objective is to adjust all coils for the same firing time, I just think it makes sense to directly measure coil firing time and make point adjustments to make all coils have the same firing time.
(Message edited by mkossor on April 27, 2016)
Just to amplify upon Royce's comments.
Contrary to statements made in yesterdays version of never ending ECCT Infomercials posted here, in the hands of a competent operator the HCCT and StroboSpark will both detect and allow you to correct all three coil point operational situations cited.
Ron the Coilman
Ron makes an excellent point, operator competency is very important in achieving good results with the HCCT but apparently that still does not guarantee proper point adjustment. Here is test results of a coil I measured that was adjusted by a coil Guru with many years of HCCT experience:
I was really surprised when 3 out of the 4 coils from this coil "Expert" tested similar to this example which is a lot easier to catch on an HCCT than the more elusive point arcing maladjustment. That experience provided the motivation to research and develop a alternative method for proper coil point adjustment with results that are very easy for the operator to interpret like: Poor, Good, Excellent.
Mike the ECCT E-Timer man
A maladjusted coil is a just a maladjusted coil and one doesn't need an oscilloscope or ecct to detect the fault and correct it.
Ron the Coilman
The double spark in the 4/28/16 post would be observed on either HCCT or Strobospark. Personally, the Poor, Good, Excellent rating without quantitative values for each category is somewhat meaningless. (Like measuring with a micrometer and marking the distance with a piece of chalk.) A time to fire with mean and standard deviation would be valuable information to me. I've asked if someone would share data regarding how different the TTF values are for separate coils each adjusted to 1.3 amps on either HCCT or Stroboscope but no data has been posted to my knowledge. If you set up four 'identical' coils on the HCCT to the same current specification, how different will the four TTF values be? TTF and current specs are both set the same way, by bashing down the lower bridge to increase current or prying the rear of the bridge up to reduce it, I believe TTF is 'set' the same way. Hard for me to understand how adjusting using TTF is better given you are still bashing or prying something bolted to a 100 year old maple coil. respectfully, jb
The coil Guru's 3 double sparking coils should have been visible to the coil Guru too but apparently they were missed because that's how he sent them out!
The Poor, Good and Excellent results categories are just a quick test result summary. The more meaningful result is the actual coil firing time displayed in 1 degree increments of crank shaft rotation at a specific engine speed (1000 RPM). That way, you can easily see the timing variation impact with respect to piston position.
I agree, its really nice to have the option to measure and display the actual coil firing time in milliseconds, the peak firing current, spark energy and even primary coil inductance if those details are of interest to you.
Here is an article that measures and compares coil firing performance of 12 coils obtained from 3 professional coil re-builders adjusted using coil current as an approximation of TTF if you are interested. http://www.modeltecct.com/uploads/ECCT_Intro.pdf
Ron, I agree you don't need an oscilloscope or ECCT to correct maladjusted coils; heck; some folks even claim to it by ear. I just think it makes more sense to use modern test methods to know the coil is adjusted properly with a high degree of certainty and consistency.
I would love to try a set of perfectly working coils just once in my life time. It's possible I've been driving with crappy coils for years ... doing the best I can as is.
Need to find someone who rebuilds coils who trades for something I have of value. Cash is always a problem when it comes to buying things I don't actually need. .. but would be nice to have. Anyone need a nice brass camera lens or an expertly made tintype or ambrotype of their T. :-)
This was a first for me, but I recently tested a coil with the StroboSpark that had no capacitor reading at all for size or leakage, yet produced a vary strong spark with the correct current value.
I suspected it would not last long, but the engine had set for 20 years and there were only 4 coils available.
All the valves were set too, so there was no problem with those coils or the plan to use them and no compression at all either.
The battery used for the last attempt to start this engine 10 years ago was still installed and produced no current or voltage.
Mike: So what was the HCCT RMS current for all these coils you examined for your article? You imply the coils you examined were all set to the same standards. What were the HCCT current values for each coil in your tests (before and after adjusting the TTF)? I remain convinced the double sparks you show would be visible on the HCCT. And I also believe the Current-Time waveforms you show with varying TTF (and varying peak current) would also show visibly different RMS current on the HCCT b/c area under the curve is visibly different. respectfully, jb
James, I assume the coils were adjusted for the nominal 1.3A average RMS current but did not measure them independently. All were adjusted by professional coil rebuilders with lots of experience so makes no sense they would send to a customer if they were noticeably double sparking. I also don't doubt the average RMS currents would have varied by some amount but again, must have met the criteria within reasonable tolerance for the coil rebuilder to ship them out. I took great care to protect the point settings and did not alter them. I even sent one set back to the re-builder for a tune-up because I was so surprised with the performance. The coils came back with almost identical performance.
The average RMS current depends upon many factors and is only an indication of coil TTF if the coil current is periodic with the same duty cycle and all coils are the same. Another consideration is how abnormally slow a hand cranked magneto generates the power to energize the coil under test; producing an abnormally low voltage. Here is a simulation of HCCT output cranked at 120RPM compared with Magneto output running at 1000RPM and 12V battery.
As can be seen from the figure below, adjusting coil points energized by 12V battery (Green Square wave) much more closely mimics actual Magneto operation at moderate (1000RPM) engine speed (Red trace) compared with HCCT output at 120RPM (Blue trace). It is really amazing the HCCT works as well as it does considering these details of operation.
I forgot to mention the reason why I think the experienced coil rebuilders sent out double sparking coils; testing coils using an abnormally slow (120RPM) rising magneto output with abnormally low amplitude (4V) subjects the coil points to an abnormally slowly rising magnetic field that does not cause the double spark to occur. But when the same coil is subjected to step change in voltage (12V) that more closely mimics normal magneto output (see previous post illustration), the magnetic field builds much more rapidly, exposing the maladjusted cushion spring tension that allows the vibrator spring contact to be pulled away from the cushion spring contact causing the double spark to occur. The coil is subjected to even faster magneto output rise time as engine speed increases further (rise time is twice as fast at 2000RPM; when pulling out into traffic in low pedal for example) were consistent coil TTF is needed to reach such high speed. This also debunks the argument that coils points can only be properly adjusted operating from an AC magneto like output not DC battery.
Mike: Thanks for the responses. My question of the relationship between HCCT current and TTF remains unanswered. Assuming the HCCT values were all 1.3 A and then showing the perceived shortcomings in TTF proves nothing. I suspect few if any of us have both HCCT and ECCT and time to make the kind of comparison I request, namely the HCCT current value and corresponding TTF from the ECCT for each coil. And it would be nice if this were a blind test with sufficient coils that one might statistically analyze the data. But I would be very grateful if reliable data for even 4 coils could be posted.
In the second post you refer to 'abnormally slow' and 'abnormally low' to describe certain attributes of the ford magneto and HCCT. I don't think this is correct especially for those who crank start their Model T's on magneto. As you show, it's the nature of the Ford magneto, I don't think that's abnormal. respectfully, jb
James, my description of "abnormally slow" and "abnormally low" in reference to HCCT magneto output accurately contrasts magneto performance operating at hand cranking speed with performance at normal engine speeds; nothing derogatory about the Ford magneto operation. My point in doing so was to emphasize coil point performance can change based on rise time of the applied voltage. TTF is very forgiving at hand cranking engine speed (120RPM) in terms of timing error, not so much at top end (2500RPM). Just because a car starts right up on the fist pull doesn't mean the engine performance will achieve a top speed of 2500RPM. That will only occur if all ignition timing variation, including coil TTF variation, is minimized under that operating condition, not just hand cranking speeds.
How about an explanation we can all understand with the difference between using AC as opposed to DC if any. We've been told point wear or build-up on one contact is lessened by using AC because it transfers the metal back & forth from one contact to the other and DC will allow a build-up on one contact because the direction of flow doesn't change. Any one familiar with a DC distributor system has seen this. Were all those condensers bad or does the condensers job actually involve reversing the current flow through the coil windings causing a collapse of the magnetic field and inducing high voltage?
As an electronic dummy I don't understand everything I've read here, but I do understand that compared to previous discussions of coil testing this one contains very little of what I would call ad hominem sniping. Thanks to all who have resisted that urge.
The main reason the points wear faster on DC is due to the amount of time they are in contact when the timer turns the coil on. On mag you may get three sparks occur rather than one continuous buzz for the period they are in contact. That long buzz would be hundreds of spark events as opposed to three.
You know I just looked over my 1960's Automotive encyclopedia concerning condensers and it doesn't say a thing about reversing the current flow through the coil just point wear as discussed here so now I'm trying to figure out where the heck I got that from.
That is because by 1960 the assumed use of the condenser in automotive application would have been as a point bypass on a distributor system and the polarity would have always been the same way for every firing. The T and other low voltage magneto systems were using AC while cars after the T were using DC and running distributors for the most part.
James, the information you seek "I've asked if someone would share data regarding how different the TTF values are for separate coils each adjusted to 1.3 amps on either HCCT or Stroboscope " is tabulated in the reference article I posted:
In each set of 4 coils adjusted by professional coil rebuilders, the TTF variation was about 0.65ms. That equates to 4 crank shaft degrees at 1000RPM or 8 crank shaft degrees at 2000RPM and 10 crank shaft degrees at 2500RPM.
Mike: Thanks, I carefully read the report and noted peak amps and TTF, I'm still not convinced all the coils were set at precisely 1.3A at the moment you tested them.
I'm more involved in the discussion than I had planed to be, but since I am, can you explain how the 12 VDC step excitation and Ford magneto ramp excitation affects performance. For those of us using Ford Mag and not 12 VDC the HCCT may be the gold standard for setting our coils. Is it not possible the HCCT coils set by your rebuilder were also fine but hammering them with a 12 V step excitation caused the double sparking you see on the scope? respectfully, jb
I believe you are talking about false indications of double sparking due to using DC voltage to test a coil, versus a Hand Cranked Coil tester using an actual Ford magneto that produces AC voltage as the car would.
James, let me first be clear; The HCCT is a time proven coil test tool. Its not my intent to condemn it or criticize folks who prefer to use it. I noticed the HCCT output voltage rise time is much slower (16ms vs 1.9ms)and reaches a much lower value (4V vs 18V) compared with magneto output at normal engine speeds. I think 12V step excitation more closely mimics normal magneto operation (not just battery operation) at normal engine speeds as I tried to explain with the graph in my previous post (see the graph of voltage vs time) the HCCT output voltage (blue trace) takes more than 8 times as long to reach its voltage peak that is less than 1/4 the value of magneto output voltage at 1000 RPM (Red trace).
I think you can adjust all 4 coils for the same dwell time to fire with an HCCT, just not as accurately and consistently as a method where you actually measure coil dwell time to fire directly.