The test or checking resistance values for a Model T coil is shown in the attached illustration borrowed from the original "Despondent over coils" discussion.
But no values for the condenser is given at E-D.
It was established that the condenser could be checked by measuring a value by using points E-D with the points open.
So what do you look for to test the health of the condenser?
Method one - using an ohm meter to assess weather the condenser is shorted or not, and to test its charge/discharge rate.
Method two - to use a capacitor test meter such as a Micronta True RMS multi-meter.
a. by establishing a value of a known orange drop capacitor at .01 micro-farad (uF).
b. open the points and measure the value of the capacitor in circuit (still connected inside the coil box). The meter indicates a value - four coils tested - of 1.4 uF to 3.5 uF micro-farad.
I can say that the condenser is not shorted. But are these values of the condenser in line to say that the coils are "good?"
Ron's diagram clearly shows the required specs.
This chart will give you better answers to your questions. To check the capacitor you will need to use an analogue multi-meter. Digital will not perform the function. Set the meter to read 10K ohms, Place one lead on contact "C" and the other on contact "E" The needle should jump up and then return to infinity. To test again you will need to reverse the leads. If the needle does not return to infinity that indicates that the capacitor is leaking. The points need to be open to perform this test, place a piece of paper between the points.
(Message edited by Mike Vaughn on May 27, 2016)
I got it correct the first time. Now has any one else measured the capacitance value of the condenser in circuit? What values have you obtained?
Mike, what is the rational behind the 0.010-0.012" (recommended) cushion spring gap? I am very interested to know the source of that recommendation because the data I took supports Ford's recommendation of 0.005" gap (aka cushion spring travel) is optimum.
The data shows the dwell time to fire (and peak coil current) does not change significantly if the cushion spring travel is more than 0.006". I have also noted that points with excessive cushion spring travel (> 0.005") can be very difficult to adjust for good firing consistency (ie minimum cylinder to cylinder timing variation). That is because the upper and lower point contacts separate after approximately 2.1ms regardless of the cushion spring travel. The additional travel beyond 0.005" presents the opportunity to delay firing with misadjusted points or slowly open, producing an arc which also delays firing. This data was taken with an ECCT using an adjustable cushion spring travel suggested by Garrett Green shown below.
Have never seen that adaptation to allow adjusting the cushion spring travel. Pretty neat. Is there lock tight on the threads to keep it at the proper setting?
Mike, that chart was drawn up by John Carter a few years ago.
Mark, No, the upper bridge is tapped using a 2-56 tap and the nut locks the adjustment for the measurement. That was sufficient for lab measurements.
Garnet, Thanks for reference, I will contact John regarding the source of that recommendation.
I have always looked at the cushion spring travel as a mechanical time delay, which allows the coil core to saturate. Seems that most new contact sets have a cushion spring travel well in excess of 0.005", while most older sets I've seen are closer to the 0.005". I'm sure you have or know how much time it takes to fully saturate the Model T coils core, how fast the contacts move and can easily figure out whether the 0.005" is enough time to saturate a core. That known why do the vendors keep making contact sets with such large cushion spring rivet spacing's? You would think that the wider the spacing, the more likely the spring could hang up a few times someplace on the longer rivet. I like the little screw you have to adjust the spacing. I think I may do that just to prove to myself the advantage of a smaller spacing. Mike
Yea, but getting a teensy-tiny screw driver under there is the real hard part !
Mike, I agree, the points vendors have been producing points with too much cushion spring travel. This has been a problem for some time know and known Ron Patterson (et al) has contacted them to convince them to put better controls on limiting the travel. I don't know where the heck the recommendation that the travel can be more than 0.005" came from but I have measured points with travel exceeding 0.020"! Many of such points can be a real bear to get to fire consistently, especially at firing rated of 2000 to 2500 RPM where low pedal power is really needed to pull out into traffic.
The cushion spring does function to provide a time delay in opening the points. The time it takes the coil points to open depends upon the operating voltage; the higher the voltage, the faster the points open. On 12VDC, properly adjusted coil points will open in 0.002s (2ms) when the primary coil current reaches 6A, which is below the primary saturation current but produces a spark with sufficient energy to ensure good combustion. Adjusting the points for higher firing current takes the points additional time to open, which makes it more difficult to hand crank.
Enjoy the discussion on fine tuning the cushion spring.
But the original question is this - has any one of the experts who rebuild coils or test coils ever established a value for the condenser in place in the coils? In this case, an original coil that has not been rebuild. I have found values with open points and measured the value of the capacitor in circuit (still connected inside the coil box, between 1.4 uF to 3.5 uF micro-farad.
Those that test and rebuild, what value have you found with the condenser in place?
An article from Vintage Ford indicates that the original value of the condenser is between 3 uF to 4 uF (micro-farad).
Bonus question -- if the original value was between 3 to 4 uF, why are coils being rebuilt with a condenser (capacitor) of .47uF?
George, I have seen the original Ford specification for the capacitor and it notes the size to be .4 to .45 microfarads for cars made between 1914 and 1926.
I did radio and TV repairs for about 7 years and in almost every case the circuit resistor and capacitor values could be halved or doubled and they would still work. So anything close to original was often used as a replacement.
.47 has been a popular size for many years and it was always readily available.
The capacitor also has to be removed from the circuit for most capacitance testers to give a good reading.
I bought a StroboSpark coil tester about 3 years ago and was about ready to send it back as being defective, as the capacitor always tested bad.
Talking with Coilman, I learned that the capacitors were always bad and replaced in the coils that he rebuilds.
I also learned that those coils may all look alike, but the were made for many different uses, like igniting flame throwers in WW-2, igniting home furnaces, and providing sparks for the little yellow Fairmont Rail Road repair car gas engines.
They were also made to different voltage and capacitor specifications for the different users.
Some were only designed to work on 12 volts DC and never tested with a Model T Magneto.
Because .47 uf is one of many standard capacitor values and happens to be closest to the required value. There's nothing magical about .47 uf - you just can't readily buy a .45 or .42 uf capacitor or whatever oddball value you want. As James alluded to, there are allowable tolerances.
Not all Fairmont Railway Motor cars were yellow, out here on the SP, they were orange. Usually the Fairmont coils were made by Pontiac Coil company and had one side longer on both ends--and knurled nuts for the electrical contacts.
OK, I found the (a) rational for the cushion spring travel greater than Ford recommended. It was Ron Patterson who said: "We have also found that by increasing the .005 cushion spring travel to .010-.012 the coil works better".
No data, No performance metric cited, it just "works better". Right.... I think Ford Engineers knew what they were doing when they specified 0.005" cushion spring travel. Any greater and point adjustment just gets more difficult in my experience. The point contacts must break without arcing and without the benefit of the abrupt stop of the cushion rivet head. The operator has to get the point gap, vibrator spring tension and cushion spring tension balanced just right to make that happen without drawing an occasional arc that prolongs the firing time, introducing cylinder to cylinder ignition timing variation. This is known by actually measuring and displaying the coil dwell time to fire as plotted in the previous chart (not a wet finger in the air!).
George, I doubt the larger capacitor values you cited will result in a more optimal spark. The capacitor value determines how fast the magnetic field collapses which determines, in part, how high the voltage induced into the secondary winding will get. The capacitor value is also specifically chosen to maximize the energy transfer from primary to secondary winding at its resonant frequency. Making the capacitor value much larger than 0.47uF will slow the collapse of the magnetic field so the induced voltage will be lower and the energy transfer will not be as efficient. i.e. weaker spark that is not as "hot".
I found a reference on coils and a description on the condenser attributed to being in Volume 15, Number 3 of the Vintage Ford. It appears to be copied from a revised 1920 edition of the Ford Service Course. There is a mention of comments from Murray Fahnestock a writer for the Ford Owner and Dealer, with additional information on coil repairs from Bruce McCalley
The condenser is described as being composed of two pieces of tin-foil seven feet long and 3% (sic three inches wide) wide. One piece of this tin foil is placed on the other one but 1/8 to one side, with two layers of glassine paper insulation between one layer on top and one layer on the bottom. …. The condenser must test from three (3) to four (4) micro farads (uF)….. The condenser is used to absorb the current of the primary windings at the breaking of the contact points and thus prevents it from arcing across the points, which would soon burn them. As soon as the condenser is charged it seeks the path of least resistance to discharge or neutralize itself, which is through the coil in the opposite direction. This causes the magnetic field about the coil to collapse very quickly. The more rapid the fall of the primary coil the greater the force of the induced current in the secondary winding.
With in the continuation of the article a brief discussion on the repair of coils by Bruce McCalley. He goes on to describe that a replacement capacitor should have a rating of .47 to .5 microfarads voltage rating of 200 volts. He also states that testing original capacitors their range is 1.5 to 3 microfarads
For the complete article see:
George, Thanks for the reference. Interesting information, but unfortunately, not entirely correct. It is true "the condenser is used to absorb the current of the primary windings at the breaking of the contact points and thus prevents it from arcing across the points, which would soon burn them" However, the condenser (aka capacitor)actually SLOWS DOWN the collapse of the magnetic field rather than makes it collapse very quick. The capacitor does this by allowing primary current to continue flowing in the same direction through the coil primary winding but decreasing value for a brief interval after the point contacts open. This does two things; first, limits how high the resulting voltage that develops across primary and secondary windings. Second, it allows the point contacts to move physically apart, making it more difficult for electrons to jump across the contacts and arc as the primary current reaches 0A and magnetic field completely collapses. At that point, the primary (and secondary) voltage reaches their maximum value which causes electrons to jump the spark plug gap and spark. The secondary voltage actually begins to decrease in value as the capacitor charges up in the opposite direction then forces primary coil current to flow in the opposite direction, regenerating the magnetic field and resulting secondary voltage across the spark plug gap of nearly the same high voltage. This process happens again and again, oscillating back and forth but with exponentially decreasing amplitude which functions to prolong the spark duration and ensure complete combustion. The frequency of these current oscillations is determined in part by the value of the capacitor across the points which is why its value is carefully chosen to be the value it is; 0.47uF.
I think the capacitor value is a balancing act for more reasons than what has been talked about here. When the points open, you have a series LC circuit that resonates at a certain frequency. For a T coil that is typically around 3800 hz with a .47 mfd. cap. A .1 mfd. cap would put this frequency to about 8000 hz. This is significant because in my experience, the q factor of original Ford coils falls off significantly as the frequency rises. So, the higher the frequency, the weaker the spark, all else being equal.
Tom, now you've done it; you went and got all technical. Yes, that is exactly what I was alluding to with the description of circulating current oscillations between capacitor and coil (parallel resonant RLC circuit) which functions to prolong the spark duration and improves the efficiency in which energy is transferred from primary to secondary windings. The value of Capacitor is chosen such that the frequency of oscillation with the primary coil coincides with the self resonant frequency of the coil secondary winding which resonates with the parasitic winding capacitance to be around 4kHz.
The theory and operation of the Model T coil is far more complicated than many realize. The more I read about coil operation from the "Experts" the more surprised I am how few truly understand it.
It would be cool to see a close-up of model T points in action....in super slow motion.
So the original Ford coil using a 3uF to 1.5 uF capacitor and the coil having an inductance between 3.7 and 3.95 millihenry ( using the value of 3.82) would have a frequency between 47.01 Hz to 66.488 Hz. Whilst the magneto is producing a frequency between 26.4 Hz (at 5mph) to 160 Hz (at 30mph).
Calculations for the L/C circuit : http://www.1728.org/resfreq.htm
George, you used the correct formula but a few caveats; First, the units you chose were in Henrys in stead of 3.82 milli-Henrys to get 47.01Hz. Correct the units to milli-Henrys and you should have calculated 1.4867e+3 Hertz (1486.7 Hz) to 2102.5Hz for 3uF to 1.5uF. Second, the formula used is only valid if there is little or no resistance associated with the coil. In reality, the resonant frequency will be different because there is resistance associated with the coil but a rough estimate. The frequency of the magneto output is not as important factor in determining the frequency of oscillations in the ignition coil as is the inductance of the magneto winding which is effectively in series with the coil primary winding. That is not the case when operating on battery and is another factor in the distinction between magneto and battery operation. The resonant frequency would be between 3350Hz and 4040Hz for magneto and battery operation respectively, again ignoring coil resistance.
Ok, wrong value for Henry. Point to make - the original coils were resonating at a lower frequency with the 3uF capacitor than with the newer .47uf.
Yes, that is correct; the frequency of coil oscillations gets lower as the value of the capacitor gets larger. The relationship is proportional but not linear due to the square root in the denominator of the equation.
Also recall the value of the primary and secondary voltage will not get as high using a 3uF because the larger capacitor slows how fast the primary current turns off when the points open which in turn slows down how fast the magnetic field collapses and thus produces a lower voltage.
Here is a simulation of Model T coil secondary voltage to the spark plug with capacitor values of 0.47uF and 3uF.
Note that with C=0.47uF, the output voltage reaches ~30KV and oscillates at ~4000 Hz. Increasing C to 3uF Lowers the coil secondary voltage to the spark plug to only 13KV and lowers the oscillation frequency to ~1600 Hz.
Here is another simulation of a Model T coil on magneto (~700RPM) with C=0.47uF that displays the coil primary current (green trace, 1A/div) reaching just over 6A before the points open causing the current to fall rapidly, controlled by the value of C=0.47uF. Note as the secondary voltage to the spark plug (red trace, 5KV/div) rises as the primary current falls; reaching a maximum voltage value of ~30KV when the primary current falls to 0A; (total magnetic field collapse). Then the capacitor discharges through the primary coil causing the current to flow in the opposite direction (negative value) until it reaches about -5A, producing a secondary voltage at the spark plug of just over 20KV of the opposite polarity and the cycle repeats at decreasing values. The spark plug arc continues for the majority of these oscillations until they decay in value to the point too low to sustain the arc.
Here is a slo-mo video Mike Robison made.
It would be fun to the points at about 10,000 frames per second. I'll have to keep my ear to the ground and see if anyone I know has a high speed video camera.
fun to "see" the points
Tom, I have that on my to do list. Considering the dwell time to fire takes 2ms and recovery time a little longer the total spark event takes about 5ms operating from 12VDC. A high speed camera with 10,000 frames per second would capture the event in 10,000f/s x 0.005s = 50 frames which seems like fairly course movement, especially if trying to see a double spark event. Interesting topic which I am just starting to learn about so these calculations may be off.
There's those two guys that do slo-mo on YouTube. Send them a coil .... although most of their stuff is destruction !!