Folks wondering what the new E-project is all about can finally read about it here: http://modeltetimer.com/uploads/ECCT_Intro.pdf
Hopefully this project will earn its place in the hobby without creating an unintentional divide; it uses modern technology to get the most out of vintage technology.
Folks going to Hershey 2012 can see a prototype of the new ECCT up close at Red field space RCC-33.
Interesting
How does the ECCT differ in its coil operation analysis abilities from Fun Project's Strobo-Spark?
http://www.funprojects.com/products/CT-1.cfm
http://www.mtfca.com/discus/messages/257047/298518.html?1341567795
Mike,
The ECCT is quite clever. It seems to answer the question I had in:
http://www.mtfca.com/discus/messages/179374/242895.html?1321748342
I had asked, "Without an oscilloscope and only a HCCT or strobo spark how would I know if I was getting a #1 or #2 (in my diagram below) single spark? And if there is no way to really tell, why would a single spark #1 be any better then a double spark?"
On the other hand, I have to wonder if a person is tooling down the road will notice any improvement in performance with a coil set with an ECCT over a HCCT or Stobo-Spark. I can see another dyno test down the road.......
Maybe the main advantage will be the cost of ECCT over the other options. My HCCT cost me $100, will the ECCT be cheaper?
Jim
Mike,
Could you post a short video showing your tester in operation?
Thanks.
Joe
A double spark degrades performance because the total useful energy in 1 spark is divided between the 2 sparks. You WILL notice lesser performance when under a load as when climbing a hill. Probably NOT when just taking a ride around the block.
Very interesting. And promising.
I'm wondering if it matters that the device sets up the coil with DC, not the up-and-down output (I believe it's far from a sine wave) from the magneto. How about putting your device onto a coil connected to a Strobo-Spark which does generate pulses??
I have doubts that 4 degrees difference between cylinders matters in a Model T. But this has to be added to the variation resulting from the imperfect timer, so maybe it can.
Have you any thoughts on why the timing varies? Could it be differences in cushion spring travel perhaps? There does seem to be a correlation between the peak current and the lateness of the spark, so more 'on' time equals more spark - and the coil stays 'on' until the cushion spring is stopped by the rivet head.
The other conclusion I draw is that we should go to Vendor 3 for our coils!
....and did you actually check all your 12 coils on an HCCT or Strobospark and confirm that they appeared to be OK on that equipment?
Mike, this is very interesting. I will have to take some time and digest this. I have a ton of scope traces just like you've shown that I produced from a triggered square wave. This was how I set my coils up for the Montana 500 for many years, and is certainly a good way to do it. In fact, probably good enough for just about any application. But, the firing time when running on mag is tied more to the action of the timer (timing, bounce and such) and even more important, the form factor of the magneto wave. How a coil behaves with a square wave,(essentially what a coil "sees" when running on DC) and how it behaves with an AC waveform full of harmonics, is a different thing.
Jim, Thanks for your comments. I did read your original thread and did ponder similar thoughts. Mainly about how consistent coil firing timing is by using RMS current as the indicator. Please let me know when you decide to see your HCCT, I will gladly take it for $100. The going price for an HCCT that I have seen ranges from $700 to $1350 depending upon model and condition. The ECCT will be a fraction of that cost and does a lot more with much greater accuracy hopefully.
Joe, I hope to eventually post a video of ECCT operation but may not happen for a while, I need to learn the process.
Chris, The AC versus DC stimulus question is a sure to be questioned and debated. The DC step response is identical to what the coil see when starting on battery but is different from how a coil operates when running on magneto which is typically a triangular wave of varying amplitude and frequency. Still, I believe use of a DC pulse will be very effective in setting up all 4 coils to respond with a first spark with the same timing which is the main objective that the popular methods of coil alignment presently ignore and rely solely on RMS current as an indicator of spark timing attributed to coil point adjustment. Time will tell.
Tom, Thanks for sharing your observations based on your measured data. I agree timing variation associated with coil point adjustment is not the only component associated with ignition timing variation but it is a significant component which should be well worth eliminating. The fact you set up your coil points similar is a good indicator I'm on the right track considering your success. I don't think the present coil point test methods commonly available to the average hobbyist are as effective as they could be for minimizing timing variation. Regarding timer and other sources of timing variation, I do have to save something to do on a rainy day :-)
Regarding coil characterization from different vendors, the varying results I measured should not be interpreted as an indictment or endorsement of one vendor over another. Some coils I received directly from coil vendors and others through 3rd parties which involved handling and shipping. I did test the coils on an HCCT but don't think that was a fair test because I'm not very experienced in the use of an HCCT and I knew exactly what I was looking for and spent the time necessary looking for double sparks. In some cases I was able to produce them on others I was not.
Mike, This looks great and would be a break thru for the many of us that don't have access to an HCCT. Keep up the great work that you're doing for our hobby.
I can't wait to hear more info and when you may have these available.
Mike, I have one Bob Scherzer repro HCCT and two originals about completely rebuilt and ready to test coils, along with a Fun Projects Strobo-Spark tester.
The Strobo-Spark provides two important bits of information that the others do not offer. So far I have only tested about 40 coils for compares, but the only 4 good ones were rebuilt by Ron Patterson. All the others showed bad or badly leaking capacitors and all the newer ones showed capacitors that were half size and still leaking.
You may want to consider having your ECCT test the capacitor for size and leakage.
Mike,
I'm interested in setting up my scope to look at my rebuilds.
In the original set up, what do you use for the "programmable pulse generator"?
Could a PLC be used as the electronic switch?
Have you give any thought to creating an interface between
the Model T ignition system and a lap top?
That way all four coils could be checked under actual road conditions all at once as well as the "Achilles Heel" itself........the conventional timer.
A conventional timer's "bounce" would be visible.
The system could also possibly show a non-concentric timer condition. Wow! The possibilities are exciting.
I realize that the well seasoned Model T mechanic can determine the gap on #3 plug just by the way the engine sounds as the car passes by in a snow storm, but from those
of us who can't I bid you a very sincere Thank You Mike for your efforts.
Thanks for the feedback and words of encouragement; both useful to make the ECCT successful.
Jim, the ECCT does test the capacitor (condenser) for value, open, short and leakage. It also measures the battery voltage and coil current at the instant of firing, not just an RMS or average reading that can easily be skewed by double sparking. Most importantly, the ECCT accurately measures how long it takes the coil to fire a spark when commanded. After all, it is the SPARK TIMING that matters, not the coil current (provided it is within a rather broad range of acceptability). Precise measurement of coil firing time (dwell) was not easy to do in the day so the HCCT utilized coil current as an indicator of dwell time. There really is no reason why folks should settle for that today thanks to modern electronics.
Mike, You can use a commercial pulse generator with 5V output, 4ms width and single pulse each time you press the trigger button. You could also build a simple one shot with a 555 timer too. The electronic switch could be MOSFET or Bipolar transistor driven by the pulse generator.
Spent some more time studying the HCCT with one of the coils that consistently double sparks when tested with an ideal 12V timer contact to ground pulse.
The HCCT was monitored with an optical tachometer to accurately monitor cranking speed. The coil begins to spark at 100RPM when the current reads 1.55A RMS. Cranking speed was increased slightly to 120 RPM to ensure consistent sparking. The amp meter reads 1.2A with variation of +/- 0.25A. No double sparking is evident on the indicator ring while cranking and observing for several minutes. Increasing the cranking speed to 155 RPM produces the first sign of double sparking; the current reading was almost identical; varying between 1.2 to 1.25A.
Here is what the same coil displays on the digital storage scope when operated on 12V battery and commanded to spark with a single, ideal timer contact to ground that lasts 0.004s (4ms).
A properly adjusted coil typically sparks after 2ms dwell time on 12V when the primary current is 5.5A. This coil sparks much sooner, 1.55ms which is 4 degrees early at 1500RPM when the primary current significantly lower; producing a cooler spark.
Here is the same coil again operating on 6V battery and commanded to spark with a single, ideal timer contact to ground that lasts 0.006s (6ms):
Did I miss a clue from the HCCT current measurements or cranking speed that indicates the coil points are not adjusted properly?
Misread the amp meter scale divisions. The current variation was +/- 0.05A NOT +/- 0.25A! That would be bad indeed..... 0.05A seemed pretty steady considering my hand cranking speed variation.
Some thoughts.
When you were turning your HCCT at 155rpm the pulses were happening every 24ms, so there's plenty of time to get multiple sparking if the power is available. But is it? 155rpm is a very low speed, and maybe below this figure, the output was only just high enough for long enough to coax one spark out of the coil before it fell away.
You could gradually reduce the DC supply voltage to your coil to determine the minimum which will fire the coil - and produce a spark, and then compare this with the voltage waveform of the HCCT at various speeds. This might tell you whether my suggestion is correct.
You will have noticed that even perfect coils sound different to each other, especially on a 'buzz box'. Some buzz, some screech and others just vibrate. I was unable to correlate any of these noise changes on a buzz box with double sparking on an HCCT, but maybe I was hearing different multi-spark waveforms which occur with DC.
One feature I have noticed when checking and setting up coils is that on both the HCCT and the Strobospark, if a coil exhibits SOME double sparking, it is very often on the same position on the HCCT ring each rev, or on the same one of the three Strobospark positions. So I conclude that the waveform (shape/amplitude) of the supply pulse does affect the coil behaviour. I remain sceptical about testing with DC.
You said earlier that the magneto output is approximately triangular. Can you generate a triangular pulse with your test equipment?
Now, if you only had your modern kit, you would probably say that this coil is faulty. But what if you only had the HCCT? 155rpm is high for hand-rotation. The coil is sparking at a rate equivalent to almost 5000rpm in the engine, but it is being fed with pulses of length and amplitude equivalent to only 155rpm, so it's hard to know.
Maybe you need to connect your 'scope to a running motor (with a perfect timer - I wonder what that might be?!) and see how this coil and others behave in real conditions?
But don't lose sight of the 'fact' (or at least a subjective result obtained by hundreds of T drivers individually) that coils which perform correctly on an HCCT or Strobospark ARE better than those that exhibit double-sparks. And my estimation is that if you carefully set up coils with a buzz box, about 85% will then be fine on an HCCT. Which means the odds are only evens on having 4 perfect ones!
Chris said: "Can you generate a triangular pulse with your test equipment? "
I have tried this by various means. The voltage waveform of a T magneto is sort of trianglish, but is actually a sine wave with harmonic distortion. One of my methods was to drive the coil with a recording of the mag pulse through an audio amplifier. The back emf wreaks havoc on the amp. The circuit impedance is of course wildly different from a T motor, so you don't have an apple to apple situation.
Chris said: "Maybe you need to connect your 'scope to a running motor (with a perfect timer - I wonder what that might be?!) and see how this coil and others behave in real conditions? "
I think I have the setup for this. It is a T magneto driven by a variable speed DC motor with a fully adjustable timer made from a conventional set of car ignition points. No ostensible bounce, virtually 100% contact.
Back to Mike's ECCT. I really think it will give a meaningful indication in its present form. As I consider it more, I think DC (a square wave) may be better than AC (a rising voltage of some shape or another) because the form factor of a square wave is going to show the maximum timing disparity due to coil-to-coil point adjustment imbalance, impedance variations, eddy current variations and such.
We already have two good coil testers available. The Hand Crank and the Strobospark are quite expensive to purchase. The only advantage to having a third option would be, in my opinion, a lower price. If the ECCT does at least as good a job as the other two, how does it compare in price?
Norm
Norm, other advantages to having a third is maybe better ease of use or more definitive readings.
"with a perfect timer - I wonder what that might be?!"
I assumed he would say the E-Timer!
However, Tom's solution is probably easier.
Mike,
You wrote, "The HCCT was monitored with an optical tachometer to accurately monitor cranking speed. The coil begins to spark at 100RPM when the current reads 1.55A RMS. Cranking speed was increased slightly to 120 RPM to ensure consistent sparking."
I wonder if you have a weak HCCT or if there is something about the coil that is not correct. I see and others have reported that they get a consistence spark on a HCCT at about 60 RPM.
Jim
The question remaining in my mind is this:
Mike has 12 coils rebuilt by reputable people.
Some of them look good on his tester; others not so good.
He has taken one of these and his tester would indicate that it double sparks - not fit for purpose.
When he tries it in an HCCT, it's fine at the speeds most people would check (120rpm max?), but does double spark at 155rpm - a very fast rate of turning the handle. Most people would not get there.
So, is it OK or not? My suggestion is to check it in a real engine with an oscilloscope to check whether it double-sparks or misbehaves. Tom's idea comes close, but the spark plug will not be in a combustion chamber.
I understand Tom's point about a square wave allowing good measurement of timing etc, but it may be that it in other respects, it induces an unrepresentative response from the coil.
If this coil (and others like it) does perform well in a car then, although I applaud both the idea of the ECCT, and Mike's excellent efforts in making it - and I am a fan - it seems to me that it is not (yet) giving a completely useful result.
Incidentally, what does a coil which definitely fails on the HCCT and Strobospark do on the ECCT?
Wow Don't understand too much of this but it Sure is interesting reading. I never imagined how technical our coils are and do you suppose they had all our modern day equipment back in the day when these were being developed?
Keep up the great work guys!
I was going to say the same thing as Gene, I have no idea what they are saying, Understanding it is a wash. As my old boss used to say,,, Boy, Thats way past your pay grade!
Jim Thode : Possible that the cushion spring tension is questionable ??
I'm sure he's using the same HCCT for all his test coils . Mike observes scientific procedures, this is why I believe he's testing on a single HCCT.
Just my $0.02
Jim, also he's reporting the same result on his electrical equipment ???
I took some pictures of the HCCT testing the coil that consistently double sparks when tested with the ECCT 12V 4ms pulse. The HCCT was operated at 125 RPM with the following characteristics:
Top trace: Coil Current; Vertical 2A/div, Horizontal 5ms/div (125 RPM)
Bottom trace: Coil Voltage; Vertical 2V/div, Horizontal 5ms/div (125 RPM)
Notice that there are 2 subsequent sparks following the first spark. These are NOT double sparks since the first spark is of normal current amplitude. They could, however, generate additional sparks visible on the HCCT display if they are of sufficient amplitude. I was unable to capture a true double spark unless I cranked the HCCT faster; about 155 RPM.
Here is the HCCT display seen:
One double spark is visible at the 3 O'clock position which is easily missed while cranking.
The ECCT 12VDC pulse test consistently flags this coil as firing the first spark sooner in time with lower current immediately followed by a second weak spark in the same time frame a normal coil fires 1 spark; The common definition of a double spark documented above. The question remains; how would this coil function when installed in a car. Fortunately, that means more road testing in the future. Too bad we are approaching the end of the touring season here :-(
Hi Mike,
This has been an interesting read.
I do want to ask about your last post and the picture of the HCCT.
Why are the sparks longer at the top of the circle and shorter at the bottom? Does it have to do with camera angle, or is the electrode off center? If the latter, would that make a difference in how the coil tests out?
Thanks for any clarification.
Joe
Mike,
I think the one "double spark" on your HCCT display photo is more likely just a single spark viewed on the start of the second time around. Your photo shutter time is 1/2 second and if you were turning and 125 RPM (= 1.04 revolutions in 1/2 sec), then one spark location could show up in the photo as a double spark in the same exposure. Really it is more likely just two single sparks.
Jim
My experience with double sparks has been two parallel sparks roughly 1/4" to 3/8" apart, not a triangular arrangement like the one at the 3:00 position above. I'm not sure that is a double spark, either. Maybe, but I've never seen one like it.
Joseph, The sparks appear longer/shorter simply due to camera angle.
Jim, Good point I had not considered but how do you know shutter time is 1/2 second? The camera is a Fugifilm Fine pix A100. The spec sheet says shutter speed ranges from 1/4 to 1/2000 sec in auto mode. I'll double check the operating mode, I thought it was auto but then only half the display should be visible.
I guess the main point is that a double spark is not readily apparent from the display yet the coil current registered on the HCCT amp meter clearly has multiple components: the First spark followed by 2 subsequent sparks which skew the coil current reading on the amp meter and have absolutely nothing to do with indicating the quality of the First spark which is responsible for combustion. The HCCT also does not provide the operator with any indication of the spark timing with respect to the time a spark was commanded. I believe both are serious limitations which can result in less than optimal engine performance. The ECCT addresses both of these limitations with a high degree of accuracy then clearly communicates results to the user without further interpretation, skills or experience. Beta test results should confirm or refute its effectiveness.
Mike,
You can pick up EXIF (Exchangeable Image File)viewers online and if you look at the EXIF data that is part of the image file you can see:
Camera Make = FUJIFILM
Camera Model = FinePix A303
EXIF Sub IFD
Exposure Time (1 / Shutter Speed) = 10/20 second ===> 1/2 second ===> 0.5 second
Lens F-Number / F-Stop = 280/100 ===> ƒ/2.8
Exposure Program = normal program (2)
ISO Speed Ratings = 100
EXIF Version = 0220
Original Date/Time = 2012:10:03 19:30:40
Digitization Date/Time = 2012:10:03 19:30:40
Components Configuration = 0x01,0x02,0x03,0x00 / YCbCr
Compressed Bits per Pixel = 15/10 ===> 1.5
Shutter Speed Value (APEX) = 100/100
Shutter Speed (Exposure Time) = 1/2 second
Aperture Value (APEX) = 310/100
Aperture = ƒ/2.93
Brightness (APEX) = -240/100
Brightness = 0.19 foot-lambert
Exposure Bias (EV) = 0/100 ===> 0
Max Aperture Value (APEX) = 300/100 ===> 3
Max Aperture = ƒ/2.83
Metering Mode = pattern / multi-segment (5)
Light Source / White Balance = unknown (0)
Flash = Flash did not fire, compulsory flash mode
Focal Length = 570/100 mm ===> 5.7 mm
Maker Note =
FlashPix Version = 0100
Colour Space = sRGB (1)
Image Width = 2048 pixels
Image Height = 1536 pixels
Focal Plane X-Resolution = 3884/1 ===> 3884
Focal Plane Y-Resolution = 3884/1 ===> 3884
Focal Plane X/Y-Resolution Unit = centimeter (3)
Image Sensing Method = one-chip color area sensor (2)
Image Source = 0x03,0x00,0x00,0x00
Scene Type = directly photographed image
Plus some other stuff.
Jim
Wow! I did not realize the picture file recorded all that detailed data. Really neat. Thanks for the information.
Mike,
Straight from the (digital) camera all image files have EXIF date included. Depending on how images are resized and saved when edited, images may or may not include the EXIF data. For example in Photoshop if an image is "Saved for Web" it will strip the EXIF date off the file but if it is saved normally it will include the EXIF data.
Jim
Does this scare anyone else?
Yes, Fred, our privacy is being wiped out at warp speed.
I applaud Mike for continuing in his efforts to improve the Ford, "Steady by Jerks" ignition.
I suspect that I have already said too much here, but cannot resist one final (?) point.
Consider how the points behave.
When current is applied, the coil magnetic force builds up and the lower point is pulled down. The upper point is mounted on the cushion spring and, because this has a preload, it follows the lower point down and maintains contact. The smooth V and A rise on the traces shows this is working well.
Double sparking occurs when the upper point doesn't maintain contact and, as someone has said, you then see two or more clear sparks with 1/8" to 1/4" between them.
The pull-down continues until the cushion spring is halted by the rivet head. The lower point keeps moving, but not far, because the circuit is immediately broken. Then, the lower point and the cushion spring make their separate return journeys upwards again. The precise way they return depends on their mass, stiffness and pre-load. However, given that they only separate by a tiny amount when the coil fires, it is quite possible that on their way back up they can briefly touch - and that would give a trace as shown. These further twitches on the trace are very small in size - and although the coil has been briefly re-energised, I doubt that there is enough charge to make a spark which will jump the gap.
I'm not convinced that this coil is misbehaving here.
Having said all that, I do agree with Jim that, usually, a good coil in an HCCT will spark from about 60rpm. Mike says this one needs 100rpm or more. Furthermore, although the current peaks at about 3.5A, the voltage only reaches 3.5V. At 60rpm it's probably under 2v. so maybe the HCCT output is low, as he suggests.
Mike,
I certainly applaud all of your efforts.
If manufactured, I'm positive I'll own one of these coil testers.
I do have an "off the wall" question.
The term HCCT stands for Hand Cranked Coil Tester as we all know. I'm assuming (a dangerous thing for me) that ECCT stands for Electronic Cranked Coil Tester which seems to an incorrect description of the device. Wouldn't EPCT (Electronic Pulse Coil Tester) or EPCC (Electronic Pulse Coil Calibrator) be a more accurate name. Just a thought.
Chris, Good points. The coils which clearly exhibit double sparking under 12V pulse test conditions have not been confirmed to result in poor engine performance. The data so far only confirms significant coil to coil firing variation (~4 deg at 1K RPM) which I believe will detract from engine performance. I question the 60 RPM HCCT cranking speed typically cited. That is pretty slow and would like to get confirmation if it is accurate.
Michael, Thanks for the feedback. Will consider your input.
Mile,
I think many folks can verify that it is possible to get a constant spark at about 60 RPM. Here is a photo at 1/2 second exposure with 8 sparks. That calculates to an RPM range of over 52.5 RPM and less then 67.5 RPM depending on when the exposure started and ended.
Jim
Chris Barker:
The HCCT has low AC voltage at 60 RPM's. Agreed.
The important fact is that the Ford ignition coil is an amperage adjusted unit, regardless of voltage.
Cranking the HCCT at 60 rpm's , with the coil firing 16 times every revolution, is approximately equivalent of the coil firing once every two engine revolutions in real time at about 1800 engine rpm's.
Bob, The frequency of magneto pulses is greater using the HCCT due to the absence of a timer. The coil is subjected to 16 pulses per crank shaft revolution as opposed to 1 pulse every other revolution. The actual rise time of the magneto voltage (and resulting coil current) still reflect 60 RPM engine speed which the coil will never operate the engine at that speed. This is has been noted by others and I agree it will this can skew coil performance testing.
Another consequence of not using a timer cited in the past is that the resulting magneto voltage presented to the coil is continuous and not switched by a timer. The coil under test is in a continuous state of change alternating between positive and negative alterations of the magneto output and never allowed to relax in between firings as is the case in car when wired to a timer. The timer interrupts the connection to the magneto and permits the coil magnetic field and points to relax to a steady state condition. Each HCCT spark is not representative of a "First Spark" energized from the coil's relaxed steady state condition. The ECCT pulse test, on the other hand, does test the coil from its relaxed state to characterize the coil's "First Spark" so I believe it is more representative of coil performance in actual use in the car.
Mike:
Thanks for the explanation, will be looking forward to your Hershey demo of your prototype.
Mike listed his Hershey space : "Folks going to Hershey 2012 can see a prototype of the new ECCT up close at Red field space RCC-33."