Just curious, any comment about this ignition devise (other than "not original"). I did not read the article.
http://www.ebay.com/itm/Model-T-TT-Ford-Master-Ignition-Vibrator-for-1909-1927-M odel-Ts-with-4-Coils-/111236572816?pt=Vintage_Car_Truck_Parts_Accessories&hash=i tem19e6376690&vxp=mtr
I myself wound never use it but Is be interested if it works or not. Seems like there was another fellow a while back that made something like that.
I have one installed on my TT and am very impressed with the results. The points and condenser are by passed other than that the ignition remains the same.
i guess that means no buzzzzzzzzz ?
My mind ...
Tis' in the gutter I fear this evening ....
The ignition is not quite the same. The original system has the coil firing synchronously by pulses from the magneto and those pulses are precisely timed with flywheel position and thus piston position. The MV linked is basically an asynchronous system with the MV putting out a stream of pulses but they are not slaved to flywheel position. The frequency of the asynchronous oscillator used is chosen to be a compromise for the typical useful RPM range of the T motor. The faster the RPM the greater the error in timing but since the T is not a high performance setup, the thing will function. The timing of the actual spark event is determined totally by the timer and the next pulse to show up after the timer lands on contact. This is similar to the Ford method except the exact time that the pulse shows up is randomly chosen from a pulse train that is continuously running without regard to flywheel or piston position. I have no personal experience with building or using this sort of thing but I understand the theory of it as posted in the MTFCI article.
John, Thanks for the explanation.
Mark, be careful what you say about what you use on your TT, you will be tarred and feathered if not in line with the general consensus.:-) KGB
If my explanation came off as being critical of the system then I am truly sorry since that was not my intention. I simply wished to explain that there was more to it than simply bypassing the points and condenser and otherwise being the same as original. The magneto based "timing" being defeated is a key element of this approach but that is not stating that it does not work nor that it should not be installed. If it reads differently than that - it was not intended.
I have a folder with 6 files in it about the Model T electronic master vibrator. Pictures, enough info to build your own if you wish and a word file describing operation and theory. If you want it, PM me with your email address.
In theory because the incoming supply is rectified and filtered, the oscillator sees nothing of the magneto waveform.
But given that the filtering won't be perfect, it's just possible the oscillator will synchronise with the magneto waveform. 2200uF won't remove all the ripple at the current drawn by the coils.
Either way, still probably better than a set of misadjusted coils.
I don't think so. The T coils when running on magneto are operated every 22.5 degrees but the pulses alternate in polarity and D1 in the circuit will only conduct on positive cycles so half of the magneto pulses are going to do nothing for sure. The frequency of the MV is going to produce a square wave with pulse widths somewhere in the vicinity of 3.3 mSec on and 3.3mSec off and I don't see how applying half wave ripple at the low frequency of the magneto is going to do anything beneficial to improve the timing since the timer will connect up this circuit at any random time during the MV square wave. But perhaps I am wrong.
I think JohnH is essentially correct. The 2200mf cap should hold enough energy to power the circuit independent of the magneto waveform. A 2200mf cap should hold about a joule of energy, which is one watt/second. That would be 333 watts in 3ms. That should be plenty of power. The flip-flop is triggered by the timer, so there is no randomness. The coil should fire 3ms after the timer makes contact. The timing will still be subject to timer vagaries. Also there will be some timing retardation at higher rpm's, but less pronounced than with a 6 volt battery. As an example, at 500 rpm (idle) the timing would be retarded 9 degrees. Assuming that you advanced your lever to make this the correct timing for this rpm, there would be an additional timing retardation of 9 degrees at 1000 rpm (25 mph). This is about the same as running with a 12 volt battery.
Here is a Spice simulation of the MV circuit for those interested:
This is operating at approximately 500 RPM. In addition to the asynchronous operation with respect to the crank shaft, note the variation in coil current due to the voltage droop as the coil charges and discharges during the off cycle when the magneto reverses direction and the diode stops conducting. So the simulator predicts ripple but the coil current (right scale) always charges to sufficient current to produce a good hot spark. However, there is a consequence of operating the coils based on fixed dwell times (which is ~2.2ms according to the simulator) is the coil current increases as the magneto voltage increases (i.e. engine RPM increases). Here is the simulated performance at 1000RPM:
Note the peak coil current is pushing 15A! or nearly 3 times the normal peak coil current. Also wonder why Q4 was added since IGBT Q3 should be capable of driving the coil.
Disclaimer: The circuit simulations are only as good as the model. Magneto internal resistance was considered ideal so the output will be somewhat reduced. Thought this would be useful to aid in the discussion underway.
I don't think it works that way. This thing is powered on and stays that way with the MV free running and supplying pulsating DC signal to the coils while the other side of the coil is then hooked to the timer going to ground. The circuit it not a one shot MV. When the timer grounds the coil either side of the MV can be in the on or off state and the timing may have to be delayed for up to 6.6 Ms if the off state of the MV is just starting when the timer grounds the coil. The circuit is a text book free running multivibrator (MV) circuit and upon initial application of power either side may start in the on or off state at random but from then on the circuit puts out a square wave whose duration in each state is the result of the time constant determined by the 33K and .1uf timing for that side. It will run continuously until power is removed. It has no trigger input. You can't reliably slave the frequency of this sort of circuit by applying power that contains a ripple frequency at half the desired sync frequency. Am I missing something?
John R., you're right, it is free running. Therefore the timing could vary by one MV cycle (2.2ms according to Mike's simulation). It seems to me that a higher MV frequency might be better for both timing and current limiting at higher rpm's, but I think Mike's simulation is overstating the current, so maybe the designed frequency is a reasonable compromise.
For those of you who don't remember and/or don't know, I have been making electronic coil conversions.
From my experience and my electronics background, I can say that this design will work, but it has some drawbacks. For sake of simplicity, I'm only going to be analyzing it assuming that it is running off of a battery, not off of the magneto, as my coils are meant for people who don't have a working magneto and must run off of a battery. Still, running off the magneto should not change how the circuit functions noticeably.
From this design, the oscillator runs continuously as long as the key is on. In other words, the oscillator runs independently of the timer and the engine's RPM. Therefore, the point from which the timer first makes contact to when the coil fires will not be consistent, and neither will be the timing.
For instance, through each revolution of the timer, the coils might fire at (that is, after the timer makes contact). These numbers will change from one complete revolution to the next:
From these numbers, we can figure out how retarded the spark will be in degrees from after the timer makes contact. Here's how I did it, assuming that the engine is running at 1,200 RPM:
1200 RPM / 2 = 600 RPM, timer speed.
600 RPM / 60 seconds per minute = 10 Revolutions Per Second
10 RPS * 360 degrees = 3,600 degrees per second
3,600 DPS / 1,000 milliseconds per minute= 3.6 degrees per ms.
Thus, we find from the numbers above, we get:
This will certainly make for a rougher running engine, and these differences will be even more pronounced at higher RPMs. Would it work? Yes. Would the engine rough smoothly? Perhaps not so much, but I can't say if differences like these would be enough for you to tell, or enough to harm the engine. I can safely say that more precise timing would be advantageous.
My coils, on the other hand, feature one oscillator per coil, and thus, they fire at a very specific and uniform time from which the timer first makes contact. This allows for a very smooth running engine.
I will bow out since the explanations are getting too long and are redundant at best and beyond the expertise of most of the hobbyists. For those of you out there that don't understand all of this, not to worry. Pick the simplest explanation and it will suffice you since all of us are saying basically the same thing. The device will work but it may not improve the performance for you much unless you have coils with poorly adjusted points in the first place. It works on both battery and magneto but does require some wiring changes to your ignition switch or coil box to insert the MV into the power side of the coil box.
Ill do my best to explain it while avoiding as many technical terms as possible.
Yes it will work with the magneto, and it will also work with a battery.
However, it will cause inconsistent and inaccurate timing, which may lead to a engine that runs rough.
The coils should still buzz. If not, the points may need to be fiddled around with. The circuit doesn't actually use them, so it doesn't matter how they are adjusted.
Cameron, in my opinion this forum has room for both hobbyists and technicians and everyone in between. Feel free to say what you want to say. Those that have an interest can chime in, those that don't can bow out.
Well said, Tom, I agree.
The bottom line is that the electronic master vibrator will hurt performance of a Model T that has properly working Ford ignition system. It is designed to "crutch" a set of poorly adjusted coils into making the car somewhat drivable.
I would say it this way:
The electronic master vibrator may not have as good of performance as a well adjusted Ford ignition system, but it could be used as a tool to allow a car with poorly adjusted or difficult to adjust coils to be made drivable.
For instance Heinz coils (as Royce well knows). If you use the T mag to charge a 12 volt battery it would seem to eliminate some of the alleged short comings and then you can run signal and brake lights on a "wind-up" T as well
Well I applaud R.L. Merz and Craig D. Merz for their creativity and initiative for designing, developing and sharing their work with others in the hobby who may enjoy and benefit from experimenting with their electronic Master Vibrator.
At least their effort was a constructive contribution to the hobby encouraging experimentation which is far more laudable in my view than those who cast negativity, criticism and discouragement on others.
This was a very positive discussion and it was a joy to follow it.
This is a great discussion. I always learn allot here. I have one of these timers and it works good on my '26 Roadster. I ran it on 12v to check it out and it runs and idles smooth. PK
The full write up and details of this item can be found at:
Operation on a fixed 12VDC has the advantage of fixed peak coil current that would be approximately 6A capable of producing a sufficiently hot spark for reliable combustion without over stressing the coils especially at higher engine RPM.
Care should still be observed to avoid prolonged coil sparking with the engine off and coils buzzing while operating on 12VDC to avoid over stressing (heating) the coils; same as operating the original Model T ignition system on 12VDC.
If it is of sufficient interest, I could run a performance comparison using the Model T Data Acquisition System (TDAS) I built that simultaneously monitors Engine RPM and vehicle speed versus time to see how the Electronic Multi-Vibrator (EMV) performs with respect to professionally re-built and adjusted Model T coils in the spring. This type of performance comparison is exactly why the TDAS was designed, developed and built. It provides solid quantitative data that permits A/B performance comparisons void of speculation, conjecture and bias with all other operating parameters held constant.
An article detailing the TDAS design along with some performance comparisons of frequently contrasted operating conditions was written and submitted for publication. Hopefully, it will be accepted and shared with those interested.
I would really appreciate you doing that. Trying to get "objective" information can be a struggle at times!!!
Mike, if you take away all the speculation, and conjecture and bias ... what fun are we gonna have? Some folk will just plain explode if they can't purvey falsehoods and such on a regular basis!
(all in good fun)
So you are saying what John and Tom said above is false? What would cause you to think that? Please explain where their fundamentally simple explanation of the circuit is in error?
I had followed this thread from a distance (due to Heinz coils on our 09) and just noticed it had "resurfaced." Interesting to see your post, Royce.
Some people act as mediators when others disagree or seem to have different opinions. Apparently not in this case.....
Guess that's why your not back on the 1906 thread. I'm still waiting there for you to acknowledge you were incorrect about 1906 Stanhope Pierces, and that Ford did not order Model K chassis after the initial contract. I'll be there when you get to it.
Stop being a troll.
As has been a longstanding tradition of men who can't agree or get along, I propose that Rob and Royce fight - winner is right forever and loser has to keep his mouth shut about winner's posts. Step forward, and battle as MEN!
Glad you didn't suggest dueling.
I don't care who's "right," I just don't care to be told I'm wrong frequently.
In all seriousness, I've had nothing but trouble with our Heinz coils and on one occasion used a master vibrator to get by while waiting for a coil to be repaired. I did not fix the vibrator to the firewall per manly, but have thought of doing so. This sounds like a viable option.
Heinz coils are well known to not work well. The reality is they NEVER worked well. Which of course is the reason the KW style points were invented.
Like you, I have a original Heinz box and coils. This seems like it might be a really good option. Consider
1. You don't need to drill any new holes ANYWHERE!!
2. Totally and easily reversible.
3. Certainly cost effective.
4. And as you don't have a flywheel magneto, you don't have to be concerned about some of the issues that have been suggested.
All the best and Merry Christmas
Yes, sounds like a great "stocking stuffer" for our Model N or 1909 T (I inadvertently "killed" the mag on it).
Now Rob is a troll, too.
That makes both of us trolls. So glad to be in the company of trolls.
You do have red hair don't you Rob?
Rob, have you tried an in car recharge?
It seems a shame to try to get this thread off of a discussion of trolls and back to Model T stuff but I'm gonna try.
Neglecting magneto operation for a moment, can anyone tell me why a T requires multiple sparks to operate on a battery?
It would take repositioning of the timer, but when the brush leaves the timer contact, it breaks a circuit and that should be enough to produce a single spark to the plug, much like in a modern car. You would still need the timer and 4 coils but no need for coil points or multiple sparking.
Dave,Acording to those Yoopers anyone who lives below da bridge is a troll.This time of year da beer is frizen so what do day no?? Bud.
There was a thread years ago on this forum where someone mentioned a set up just like that, and it did work.
My only query is that with no capacitor from each of the timer contacts to ground, the performance of the coils might be a bit indeterminate. It would be easy enough to add 4 distributor type capacitors if one wanted to do it correctly.
In simple terms that is how an E-Timer works. When power is removed, the coil fires. As John said the problem would be good contacts/points in the timer and a capacitor would be required on each contact. Of course it would not work on mag power.
With the all the ideas that have been developed to "improve" the Model T ignition system you can bet that if it was a good idea, it would be common today.
The problem with using the timer contact as the mechanism to control charging the coil is the coil charge time (dwell time) will vary depending upon engine RPM. That's why the timer only enables a coil to begin its process of generating a spark for its corresponding cylinder in need of ignition at the proper time but the coil and points are responsible for generating the spark optimally and consistently irrespective of engine speed.
In a normal distributor ignition system the dwell angle is set but the dwell time varies based on the speed. Why would the dwell time be a problem for Model T coils and not for "normal" coils?
I would think that using the commutator to cause the spark would require it to be re-positioned so that it breaks at TDC and the coils adjusted so they do not vibrate. I suspect it would only work on battery, my mind is a little fuzzy on why it would not work on magneto. I really don't see any theoretical reason why it would not work, maybe someone should try it. There's a nice little task to keep you busy over Christmas.
Not sure if the Model T coil can withstand the prolonged dwell time of slow engine operation. The dwell angle of my New Day Timer is 45 CAM degrees so at 500 RPM the coil dwell time would be 30ms; the typical dwell time is 4ms operating on 6VDC (2ms on 12VDC). Reducing the dwell angle to 30 degrees would reduce the dwell time to 20ms at 500RPM and still provide 4ms at 2500RPM but again, don't know if it could sustain the higher power dissipation associated with a 5X dwell time. If it could, it should be possible to operate the Model T ignition system as proposed provided the timing associated with when the timer ends the dwell interval could be accurately controlled.
Worse yet Mike, when you turned the switch to battery to start the car, if the timer was grounding a coil, there would be full voltage across the primary winding for 15 to 30 seconds while you walked to the front and cranked her.
If 30 ms would possibly be a problem, 30 seconds could be even worse! I guess standard coil/distributor systems must have the primary winding sized to handle the current.
Here is a previous thread where running coils without points was discussed.
Yes, of course, that would indeed be a worse case I forgot to consider. That is why breaker point ignition systems typically employ a ballast resistor to limit current through the ignition coil to a safe value. The ballast resistor is bypassed (shorted) during cranking to compensate for battery voltage droop that occurs during cranking as the starter motor draws heavy current. Coil current is not continuous during cranking as the CAM is turning and the points are opening and closing.
The problem with adding a ballast resistor to Model T ignition system to prevent excessive coil current draw and excessive power dissipation is the added resistance will increase the necessary coil dwell time. This will require additional spark advance (more timer rotation) as engine RPM increases to compensate and maintain engine performance.
I did an "in car" charge and the result wasn't great. It will run on mag, but not as well as on battery (after I tried an in car charge). The mag worked fine before, but I accidentally shorted (for just a moment) the brass switch cover and mag and battery wire when working on the switch (stupid of me, I should have disconnected the battery first ).