Has anyone ever checked the voltage output on a mag with a voltage meter? If so explain how and what your readings were. Just curious of the ways people have done this. If you do this another way other than a standard voltage meter tell me. Pics are great!!! Thanks. Matt
It is possible to check the Model T magneto output with a voltmeter, but drawing conclusions of the results requires some knowledge and experience.
The best way to check a Model T magneto is a tool specifically designed for this test. That tool is the FunProjects reproduction St Louis tester.
You can find this tool at: http://www.funprojects.com/products/MT-1.aspx
With the St Louis tester you can quickly and comprehensively tell the health of the magneto and if it has sufficient power to operate the original ignition coils.
Ron the Coilman
I use this st Louis tester that I got from a good friend from the US.
can a mag. be tested with out running car? with a volt meter? I would like to test mag just with hand cranking what voltage should it put out
This thread may be of some use:
Another instrument capable of comprehensive magneto testing is the Electronically Cranked Coil Tester (ECCT). This optional function tests the magneto magnetic field strength similar to the St. Louis Magneto Tester under loaded conditions using the same familiar scale indication.
Software is available to perform more comprehensive testing that includes measurement of the engine speed (RPM), magneto RMS output voltage under load and no load conditions, the magneto internal resistance (including the mag post contact resistance) and a numerical value of the magneto field strength factor.
More information about the ECCT is available at www.modeltecct.com
With the St Louis Electrical Works Meter, are you checking voltage or are you checking amperage?
The St. Louis Magneto tester is not checking voltage nor amperage, it provides an indication of magneto magnetic field strength.
Testing a magneto by its output voltage or current would require operating it at a specific speed (RPM) with a specific load applied. The St. Louis Magneto Tester, developed by Edwin C. Ballman in 1916, was an ingenious approach that which does not require the magneto to operated at a specific speed while providing an indication of its ability to operate a Model T coil by indicating magneto magnetic field strength.
The ECCT takes a slightly different approach to measuring the magneto magnetic field strength using a 2 point calibration where the magneto output is measured at 2 different speeds. The specific speeds does not matter similar to the St. Magneto Tester method.
Tom Carnegie published a circuit plan and description for "A Poor Man's Mag Tester" based on the St Louis Electrical Works magneto tester. It includes an ampere meter. Is there a difference? How and why?
Tom's article provides some additional insight into the St. Louis Magneto tester basis of operation, however, under emphasizes the critical importance of the inductor utilized for the application. The inductor value and construction are critical to the accuracy of the St. Louis Magneto tester.
The use of a single Model T coil as the inductor would raise the equivalent operating current to 5.0A which is sufficiently high to introduce significant distortion due to core saturation that yields a much lower reading; ~3A or so according to the article. Furthermore, Model T coils vary rather widely in their inductance value, +/- 15% range based on my test data, which would impose a corresponding +/- 15% variation in the reading alone of the magneto tester readings built using a Model T coil.
The St. Louis Magneto tester uses a precision inductor with tightly controlled value that is much higher in value than a Model T coil. The core is also specifically designed to operate far below saturation at the operating current. Another important detail is the type of AC ammeter used to measure the test current. An RMS ammeter must be used to accurately represent the magneto magnetic field strength since the waveform is not sinusoidal. An AC voltmeter calibrated to read the RMS value of a sinusoidal waveform will introduce an additional error of ~23% from the expected meter reading.
Be sure and set your volt meter to AC. You will get "funny" readings with it set to DC.
You know guys, I'm betting the John Regan magneto test with the 1156 bulb would be sufficient for what Mathew was asking about.
My "Poorman's" series in the Montana 500 newsletter is a bunch of procedures directed toward the Montana 500 group that mostly border on the edge of being micky-mouse (the procedures, not the group). There are likely better ways to do most anything I describe if you are willing to spend some money. Having said that, my micky-mouse mag tester actually works quite well if you are just going for an A/B comparison. That fact that coils vary plus or minus 15 percent doesn't matter if you use the same coil each time. Same with the volt or amp meter used. The St. Louis meter has no significant scale on it anyway. What does 1915 mean? Does that mean a mag is good if it reaches that mark? What if you are testing a 1923? Where is the mark for that? In my experience, most folks that use the St. Louis meter will say something like "my mag puts out just above the N in MAGNETO, and works just fine". As far as I am concerned, one subjective test is pretty much as good as another subjective test.
I tried hand cranking with a a-c digital meter and I didint get any reading but the scale is 750 volt I will try with a analog meter when I can find it. should the mag lite a small bulb hand cranking?
"You know guys, I'm betting the John Regan magneto test with the 1156 bulb would be sufficient for what Mathew was asking about."
Yes, probably true, but then all that time, research & money spent on the Large Hadron Collider would have all been wasted.
I have to agree with Toms comments regarding the poorman's tester, as a comparison it would be fine. Out of interest does anyone have (or willing to supply) the inductor value and its construction as fitted to the St Louis Magneto tester?
You could just use a multimeter and you will find about 9 volts at idle and about 30 volts at high speed.
But will it do it with a load?
To clarify my comments regarding Tom's article, I applaud the effort to experiment and improvise.
My point was that the St. Louis Magneto tester is a very clever and elegant design but must be built with the proper components to earn that distinction. It can then provide a repeatable standard in which magneto magnetic field strength can be reliably indicated independent of RPM, albeit with a seemingly arbitrary scale - an unfortunate choice. The elegance of the design is totally lost if the wrong components are used to construct it because the principals of operation are totally changed, altering the ability so serve as a reliable performance standard the properly built tester is.
In the arena of experimenting and improvising, I posted the following on the Forum Monday, November 10, 2014 - 11:24 am:
""(I) Finished the Magneto Current Tester (Ampere meter). There has been past discussion about measuring the voltage and amperage out put of the magneto. The main interest that I have is amperage at cranking speed. According to the notes from Ford Service Bulletin for 1919 at 200 rpm the output voltage is 5-volt AC and 6.1 amperes AC. Measuring voltage is relative easy, but to measure amperage one needs something similar to the St Louis Electrical Works magneto tester. The voltage can be measured with any AC voltmeter. But measuring the amperage has been a challenge. Tom Carnegie published "A Poor Man's Mag Tester" based on the St Louis Electrical Works magneto tester. I was looking for a more period instrument.
I built an amperage tester based on a swap met 10 Amp Jewell Portable AC Meter ($20.00) and a coil core from Snyders ($19.25). In the diagram found in Dykes the St Louis meter used a "reactance coil" to provide a constant current. So why did I decide to use the coil from Snyders as a reactance coil? According to a 1920's publication, Automobile Catechism, a reactance coil is a kick coil. And the kick coil is a kind of spark-coil used for low-tension (make and break) ignition when electrical current is supplied by a battery. Based on the description found in the text the coil from Snyders is a reactance coil.
I have provided a diagram of the finished circuit. The black box contains both a one-ohm resister and the reactance coil. Coil has a resistance of .7 ohms. The meter with shut is for 10 amperes ac."" (note: Inductors do not behave the same as resistors. DC current resistance. AC Current inductance.)
I am not going to make any statement about how it works for me. I let this up to those who enjoy a challenge.
If I remember my history of design for the magneto, the re-engineering of the magneto pick-up from 1914/1915 and 1915/1916 is one reason the St Louis Magneto tester indicates for 1914 and 1915. There was a significant improvement or amperage/voltage in the later magnetos (1915).
George, thanks for the info, in the interest of experimenting and improvising I coupled one of these together, seems to work well. To calibrate I will use a good rebuilt engine / magneto output and a known magneto with poor output. Not very scientific but usable in the future testing of magnetos.
An important detail not provided about the St Louis Magneto Tester is the fact that the voltage numbers are not as important as a constant reading across the normal engine RPM range.
Anthonie's reading in his photo is showing a good Magneto, if it remains relative constant over a wide RPM range of engine operation.
The early test instruments for the magneto were go or no-go variety. The St Louis meter's range was up to 1914 or after 1914(1915). The Sterling Magneto-Meter registered the output as low (0-11 v), fair (11-17 v),good (17-24v), and high 24-30v), and then the Hoyt Magneto-meter with a range of P-M-G-E - poor (5v), marginal (10v), good (15v), and excellent (20v.) The KRW magneto tester has a scale that state the condition Starting or Running. With the additional information of poor, fair, good, and extra. The other alternate is using the provision allowed with the design of the Hand Crank Coil Tester. There are provisions to connect the meter to the magneto with the engine running and measure voltage produced. If a reading of .8 is achieved it is within the go or no-go tolerance.
To answer the original question - what is the correct voltage? -- any reading in the middle of the meter is acceptable. But if a 1914 (original) a bit to the left is ok. After 1915 (original) in the middle and a bit to the right is ok. - as long as the coils are buzzing.
Here's the 1156 bulb test: http://www.mtfca.com/discus/messages/331880/MAGNETO_TEST_58163152-332508.pdf
Jeff thank you for the other go or no-go test.
The simple way to measure the output amperage would be to put a 0.1 ohm resistor in series with your coils or other load and measure the voltage drop across the resistor. The amperage would be the voltage drop divided by the resistance. At the same time you could also measure the magneto output voltage. You could also borrow someone's strobe light and measure the rpm.
Unless you actually own a St. Louis Tester and use it for awhile or understand it, you will not learn to rely on it as the only really accurate measure of the EXACT health of the magneto on an absolute basis without comparison to any other magneto or without the need for any other device. People don't want to spend the money to have a good tool. Sometimes you are then simply stuck. When seeking an absolute determination of the magneto health the actual voltage reading of a magneto is meaningless unless you can establish the exact RPM and exact load current that is connected to the magneto at the instant that the voltage is measured. AC volt meters alone show you voltage or "LOW, FAIR, GOOD, and HIGH" and sure you can use a bulb and an AC volt meter and get some idea if the thing is working but if you want to know for sure why a car won't run right on coils and magneto you need some way to determine absolute health of the magneto. All of the AC voltmeter types include in their instructions to compare readings on a known good magneto but fail to define what a known good magneto is. The St. Louis tester was mentioned a lot in Dykes early service books since it gives one and only one reading on a magneto that is working and also tells if that is enough to run coils. I make the ST. Louis tester and at first I was going to mark the scale as "1915 and Later" rather than just the "1915" mark but I wanted to reproduce the whole meter exactly as it was made so I left the marking as it was but if you follow the history of the magnets you will see that 1915 and later are the same strength The slight difference in DC resistance of the later mag ring is essentially a non event to the St. Louis meter and the coils too. The heart and soul of the device is a very very wide band precision inductor that does not change with excitation level or frequency (RPM of engine changes the frequency). Then an accurate true RMS responding meter to indicate the strength. Inductors are hard to make today that fit the bill but are actually way easier than in 1916 and I was amazed at the original inductor's ability to work very well considering what it had to be. OK the box is finger jointed red oak as original and it has real glass lens and metal front frame. I could have cheapened it to a nailed together bunch of poplar pieces but I duplicated every detail that I could duplicate. It is a very accurate instrument for those who want such things. For those who think its not worth it - don't buy it but don't kid yourself that a T coil is just as accurate. If it was - I would have used a T coil winding since I have lots of them. Sometimes you guys spend more money and time just to jury rig something that you think is clever when in fact the original idea was way more clever in the first place. I take no credit for the engineering of this fine 1916 instrument and I believe it was the best magneto tester ever made but Hoyte came out with their "toy" meter that said poor, fair, good, and extra but then told you to compare it to a good working magneto. Why not then just put the coils into that car with the good working magneto and if it runs OK the mag is the problem - or is it? I can find out with my trusty STL Mag tester.
I got 35 volts on one of my cars, with a late style mag.
Now I wish I didn't sell mine. (LOL)
I know the guy you sold it to and my bet is he will not keep it from you nor want a profit in your getting it back.
I have a Fun Projects St. Louis meter. I sell Fun Projects St. Louis meters. I love my St. Louis meter and recommend them to any mag-centric Model T person. I use my St. Louis meter often. I am glad that John duplicated it as close as he possibly could to the original, even to the markings.
Now, having said all that, I disagree with most of what Mike Kossor wrote above, and I promise you that my poorman's meter works as stated in my article.
People make fun of my poorman's stuff. That is to be expected, I suppose. But many people have done stuff that I've written about and thanked me for the ideas. These are the people that I'm writing for.
Tom, thanks for sharing your observations of measured data on your Model T coil based magneto tester. Measured data is difficult to dispute and prompts me to do some experimentation because the observations you reported are contrary to theory, even with several ideal component assumptions.
I do concur Model T core saturation should not be significant factor affecting performance. The main issues using the Model T coil for the inductor is the ratio of inductive reactance (XL) to resistance (R ) and the frequency range the value of L remains constant since the magneto output is not sinusoidal but close to a triangular wave which includes all odd harmonics of the magneto output frequency.
Assuming the Model T coil inductance (L) does remain constant over a wide frequency range (30 to >6000Hz), the linear model I used to predict performance indicated the total series resistance would introduce non-linearity which would limit the useful range of accuracy. Here is a predicted performance performance comparison of using a linear inductor versus a Model T coil as the inductor:
Linear Inductor model assumptions:
L=25mH, RT = 2.35 Ohms (R=1.6 Ohms, Rmag = 0.75 Ohms)
The St. Louis Magneto tester reading (current) of a properly functioning magneto (coils, gap, magnets) will remain constant (< 2% variation) over the full typical operating range of the engine (see top chart). The value of current indicated on the meter will only be higher or lower if the magnet strength of the magneto is higher or lower respectively. Fluxuations in meter reading could also indicate variations in the magnet gap (crank shaft end play).
Model T coil model assumptions:
L=3.3mH, RT = 1.05 Ohms (R= 0.3 Ohms, Rmag = 0.75 Ohms )
The modeled results of a magneto tester using a Model T coil as the inductor is not as linear (even assuming ideal L value with frequency) just due to the much lower ratio of inductance to resistance. That means a constant reading (<2% variation) would not occur over the full typical operating range. Much greater variation will occur at lower RPM, incorrectly indicating possible problems as gap variation or other problem exists.
Will be interesting to determine why you do not observe similar variation in meter reading as engine speed varies, especially at lower RPM.
Very good work Mike. My simulations showed a lower effective range for both the SL (St. Louis) meter and the PM (poorman's) meter than yours did. I used a total of 1 ohm of resistance, which is probably too low. This next line is mainly for Mike's edification. For my AC source I used two sources in series, the second at 3 times the frequency, down 15db, 120 degrees out of phase to simulate the 3rd order harmonic content. My simulations as yours, show that the SL meter had a much lower effective range. I assumed that this was because the PM meter requires about 8 times the power to operate compared to the SL meter.
I am going to hook up a PM setup onto my car and try to see when it appears to go linear.
Ok, I hooked it up. From 522 rpm (which is slower than my T wanted to idle) to 1240 rpm (which is really racing the motor). The readings that I got were essentially linear (within .2 amps on a 5 amp scale).
Mike, above I said 120 degrees, I meant 180 degrees. Sorry 'bout that.