Mercotronic Model 98 and Model 9800 Information

[29Chev]

Here is the second picture that identifies the contact set

[joecb]

I have been following this thread with interest, lots of good info and thanks to all that have contributed.  Relative to the issue of the vibrator and possible replacements, I just stumbled across a neat motor gear box in the latest catalogue from “American Science & Surplus”  ( item 49163 , $5.00) that just might work if the proper lobed cam were installed on the output shaft. The motor is 1.5 to 12 VDC draws 250 to 750mA,  at 9 volts the output shaft is 75RPM,  Nice compact unit 3 X 3-1/2 X 2 1/2. … output shaft is 5/16 dia.

Might be worth while checking it out if one is in need of making a replacement vibrator.  Someone smarter that me will have to figure out the cam lobe design at 75 RPM to give the desired pulse frequency.

Joe B

1 user thanked author for this post.

29Chev

[29Chev]

joecb wrote:

I have been following this thread with interest, lots of good info and thanks to all that have contributed.  Relative to the issue of the vibrator and possible replacements, I just stumbled across a neat motor gear box in the latest catalogue from “American Science & Surplus”  ( item 49163 , $5.00) that just might work if the proper lobed cam were installed on the output shaft. The motor is 1.5 to 12 VDC draws 250 to 750mA,  at 9 volts the output shaft is 75RPM,  Nice compact unit 3 X 3-1/2 X 2 1/2. … output shaft is 5/16 dia.

Might be worth while checking it out if one is in need of making a replacement vibrator.  Someone smarter that me will have to figure out the cam lobe design at 75 RPM to give the desired pulse frequency.

Joe B

At 75 rpm the motor would need to have multiple cam lobes so that the points opened and closed at a respectable speed in my opinion.  A point set on an engine running at 3000 rpm would be opening and closing 50 times a second if an engine was a single cylinder and the points were driven off of the crankshaft.  If the points were driven off of a camshaft they would be opening and closing at 25 times a second since most 4 cycle camshafts rotate at half the engine speed.  If the cam on a motor operating at 75 rpm had 8 lobes it would be opening the point set at 75 x8 or 600 times per minute which would be 10 times per second.  This would equate to a very low rpm test of an ignition coil in my opinion and might not show up a problem that could be present at higher engine speeds.  To increase the number of lobes on a cam to a higher number than 8 would probably result in a very low dwell number (the amount of rotation duration in degrees that a point set remains closed).  A low dwell number could result in the point set not being left closed long enough to properly saturate the primary winding of an ignition coil before the point set opened and the coil fired.  If you could find an electric motor that operated at a much higher rpm speed it might be better suited as a possible replacement for a bad vibrator.  The other consideration is that there is not an abundance of room inside a Model 98 or Model 9800 unit for an electric motor.

Just as a matter of interest I will mention that in testing the original mechanical vibrator we found that the vibrator operated at a frequency of approximately 100 hz (100 cycles per second) and that during each cycle the two sets of contacts each opened and closed once.  The duration that the contacts were closed created a duty cycle (closed time) of approximately 3 ms and an open time of approximately 2 ms during each cycle for each contact. This may sound confusing but the end result is that the vibrator emulates a set of points opening and closing  approximately 200 times each second  which would equate to a set of points opening and closing in an engine running at 12000 rpm if the points opened once every revolution driven off the crankshaft or 6000 rpm if the points were driven off of a camshaft on a single cylinder 4 stroke engine.

The electronic timer circuit I have been experimenting with and the circuit the other chap on the Smokestack website has actually working in his Model 9800 Mercotronic emulate the original mechanical vibrator very closely as far as frequency and duty cycle.  The circuit that I have been working with has a frequency of between approximately 195  and 215 hz (depending on the voltage input 7.0 volts – 12.5 volts)  and the on (points closed equivalent) duty cycle is approximately 3.5 ms and off for approximately 1.9 ms – this too varies slightly with the voltage level being used to test the coil. The other chaps circuit uses a small 555 timer module that can be purchased for less money than the individual components which is why he chose to go that route I believe.

One other thing we did learn from documenting the Model 98 and Model 9800 is that the original Mercotronic vibrators purpose was to emulate a set of contacts opening and closing such as what would be encountered in a points and condenser style ignition system.  The two separate contact pins of the vibrator that feed interrupted current to the coil under test are actually connected together physically with a jumper wire in the vibrator socket.  This is one thing that differs from how a radio vibrator is used as in a radio application one set of contacts powers half a transformer primary winding and the second set of contacts powers the other half of a transformer primary winding to create high voltage AC current at the secondary winding.   We believe that the original vibrator that is used in the Mercotronic was designed to work in a 6 volt positive ground circuit.  Since the original mechanical vibrators are no longer being made new as far as we know some people may be tempted to try a new all electronic version of a vibrator in a Mercotronic that is designed to work with older radios where the vibrator was used along with a transformer to create high voltage AC to power the tubes.  I do not think they would work properly in a Mercotronic since most are designed to switch two transistors on and off – one transistor turns on for a certain amount of time and turns off and at the same time the other transistor turns on for a certain amount of time and then off and the cycle repeats. The time period between when the first transistor turns on and off and the second transistor turns on and off is so instantaneous that there is very little if any off time between when the first transistor is on and the second transistor is on in most electronic vibrator versions designed for radios from what I have been able to find out.  This off time is needed in the Mercotronic to allow the ignition coil to actually have time enough to fire and oscillate properly before the primary windings begin to have current supplied to them again which is why we used a 555 timer circuit to power a single Mosfet transistor on and off – this allows for the proper off time the way the original mechanical vibrator contacts provided.

Hopefully this post does not bore readers but I thought I would provide a bit more insight into the original vibrators purpose which is slightly different that what a vibrator in a radio does.

[Tom]

An alternative to mechanical vibrator tube or an electronic replacement is the old school method of using the primary circuit of a Model T buzz coil.  These coils are readily available and you don’t need one with a good secondary because you just ground that out anyway.  Ground the side terminal to the bottom terminal, and insert the coil in the test circuit using the top and bottom terminals.  Throw a rheostat and an ammeter in there and you have a coil tester that lets you adjust voltage to see the minimum amperage the test coil will fire at.

T

1 user thanked author for this post.

29Chev

[Tubs]

Thanks so much for taking the time to attempt
to diagnose the issue with my unit. If my testing
equipment appears to be amateurish it would
be because I am.
Took a copy of your test suggestions to the boat
house and below are the result.
Tubs.

A "Boathouse Repair" is one that done without having tools or the skills to do it properly.

[jeff-register]

I have worked on the vibrator type mot. Using a burnishing tool, I’ve added material to the contact points with good results. I have thought about using  a “555” chip to form the polsed signal to replace the electro-mechanical viberator. Use that with a darlinton transiator for output.Unfortionally I have been very sick for the last year, almost died. Also thought of using a vibrator from an old car radio as they fed a pulsed signal to a transformer to “UP” the voltage to run the heaters in the vacuum tubes. It’s just a matter of matching the pin-outs on each vibrator. This thread gives me incentive to do research using electronic “555” chip based unit to help others, Using a R–C timing math to acheave the correct frequency. A capacitor, commily called a condenser is the timing  devise controlled by resistor bleed down time (varibal) to get the required pulse timing to match the MOT vibrator.

CHEV 29: Question. Have you ever used an O-scope to find the electro-mechincal vibrator? Please advise if you know the frequency. I would thing at WOT RPM would be fast enought for testing. Being mechnical it must be slow. I did find a suitable darlington transistor used by Ford for ignition we could use but it was out of production, I was a few years late sadly P.S. A 555 & 556 chip is one that is still in production from the 60’s. That’s rare for chips! Thank you everyone ahead  of time, feels good to be back, Still in a wheelchair but breathing still.

Jeff

 

[jeff-register]

Yes please check you capacitors ( condensers) in the mot because they go bad like the ones used in a magneto. Please remove on lead from the circuit to avoid false readings! Testing with a meter is also not “under load” test, only face value. The newer capacitors have thinner & better insulators making them much more better performers. I’m feelin good enough to start a new patent again, Had West Nile, Covid & C-Diff all in a row. lucky to still be breathing!

A darlington transistor is two built into one. It uses a very small signan to turn on a second  which handles that handles the load.

Best Jeff

[29Chev]

Thanks for the readings – blue band = 6 (first digit of ohm value), grey band = 8 (second digit of ohm value) , the black band = 1 (multiplier) and the silver band is the tolerance which is 10%.  So the resistor value should be 68 ohms with a tolerance of 10% (6.8 ohms).  A measurement reading of 74.8 ohms shows it is right on the edge of the higher resistance tolerance but should still work ok.  This is all assuming that it is the equivalent of the shunt resistor shown in the schematic which I cannot say for exact certainty.

The capacitor readings that you show on your meter are in nF (nanofarads) looking at your pictures.  Capacitors (and condensers) usually found in ignition circuits can be measured in nanofarads or microfarads.  Smaller capacitance values will often be stated as picofarads and if you do a google search you can find conversion calculators online to enter one reading value and have it converted to the other two values as the multiplier factor is all that is changing.  A reading of 95.5 nF is equal to .0955 uF (microfarad) and looks like you are measuring a new capacitor with a value of 0.1 uF so the meter may be off a smidge or the test leads may not have been zero’d before taking the reading.  Not criticizing – just observing as some capacitance meters need the leads to be connected together and zero’d before doing a test on the actual capacitor so any capacitacne in the meter leads is removed from the measurement.  The reading you got on the 0.15 uF capcitor in the Mercotronic of 196.5 nF  – this would equate to a value of 0.1965 uF.  If we compare the two readings that you got on the two capacitors and assume the first reading on what looks like a new capacitor of 0.1 uF is off just a smidge then the one in the Mercotronic may be a little off its correct value at almost 0.2 uF.  You could use a scotchbrite pad on the two leads of the capacitor in the Mercotronic to remove any tarnish on them as that could affect the reading accuracy as well.  When measuring electronic components connection cleanliness is a must to obtain accurate readings.  Given the age of that capacitor you could consider replacing it with a new one of the same rating and value but that is entirely up to you.  I have experienced capacitors (which are the same component as condensers electronically) test good or just slightly off value using a tester such as you are using and still be bad in an ignition circuit as stated by the other poster in the previous post.  The oscillations that occur in a primary ignition system when the ignition coil fires can result in voltage peaks of close to 200 volts – only happens for a brief moment in time during each firing but these voltage peaks can jump across the condenser connections internally if the condenser is starting to fail or is partially shorted.  Given the value that your meter is showing the original condenser may be partially shorted internally.

One thing you could try is to use a pair of side cutters and clip one lead about halfway between the capacitor body and the connection where it joins to another component – before doing this clean the lead with a scotchbrite pad or fine emery paper so it will be clean to solder back together afterward if you do not plan on replacing the capacitor.  This would remove it out of the circuit temporarily and you could then use two test leads to connect a new capacitor of the correct value and rating 0.15 uF 400 VDC (or condenser) to the circuit where the capacitor is now connected and see if the meter reading changed when testing a coil. If it does then this would indicate the capacitor is probably affecting the meter reading.  If the meter reading does not improve then I would check for resistance in any connection point in the circuits that go to the meter terminals (such as the switch terminals or solder or crimped joints).  If the new condenser temporarily connected does not change the meter reading and you chose not to replace the original condenser the lead ends that you cut with the side cutters can then be placed back together end to end and a thin strand of wire can be wrapped around several times on the lead on each side and over the join area – this will enable you to solder the lead back together easily and insure a good joint as the wrapped strand of wire will allow the solder to flow and form a good connection. Any electrical connection should be done with a rosin core solder or equivalent – do not use acid core solder – you probably already know that but just thought I would mention it.

Hope this makes sense and  helps you

[29Chev]

jeff-register wrote:

I have worked on the vibrator type mot. Using a burnishing tool, I’ve added material to the contact points with good results. I have thought about using  a “555” chip to form the polsed signal to replace the electro-mechanical viberator. Use that with a darlinton transiator for output.Unfortionally I have been very sick for the last year, almost died. Also thought of using a vibrator from an old car radio as they fed a pulsed signal to a transformer to “UP” the voltage to run the heaters in the vacuum tubes. It’s just a matter of matching the pin-outs on each vibrator. This thread gives me incentive to do research using electronic “555” chip based unit to help others, Using a R–C timing math to acheave the correct frequency. A capacitor, commily called a condenser is the timing  devise controlled by resistor bleed down time (varibal) to get the required pulse timing to match the MOT vibrator.

CHEV 29: Question. Have you ever used an O-scope to find the electro-mechincal vibrator? Please advise if you know the frequency. I would thing at WOT RPM would be fast enought for testing. Being mechnical it must be slow. I did find a suitable darlington transistor used by Ford for ignition we could use but it was out of production, I was a few years late sadly P.S. A 555 & 556 chip is one that is still in production from the 60’s. That’s rare for chips! Thank you everyone ahead  of time, feels good to be back, Still in a wheelchair but breathing still.

Jeff

 

The vibrator of a working Mercotronic unit has been scoped.  As I mentioned in my original post I have worked with a chap in Australia who has documented both the Model 98 and the Model 9800 and we have compiled a working schematic as well as other information on each model.  It is posted on the Smokestack website at the links I provided.

If you check out this link to a thread and go to post # 15 you can view the waveform that is produced on an oscilloscope of the vibrator contacts using a coil that is being supplied a low current so it is not actually firing.  Once the coil starts to fire the frequency reading displayed on a scope are not always accurate due to the coil oscillations but if you observe the grid and check the time factor the frequency does stay constant at a given voltage level.

https://www.smokstak.com/forum/threads/merc-o-tronic-model-9800-schematic-yellow-circuit-board.227244/

The oscillation frequency of the reed is around 100 to 110 hz depending on the voltage used testing the coil – 7.5 volts or 12 volts.  Since the vibrator has two sets of contacts this results in a contact opening and closing frequency of approximately 200 to 220 hz since each contact will open and close once during each reed cycle.  Given the unique way that the meter circuit is configured to indicate current flow during a coil test I would suggest that any replacement vibrator (mechanical or electronic) operates at a similar frequency and duty cycle to insure the meter reading accuracy.  The Mercotronic Model 98 and 9800 offer a unique challenge to an electronic switch since the coil supply circuit is switched on the high side of the circuit.  I originally started experimented using a relay to replace the contacts of the vibrator but was not getting good results at the high frequency that the vibrator operates at so I then pursued using a P channel Mosfet to act as the electronic switch which appears to work well.  Another chap who is also located in Australia (I am located in Canada by the way) also became involved since he had a Model 9800 with a vibrator that was in poor condition.  He started off using a 55 timer circuit board and a Darlington transistor set up they way you have mentioned but found that it and the heat sink it was attached to got quite hot once the amperage going to the coil under test got around 1.5 amps.  He then switched to a 200 volt rated Mosfet similar to the one I was experimenting with and has now got his working well in his Model 9800.  He mounted the Mosfet and the circuit board inside the original vibrator can and is using the can as a heat sink.

You can view the complete thread at this link if you are interested

https://www.smokstak.com/forum/threads/possible-replacement-for-a-merc-o-tronic-model-98-model-9800-vibrator-suggestion.227436/

I would suggest using a Mosfet over a Darlington transistor as the switch because the Mosfet appears to do the switching with less heat produced and since a Mosfet acts more like a resistor than a diode (the way a Darlington transistor would) there is less voltage drop across the switch which was also a consideration when creating an electronic vibrator circuit.

I have posted a few links to some youtube videos of the circuit I am experimenting with using it to fire a 12 volt ignition coil in the thread I linked to on the Smokestak website.  I believe the circuit I have been using would work well in a Mercotronic but at this point in time I do not have a Mercotronic unit to test it with.  It uses a heat sink and circuit board that can be constructed with basic soldering, mechanical and electronic skills if someone wanted to make their own.  The circuit board and the Mosfet that the chap in Australia has working in his Model 9800 might be better suited for someone who does not enjoy the challenge of mounting a 555 timer and the related components on a circuit board.  With his unit the circuit board, Mosfet and a few of other components are all that are required to be connected together and mounted into the original vibrator can.

Sorry to hear about you health issues and hope you get well soon.

[crosbyman]

deleted

 

 

Joining AOMCI has priviledges 🙂

[Author]

Posts