Puzzling behaviour on MAG vs BAT

[Chris Bamford]

Hello all — I've been chasing a rough-running-on-mag problem with my '24 Speedster and am puzzled about the cause. It sure would be great if the hive mind here can weigh in with more experience and understanding than I can muster!

There has been a slight miss at high RPMs on MAG for some time. A few days ago, I may have accidentally flashed the mag circuit with a smart-charger battery clip; while I assume this could have de-magnetized the magnets (and the engine does run much rougher on MAG than it did before), the fact is that the Magneto still puts out 30+ volts at high rpm and can muster a 2-amp charge rate at those same high RPMs.

So, are the magnets de-magnetized or not? I'm game to do an in-car recharge if required, but would like more clarity, first, on if it is actually necessary or perhaps there is some other issue?

I run recently-tuned coils (1.95-1.99 TTF, good waveform and consistency), a clean TW timer, the coil ring was rewound by Wally S in 2010 and the magnets were recharged, individually, in 2020. Checked the mag post pickup yesterday and it's clean as a whistle. I have experimentally by-passed the magneto wiring and ignition switch with a direct jumper wire: no change.

The linked Youtube video HERE has a detailed timeline of what to watch for in the description, timeline also shown below.

Basically, the car runs strong and smooth on BAT and rough and ragged on MAG.

Having said that, while running on BAT, my just-built magneto charger pumps out 0.9 to almost 2.0 Amp depending on RPM, and the magneto output voltage is 30+VAC. This magneto output voltage drops by about 40% when switching from BAT to MAG and the reading is jumpy and uneven.

Is it reasonable to see weak magnets still put out 30+ volts and 2.0 charging amps at high RPM, but falter when they have to drive the coils? Could there be another cause for this behaviour? As stated above, I'm fully prepared to do an in-car recharge if necessary, but would sure like a much better understanding first.

Video timeline:

0:00 Mag-charger lites up with revs (running on BAT)
0:13 Magneto meter reading on BAT then MAG (note: meter reads 5 VAC high)
0:55 Mag-charger charge rate various rpm (running on BAT)
1:30 Magneto VAC reading running on BAT
1:50 Magneto VAC reading running on MAG
2:04 Magneto VAC reading running on BAT again
2:10 Bypass mag circuit and switch to confirm wiring/switch OK
2:23 Rough running and lower/jumpy Magneto meter reading on MAG vs BAT

HELP!

[TXGOAT2]

Are you certain it's the magneto? Maybe you have a faulty coil that is breaking down at the higher input voltage provided by the magneto. Intermittent sticking coil points or a shorting coil primary might give that result.

Does the magneto output get jumpy when running on battery? Something could have damaged your magneto coil. If you can get a jumpy output reading straight off the mag post with nothing but the meter connected, and connected securely, it would almost certainly be magneto related.

NOTE: It seems to be common for digital meters to misbehave around Model T engines due to high RF radiation from the ignition system.
An analog meter with short leads may be a better choice for checking the magneto output.

30 volts maximum or anything near it at the mag post indicates good magnets, but it doesn't rule out an intermittent poor connection at the mag post or an intermittent internal short or open circuit condition.

An intermittent external problem in the coils/coil box/coil wiring could load magneto and give the jumpy reading.

[Chris Bamford]

Thanks Pat for your comments and questions. My responses follow.

I am not certain it's the magneto, but on MAG setting is where the trouble lies.

The coils, as mentioned, are recently set on a FACT test rig, with new caps and points as required. Maybe one or two are at fault? Dunno, it's hard for me to tell from here... any suggestions?

Magneto voltage is not jumpy when running on BAT (as per the video).

I understand that a digital multi meter may be erratic around the T mag... mine is pretty calm when running on BAT, much jumpier on MAG. Also check out the video from 2:23 onward: analog Magneto Meter is calm on BAT, jumpy on MAG.

Had the mag post out yesterday to examine the pickup... it is fairly new and looks OK, appears to have full continuity thru the range of pin motion. Connection is solid. Also tried bypassing the mag wiring and switch assemble completely, no change (Video 2:10)

Re the coil box, I forgot to mention that was my suspicion several days ago. The wood was original with no obvious leakage tracks but I replaced the wood and refreshed all the connections anyway.

Still stumped!

[Moxie26]

Chris : .... Your video link posting does not work on my 'puter.. Chances are if your magneto post is one with a oil line connection, they have a history to fail .
I suggest using a new mag post . Chances will be magneto problem will be gone.

[Scott_Conger]

That is an awful load/draw that seems to be showing up on MAG when you switch to MAG ignition and on the surface of it, just doesn't seem right.

I would normally suspect the switch, but your troubleshooting appears to be sufficient to exonerate it (?)

EDIT: Well, I type slow so the coilbox has been sorted, but I'm too lazy to delete unuseful advice . This still leaves the coilbox or wiring as somehow suspect under the assumption that the coils ARE ADJUSTED CORRECTLY. I have never used a FACT so do not know one way or another if there can be a "good" reading and that reading be "good" for 6VDC AND good for 30VAC. With traditional monitoring equipment, you can create coils which are good on BAT and not MAG, good on Mag and not BAT, or good on both, and of course each requires different adjustment for each condition (and will still show the exact same "test" parameters specific to the equipment used), meaning, they will test "good" and still not necessarily operate correctly under ALL conditions. This is often the downfall of folks new to the game...perhaps the FACT closes that little gotcha and points it out...I just don't know. The FACT designer will be the authority on that and if you ask him, then we will both know!

Edit: correct type of tester referred to (do not want to muddy waters with incorrect info)! Thank you for the correction, Chris.

[Chris Bamford]

Robert — apologies that video link does not work for you. I can tell you that the mag post is recent and does not have an oil line.

Scott —agree that is a big pull when switching to MAG and seems very wrong. I ran this same test on my ‘26 touring, and the magneto voltage went up upon switching to MAG (related, I’m sure, to the immediately-higher RPMs.)

To clarify, I am not using an ECCT, rather the FACT mini-oscilloscope tester from Matt Graff in CA, a forum contributor.

At a family do the rest of the day. Will switch out my Speedster and Touring coils tomorrow AM and see what happens. The Touring behaves exactly as expected running on BAT vs MAG.

[Norman Kling]

A few things you might do. Adjust each spark plug to 25 thousandths. Try looking under the hood at night with lights off and see if there are any sparks visable to the eye. If so you have something arcing. Might only arc at high voltage. If so, repair the problem. Another thing could be in the timer. On battery, you produce the spark when the rotor first contacts the segment, but on magneto it sparks when the magneto voltage builds up which would be somewhere in the middle. Is there any change with slight movement of the spark lever" If you have a few spare coils which you know to be good, try substituting one coil and try to produce the misfire with the good coil in each position in the coil box. If you find it is in one position and runs good with the good coil in that position, you have a bad coil. If none of these suggestions work, maybe someone else can help you find the answer.
Norm

[Moxie26]

Chris, please repost the link for your YouTube video. Thanks

[TXGOAT2]

Note that good coils can go bad, and a good coil or mag winding, or just about any other electrical device with an intermittent problem can be difficult to test.

Substitution, where practical, eliminates on again, off again gremlins in specific components.

[TXGOAT2]

I don't think a momentary contact with a typical smart charger lead would affect the magnets or any other part of the system. Most smart chargers have a lot of protection built in, and a low amp charger wouldn't be much of a threat anyway. Regardless of all that, your readings indicate strong magnets.

[speedytinc]

Mag is fine.
How bout bypassing the switch? Wire the mag post directly to the coil box. (eliminate the switch as the problem) Of course remove any possibility of DC getting to the mag.(remove DC input to the coil box)
Glad you explained the FACT. If properly used to tune coils, its supposed to be equivalent to an ECCT.
Didnt catch - you charging a battery with the mag? 6 or 12V?
Are your coils rebuilt with new caps?

[Norman Kling]

One thing I thought of after I posted. Test each coil to see how many amps it draws when it sparks. They should all be equal. If one sparks at a lower amperage than the others, it would be first to spark when the magneto voltage rises. If one takes a higher amperage to spark, it would be later as the voltage rises. so you might have one sparking early or late. That would still run but that cylinder would either be early or late and it would run a bit rougher.
Norm

[Scott_Conger]

Norman

he has tuned his coils to a specific time-to-fire.
measuring current will provide no information other than they may well not all fall into "ideal" current readings

your point is certainly valid and good advice if using traditional methods, but that is not what Chris is using

[speedytinc]

The coils are tuned for TTF. There will be a slight variation of amp draw. Wouldnt hurt to check the amp draw for each coil looking for a very high or low draw. Again assumption - coils have good new caps. I believe the FACT tests this also.

Rule out the ignition switch.

Run in the dark looking for external errant sparking during the rough running.
There may be internal errant sparking in the coil box. I would expect you checked the coil/box contact points looking evidence of poor/intermittent contact.

This is an interesting case. Do let us know what you find.

[Chris Bamford]

So much interest — thanks, guys. I have several avenues to pursue in the morning.

Robert, sorry, I can’t repost the video link this evening, and it would likely still misfire for you.

Could someone please PM the link to Robert in hopes he can open it that way?

[speedytinc]

Another thought.
Put a modern timing light on each plug wire. Watch a spot on the crank pulley. You may need an external 12V battery to fire the light if your system is not 12V.
If a coil is firing wack you will see the spot move. You can narrow down the problem to a specific cylinder or if all are wacking out that gives you another diagnostic tool to narrow your search.

[TXGOAT2]

Quote: "Having said that, while running on BAT, my just-built magneto charger pumps out 0.9 to almost 2.0 Amp depending on RPM, and the magneto output voltage is 30+VAC. ***This magneto output voltage drops by about 40% when switching from BAT to MAG and the reading is jumpy and uneven."*****

It sounds like something in the ignition system is causing a heavy current draw.
It would almost certainly be something in the ignition wiring, timer, or coil primary circuit, including the coil primary windings and perhaps capacitors.

Running on battery would limit ignition input voltage while providing a huge reserve pool of amperage, which might explain why it runs OK on battery.


I'm thinking a coil or coils may have a problem.


Have you tried disconnecting the mag battery charger? Perhaps it is somehow loading the magneto output, perhaps due to a failing diode.

[Moxie26]

Clicking on your link, this comes up, not your YouTube video,............. " The linked Youtube video HERE has a detailed timeline of what to watch for in the description, timeline also shown below. "

[Poppie]

Hi All,
I believe the problem is an excessive current draw on the model T magneto.
I would make sure that the battery is reasonably well charged and I feel that the coils being set up with on electronic coil tester (FACT) is drawing too much current ( Try a higher ttf ) It would be interesting to know what the current draw would be if the said coils were checked on a HCCT, ....n

[Chris Bamford]

Just back from our family shindig and I took advantage of the darkness to look for stray electricity running high RPM on MAG: No sign of random sparks escaping and no amount of wire jiggling had any noticeable effect. Darn it.

I see that my YouTube video earlier today had almost no views, so Robert was not likely the only one who couldn't connect. Sorry about that. Please try this https://youtu.be/VQqdXyAVbzQ. Fingers crossed.

Whilst out in the garage just now, I shorted out the plugs individually to look for localized trouble. Click https://youtu.be/nwmfob_ofvw and re-cross your fingers. First half of the video is high RPM on BAT, and the four cylinders act similarly, #4 a bit more noticeable. Switched to high RPM on MAG at 0:38 and was surprised... Plugs #1—#2—#3 were comparable, but shorting our Plug #4 brought the RPMs down dramatically!

This must surely be a clue to my problem, but how it relates needs some serious sleeping on it. More tomorrow. Thanks again for the quality discussions.

[TRDxB2]

......
Having said that, while running on BAT, my just-built magneto charger pumps out 0.9 to almost 2.0 Amp depending on RPM, and the magneto output voltage is 30+VAC. This magneto output voltage drops by about 40% when switching from BAT to MAG and the reading is jumpy and uneven.
.....

Out of curiosity, have you tried running on MAG with your charger disconnected?

[TXGOAT2]

In addition to the above, I would re-check the timer very carefully.

[Chris Bamford]

Good morning all. I see the new video links were a definite improvement.

So, unhook the mag-battery charger they said... examine your timer they said... swap in some known-good coils they said...

The mag-battery charger was easy: it was already off and had only been a tool to indicate magneto output. I have a set of known-good coils in my '26 Touring (also set up on the FACT) but the Speedster coils are tucked up under the cowl and awkward to access. The easier task was to swap out the TW Timer with my backup roller timer.

Turns out they were right!

I found significant oil inside the housing (perhaps the cam seal is failing?) and obvious schmears of oily carbon residue around the contact ring. I cleaned this timer only a few hundred miles ago and it was reasonably dry, so something has changed.

First photo is the TW partially cleaned but showing definite greasy carbon tracks between the contact segments. Second photo shows some pitting/erosion on the timer ring contact segments and extending into the insulation.

The engine performed as expected with the roller timer: RPMs picked up a little on MAG and the magneto voltage tracked RPMs nicely. Magneto voltage is well over 30 VAC at speed for both BAT and MAG. Rough running on MAG is gone. I assume the oily carbon schmears in the TW shorted out the high-voltage/rpm magneto current and jumbled up the timing.

Not out of the woods completely though — I'll have to sort out that oil in the timer housing (or switch to a roller) and I've been in touch with our club master machinist about cleaning up the TW Timer contact ring.

Thanks to everyone who took an interest and offered assistance... particularly Pat M and also Scott C. Much appreciated, my friends.

[TXGOAT2]

I'm glad you isolated the problem.
It seems far-fetched, but I suppose the higher mag voltage was able to ground randomly through the goop in the timer.

If it was poor contact in the timer, I'd think the magneto would have done better than the battery.

[Moxie26]

Looks like the unevenness worn into the plastic between the segments caused the roller assembly to bounce during high speed. You may want to consider a new commutator. Your present commutator possibly could be turned in the lathe to take down the high spots to match the worn areas.

[Chris Bamford]

Hi Robert , I have had that problem with roller timers, and is a reason I avoid them.

The subject timer, though, is a TW with carbon brush with no roller, despite the wear pattern. IIRC, Tony Wiltshire attributed the pitting etc to minor arcing from the brush, at both the point of contact and the point of disconnect.

[TXGOAT2]

When you get everything back together and the oil intrusion stopped, running on magneto as much as possible will probably extend the life of the timer.

I'd check it every 100 miles or so for a while, just to be sure it's staying clean and dry. If it stays clean, extend the run periods.


Properly adjusted coils will also help keep the timer in good condition.

A problem with coil adjustment, even on just one coil, can ruin a new brush type timer in very short order.

[Moxie26]

Any progress with your problem?

[TXGOAT2]

The problem seems to have been oil intrusion causing timer issues.

[Scott_Conger]

Well, never having dealt with a carbon-brush timer, I would NEVER have guessed it to be the culprit, but in hindsight, I certainly can add up the fact that oily carbon smut spread around the case could exacerbate trouble as voltage increased and unwanted conductivity played it's havoc.

I am really glad to have followed this thread to conclusion(?) to learn something new (that really should have been obvious, but wasn't at the time). But for the oil contributing to the agglomeration, I doubt that just the carbon dust alone would have created trouble for many many more miles down the road.

thanks Chris

[Moxie26]

REALLY , PAT & SCOTT ???!? .... I believe those who are following this are really waiting for Chris's response, not yours.

[Scott_Conger]

Bob

suffice it to say, that what I'm thinking right now would have me banned for life if it went to print

[Moxie26]

Scott... Doesn't surprise me at all. Go right ahead. We're all waiting for it.

[speedytinc]

Good morning all. I see the new video links were a definite improvement.

So, unhook the mag-battery charger they said... examine your timer they said... swap in some known-good coils they said...

The mag-battery charger was easy: it was already off and had only been a tool to indicate magneto output. I have a set of known-good coils in my '26 Touring (also set up on the FACT) but the Speedster coils are tucked up under the cowl and awkward to access. The easier task was to swap out the TW Timer with my backup roller timer.

Turns out they were right!

I found significant oil inside the housing (perhaps the cam seal is failing?) and obvious schmears of oily carbon residue around the contact ring. I cleaned this timer only a few hundred miles ago and it was reasonably dry, so something has changed.

First photo is the TW partially cleaned but showing definite greasy carbon tracks between the contact segments. Second photo shows some pitting/erosion on the timer ring contact segments and extending into the insulation.

The engine performed as expected with the roller timer: RPMs picked up a little on MAG and the magneto voltage tracked RPMs nicely. Magneto voltage is well over 30 VAC at speed for both BAT and MAG. Rough running on MAG is gone. I assume the oily carbon schmears in the TW shorted out the high-voltage/rpm magneto current and jumbled up the timing.

Not out of the woods completely though — I'll have to sort out that oil in the timer housing (or switch to a roller) and I've been in touch with our club master machinist about cleaning up the TW Timer contact ring.

Thanks to everyone who took an interest and offered assistance... particularly Pat M and also Scott C. Much appreciated, my friends.


24T TW Timer 1.jpg

24T TW Timer clean.jpeg

Chris's response!!!.

The carbon/oil "paste" was the problem manifesting @ the higher mag voltages.

Lesson to learn : when running a brush, have a good oil seal to keep the oil out. (like a new day)
This failure can also happen with a roller brush. The conductive "paste" is from worn metal, but takes longer to build up than carbon.

[Moxie26]

Good morning Chris... Given the suggestion of cleaning commutator contacts and roller and your suggestion of switching your known good set of coils, which gave you better results?

[Chris Bamford]

Robert, I had some confidence in my coils already, so I chose to swap in a roller timer in place of my TW unit. This identified the problem and I didn’t pursue the coils.

[George Mills]

Angles and photos show strange reflections!

Had you not said it was a TW sideways brush timer, i would have suggested the timer housing was one of those ‘oops’ of some 15 years ago on a repo Tiger where when there was roller bounce, even in the slightest, the plastic wallowed away when it reseated and caused very erratic operation.

Convince yourself it is smear to continued ground and not a wobbbly cam end. Good old Royce use to advocate filling the timer housing with non conducting grease. I did that in one and other than taking a little more to get going…has been flawless for years. I got around the start by dry cranking a few turns before ignition attempt.

FWIW

[J1MGOLDEN]

About 30 years ago, I read a story in one of the club magazines about a long tour.

Every morning, the writer would drain about a half cup of gas out of the carburetor drain to eliminate any water that might have settled out of the gas during the night, as the gas often had some water, perhaps from moisture gathering in a low gas tank.

Then, remove the Timer and use that fluid to clean out oil, grease or filings.

Then, wipe the Timer dry and clean.

Then, replace the Timer and add fresh oil in the small oil hole.

The writer noted that this sequence of events always insured a good daily drive.

[Jerry VanOoteghem]

REALLY , PAT & SCOTT ???!? .... I believe those who are following this are really waiting for Chris's response, not yours.

I always look for and value Scott's responses. Yours, not so much...

[Chris Bamford]

Now, that’s a pretty AI sounding post…

[TXGOAT2]

Many hobbyists do not realize how beneficial a quick trip to a casino can be when it comes to dealing mechanical or electrical issues with older cars.
Even professional mechanics find that an impromptu casino visit can solve tough mechanical and electrical problems. For instance, Earl T., a certified technician at a major auto dealership, reports that he was having trouble diagnosing a customer complaint of poor fuel economy in a mid-priced mini crossover. Frustrated, and feeling that he might be suffering from a touch of Inadequate Gambling, Earl decided to take a break at the Golden Taj Casino Royal, and after returning to his job, he took the vehicle out for a test drive. When he returned to the shop, he was surprised to find that although he had left with half a tank of fuel, after the 5 mile test drive, the gas tank was full! His shop foreman gave him a hefty pay raise on the spot, and his customer was very pleased, to say the least!

[John Codman]

About 30 years ago, I read a story in one of the club magazines about a long tour.

Every morning, the writer would drain about a half cup of gas out of the carburetor drain to eliminate any water that might have settled out of the gas during the night, as the gas often had some water, perhaps from moisture gathering in a low gas tank.

Then, remove the Timer and use that fluid to clean out oil, grease or filings.

Then, wipe the Timer dry and clean.

Then, replace the Timer and add fresh oil in the small oil hole.

The writer noted that this sequence of events always insured a good daily drive.

If I had to do that to my T every morning, it would be out front with a "For Sale" sign on it, and I would be looking for an A.

[TXGOAT2]

If you follow the owner's manual and/or shop manual, you'll be OK. Believe it or not, Ford Motor Company knew a little about these cars and how to handle them.