Welcome to our resource for MG Car Information.



MG parts spares and accessories are available for MG T Series (TA, MG TB, MG TC, MG TD, MG TF), Magnette, MGA, Twin cam, MGB, MGBGT, MGC, MGC GT, MG Midget, Sprite and other MG models from British car spares company LBCarCo.

MG MGB Technical - Dwell - what's it trying to tell me?

The dwell on my GT, fitted with a 25D4 distributor, drops by about 10* as the revs increase. It leaves me wondering whether to set it at idle or higher rpm and if this likely to be a sign of wear in the distributor or just the analogue meter not keeping up?

Steve Postins

Probably the meter. Dwell is a mechanical function and should not actually be changing with rpm at all (OK, at stratospheric rpm when the points start to float does not count). Find a friend with a pertronix installed and check it against his.
David "I don't think my dwell meter works any more" Lieb
David Lieb

It's a bit rash blaming it on the meter, which simply displays the duty-cycle of the points i.e. the length of time they are open compared to the length of time they are closed. I'd be more inclined to think the points were bouncing or something as the revs increased. With points and certain electronic triggers the dwell is constant as it is totally governed by the shape of the cam. With other electronic triggers the *open* time is fixed and hence the closed time is relatively longer (giving a higher dwell) at low rpms compared to high rpms. I think Pertronix do two versons one fixed dwell and one variable dwell so you would have to know which one was fitted. Easier to use the meter on two points-driven engines, or two meters on your engine. At the very simplest you can use just a voltmeter, although this will vary with system voltage as well as dwell. To ignore that do the measurements with the alternator always charging i.e. between about 1500rpm or so upwards, or unplug the alternator altogether (for the duration of the test!).
Paul Hunt

Or it could be a worn shaft bushing or advance mechanism in the distributor. I agree with Paul, the meter would be the last thing I would suspect. Cheers - Dave
David DuBois

I have to agree with David. A badly worn distributor action shaft bushing would cause the dwell angle to decrease as engine speed increases.
Steve S.

Are you checking the dwell with the vac. hose connected and how far up the rev range are we talking here.
If you have the hose on ,it could be that the advance plate in the dizzy is worn and altering the points gap as vac. is introduced when the engine revs increase above idle. Try checking with the hose off, this will eliminate this as a cause.Dwell change caused by this usually shows up just above idle. If you still have a problem it would point towards wear between the two shafts of the distributor (unusual) or more likely the shaft loose in the dist. housing, Also check for excess up & down movement between the two shafts and wear on the dist. cam. It could be possible that the lobes are worn and as revs increase the shafs moves upwards and the points rub on a different part of the cam causing a variation in point gap (dwell) Hope this helps Gary

Many thanks for the suggestions. I tested with the vac. disconnected (points system). I'll have a closer look at it this weekend and will try a comparison with my Roadster as it comes out of hibernation.
Steve Postins

A useful test from Gary, but just to clarify whilst vacuum being connected will change timing as the revs change, it shouldn't change dwell. Disconnecting the vacuum pipe from the carb or manifold and sucking on it *should* increase rpm a little, but shouldn't change dwell.
Paul Hunt

I have done some searching, and I have not found out exactly how a dwell meter works. Yes, I know what dwell angle is and what it means to the ignition system, I just do not understand quite what the meter is using to determine the reading. Everything I seem to recall indicates that it should be done at idle, implying that, although the dwell angle is not supposed to change as the rpm goes up, the ability of the meter to accurately reflect the correct dwell angle could be impacted. So, who can splain at me just how a dwell meter really works?
David "curiouser and curiouser" Lieb
David Lieb

this should be good....My guess it measures resistance but like you david, I never really grasped the concept...just set the points. gap it, start and adjusst is all I have ever done.


AHHHH.... the dwell is the angle of the plate that holds the points and condenser as it relates to the rotor not only do you gap the points and adjust the plate forward or backwards...then adjust the dist. for better timing....that makes how do you adjust the plate on a 1275 A-series....yeah wrong board but was curiuos as to what type of trouble david lieb was causing over here....


HAaaa, so it is measureing resistance... the less retard the more resistance....the more the advance the less resistance...resistance is the degree at which the plate is set....Im so cool!!!

in a nut shell a degree of dwell is a unit measurement of resistance


No, Prop, it isn't. Dwell angle is the number of degrees that the points "dwell" in the closed position each cycle. It has nothing to do with the position of the plate and everything to do with the points gap. It is the best way of measuring the effective gap, since points, especially after they have some miles on them (and need re-gapped because of wear of the rubbing block), are no longer flat and a feeler gauge is measuring the distance between the high points of their surface. Dwell indicates the number of degrees during which the points are conducting electricity, I am just not yet sure what they are really looking at. The needle is genmerally quite steady, so that eliminates a lot of possibilities.
David "I miss the window in my 72 Chevelle's distributor cap" Lieb
David Lieb

As David says, dwell is simply the length of time the points are closed in each close/open cycle. This can be expressed as a percentage, a ratio, or the number of degrees of crankshaft or distributor rotation, usually the latter. It is changed by setting the gap, and it is the lift and duration of the cam at a given gap that determines the dwell. For a given cam as you change the gap the dwell changes and vice-versa. Dwell is a much better method of checking points gap for the same reason that a timing light is better for checking timing i.e. dynamically than a test-lamp i.e. statically. By checking dwell not only do you check the gap, but you can also be made aware of defects in the system that gap setting wouldn't tell you, which is what seems to be happening with the original poster. Once you have set the gap on new points there should be no need to use a feeler gauge again, just connect the dwell meter and if that shows in spec then leave them alone. I've just changed a set of points that hadn't been touched in 15,000 miles, they were still within the +-5 degrees for a 45D. I had deliberately left them in longer than the normal 10k change intervals just to see how long they *did* last. They still hadn't failed or gone out of spec, but recently it did seem that the roadster was taking a bit longer to start than usual, with new points it has gone back to firing up in about 3 seconds, even after standing for several weeks.

You can make a very simple dwell-type meter by simply connecting an analogue voltmeter between the coil -ve or CB terminal and ground, and switching it to its 12v scale. As the points open and close on a running engine the voltage will switch between 0v and 12v. Because the meter needle can't respond rapidly it settles at an average of the two, which is governed by how long the points are closed compared to how long they are open. The longer they are open the higher the voltage will be. The dwell for a 45D4 is 51 degrees +- 5 degrees, call it 50. This is relative to 90 degrees of distributor rotation (being a 4-cylinder), and 50 degrees of that means the points are closed for 50/90ths and open for 40/90ths, or 56% and 44% respectively. At 12v system voltage 44% of the time being open would cause the meter to show an average of 5.3v dwell voltage. But it all goes wrong at this point because the system voltage varies anything from 12.5v to 14.7v, and you would need to know the instantaneous system voltage as well as the dwell voltage. The dwell meter is designed to ignore voltage changes.

However you can use the voltmeter method to indicate whether a *varying* dwell is down to the dwell meter or the ignition, by either disconnecting the alternator so there is a relatively constant system voltage (measure it), or leaving it connected and only measuring off-idle where the alternator is giving a relatively constant output across a rev range.

You can also use an analogue ohmeter to indicate dwell, as it switches between full-scale deflection when the contact is closed and zero deflection when it is open. If you have a 100v or 100 point scale on the meter you can read the percentage or time the contacts are closed directly off that. However in the case of a distributor there would need to be some other device driving it i.e. a variable speed drill, as the voltage and current when the points are running the engine renders the readings from an ohmeter unusable.
Paul Hunt

Just to set my mind at ease.

If the points gap is too wide then the charge in the coil would be low as it only has a short time to charge up, but what is the effect of a points gap that is too narrow? Is it a case of the points arcing more at lower gaps than normal? I've been trying to work this out for a couple of days and can't remember why small point gaps are not good.

Tony Oliver

Too small of a gap leads to arcing and burned points, greatly compromising the life-expectancy of the points.
David "probably impacts the life of the condenser, too" Lieb
David Lieb

OK, a couple of observations:

1. The meter responds in the same way with my Roadsster as the GT, recording a fall in dwell as the revs rise from idle to 4000+rpm.

2. The meter also reads volts, so taking Paul's suggestion, I switched scales. At idle the meter sits at exactly the same position as the dwell setting, but as the revs rise the needle sweeps sharply up the voltage scale (= down the dwell scale).

What does it mean? No idea, but I'll leave it alone and set dwell at idle.
Steve Postins

Steve. What this means is called "points bounce". An old term, but I am an old man.

Think about why we advance the ignition timing as the rpms increase. The burning rate of the fuel is a fixed time and we want to fire the fuel so that it is fully burning at a time that will give us the best engine performance. So, we fire the mixture just before the piston is at top dead center so that the fuel will be burning properly just after the piston has advanced beyond top dead center. Because the burning time is a fixed rate, we increase the point where we light it off (i.e. fire earlier before TDC) because the piston is traveling faster as the rpms pick up.

The spring rate on the points is also a fixed rate. The contact surface of the points comes off the high lobe of the cam and takes a fixed amount of time to come back onto the cam as the points are closing. (The points are in constant contact with the cam when the distributor shaft is rotated by hand. The "spring effect" takes place when the cam is being rotated by the camshaft.) Here again, we have a fixed rate working in conjunction with a non-fixed rate. As the cam is turning faster, the points contact surface returns to the cam a hair later in the cycle. This means that the points remain open slightly longer which results in a lower period of time in which the points are closed--and the length of time the points are closed is what defines dwell. So, all you are seeing is the well understood doctrine of points bounce.

Yes, it is possible to produce a set of points having a stronger spring arm which would allow the contact surface to be in constant contact with the points cam. In such a case, the dwell would be constant at all rpm ranges. It would, also, result in rapid wear of the contact arm of the points set.

Thus, we sacrifice a small amount of dwell to achieve a longer life of the points set.

Les Bengtson


Thanks for setting me straight, Its been scence almost like ever that I last had to use points and condenser for any length of time....I have a mallory duel point dist. for my 1275, for some reason I cant get pertonix chips for it, so Im going to be doing the ol match book and screw driver thing when ever I get around to intalling it, and have been looking for a better way...appears I have found it...thanks have been vary informitive


I would have expected that as revs rise the voltage will go up as the alternator puts out closer to its voltage limited max. If the dwell and voltmeter readings are = it would appear that your meter does not have an internal voltage clamping circuit to take this effect out. This would be OK for setting at idle but it means that most of its usefullness as a diagnostic tool is gone. you will also see the primary side peak voltage at this point. I had a look around and found this.
The waveform is interesting showing a spike to ~200v (this transformed up by the coil is the HT)then staying at around 40v until the points close and clamp it back to ground. As the article notes although dwell time as a percentage remains constant in a simple points system the stored energy in the coil will be less as the absolute value time to build current falls with increased revs so induced voltage will drop. Maybe the makers rely on these opposing effects to allow them to avoid adding components, which cost a few pennies? Anyone on here the owner of a scope? Oh and an attenuator I think :-)
Stan Best

Stan - the dwell and volts are aligned at idle but as the revs increase the volt meter rockets up, as you'd expect, whereas in comparison on the dwell setting the needle shows a slight deflection, albeit in the same direction.
Les - thanks for the explanation, that makes a lot of sense.
Steve Postins

If the points gap is way too small then it will still be arcing when they start closing again. It isn't until they *stop* arcing that the spark actually occurs, so too small a gap, even though there *is* a gap when measured statically, could mean that you won't get any sparks.

The meter responding exactly the same in two cars is interesting, and it isn't something I have noticed on either of my cars. The advance in timing as the revs increase isn't really relevant here, but the speed the heel of the points moves up the cam is. If the heel of the points lifts off the cam at the peak and doesn't come back down again until *after* they would have closed if the heel had followed the cam, that isn't points bounce but overshoot. Points bounce can occur when overshoot occurs, and is the heel of the points bouncing off the cam sufficient to open the points again. This generates double sparks, and while overshoot can reduce coil charge time, bounce can reduce it even further. Whether you are getting overshoot, or overshoot plus bounce, can really only be determined with an oscilloscope. I've done a lot of bench testing with an oscilloscope in the past and never seen either, if the cam and points are working as they should it shouldn't happen. V8 points open and close twice as often as 4-cylinder points, and V12 points 3 times as often for much the same peak rpms. If overshoot and bounce were a common problem on 4-cylinder cars the others would be undriveable.
Paul Hunt

Thats very interesting, constant car voltage you would expect increasing dwell to give lower voltage, the meter input spending more time clamped to ground, this plus the fact you get different readings means there is either a clamp circuit in there, or its a digital meter and is doing some sums. Call me old fashioned but if it's the former then I believe what it's telling you that your dwell goes up by betwen 8 and 12% ,if its doing some sums then I would ignore it as very often you find equipment limitations (a 50cent meter chip in this case) mean so many corners are cut that the answers are not useful. I speak from hard won experience here in my professional life.
If in fact you are getting a bit more dwell as revs go up then in itself thats OK as it gives more time for a current "charge" to build up in the coil. It will of course affect timing and maybe give you some unwanted advance. You would have to dial this out by retarding slightly at idle, or recalibrating the springs. However if your dizzi top bearing has no play when you rock it from side to side, and you cannot pull it out along the skew gear , which is a scary fault when you see it, then I think it's the meter and it can't add up. The reason I say this is the dwell in a mechanically good dizzi wil be determined by the cam form and how well the contacts follow it with revs. The inertia of the points might delay the opening, but for symetrical ramps it will be cancelled by closing. Other than that I can only think of contact bounce which does happen, see any relay data sheet especially an expensive mercury weted one which will be full of why it's worth it because it has no bounce, toggle swithes do it as well, I had to calibrate one with a scope once and the waveforms were amazing. With the 0.1 to 0.2 uF across the points you will not see the voltage fluctuating just an extended closed period. You might see 10 to 100 ms of bounce(just a guess based on past experience, and because it happens there does not mean it happens with the unique spring blade configuration of the points) Now I reckon that at 1000 rpm you have about 40mS of dwell (order of magnitude sum) and this lines up well with what you measure. At this point without a scope I am all done, but anyone is welcome to comment and check my sums.
Stan Best

The meter is a distinctly old, old-school Speedograph. It seems accurate, but you get what you pay for, and a fiver wasn't that much!

Steve Postins

Stan Best,

The scope method is the right one to accurately measure the dwell. Provided you use it with an external trigger to get a convenient synchronization independantly of speed. In this way you can read the closed/opened periods and compute the ratio. Converting from points to electronic switches reduces to a minimum the spring bounce phenomena. Then only thefree end play combined with rotational inertia may make vary the dwell in a sensitive manner.

I've just checked my roadster dwell with my meter and it goes up by maybe half a degree between idle and 3000 rpm. Going by that 10 degrees definitely indicates something is wrong.
Paul Hunt

Thanks Paul, interesting and useful post. This is what you would expect and it's nice to have that confirmed.
Regrettably it does not tell us if the problem Steve is seeing is in the dizzi or the meter. Other than putting a scope on the points you could drive the meter from a function generator, even a fixed 50% duty cycle square wave that could be varied in frequency up to about 100 Hz would show you if it's readings are frequency dependant.
Stan Best

Easier to borrow a friends dwell meter I would have thought, and the more friends with one you have the greater the odds of working out which is at fault :o)
Paul Hunt

This thread was discussed between 04/07/2008 and 20/07/2008

MG MGB Technical index

This thread is from the archive. The Live MG MGB Technical BBS is active now.