Starter Motor Current

[dmr]

Its a wet start to the new year so here's something for the forum techies to talk about.

 

This is starter motor current and "instantaneous" engine speed during cranking. This is my Beta JD3 on a cold morning. The engine water temperature was about 10 degrees and the injection pump solenoid was not activated as otherwise the engine starts almost instantly.

 

 

 

[Flyboy]

It looks like what you would expect. The peaks are due to the increased load due to each compression stroke. The graph shows that your starter motor has a stall current of 1100 amps dropping to 300/400 when steadily cranking. You stopped cranking after about 11.3 seconds & came to a complete halt about 1.5 seconds later. The cranking speed was about 200 rpm.

Edited January 1, 2017 by Flyboy

[MoominPapa]

Interesting that during the "coast down" phase, after the starter is off, the energy stored in the air compressed the cylinders is enough to accelerate the crankshaft momentarily.

 

MP.

[dmr]

Interesting that during the "coast down" phase, after the starter is off, the energy stored in the air compressed the cylinders is enough to accelerate the crankshaft momentarily.

 

MP.

 

I am measuring engine speed from an inductive probe on the starter ring, and getting this speed by timing the interval between tooth passes. The limitation with this is that it can't tell which way the engine is turning and so I am almost certain that the crank is actually oscillating backwards and forwards here as it "bounces" off the compression peaks. I reckon the section between 11.5 and 12 seconds is the engine going backwards.

 

..............Dave

[MoominPapa]

 

I am measuring engine speed from an inductive probe on the starter ring, and getting this speed by timing the interval between tooth passes. The limitation with this is that it can't tell which way the engine is turning and so I am almost certain that the crank is actually oscillating backwards and forwards here as it "bounces" off the compression peaks. I reckon the section between 11.5 and 12 seconds is the engine going backwards.

 

..............Dave

 

Quite possible. Ye Olde FR2 certainly sometimes sets off backwards for a turn or so as it stops. I've had to learn not to use the raised hand-start shaft as a convenient hand-hold as I lean over to get to the stop lever - when the crankshaft goes backwards, the ratchet engages and the hand-start rotates too!

 

MP.

[nicknorman]

You haven't told us how you measured and recorded the data!

 

Anyway it's interesting to see the 1100A peak, just goes to show that a tiny bit of extraneous resistance from a bad connection is going to have a big effect.

 

On the direction of rotation thing, you could add a second inductive probe such that the teeth passed just slightly out of phase with the original probe. This would allow the software to work out the direction of rotation and invert the sign if it was going backwards.

Edited January 1, 2017 by nicknorman

[RD1]

It would be interesting to know how much of the current limit is caused by the starter and battery connection cables and contact resistance of inline switches and terminals.

 

If you know the cable size and length of the interconnecting cables then you can find the resistance of the cable, but the most limiting factor could be the internal resistance of the battery, this is where the Cold Cranking Amps CCA comes into effect.

 

To check connection and cable resistance, it would useful to monitor the voltage at the battery lead terminals, and the voltage on the connection bolts on the starter, then all connection points in between, any points of low or poor resistance can be adressed if necessary.

 

Your graph looks a very healthy indication of a good set up.

[dmr]

You haven't told us how you measured and recorded the data!

 

Anyway it's interesting to see the 1100A peak, just goes to show that a tiny bit of extraneous resistance from a bad connection is going to have a big effect.

 

On the direction of rotation thing, you could add a second inductive probe such that the teeth passed just slightly out of phase with the original probe. This would allow the software to work out the direction of rotation and invert the sign if it was going backwards.

 

There is a very convenient tapped hole in the flywheel housing in exactly the right place to insert an inductive probe (more than a co-incidence I suspect) but getting a second probe in would be much much more difficult. I know my engine wobbles forwards and backwards as it stops, I can sometimes see it. The current was measured with a mid range clamp on current probe, and this was quite close to the starter motor. I know this forum is suspicious of the accuracy of clamp-ons so I might investigate this. It is also only rated to 400amps but the signal at 1100 amps looks reasonable to me, I was a bit surprised by the 1100 amps. I would have thought that the cheapo isolator would have had enough resistance to limit this further.

 

I had the use of a 24bit multichannel data acquisition system recently which I used to record this stuff, sampling at just over 50000 samples/sec.

 

................Dave

[dmr]

It would be interesting to know how much of the current limit is caused by the starter and battery connection cables and contact resistance of inline switches and terminals.

 

If you know the cable size and length of the interconnecting cables then you can find the resistance of the cable, but the most limiting factor could be the internal resistance of the battery, this is where the Cold Cranking Amps CCA comes into effect.

 

To check connection and cable resistance, it would useful to monitor the voltage at the battery lead terminals, and the voltage on the connection bolts on the starter, then all connection points in between, any points of low or poor resistance can be adressed if necessary.

 

Your graph looks a very healthy indication of a good set up.

 

The cables are all quite short as the batteries are right next to the starter motor side of the engine, but they are only 35mm. I suspect the isolator is the weakest link. I have never swotted up on exactly how CCA is measured but there must be a time factor in it, I suspect a battery can deliver more than its CCA for a very short time.

 

 

This is the initial current rise, You can see the solenoid current and then the starter motor current and what looks like a fair bit of contact bounce.

 

..............Dave

[by'eck]

CCA is time, temperature and voltage related but only likely achievable in a laboratory or possibly for the first few occasions the battery is used.

 

Although CCA and amp hour capacity of a battery go hand in hand, it seems to me the latter is a better parameter to be considered long term when choosing a starter battery, subject to its ability to supply adequate current in the first place. No point in having a high CCA battery if it fails to have the remaining capacity to supply the required current long enough to start a stubborn engine.

 

I'm sceptical therefore of Lithiom Ion baterries for start applications. Given their low weight they seem to be a popular choice in car racing applications supplying high CCA but with relatively low capacity. Interesting also that many hybrid cars whilst employing Li-Ion for main storage still use a lead/acid battery to start the engine.

[MoominPapa]

I'm sceptical therefore of Lithiom Ion baterries for start applications. Given their low weight they seem to be a popular choice in car racing applications supplying high CCA but with relatively low capacity. Interesting also that many hybrid cars whilst employing Li-Ion for main storage still use a lead/acid battery to start the engine.

 

There may be other reasons than suitability of Lion for starting. Since you'd want to be able to start the engine even when the main bank is dead, you'd need a separate battery anyway. Lead-acid based starting systems are sorted technology and very cheap, why change for the sake of it?

 

MP.

[Arthur Brown]

It's a simple process for a recovery company to use a 12v jump starter supply for a flat battery on an ordinary car so probably it's just as simple on a dual fuel car. -Won't start -apply 12v, get engine started. -lets hybrid car users be aided by an established recovery network (RAC, AA, and lots of others.)

[Tiggs]

Appears to be a lot of current. Would be interesting to see the battery volt drop too.

Edited January 2, 2017 by Tiggs

[Chalky]

What would have been interesting as well would have been the battery voltage. Pulling +1000A will bring it down a lot - I've seen voltages as low as 3V when cranking a 4.6 litre V8 petrol engine at -40C (yes it did start!). The current profile and magnitude look about right.

 

There may be other reasons than suitability of Lion for starting. Since you'd want to be able to start the engine even when the main bank is dead, you'd need a separate battery anyway. Lead-acid based starting systems are sorted technology and very cheap, why change for the sake of it?

 

MP.

 

Big problem with Li-on batteries is that their performace is poor at low temperatures and they need to be warmed up to get the best from them (or cooled if too hot). Capacitive starting using a buck boost converter to charge them is a better option.

[BEngo]

These amps appear to be very large - are you sure the calibration of the measurement is correct? I would have expected 150 max, about 60 continuous. I'm estimating this on the basis of what I have measured starting car engines and that 1 hp is 740 W which is about what you need to start an engine otherwise a person would not be able to hand start it, and other posts about starting currents.

Those figures look about OK for a diesel starter motor. A boat/tractor diesel is a different kettle of bolts to any petrol car engine- heavier flywheel, higher compression to start with.

 

However, it's not power you need to start an engine by motor, it is torque. In both petrol and diesel engines you have to force it over compression. In the petrol engine the spark and the combustion will then take over- that's why the old-style petrol starting handle worked. One did not wind away until the engine started, the technique was to pull/push the engine smartly over compression (keeping the thumb out of the way of a kickback) and expecting it to start as soon as combustible mixture was in the combustion chamber. The same technique also works on things like light aircraft engines by swinging the propeller.

 

Hand starting a diesel usually involves giving the flywheel enough energy to get the engine over compression. You wind away with the decompressors out and then drop them in once the flywheel is going fast enough, possibly pushing hard to help the engine over the first compression when hopefully the air is hot enough to fire the diesel. The rate at which you bring the flywheel up to speed doesn't matter, so relatively low power is acceptable. That is not the case for a starter motor which must spin a diesel fast enough to get/keep the air hot(too slow and all the compression heat goes into the chamber walls and piston) and generate enough torque to force the engine over TDC.

 

N

[Chalky]

With modern electronic engine management the engine needs to spin at several hundred rpm to allow the crank sensor to correctly sync and work out where the crank is. Depending on the control method used it may be necessary to sync cam position as well.

A diesel requires more torque to turn it over but most require a minimum speed to get them going.

[Sir Nibble]

I've mentioned this before but take a look at the solenoid current. I have found typically around 35A which seems to apply to your graph. This is why voltage drop in the solenoid supply is a not uncommon cause of failure.

[Sir Nibble]

I think also that the dip at about 6.75 is the contact settling but the second dip may well be backlash taken up as the pinion hits the stop collar and drive begins to be transmitted.

[dmr]

I think also that the dip at about 6.75 is the contact settling but the second dip may well be backlash taken up as the pinion hits the stop collar and drive begins to be transmitted.

 

Sounds like a good theory.

Here is a bigger zoom of the initial current rise which better shows the smaller bounces and the single big interruption to starter current.

 

 

..........Dave

[1st ade]

With modern electronic engine management the engine needs to spin at several hundred rpm to allow the crank sensor to correctly sync and work out where the crank is. Depending on the control method used it may be necessary to sync cam position as well.

A diesel requires more torque to turn it over but most require a minimum speed to get them going.

 

I've a vague recollection of a vintage car engine which (at least when warm) could be started from nothing by moving the ignition advance lever between the extremities? I assume that at least one cylinder had fuel vapour in it when stationary; racking the advance / retard back and forth fired the spark and turned the crankshaft just enough for the next cylinder to take over.

[WotEver]

 

I've a vague recollection of a vintage car engine which (at least when warm) could be started from nothing by moving the ignition advance lever between the extremities? I assume that at least one cylinder had fuel vapour in it when stationary; racking the advance / retard back and forth fired the spark and turned the crankshaft just enough for the next cylinder to take over.

 

My dad told me that was quite a common wheeze. When stopping the engine you would pull the lever to full retard and then if you were lucky advancing it to full advance would indeed start it. Obviously this only worked on a car with coil ignition systems, you were out of luck with a magneto system.

[OldGoat]

dmr, on 01 Jan 2017 - 4:34 PM, said:

Its a wet start to the new year so here's something for the forum techies to talk about.

 

This is starter motor current and "instantaneous" engine speed during cranking. This is my Beta JD3 on a cold morning. The engine water temperature was about 10 degrees and the injection pump solenoid was not activated as otherwise the engine starts almost instantly.

 

starter current.jpg

 

 

OOO - I'm about to do some work on the dreaded mains fridge startup current, about which there's zilch specific info available...

What sort of kit is required (I assume it's very 'professional' and expensive.

Could you point me in the right direction - costs, availability, calibration, whatever.

 

(I'd be most grateful....)

[Chalky]

 

My dad told me that was quite a common wheeze. When stopping the engine you would pull the lever to full retard and then if you were lucky advancing it to full advance would indeed start it. Obviously this only worked on a car with coil ignition systems, you were out of luck with a magneto system.

 

There is one (petrol) engine that I've come across (about 18 years ago) that used a wasted spark ignition system and had to be cranked for a few seconds to purge the manifold and cylinders of fuel vapour before the ignition was turned on. If it wasn't purged the results were "spectacular".

[dmr]

OOO - I'm about to do some work on the dreaded mains fridge startup current, about which there's zilch specific info available...

What sort of kit is required (I assume it's very 'professional' and expensive.

Could you point me in the right direction - costs, availability, calibration, whatever.

 

(I'd be most grateful....)

 

My stuff is expensive because I use it for serious work projects, but there are cheaper options. Probably best we do the details by PM.

As you are doing mains stuff a clamp on current probe would be a whole lot safer than using a shunt. I used something similar to this at about £200:

 

http://uk.rs-online.com/web/p/current-probes-clamps/3155204/

 

You need to "sample" quite quickly to capture the transient. You could get by with a digital storage 'scope but a PC and USB data capture system might be better. Something like this might just about be good enough:.........

 

https://sine.ni.com/nips/cds/view/p/lang/en/nid/201986

 

However if you are only measuring AC mains you could possibly get by with a USB sound card and a bit of ingenuity. I could probably help with the software side of things. If you are setting up a long term capability rather than a one off job then there is better stuff available but it rapidly goes well over £1000.

 

................Dave

[dmr]

or maybe a digital storage scope like something from the Hantek range would be a good way to go. These cost £300-500 and are on eBay. They have a USB connection so you can get the recorded data into a PC, but the supplied Chinese software is not pleasurable to use.

 

............Dave