voltage loss data

[focus frenzy]

working on the project bike and gathered some interesting data on voltage loss in a GS ignition system.

the project bike has a new wiring harness so the numbers I found can be worse on a bike with the original harness.

after stringing the new harness and much toil and trouble cleaning carbs someone installed pilot jets in with a hammer it wouldn't start, just the random backfire.

voltage readings with new battery, ignition on.

at battery.... .................................................. ..........12.60 volts
at hot side (to wards front of bike) of master fuse.............12.45
at "out" side of master fuse. (voltage lose through fuse)....12.30
at connector from fuse box into main wiring harness..........12.15
at main harness plug going back to individual fuses............11.80
at + terminal on each coil.............................................. 9.60

for a grand total of 3 volts lost!!! and this is a brand new shiny harness!!
I did not measure back across the ignition fuse for loss there as it was getting late and Stargate Atlantis comes on soon but will test tomorrow.

[Junkie]

Wow, that's really bad. I'm planning on complelely redoing the wiring harness on mine in a few weeks, I'll draw up a diagram when I do so.

[SPARKSS]

...snipped....
the project bike has a new wiring harness so the numbers I found can be worse on a bike with the original harness.

voltage readings with new battery, ignition on.

at battery.... .................................................. ..........12.60 volts
at hot side (to wards front of bike) of master fuse.............12.45
at "out" side of master fuse. (voltage lose through fuse)....12.30
at connector from fuse box into main wiring harness..........12.15
at main harness plug going back to individual fuses............11.80
at + terminal on each coil.............................................. 9.60

for a grand total of 3 volts lost!!! and this is a brand new shiny harness!!
I did not measure back across the ignition fuse for loss there as it was getting late .....

Quite an eye opener wasn't it! Did you use the battery negative post for reference always or some point on the chassis? Often forgotten is to check that lead (and any other ground side wiring if running through connectors and having bonding screws etc).

Your measurements (key-on, engine off I'll assume) look not untypical up to the large loss seen at the ignition coils (so a place to start). If only people would realize how important it is to have clean, protected and tight connections. Not your case, but old wiring can surely have significant resistance and when doing voltage drop testing like this, needs to be measured as well.

Good job sir!

[focus frenzy]

thanks!
the key item, and one I was ignorant to before buying and fixing my G and learning the hard way. Brass is a decent conductor, brass tarnishes, brass tarnish does not conduct electricity.

its the tarnish that causes stator connections to heat up and burn up. it causes a assortment of issues and headaches.
with my 80 1000G it would totally and randomly fail to start, it would crank over fine but not fire and in the process of figuring it out I would inevitably check fuses and of course the fuse would be fine and the bike would then start.

well I had enough of that and started taking a test light with me and one day it did it and I was ready and discovered the fuse had power when probed but when I touched test light to the clip that held the fuse there was no power there till i just barely nudged it.

I now have a aftermarket fuse holder that takes mini blade fuses and doesn't use brass for connections.

that is likely what I will do with the project bike.

[posplayr]

A couple of comments

I agree with your analysis on the brass connectors. And given your results on the various fuze box measurements , I dont feel so anal for having done my fuze box refurb. I posted the link and pics a while back.

Used Jasco metal etch to clean the fuze box and then soldered all of the crimp connections inside. The cimp connctions are probably as worse (hidden) problem than the bullet friction connections. I have tried to heat and get solder to flow into the alot of the crimps but with too much corrosion it won't stick. Better to just cut it off and replace. If you like and want to retain the bullets the Jasco metal etch will work. Probably follow with dialetric grease/

One comment relative to the coil voltage is that might be not all cable/connector drop. Not sure if you have points or inductive pickups but the + side of the coil is not exactly the top side of the battery.

[Edit 3/11/08: I was thinking in term of the ignition system but later realized that the coil is low side controlled and in fact the + side of the coil should as close to teh battery as the 4 amps will allow]


Posplayr

http://www.thegsresources.com/_forum/showthread.php?p=743425#post743425

picture is a little hard to see but each crimp was heated so solder would flow into it. This basically will maintain good connectivity in the crimp regardless of surface corrosion. Atill got to watch the bullet part of the connections

[SPARKSS]

..snipped.... If you like and want to retain the bullets the Jasco metal etch will work. Probably follow with dialetric grease/

One comment relative to the coil voltage is that might be not all cable/connector drop. Not sure if you have points or inductive pickups but the + side of the coil is not exactly the top side of the battery.

picture is a little hard to see but each crimp was heated so solder would flow into it. This basically will maintain good connectivity in the crimp regardless of surface corrosion.

Hello Pos.....I keyed on a couple of points you made, first that the idea of cleaning and using grease on those under-box crimps is a good idea. The suggestion to solder such connections can be a mixed bag - corrosion as you pointed out is one but melting the retaining plastic or overheating the spring fingers which makes those connections can also easily happen and should be avoided - and would likely happen with even moderately skilled soldering.

Now I'm curious as to why you wouldn't think the feed voltage (with respect to ground) would normally be significantly less than battery voltage? I admit I was looking at the diagram for my 79 GS1000 which does not have any resistance wire in the coil + feed. Perhaps his 80 or later does which would explain your comment. Otherwise, I'd suspect that the measured drop at this point could indeed be the usual, measureable suspects (connectors, wiring) and that darn kill switch.

.02 more for the discussion

[posplayr]

Sparkss

Sorry to take so long to answer. I just happened to see your post. I'll try and answer.

Actually as yet I have not greased anything. I have the grease and will get to it. As far as the soldering the fuse box I have to admit that I do have a quality electronics Weller soldering station (not your typical automotive grade soldering iron) so you make a good point for the more general populace. However, with a reasonable quality soldering iron, after having used the Jasco metal etch, it took very little heat to get solder to flow into the crimp. And maybe I was not clear on this point, I was only looking to provide corrosion resistance to the high current paths through the fuse box. Having ohmed some of these bullet connections out much of the resistance seemed to be there even after having cleaned the bullet.

So I guess in the final analysis, I would still check the fuse box for resistance(to get an idea of how bad it was), definitely clean it with metal etch as overheating the connections to try to get solder to flow will likely damage the fuse box plastic. After the cleaning, it really went much easier that I thought(see pics of the bright shiny copper) to solder the crimps only.

OK on the voltage to the coils, I have to claim a certain mount of ignorance as to all of the points configurations, but here is my thinking:
If the points are open, (ok brain fart), you should see all the battery voltage, but this is not really a test of continuity becasue there is no current flowing. If the points are closed, if you have a ballast resistor, there could be a big drop as would be expected from a voltage divider. If there is no divider an if the points are closed, and if there is a voltage drop to the coil + (measured relative to the + batt) this might be an indication that there is an excessive voltage drop to your coil. However, You would need to be careful as the ignition could very easily upset any digital volt meter with the AC coming from the seconday. So now we will additionally assume you are using a analog volt meter with heavy damping, yes you should see close to full batter voltage at the + side of the coil. But what if it is 11 volts is it too little? Probably not. The bigger issue is where is the resistance coming from.

So stricly speaking, I was probably misspeaking, however thinking of the general populace, with the various caveats identified, there could be plenty of head scrathing involved with trying to measure coil voltage. I always look for blue spark (which was shocking the first time I saw the feeble yellow spart from my GS750).

Drawing on my previous Chevy experience, I just never had to verify voltage to the coil, and I was probably thinking in terms of getting a dwell reading(negative side) or an ignition related AC voltage reading coming back from the secondary.

You have to also realize, I was trying to get my voltage drops between neg and pos on batt to R/R to less than 0.1-0.2 volts at 4000 rpm. Given that the this can only be done while the engine is running, and depending upon the quality of the voltmeter, it just is not clear how well anybody will ever be able to determine battery voltage from looking at the high side of the coil while running (see comment re:secondary coupling).

The bottom line is , if the engine is off and the points are open, then I would expect the full voltage because there is no current flowing through the primary (however this is not a very good test of resistance). If the motor is off, but the points are closed, the voltage you see will be dependent on whether there is a ballast resistor, but with no ballast resistor and not running but with the points closed then you should see full battery voltage.

In actuality what I was thinking about was the variables in establishing the resistance in the ground path (through the points or other electronics module; without adding that to the circuit what are you really testing??)

I guess all of these caveats left me a little uneasy with expecting perfect battery voltage for the uninitiated. Certainly having an extra 0.5V voltage drop to the coil should not hurt anything when compared to the impact to the charging system, but your point is well taken it is possible, you just need to understand the issues outlined.

Thanks for the question; Nice discussion.

Posplayr

[SPARKSS]

Good discussion and points POS. Hopefully this sort of dialogue adds to the knowledge of people searching for answers within the site

[oldryder]

Theres a great solution for this. On many bikes you have 2 or 3 or even more sets of contacts between the primary side of the ignition coils and the battery. (ignition switch, kill switch, clutch interlock, sidestand interlock, maybe something else).

In industrial controls low voltage DC switch and relay contacts are gold plated so they don't corrode and introduce more and more voltage loss over time. motorcycle switches have no such gold plating and over time they degrade.

On an older bike it's common to see the voltage at the coil primary 2 or more volts below the voltage at the battery. Typical coils step up the voltage by about 2500 times so a 2 volt loss is decreasing your spark voltage by 5000 volts.

Looked at another way 2 volts lost out of 12.4 is a 16% loss. More than enough to make a bike hard to start particularly with 25 year old coils

Fortunately there is a very simple solution: A relay between the battery positive and the coil primary. A $3 horn relay from NAPA works fine. If you use the existing wire going to the coil primary to energize the relay all your interlocks will still work. You connect a fused lead at the battery positive, go thru the relay contacts, and on to the coil primary terminals.

total cost including a fused terminal lead is about $5. install takes about 20 minutes. your bike will start much better.

I often see guys spend hundreds of dollars on aftermarket coils when all they need to do is add a relay as described here.

[posplayr]

oldryder

Well, I guess if you have a bunch of bad connections getting to the plus side of the coil, then adding a selonoid would get you better voltage to the same point (assuming you have good enough connections to control the selonoid).

However, you have added back in several connections and on my GS at least you would have only removed the key and kill switch. With new wire and connections the coil primary is going to be in better shape, but that is not going to help much on the rest of the electrics.

I guess I would consider this solution a last resort if you could not clean the connections. You end up using about a 50 amp relay to control 4 amps (or less) of primary current. I certainly would not suggest adding hotter coils to fix a bad set of connection, although that might also be able to help, it is masking the real problem.

Thanks
Posplayr

[BassCliff]

Fortunately there is a very simple solution: A relay between the battery positive and the coil primary. A $3 horn relay from NAPA works fine. If you use the existing wire going to the coil primary to energize the relay all your interlocks will still work. You connect a fused lead at the battery positive, go thru the relay contacts, and on to the coil primary terminals.

Fascinating! Mr. oldryder, do you think you could draw up a diagram? (I like pictures.) :-D

Have you done this to any of your bikes?

Thank you for your indulgence,

BassCliff

[oldryder]

Basscliff: I am an electrical engineer with as lot of experience with low voltage DC controls so the problem was a familiar one for me. I have done this as described on a few bikes. The effect was most dramatic on a KZ1300 I had. Those bikes are notorious hard starters and coil upgrades are very common but I didn't want to spend that much money. I had a 2.4 volt drop from the battery to the coil primary. After installing the relay the bike started easily all the time even on cold MN mornings.

A nice feature is that by using the wiring harness wire to the coil primary to energize the relay you maintain all the functionality of the various interlocks so the wiring change is "invisible" to the rider.

I've got a drawing is someone can tell me how to post it. I can embed it in a word document so if someone wants to PM me I can also send it in an email.

[posplayr]

oldryder

Any ideas where the drop is coming from? I just checked out my 1981 GS750EX schematic and to get that kind of drop would seem most likely to be coming from bad contacts in the ignition or kill switch. If you can not clean those contacts, then this would be a much cheaper solution than switch replacement which would be close to $100 for both. If it was in the fuze box I would try and clean that.

One modification I guess I would throw out there is that if you are really dealing with poor switch contacts that can't be cleaned, then the entire ignition keyed side of the +12 should be relayed and not just the coils. If the problem is only in the starter button then you can do much better than your solution short of started switch replacment.

Regardless a $3 relay is a pretty simple way to insure good voltage to the coils thanks for the idea.

Posplayr

[focus frenzy]

wow this thread came back from the dead.
I had taken my initial data and also posted it in general GS and also posted revised readings after discovering a nasty error I had made (oh the shame)
I also ran the test on my own 1000G that has a assortment of relay mods.
I have taken my switches apart and cleaned contacts and greased contacts and it did not make much differance.

with cleaned contacts I still have .6 volt loss across kill switch and surprisingly I also have a .65 volt drop across the contacts in the relay!

I also have aftermarket quality coils but I went to them as they are simpler to change out old dryrotted cracked leaky wires on.
they also put out a very healthy fat and hot spark. motorcycle coil technology has come a long way in the past 30 years.

[posplayr]

Hey Guys

We're only talking about 4-5 amps max going to the coils. So the resistance has to be somewhere .

Ok lets say we have a 3 ohm coil as 12 volts. We are trying to carry 4 amps. It only takes .15 ohms to generate a 0.6 volts drop at 4 amps. We are getting to some pretty low resistance readings. A garden variety ($10) meters might only have 0.1 ohm resolution). So a few 1/10 of an ohm could easily be sprinkled around and be easily overlooked. Basically even if you can NOT measure the resistance very well the voltage drop is telling you the same; you have excessive resistance.

As I mentioned earlier, I think corrosion gets into the crimps of the connections (I tested a few to confirm it happens). You can't clean those with sand paper and short of replacement, a little Jasco metal etch will clean them up just fine in about 1-2 minutes. The metal will then be bare so cover it with something. I used solder to flow into the crimps. Following with dielectric grease on spades or bullets wont hurt.

As shown in the previous pictures I even did this to all of the crimps in the fuse box (with a Weller small tipped iron). During charging tests I don't think I had much more that a 0.1 volt drop (at 5000 RPM) between R/R high side and battery and corresponding same low side measurements.

So my new recommendation above and beyond replacing bullets with spades is to etch and solder flow any crimp connections if you find voltage drops on them.

Posplayr