Nasty Hiccup

[docc]

Hmm, I see how that comes out to 49 amps. Doesn't that seem like immense current to energize a solenoid? Seems like you could run the starter motor with that kind of current.

[Ryland3210]

Hmm, I see how that comes out to 49 amps. Doesn't that seem like immense current to energize a solenoid? Seems like you could run the starter motor with that kind of current.

 

According to the manual, the starter motor is rated at 1.2 kW at 12 volts. That translates into 100 amps if it were 100% efficient. It's probably somewhere between 75 and 85% efficient, so the actual current draw would then be somewhere between 117 and 133 amps. The current with no load is given as 600 A no load and 230 A with load. I think the manual has it backwards. The 600 amps is most likely the current drawn if the rotor is locked. At 230 amps, the input from the battery would be 2.76 kW, which implies an efficiency of only 43.5 %. Maybe so, but that's pretty poor performance.

 

In any event, it is not unreasonable for a solenoid to draw 49 amps to operate a 230 amp set of contacts, plus operate the solenoid to engage the starter motor mechanically.

 

When I first measured the resistance of the solenoid with an ohmmeter, I had difficulty believing it when it read 0.2 Ohms since its resolution is + - 0.1, so I measured it to three decimals using a current source and measuring the voltage across the solenoid. Since two independent measurements agree, I'm confident it's around 0.25 Ohms.

 

Yet, the fuse running it is only a 20 amp fuse. I see three possibilites: either the fuse can withstand much more than 20 amps for several seconds, or the solenoid current draw decreases as it pulls in, or both. The manual is emphatic about only attempting starting for 5 seconds, and then waiting 10 seconds before trying again. This would give the 20 amp fuse and relay contacts some time to dissapate the heat generated. As far as the solenoid current draw decreasing as it pulls in, that is normal for an AC solenoid, but not for a DC.

 

One conclusion I have reached so far, is that the 20 AMP fuse will probably do a good job protecting the relay contacts, but I plan on carrying spares with a higher rating, just in case I blow one on the road just because it takes too long to get the bike started. It also means the inrush and switching capacity of the relay contacts being above 50 amps are important, given the 49 amp load.

[Ryland3210]

This spec sheet does

http://www.geirelays.com/acrobat/AR4.pdf

 

Thanks, Dave. That fills in some blanks.

 

Here are the differences between the Omron and GEI AR4 relay NO contacts according to the data sheets:

 

GEI Omron

Switching Amps 25 60

Steady Amps 25 20

Initial contact voltage 0.2 0.15 ( both at rated current)

However, the GEI also states initial contact resistance is .05 Ohms, which translates to 1.0 volt at 20 amps)

Vibration 10-40Hz at 1.27mm double amplitude for the GEI versus 20-500 Hz at 43.1m/s^2 for the Omron.

Shock, functioning "Aproximately 100m/s^2" for the GEI, verus "100m/s^2" for the Omron.

Shock, mechanical Not specified for the GEI, 1000m/s^2 for the Omron.

Temperature, operating -40-85C -40-125C

Weight 19g 20g

 

Omron shows durability of 300,000 cycles at 40A inductive load inrush, 20A steady, 300,000 cycles at 100A inrush, 20A steady for lamp load.

 

GEI does not specify durability ratings.

 

Omron shows overcurrent rating: 80A for 10 seconds at 14 volts, 80C, and

50A for 30 seconds at 14 volts, 80C.

 

GEI does not specify overcurrent ratings.

 

In terms of the vibration ratings, the higher frequency ratings provided by Omron are more relavant. The 40 Hz highest frequency tested by GEI corresponds to 2400 RPM, wheras the range rated by Omron corresponds to 1200 to 30,000. Although these engines do not rev that high, harmonics of the fundamental frequency will go above the redline.

 

While it appears the GEI has a higher steady current rating, the Omron has the advantage on being able to switch 60 amps versus GEI's 25. GEI's somewhat higher contact resistance means that more heat will build up in the contacts and there will be a higher voltage drop. There also is an inconsistency in its ratings. At 25 amps load, one specification indicates 0.2 volts, whereas the other translates to 1.25. At 25 amps, that's 5 watts and 31.25 watts respectively. In either case, that's a lot of power concentrated at the contacts. The Omron version dissapates a little less power in the contacts and has a lower voltage drop. It comes down to how much precious metal is used in the contacts, how much pressure is applied by the coil to the contacts, and how well the design conducts away heat.

 

It would be worthwhile to dissect one of each to see what the differences are. Meanwhile, in my opinion, Omron's higher switching current capacity overcomes the GEI steady current's 25A versus Omron's 20A because of that and because when one examines the voltage drops, the Omron shows less voltage drop. The implication is that Omron has better contacts in spite of the lower rating. It appears to be a more conservative rating. The higher operating temperature rating, vibration rating covering the entire RPM range, overcurrent ratings, and durability ratings all inspire confidence.

 

I'm looking forward to receiving Omron's samples early this coming week.

[dlaing]

What are the specs on the Omron NC connection?

The GEI is rated at 20 Amps at the 87a terminal.

When Dan and I searched for relays years ago, that was the number that we could not beat.

 

Regarding the Starter Solenoid Load, if it is too much, than a standard size relay would be the best fix, as it could easily triple the Amp rating of both the NO and NC (normally open and closed contacts).

If someone offered a plug and play kit to convert the relay size, that would be an excellent solution for the starter relay, and if the polarity could be reversed for diode protected relay use, it could be an even better solution, but bulkier.

[Ryland3210]

What are the specs on the Omron NC connection?

The GEI is rated at 20 Amps at the 87a terminal.

When Dan and I searched for relays years ago, that was the number that we could not beat.

 

Regarding the Starter Solenoid Load, if it is too much, than a standard size relay would be the best fix, as it could easily triple the Amp rating of both the NO and NC (normally open and closed contacts).

If someone offered a plug and play kit to convert the relay size, that would be an excellent solution for the starter relay, and if the polarity could be reversed for diode protected relay use, it could be an even better solution, but bulkier.

 

The Omron NC connection specs are 30 amps switching, 10 amps steady. They used to make a heavier duty relay rated at 40 NO and 20 NC, but do not want to gear up for production on it. GEI does not specify the switching current, and its steady current rating is 20 amps. I didn't focus on that because with the newer model bikes it only drives a switching relay coil that draws about 0.15 amps, and on the older bikes around 8.1 amps maximum. However, for example, using GEI's spec of .05 ohms initial contact resistance, that would permit a voltage drop of 0.4 volts across the contacts and dissapate 3.3 watts, not terribly much. That's the initial contact resistance, when they are shiny and new. It generally increases with use. If the older bikes have had problems with this contact on other brands of relays, it may have been that an even higher contact resistance would lead to more heat damage to the contacts and dimmer headlights, especially considering the effect that two contacts in series driving the lights would have. 0.8 volts can make a big difference in brightness in incandescant bulbs.

 

There is no need to reverse the polarity of the circuit driving the coil in order to use a diode. The diode can be wired in reverse instead.

 

I'm with you on using the existing relay to drive another, higher powered, relay to drive the starter in turn. I'm frankly amazed these micro starter relays don't get fried more often. I've already taped a spare next to the relay bank.

 

Here's another interesting spec. from Omron: minimum carry current is 1 amp at 5 volts.

What this is about is that, except for mercury wetted relays, dry contacts can increase resistance over time if the load is too light. What Omron is saying is that life will be shortened unless at least 5 volts is switched, and at least 1 amp is the load current. The reason is that a little load is a good thing because it helps keep the contacts clean. Therefore, the fact that the NC contacts only drive 0.15 amps is worse than if it drove 1. The fact that 12-14 is higher than 5 is a positive in this regard. The same would hold true for GEI, even though they do not show a minimum carry rating.

[dlaing]

There is no need to reverse the polarity of the circuit driving the coil in order to use a diode. The diode can be wired in reverse instead.

Isn't that the same thing??? Obviously not...

Please elaborate.

I'm with you on using the existing relay to drive another, higher powered, relay to drive the starter in turn. I'm frankly amazed these micro starter relays don't get fried more often. I've already taped a spare next to the relay bank.

I meant completely replace the small relay with a larger relay, probably with blade connectors in existing socket wire to larger socket with larger relay, and the 85 and 86 would be swapped so that polarity would be reversed.

 

As for the numbers, I really don't understand the difference between switching, steady and 'rated' rating.

On the GEI NO the 'switching' and the 'rated' are both 25A and the 'max carry' current is 30A with maximum switching power of 350W(25A*14V), while the NC is 20A rated at 14V.

[Ryland3210]

Isn't that the same thing??? Obviously not...

Please elaborate.

 

I meant completely replace the small relay with a larger relay, probably with blade connectors in existing socket wire to larger socket with larger relay, and the 85 and 86 would be swapped so that polarity would be reversed.

 

As for the numbers, I really don't understand the difference between switching, steady and 'rated' rating.

On the GEI NO the 'switching' and the 'rated' are both 25A and the 'max carry' current is 30A with maximum switching power of 350W(25A*14V), while the NC is 20A rated at 14V.

 

If a relay comes with internal diode protection, then its coil terminals will have to wired with the manufacturer's polarity. If a diode is to be added externally, the diode's cathode must be connected to the side of the coil which is positive, and the diode's anode to the negative terminal.

 

The switching current is the maximum current the relay can absorb the instant the contacts are closed, but not forever. The steady current is the continuous current it can handle thereafter. For example, lamps normally draw several times their operating current when first energized. As the filament heats up and glows, its resistance increases dramatically, so the current draw decreases. That's why both ratings are important. Rated load is normally the same as the steady or continuous current it can handle continuously without failure, but "rated" can mean different things to different manufacturers. For example, for how many cycles, at what temperature, etc.?

 

The way GEI specifies it, the rated current of 25A is continuous without time limit. The 30A max. carry is the current it can handle without damage for a limited (unspecified by GEI) period of time. The corresponds to Omron's "overcurrent"

[docc]

I thought I had wired the last additional relay to the Sport. Yet, the additional starter relay sounds good and it could be tucked up above the starter which is pretty close to the positive terminal. I've got to take the starter off to try and get my neutral switch working again; I'll look for a good mounting location. What size wire for the hot supply? Is 14 gauge enough or should I go with 12?

 

The simplest solution may be a change interval on the relays. 12,000 miles?

[Ryland3210]

I thought I had wired the last additional relay to the Sport. Yet, the additional starter relay sounds good and it could be tucked up above the starter which is pretty close to the positive terminal. I've got to take the starter off to try and get my neutral switch working again; I'll look for a good mounting location. What size wire for the hot supply? Is 14 gauge enough or should I go with 12?

 

The simplest solution may be a change interval on the relays. 12,000 miles?

 

For the starter solenoid, go with 12, multistranded copper. More strands means more vibration tolerance.

 

I'm inclined to agree with you about simply replacing the starter relay when it blows if 12,000 miles is the typical lifetime. It's cheap and easy. The Omrons should do even better with the high inrush rating.

[dlaing]

For the starter solenoid, go with 12, multistranded copper. More strands means more vibration tolerance.

 

I'm inclined to agree with you about simply replacing the starter relay when it blows if 12,000 miles is the typical lifetime. It's cheap and easy. The Omrons should do even better with the high inrush rating.

Why not replace the relays before they blow?

I guess because mileage may not be good predictor of failure.

How many miles were on Docc's GEI relay?

Quazimoto and Ouiji Veck both had GEI failure.

How many miles were on their's?

My Bosch and Siemmens failed quickly.

My GEIs have lasted since December 2003, about 18,000 miles, but the past year (~5000 miles) the current for the headlight has been bypassed. Also, I replaced the voltage regulator which could have been killing the earlier relays, but the GEIs survived 24 months with OEM regulator that went bad, and 23 months so far with Electrex regulator.

I suspect many of the relay failures are related to other electrical problems, but it is near impossible to quantify.

We do know the GEI has been much more failure resistant than the Bosch.

From reading the forum it also appears to be more reliable than the Tyco A1001-A403, which might not look bad on paper, but the NC is still a potential issue.

http://www.semitek.cz/tyco/rele/ec/v23074.pdf

[docc]

Why not replace the relays before they blow?

I guess because mileage may not be good predictor of failure.

How many miles were on Docc's GEI relay?

13,500 miles on the FI and ECU GEI relays. I did find the uncovered fuel pump connection which may have contributed to the failure.

[dlaing]

13,500 miles on the FI and ECU GEI relays. I did find the uncovered fuel pump connection which may have contributed to the failure.

 

AHAHHH! So, if you had replaced the relays at 13,499 miles you never would have known about this problem!!!!

If Quasi and Ouiji let us know their mileage, we might find a pattern.

Any others with GEI relay failures?

I don't think amp load is the problem when the relay is rated twice as high as the fuse that is rated twice as high as Ryland's amp ratings for that circuit. Or did you do some funky wiring for that heated vest? Not likely.

Or maybe it is a switching current problem and 25A rating is not enough????

Probably not on the ECU relay, but the fuel pump may require more than 25 switching amps????

Vibration plus a relay built by a disgruntled employee, could be the problem????

 

The cover pops off of my fuel pump connections all the time. I just coat the terminals with silicone dielectric grease and I don't think they have been a problem, and I may have added silver conductive to the contact points, but I don't recall.. Bolt down eyelet connections, or whatever the proper term is, should be the most reliable, atleast compared to blade and bullet connectors.

Coating with battery terminal protector goop would be ideal.

[docc]

AHAHHH! So, if you had replaced the relays at 13,499 miles you never would have known about this problem!!!!

 

Or did you do some funky wiring for that heated vest? Not likely.

 

I replaced the relays at 13,500, but had been fussing with the problem for at least a couple thousand miles. Even a change at 12,000 probably wouldn't have caught it in time.

 

My accessory jack is fused straight from the junction blocks for the battery without any relay.

[Ryland3210]

AHAHHH! So, if you had replaced the relays at 13,499 miles you never would have known about this problem!!!!

If Quasi and Ouiji let us know their mileage, we might find a pattern.

Any others with GEI relay failures?

I don't think amp load is the problem when the relay is rated twice as high as the fuse that is rated twice as high as Ryland's amp ratings for that circuit. Or did you do some funky wiring for that heated vest? Not likely.

Or maybe it is a switching current problem and 25A rating is not enough????

Probably not on the ECU relay, but the fuel pump may require more than 25 switching amps????

Vibration plus a relay built by a disgruntled employee, could be the problem????

 

The cover pops off of my fuel pump connections all the time. I just coat the terminals with silicone dielectric grease and I don't think they have been a problem, and I may have added silver conductive to the contact points, but I don't recall.. Bolt down eyelet connections, or whatever the proper term is, should be the most reliable, atleast compared to blade and bullet connectors.

Coating with battery terminal protector goop would be ideal.

 

I just cannot get my arms around the idea of using an insulating grease for the purpose of improving contact reliability. Battery terminal protection sprays and "goop" are to prevent corrosion from the outside. Useful on the outside of the connector to prevent moisture from getting in, but not on the terminals themselves. That defies common sense. Conductive paste to the mating contacts makes sense to me.

 

In cases where the load is far below the contact rating, vibration could well be the culprit. 100,000 life cycles can happen pretty quickly. 33 minutes at 3,000 RPM = 100,000 cycles. The key issue is whether the design maintains enough contact pressure to withstand the peak accelerations caused by the vibration and whether it has any resonances in the range of the engine's vibration frequency. Omron specs include the entire range of a Guzzi's vibrations.

 

Now, I'm not predicting that one can expect relays to fail in 33 minutes of riding. However, suppose vibration causes the contacts to rub against each other, wearing them out prematurely. Or, more plausible, suppose the relay has a resonance frequency within the RPM range. Then its lifetime would have a lot to do with how much time it spends at that RPM.

[dlaing]

Thanks, Dave. That fills in some blanks.

 

Here are the differences between the Omron and GEI AR4 relay NO contacts according to the data sheets:

 

GEI Omron

Switching Amps 25 60

Steady Amps 25 20

Initial contact voltage 0.2 0.15 ( both at rated current)

However, the GEI also states initial contact resistance is .05 Ohms, which translates to 1.0 volt at 20 amps)

Vibration 10-40Hz at 1.27mm double amplitude for the GEI versus 20-500 Hz at 43.1m/s^2 for the Omron.

Shock, functioning "Aproximately 100m/s^2" for the GEI, verus "100m/s^2" for the Omron.

Shock, mechanical Not specified for the GEI, 1000m/s^2 for the Omron.

Temperature, operating -40-85C -40-125C

Weight 19g 20g

 

Omron shows durability of 300,000 cycles at 40A inductive load inrush, 20A steady, 300,000 cycles at 100A inrush, 20A steady for lamp load.

 

GEI does not specify durability ratings.

 

Omron shows overcurrent rating: 80A for 10 seconds at 14 volts, 80C, and

50A for 30 seconds at 14 volts, 80C.

 

GEI does not specify overcurrent ratings.

 

In terms of the vibration ratings, the higher frequency ratings provided by Omron are more relavant. The 40 Hz highest frequency tested by GEI corresponds to 2400 RPM, wheras the range rated by Omron corresponds to 1200 to 30,000. Although these engines do not rev that high, harmonics of the fundamental frequency will go above the redline.

 

While it appears the GEI has a higher steady current rating, the Omron has the advantage on being able to switch 60 amps versus GEI's 25. GEI's somewhat higher contact resistance means that more heat will build up in the contacts and there will be a higher voltage drop. There also is an inconsistency in its ratings. At 25 amps load, one specification indicates 0.2 volts, whereas the other translates to 1.25. At 25 amps, that's 5 watts and 31.25 watts respectively. In either case, that's a lot of power concentrated at the contacts. The Omron version dissapates a little less power in the contacts and has a lower voltage drop. It comes down to how much precious metal is used in the contacts, how much pressure is applied by the coil to the contacts, and how well the design conducts away heat.

 

It would be worthwhile to dissect one of each to see what the differences are. Meanwhile, in my opinion, Omron's higher switching current capacity overcomes the GEI steady current's 25A versus Omron's 20A because of that and because when one examines the voltage drops, the Omron shows less voltage drop. The implication is that Omron has better contacts in spite of the lower rating. It appears to be a more conservative rating. The higher operating temperature rating, vibration rating covering the entire RPM range, overcurrent ratings, and durability ratings all inspire confidence.

 

I'm looking forward to receiving Omron's samples early this coming week.

Just thought I would add the Tyco V23074-A1001-A403, that is known to fail, although many are still working, into the comparison.

If I had not done the headlight modification, I would be concerned about about the NC numbers, (but the testing standards between GEI and others, are not consistent), otherwise an excellent relay:

weight 16-20g

terminals tin plated

Contacts AgSnO2

max switching current ON NC 40A NO 120A

max switching current OFF NC 15A NO 30A

Limiting continuous current at 23C NC 15A NO 25A

Limiting continuous current at 85C NC 10A NO 15A

voltage drop at 10A 20mV

Increase in coil temperature at 10A load 5C

power consumption at 12VDC 1.4W

ambient temperature range -40 to +125C

Coil resistance 124 OHMs

Must operate Voltage 7.2V (important for keeping ECU etc. functioning during starting when battery is weak)