Apparent Regulator Failures

[lragan]

My apologies for the disorganized way I have presented my data -- it has obviously caused considerable confusion.  I hope this post will clear it up.

The assumption behind all comparisons is that the relationship between heat dumped to the R/R and the temperature of the R/R is linear.  What this means is that if the heat from the engine under a certain condition causes the temperature to rise, say 10 degrees, and the heat from the electrical energy dissipated by operating the unit causes the temperature to rise, say 7 degrees, and the heat from sun load causes the temperature to rise, say 6 degrees then the temperature of the unit will be 17 degrees above the ambient temperature of the air when it is on and the engine is running, but the sun is not striking it, and 23 degrees above ambient when in the sun, all other conditions being the same.  This assumption is well established by experiment and commonly accepted by engineers everywhere.  It makes calculation and comparison relatively simple.  By invoking this assumption, we can calculate what the unit temperature will be on any given day, knowing the operating conditions and the temperature of the air (ambient).

The attached sheet shows the original data, including the reference ambient temperatures, and the columns are repeated below with the difference between measured and ambient temperature displayed.  This second set of numbers are the only ones valid for comparison.

To answer Charles' question -- today I measured the temperature of the R/R with and without sun load  Without starting or running the bike at any time during these measurements.  Just sitting in the shade, the unit was at air ambient, 35C.  Time was 6 p.m.  Parked headed north for 10 minutes, still read 35C.  Turned the bike around so the R/R was directly exposed to the sun and measured 41C.  So the 6C rise in temperature could be added to the relocated numbers parked or at low speed if the unit is directly exposed to the sun's rays at 90 degree incidence.  At noon the solar radiation density is higher, but the area in the sun is much lower due to the shape and mounting of the unit.  I think I captured either the maximum rise or close to it due to solar load.

Where does all this leave us?  It depends on the assumption of failure mechanism.  If failures are simply a matter of metal migration due to temperature, then higher temperatures over a longer period will reduce the Mean Time to Failure (MTTF) by approximately a factor of ten for every 10 degrees C the unit experiences -- the effects are cumulative.  Under this assumption, the relocation should substantially extend the life, especially for riders like me who rarely sit in traffic.

I strongly suspect that the failure mechanism involves a phenomenon known as "thermal runaway".  This accrues especially to CMOS circuits, and is due to leakage currents, which increase with temperature.  So the scenario goes like this.  A hot day, the leakage current goes up.  The circuit design may be inadequate, and the leakage current causes the circuit current to increase, so the device dissipates more heat, and then gets hotter, and the leakage current goes up even more, which makes it hotter yet... well, you can see this will not end well.  The unit will fail rather quickly.  This mechanism depends on the design, variances in manufacturing the semiconductor, and the peak temperature the unit experiences while in operation.  Without being able to prove it at all, I strongly suspect this is how they fail.  If this is true, then the temperature surge that occurs when the bike is still or operating a very low speeds in the stock configuration is the biggest killer.  If you relocate the unit and are sitting in traffic on a hot day with a direct sun on the unit, it is almost as big.  Variation in leakage current from wafer to wafer and run to run in a CMOS process is huge, so if you are lucky enough to draw devices from an unusually "tight" wafer, your unit may never fail.  If, on the other hand...

Bottom line, I can't say unequivocally that relocation will solve the failure problem.  I believe it is a step in the right direction, but I do not believe it is "the solution".


[attachment deleted by admin]

[Sledge Hammer]

We should be grateful this isn't the Aprilia Owners of Texas forum. I spoke with a guy at work who has an Aprilia Futura that not only had its regulator go out, but one phase of the alternator, and the price tag just for Aprilia parts is knocking on $350, and then there is no guarantee they will not fail the same way. Believe it or not, the fix is to use Honda parts instead (yes, I 'm struggling to keep a straight face as I type this). His previous bike was a VFR, whcih he said cooked its regulator because the tiny little cubbyhole where it was mounted received virtually no airflow, but at least it didn't call for a bank loan to finance the repairs.

[lragan]

If I had Lawrence's fancy do dad I could see how my added heat sink worked. I have one with a long probe, looks like it would get bent in the process.

Terry, I still have the meter that includes the tach.  New in the box -- I will never use it, as I have opted for the other meter.

If you will pm me your address, I will ship it to ya.

[TLRam1]

Will this do what I need it to do with the temps? PM is on the way.

[lragan]

Yes, it should take temperature readings without difficulty.  You will need to keep the display from getting too hot, or it will fade, but if you carry it in a tank bag or otherwise shield it from the sun, it will work.

I am interested in the data you take.  Be sure to get the air temperature before you start, and the "cold" temperature of the R/R before you start the bike...

I did a simple calibration on mine, just to check it.  In ice water, it should read 0C.  In boiling water, 100C.  Mine was right on at both points.  I suspect yours will be too.

[BudMan]

In view of the temperature rise from exposure to direct sunlight, is painting the radiator fins black still a god idea?  Lawrence, so you think the fins would absorb more heat from solar radiation because of the black color?  I can see it radiating better when black, but it looks to me that it will have a greater temp raise from solar exposure if it were black.
As long as the device runs hotter than ambient, it will dissipate heat to the air, but if it is setting in the sun, it may have a starting point for dissipation that is closer to the thermal run away temperature. 
I was going to paint mine black on the V-65, but the more I think bout it, I think I'll leave it alone.

[TLRam1]

Yes, it should take temperature readings without difficulty.  You will need to keep the display from getting too hot, or it will fade, but if you carry it in a tank bag or otherwise shield it from the sun, it will work.

I am interested in the data you take.  Be sure to get the air temperature before you start, and the "cold" temperature of the R/R before you start the bike...

I did a simple calibration on mine, just to check it.  In ice water, it should read 0C.  In boiling water, 100C.  Mine was right on at both points.  I suspect yours will be too.

All these strings attached and responsibility, Lawrence acts like we're married    BTW Charles did you understand the explanation from the English Major?      

[TLRam1]

They are coming close to black from the factory now.

[Sledge Hammer]

2) Devise a thermostat with suitable hysteresis to keep the thing from continually cycling on and off.

I would let it run continuously.

Depending on the fan, that's anywhere from 13 to 30 watts of extra power being drawn off the system. You may be draining down your battery doing that, depending on what other loading you have.

Looking at the charts, the fan you found uses, depending on size 1-4.5 watts, very little in the big picture considering your headlight at 55-65 watts and your tail light at ??I forgot. I run a pair of lights that pull 110 watts total and they are seldom ever cut off be it idling or traveling which again someone running 100 watts of lights for most of the life of the bike has not had a failure yet. More advantageous to cut your headlight than worry with a fan. When my battery was going dead on a trip back home it was nice to cut the extra lights.

Yep, you're right: if 4 W makes a difference, you're already in trouble! I couldn't get to the ADDA web site from my office n(y)etwork to see the power demand on the AQ1212HB-A71GL-LF. I went by data from other 120X120 mm fans which I've been evaluating recently which are power hogs but which also have twice the airflow of the ADDA. But my point is not you will have to adopt load management measures but that it something that might have to be considered. Fans in this size are all over the map in terms of airflow and power draw. Speaking of airflow, this is where we could use the guidance of a mechanical engineer with a background in heat transfer, because we do not even know what airflow is required for adequate cooling.

[TLRam1]

With a fan you would not have to worry about air flow being stopped or traveling. I would also guess even the slow fans, 2500 RPM would be sufficient, more could always be better.

I worked on 4 bikes today but this fan is something I want to pursue further when I have time to check clearance.

[lragan]

In view of the temperature rise from exposure to direct sunlight, is painting the radiator fins black still a god idea?  Lawrence, so you think the fins would absorb more heat from solar radiation because of the black color?  I can see it radiating better when black, but it looks to me that it will have a greater temp raise from solar exposure if it were black.
As long as the device runs hotter than ambient, it will dissipate heat to the air, but if it is setting in the sun, it may have a starting point for dissipation that is closer to the thermal run away temperature. 
I was going to paint mine black on the V-65, but the more I think bout it, I think I'll leave it alone.

Buddy, I don't think the difference between the dull gray aluminum anodizing and a flat black paint will prove significant -- just my opinion, based on solar absorption tests I ran decades ago on various colors,

Secondly, the thermal runaway hypothesis is just that -- merely conjecture.  May not be true at all.

My advice is the same I gave to Jim Tindall -- if you want it black, paint it black.

[lragan]

All these strings attached and responsibility, Lawrence acts like we're married    

Nothing is ever really "free".

[Charles S Otwell]

BTW Charles did you understand the explanation from the English Major?   

Yes Terry! I'm not actually as dumb as I sound

[Curtis_Valk]

BTW Charles did you understand the explanation from the English Major?   

Yes Terry! I'm not actually as dumb as I sound

Yeah, but you ARE shorter than you look on TV! ha ha ha ha!!!

Curtis

[Charles S Otwell]

Yeah, but you ARE shorter than you look on TV! ha ha ha ha!!!

Here we go again! By the way didn't Dustin Hoffman play in a movie named after you?? Indian movie of some sort ..