This is all great stuff and Paul said something that really caught my ear: namely the use of heat transfer compound to enhance the contact between the electrical components and the heat sink.

For a heat sink to actually work, it must be in intimate contact with the source of heat and that special heat transfer "glue" is an important part of the heat conduction path. Electrocial supply places carry it I believe.

Pete
 
It is also important to maximise the contact between the rectifier and the heatsink. Sheet metal, if too thin, distorts when the rectifier is bolted on and this is probably why the commercial heatsinks are quite thick at the mount point and very smooth. At some stage I would like to try soldering a Copper pad in between my rectifier and the aluminium. I realise I need the correct flux to do this but it offers certain advantages due to the high heat capacity of Copper.
If making your own heatsink, I would not use 1/16" flat plate. Minimum should be 1/8", so that there is some mass, which will help draw the heat. Sure, copper can be used in stead of aluminum, but it is just more costly. If you have an all aluminium heat sink of enough surface area, it will do an excellent job. The 4 mm thickness is very good. Also the length and width should work, no problem at all.

I think for the XS650 rectifier, a minimum surface area would be about 20 square inches, which isn't much. That's only 4.5" by 4.5", and as long as its at least 1/8" thick.
Of course as soon as the heat sink has fins, it gets much smaller in length and width.
When I made up my rectifier, I just happened to have an old heat sink from many years ago. It has about 60 square inches surface area, so it's over kill for sure. More surface area is always better. If you keep the surface area too small, it just means the diodes are going to run at a hotter temperature. With mine at 60 square inches, the diodes will be almost freezing.
 
Last edited:
Each surface of my heat sink stack has a thin film of thermal compound making full contact across the surface. Clearly not as effective as one continuous block, but hopefully sufficient.

Copper is more costly, more difficult to machine, hence not used as often. However, my understanding is that it does have the benefit of soaking up more energy than aluminum, staying cooler with that extra energy, and transfers that energy through its mass more effectively. I love my copper core pots and pans for this reason. Solid aluminum, no matter how thick doesn't transfer the heat up the sides nearly as well.

Eh, whatever, my sink is currently all aluminum. 7/16" thick with some fins for extra surface area.

5twins, I checked it again this morning after a cold startup. 13.5v, no higher.

Took her to work again. After the 4 miles: 4-5sec on the how-long-can-I-touch-it scale for the rectifier. No charge increase at all with added throttle.

It may be Tuesday before I have the chance to do it, but a thorough methodical troubleshooting session along Curly's List is next on the docket.
 
Daniel......................your lack of charging is not related to you rectifier temperature. Your heat sink is large enough.
Back to basics...........brush length, slip rings clean, rotor winding resistance, etc.
 
If you keep the surface area too small, it just means the diodes are going to run at a hotter temperature. With mine at 60 square inches, the diodes will be almost freezing

RG - where you live, damned near everything will be freezing at this time of the year! ;)

Pete (in balmy Windsor....)
 
Au contraire Pete. Actually a nice sunny day at 11 C or 52 F. It will be up to 11 C everyday this week I believe. Sometimes Calgary is
the warmest part of Canada. On the other hand, it can be -30 C here in March. Weather in Calgary is like a box of chocolates, you never know what you're going to get!:lmao:
 
And RG of course you also get the benefit/ detrement of those Chinooks. I so used to enjoy them when I lived in Calgary.
 
Well, I had a short bit of time between shifts and armed with only a VOM decided to take a few measurements. Clearly I need to begin at the basics and check everything, but I was bored and anxious to do something.

Rectifier, tested perfectly on the bench before install, this is what I have now:
  • Red lead on +, Black lead on the AC terminals. ---, ---, ---
  • Black lead on +, Red lead on the AC terminals. ~.450, ~.450, ~.450
  • Red lead on -, Black lead on the AC terminals. .000, .000, .000
  • Black lead on -, Red lead on the AC terminals. .000, .000, .000
Like I said, I'll be starting at square one with the charging system guide. Results to be posted as soon as I do.
 
This is really strange. Is that 2 bad rectifiers or is it 3......................I've lost count. Tested perfect on the bench, and now its failed??????
So what kind of numbers were you seeing on the bench test. I assume you are using the diode test position on the meter. The number 0.450 as a forward voltage drop does not seem correct. It should be more like 0.6XX. I have never used an auto ranging meter, but I have heard that they can be inaccurate at times. Maybe you should borrow a different meter, from a friend or neighbour.
Do you still have the old original rectifier, and was it working when removed? If so, bench test it and then install into the bike.
 
Sorry for your troubles Daniel, you are methodical and have a good aptitude for this type of work. I know you will get to the bottom of this.
Good luck. Bob
 
This is the second modern rectifier to be tried. I don't recall exactly what the readings were, but all six diodes showed forward voltage drop of somewhere near .45x and non-continuity in the reverse. I have another new one, I bought two, and it has the same readings. I do have the original, but it is very corroded. I may be able to borrow another meter from work.

Thanks, Mailman.
 
My original rectifier looked bad too, all caked up with green and white corrosion. I scrubbed it down with a baking soda/water mix and that cleaned it up quite nicely. Then I used it for about the first 4 or 5 years I've had the bike and it worked fine. Give yours a try, it might be OK.

Everyone says the newer modern rectifier is the way to go, that it works a bit more efficiently. I made that change first on my charging system then did some checks. I really didn't see any improvement in the charging output. But, having a new modern part just made sense. The big improvement for me was upgrading the regulator. That really helped.
 
Rectifier, tested perfectly on the bench before install, this is what I have now:
  • Red lead on +, Black lead on the AC terminals. ---, ---, ---
  • Black lead on +, Red lead on the AC terminals. ~.450, ~.450, ~.450
  • Red lead on -, Black lead on the AC terminals. .000, .000, .000
  • Black lead on -, Red lead on the AC terminals. .000, .000, .000

This shows that the top (+ side) 3 diodes are OK,
and the bottom (- side) 3 diodes are shorted.
This is the second modern rectifier to be tried. I don't recall exactly what the readings were, but all six diodes showed forward voltage drop of somewhere near .45x and non-continuity in the reverse...

If "---" is open circuit, non-continuity, non-conduction,
And
".000" is total continuity, dead short, same as shorting the probes together,

Then, your test results should be:
  • Red lead on +, Black lead on the AC terminals. ---, ---, ---
  • Black lead on +, Red lead on the AC terminals. ~.650, ~.650, ~.650
  • Red lead on -, Black lead on the AC terminals. ~.650, ~.650, ~.650
  • Black lead on -, Red lead on the AC terminals. ---, ---, ---
 
I just measured two KBP5006 bridges wired for three phase using the Diode function. I assume the readings must be millivolts? For new unused diodes from China these values seem low, but I did use my cheap meter?:
  • Red lead on +, Black lead on the AC terminals. ---, ---, ---
  • Black lead on +, Red lead on the AC terminals. 412, 410,411
  • Red lead on -, Black lead on the AC terminals. 408, 407, 406
  • Black lead on -, Red lead on the AC terminals. ---, ---, ---

  • Red lead on+, Black Lead on -. ---
  • Black lead on +, Red lead on -. 928
( --- means No Conduction )
 
Last edited:
My original rectifier looked bad too, all caked up with green and white corrosion. I scrubbed it down with a baking soda/water mix and that cleaned it up quite nicely. Then I used it for about the first 4 or 5 years I've had the bike and it worked fine. Give yours a try, it might be OK.

Everyone says the newer modern rectifier is the way to go, that it works a bit more efficiently. I made that change first on my charging system then did some checks. I really didn't see any improvement in the charging output. But, having a new modern part just made sense. The big improvement for me was upgrading the regulator. That really helped.
5twins.....................yes, the original rectifiers sure get grungy with nasty green stuff, but they are very rugged. I still keep my original in my spare parts boxes, because it still works, but I prefer my "modern" one.
 
Back
Top