Rectifier/Regulator replacement list

[Nessism]

the time tested Honda would be the choice for those with stock charging system needs, or with an unknown stator condition.

at this time I believe the FH-series R/R (FET) should be used with a known good stator with no flaking/bad coating, or a properly rewound/new stator. I plan on testing a FH010BA with a 27 year old stock stator on my '82 gs to see if I can stress the stator (make it fry). then combine it with a rewound stator sometime in the future on my '80 gs.

the FH-series, in my opinion, would make a nice choice for those seeking the maximum power out of the stock charging system when used with a proper stator, "for those that need it".

***edited*** I can't speak for ALL the cbr1000rr reg's, yet...

Why would the type of R/R add stress on the charging system? Seems to me it would add less.

[posplayr]

Ed

EDIT: I'm retracting the statements below that are in quates

I was going to explain this in detail, but then realized that the R/R and the load are in series with the stator. So the total power delivered by the stator is pretty much the same independent on how it is distributed between R/R and load.

The prior discussion/assumption was that since the FET R/R did not drop as much voltage, it was putting a bigger load across the stator. I don't think this is true. I'll check but I think it simply means there is more power available (at lower RPM) than what would be available from the lower efficiency SCR R/R.

A counter example, lets say both an SCR R/R and a FET R/R are operating at 5K RPM. Both are delivering the required load current at the same specified output voltage. Then the power drop is the same in either R/R as the stator is doing the same thing and the load is current is the same.

The FET regulator has to dissipate the same power as the SCR regulator.

This is still true. See a later post on overall operation.

The benefit of the FET regulator is that it can achieve this full output at lower RPM than the SCR R/R due to lower losses in the FET R/R.

Pos

[Matchless]

Very Nice write up.

And LOL I thought I had overlooked something

U can always update it Andre. .

Pos

Jim, basically I was first supposed to sit down and trim it down, remove repeats etc, but I have just been too lazy as it was written over a long period. I call it the "longwinded" one. Maybe I will just leave it as it is until I have nothing better to do!
Keep well.

[dennis roy]

So...if i'm reading the list right...a r/r from an 05 kawasaki zx10 would work fine...right?

[Matchless]

So...if i'm reading the list right...a r/r from an 05 kawasaki zx10 would work fine...right?

You have it! In actual fact the majority of R/R's will work as long as the current rating is not lower and it is three phase and not for a field regulated alternator.
If you have one of the FH models consider yourself very fortunate.

[rustybronco]

So...if i'm reading the list right...a r/r from an 05 kawasaki zx10 would work fine...right?

http://cgi.ebay.com/ebaymotors/ws/eB...m=250384316690 that's what I bought

***note*** the bolt hole spacing is wider than the stock R/R by about 3mm (.110).
I would suggest at this time, mounting it with one bolt and modifing your mounting bracket or finding an alternate means of tying down the other mounting hole (zip tie, metal strap, modified bracket?) ***at this time I believe the bolt hole in the FH010BA should not be modified***

if you would like, you can widen the 2nd hole in the mounting bracket, (on the bike) and put a nut, bolt and lock washer through the 2nd hole.

[bakalorz]

<snip a bunch of stuff I agree with>

The benefit of the FET regulator is that it can achieve this full output at lower RPM than the SCR R/R due to lower losses in the FET R/R.

Pos

I don't disagree with this, but wonder how much of a benefit it really ends up being.

At any given RPM, the stator output will have some curve of voltage vs current. (i.e. how "stiff" a voltage source the stator is) The slope of this curve will determine how much extra current you get for the ~1 volt less loss.

My impression is that the stator is not a stiff voltage source at all, meaning that you don't get a whole lot of extra current. (admittedly a guess, but I believe an educated guess)


Another way to look at it is that some of the reduced voltage loss simply gets eaten up in the stator instead of the diodes/fets.

To get an accurate idea of how much better you do with the FETs would require either some pretty accurate testing, or some much more involved mathematical analysis than has been done so far in the thread.
(Specifically you would need to provide a better analysis of the stator as a voltage source, as well as probably considering the compliance of the battery (and probably the loads too ... incandescents have pretty funky voltage/current charachteristics too I believe))

I'm not saying you don't get SOME gain from the fets, but I think in real terms it ends up pretty negligible, and you could do better by turning your idle up by maybe 25 rpm.
And once you get to an rpm where the regulator is starting to shunt, (which is all rpms during actual riding) there's no advantage.

I suspect that the OEMs are using the fet regulators mainly because they move a lot of the heat generated during shunting away from the R/R to the stator, which is better able to handle it.

Martin

[rustybronco]

Another way to look at it is that some of the reduced voltage loss simply gets "eaten up in the stator" instead of the diodes/fets.

I suspect that the OEMs are using the fet regulators mainly because they move a lot of the heat generated during shunting "away from the R/R to the stator", which is better able to handle it.

Martin

with voltages regulated to the same value, plus considering the stator has a fixed resistance and the impedance of the stator should be the same with the "differences in resistance" between the scr and fet r/r's.
would you please explain why?, so I may absorb your answer. (my educational value only)

***added thoughts*** the only factor that would change the power dissipated in the stator would be, if the capacitance of the changed device (R/R) was different from the OEM values, and that would be true with the Honda and FET R/R.

the series load formed by the fixed loads (lights, devices) and the battery in parallel, would decrease as the battery charged up because of it's increased internal resistance nearing a full charge. therefore limiting the current in the circuit.

if the above is correct, that would make the FET style R/R the better choice in R/R's.

[bakalorz]

Another way to look at it is that some of the reduced voltage loss simply gets "eaten up in the stator" instead of the diodes/fets.

I suspect that the OEMs are using the fet regulators mainly because they move a lot of the heat generated during shunting "away from the R/R to the stator", which is better able to handle it.

Martin

with voltages regulated to the same value,
plus considering the stator has a fixed resistance and the impedance of the stator should be the same with the "differences in resistance" between the scr and fet r/r's.
would you please explain why?, so I may absorb your answer. (my educational value only)

***added thoughts*** the only factor that would change the power dissipated in the stator would be, if the capacitance of the changed device (R/R) was different from the OEM values, and that would be true with the Honda and FET R/R.

the series load formed by the fixed loads (lights, devices) and the battery in parallel, would decrease as the battery charged up because of it's increased internal resistance nearing a full charge. therefore limiting the current in the circuit.

if the above is correct, that would make the FET style R/R the better choice in R/R's.

I only have a bit of time during lunch, so I'm not sure I can address all of it today.

First, just to be clear, my two paras you quoted were meant to apply to different aspects of the R/R operation.
The first was primarily to explain why the FET may have less advantage at Low RPM than it would seem to (although I guess it applies during shunting too)
The second was to explain why I suspect the OEMs use the FET type (apparently especially for high current ones)

Ok, let me try to explain my second para better first.
I know how the SCR type work, but prior to this thread have not seen any info on the FET type.
I assume Posplayr's description of operation in post 10 to be correct. If not, then none of this holds ...
(As an aside ... it would be possible to run an FET type in a manner where it doesn't shunt, but instead disconnects the stator leads for a portion of the cycle.
This would reduce currents (and heat) by a LOT. But it would have its own issues (the stator leads would go up to 120 volts at times for instance))

But anyway ...

The R/R has to get rid of any heat it generates.
This heat is equal to V*I losses in the R/R.

During those parts of a cycle where the R/R is rectifying (but not shunting) V will be the "diode" losses.
Where an actual diode is acting as the diode (top and bottom in a standard R/R or the top diode in a FET R/R) this V will be .7 to 1.4 Volts.
Where a FET is acting as the "diode" (the bottom of a FET R/R) this V will be .3 volts or less.
The currents will be similar, perhaps a bit higher for FET type.
(if the current is higher, the R/R will compensate by shunting a bit higher % of the cycle)
So the total loss of a standard R/R is 1.4 to 2.8 volts
The total loss of a FET type is 1.0 to 1.7 volts
So the heat will be significantly less for a FET R/R for this portion of the cycle.

During those parts of a cycle where the R/R is shunting, V loss will be:
For a standard R/R one diode drop (0.7-1.4 volts) plus 1 SCR drop (0.7-1.4 volts) total 1.4 to 2.8 volts
For a FET R/R during shunting it would be 2 FET drops (approx 0.3 volts each) total .6 volts or less
Depending on the stator's V-I charachteristics, current will be slightly to a good bit more.
So the heat generated during shunting is 1/4 to 1/2 as much as a non-FET.

Because the current will likely go up a bit for a FET type, the losses (and heat) in the stator will probably increase a bit.
So the R/R runs a LOT cooler, but the stator a bit higher ... but the stator has an oil bath to help carry away the heat ...

I probably won't be able to get to the other part till mon if you still want me to.

[dennis roy]

thank you Rustybronco. I was just worried that I wouldn't be able to wire it up. I dont mind doing a little "clearancing". I hope it helps my bike keep up with the driving lights I added when I bought it. I'm blind as a bat at night lol.

[rustybronco]

I probably won't be able to get to the other part till mon if you still want me to.

At this point no, it won't be necessary to expand on the 2nd point and I understand what you are saying.

now I have to figure out if there is any advantage to rewinding the stator with a heavier gauge wire (if possible) and seeing if there is a increase in efficiency by bringing down the stator output voltage closer to the regulated voltage and any "possible" increase in current available for those that need the extra for lights, vests ect.

yes I know that a rewind will result in different voltage and current curves per a given rpm. what I am looking to find is, what is the practical limit to power available and is there any advantage to rewinding, other than durability.

Thank you for your input.

[rustybronco]

thank you Rustybronco. I was just worried that I wouldn't be able to wire it up. I dont mind doing a little "clearancing". I hope it helps my bike keep up with the driving lights I added when I bought it. I'm blind as a bat at night lol.

your welcome. my FH010BA is under my counter with the 5 wires attached to it, ready to be installed this weekend.

clearance (I know what you meant) is not the issue, mounting hole spacing is. do not modify the holes in the R/R, modify the bracket where it bolts up. if I can help with anything more, just ask.

my point in all these discussions, has been just for situations like yours.

***EDIT*** What kind of bike?

[posplayr]

GS Charging System Design Overview

The attached figure shows a comparison of the performance to be expected from a FET based shunt R/R (i.e. FH012AA) and a more conventional SCR based shunt R/R when used in conjunction with a GS charging system.

The valves are approximate based on the recent measurements I made on my GS1100ED.


Supply and Demand Curves

First what is important to note is the load demand curve (the red curve). At idle (1250 RPM) the demand curve requires about 10 amps. As RPM increases the demand only goes up about 25% to say 12.5 amps at 6000 RPM. The curves are approximate but are based on measurements made with a DC current probe on the R/R red output wire.

The large initial demand is due to lights and ignitor which are almost constant load, and a speed dependent portion which is most likely the ignition coils. So the generator needs to put out about 10-13 amps and no more and no less.

In the lower pair of curves we see the approximate regulation duty cycles (not to scale) for the two R/R types and when they begin regulation. The FET starts to regulate at a lower RPM because of the lower drops in the R/R allow for a higher voltage to go to the battery. As the output voltage rises , the FET regulator hits 14.5V at a lower RPM and so regulation starts at a lower RPM. SCR R/R regulation is probably delayed till something closer to 3K RPM. There are approximate values YMMV.

Since a permanent magnet generator should have an output voltage which is proportional to RPM, and since we see only a relatively small change in the control duty cycle, (50% to a maximum of 45%), we have to assume that something is limiting the generator output as the R/R only seems to be doing some marginal trimming.

We have to assume that it is a saturation effects in the stator magnetic feild that are limiting the power output at higher RPM. In my earlier comments about “system design elegance”, this is what I was referring to. The stator was designed with the ability to supply only so much current before the magnetic field material saturate. This effectively caps the stator output power even though RPM keeps rising. See the link below fro an explaination of magnetic staturation.


Saturation



The principle at operation here is that the generator supply curve was designed to closely match the demand curve so that the R/R doesn’t have to dump/disapate as much current. The R/R’s responsibility lies primarily with rectification and only does a partial job of regulation.

see quote below from Post #103 in this thread

I am drawn back to a statement that a fellow GS member (I won't guess who it was because I don't remember) made here a while back about the charging system. The statement was something to the effect that "the charging system is not intended to charge the battery". I'm pretty sure that he had heard that conclusion from someone more knowledgeable about the charging as the statement was unqualified or elaborated on. So I'll guess further as to the meaning..

The system level elegance that I speak of is to take the simplest generator design and match it's output to the demand curve for the motorcycle as a function of RPM. Using this approach, the system can be designed simply and is almost fool proof as even without any voltage controls the generator and load are matched. OK so you can't quite do that so you need to have some regulation because the battery is relatively intolerant of over charging. Some form of regulation is required. Unfortunately for Suzuki, the elegance of the systems engineering did not translate into a well executed R/R design. Given the recent heating of my Electrosport it is not a whole lot better. Honda did a better job on their R/R.

Voltage Regulation

The upper curves show what the R/R outputs would be with no regulation. We can see that due to the lower drops in the FET regulator, the R/R produces a higher voltage at idle than the quiescent battery voltage (12.8V). The SCR regulator may or may not be below 12.8 at idle but it is certainly below where the FET R/R would be operating.

Power Dissipation:

I had previously measured voltage drops for the SCR R/R at approx 3.4V total and about 1V total for the FH012AA. So the FET regulator is dissipating about 1/3 as much power as the SCR based R/R. As has been recognized (back tracking on some of my earlier comments) the power not dissipated in the R/R will be dissipated in the stator. However since the stator is already dissipating close to 200W (12.5*14.5= 181W) adding an additional 25 watts is only about a 10% increase. In addition the stator is better equipped to dissipate power albeit through 250 degF hot engine oil. On the other hand reducing power dissipation in the R/R to 1/3 should yield a significant improvement in reliability due to reduction in component junction temperatures.

One still might worry that the FET based R/R is stressing the stator too much due to the hard shorting (near zero voltage drop of the FET). However refereeing back to the magnetic saturation characteristics, the stator is just incapable of increasing the current flow any more than it already is so the likely effect is negligible.


Summary

The revelation here is that the overall operation of stator- R/R-Load system is to match the generator supply curve as close as possible to the loads demand curve. In addition the supply curve should always exceed the demand curve. There are some additional conclusions we can draw:

It is hard to imagine that with the magnetic core saturating as is apparent, that rewinding a stator will ever yield an additional 20% more power. The power output of the generator was designed to be limited so that the shunt control was not over whelmed at high RPM.

The R/Rs’ primary job is to rectify and closed loop operation is actually very limited. The FET regulator supports the role of the R/R in that for the GS, it allows the generator output voltage to exceed the battery voltage at a lower RPM, and it dissipates less power resulting in a more robust solution due to lower junction temperatures.

In contrast the SCR based regulator is at best marginal at idle and a typical GS actually discharges at idle using the SCR R/R. In addition the SCR R/R must dissipate 3 times the heat of the FET design in a very similar package to the FET design. The FH0012AA is actually larger than the typical SCR design.

Conclusion:

So is the FET worth it? Given the limited role the SCR R/R has, it actually does it’s job fairly poorly. R/R’s get hot and can fail and may not charge at low RPM. These problems largely go away with the FET design. The SCR R/R will work, albeit marginally. The FET design does a better job, and in the world of “GS charging compromise”, this really is a break through technology.


Pos

[dennis roy]

it's an '82 850glz. guages off an '84 1100e, hard bags off an '84 1100gk, forward controls, saddle style seat. i'm gonna try to fit the tank off the gk also so i don't have to stop so often for gas.

[dennis roy]

Well now i'm confused. I replaced the r/r today. It charges at about 11 volts at idle and never gets over 12.5 when its revved. Stator bad maybe?