Rectifier/Regulator replacement list

[rustybronco]

Rusty thanks for pointing out the issue with the earlier readings. I understand better how these R/R's work and what they can and can't do for charging...

Pos,
I was thinking that because of the lower resistance of the newer components, more current would be available to use at higher RPMs. I suspect there would be at least a 10% improvement in usable wattage.
sorry, I should have made myself clearer as to what I had intended.
Dale

[duaneage]

duaneage: Thanks for providing such a reliable and cost effective solution (Honda Regs). For probably 90% of the GS'ers here the Honda reg is the way to go. Some people are always trying to do better and so now we have the SH012AA .

Pos

Thanks for your support, Pos.

The vast majority of my customers are people without soldering or wiring skills who just want their bike fixed. Many don't want to wait for an auction to end and then wait longer for a RR to show up they then have to modify themselves. Add to that the heatshrink, connectors, wire, etc and by the time they are done they spent almost 40 dollars already.

I ship same day or next and back up the RRs with a swap out if the rider thinks it doesn't work. I worked with three people last year to help them determine what else was causing problems with their bike by sending them a second RR. In each time the Honda was not the problem. Try that with an ebayer.

I'm happy to provide this quick RR service as long as I can still get stock and people want them. Maybe some day all GS bikes will be updated then I can retire.

Do you mean not everyone has a Fluke 97 digital portable O-scope in their garage? I consider it the ultimate tool for electrical.

[posplayr]

duaneage

Do you mean not everyone has a Fluke 97 digital portable O-scope in their garage? I consider it the ultimate tool for electrical.

I was referring to this stuff which is avaliable for pretty cheap on ebay now, but I bought it all new back in 1998 for about $4K. I have the Fluke 97 as well. That was about $120 I think.

Scope

http://www.trs-rentelco.com/Model/TEK_THS730A.aspx

Software

http://www.testmart.com/search/compare.cfm/SOFWAR/TEK/WSTRO/1.html?ad_=3697

Current Clamp

http://www.testpath.com/Items/ACDC-Clamp-on-Current-Probe-for-DMMs-and-Oscilloscopes-100kHz-100A-117-283.htm

[Billy Ricks]

Thanks for your support, Pos.

The vast majority of my customers are people without soldering or wiring skills who just want their bike fixed. Many don't want to wait for an auction to end and then wait longer for a RR to show up they then have to modify themselves. Add to that the heatshrink, connectors, wire, etc and by the time they are done they spent almost 40 dollars already.

I ship same day or next and back up the RRs with a swap out if the rider thinks it doesn't work. I worked with three people last year to help them determine what else was causing problems with their bike by sending them a second RR. In each time the Honda was not the problem. Try that with an ebayer.

I'm happy to provide this quick RR service as long as I can still get stock and people want them. Maybe some day all GS bikes will be updated then I can retire.

Do you mean not everyone has a Fluke 97 digital portable O-scope in their garage? I consider it the ultimate tool for electrical.

I applaud you for the service you provide Duane, for all the reasons you just stated. People that can't troubleshoot automatically think R/R failure. That's why we get so many of the same questions over and over again. For quite a while I got tired of all the redundant questions. I think I may be getting over that and might start posting more answers in the tech section.

[posplayr]

Rusty

Well it might not be obvious, but all things being equal if there is less power drop in the R/R then there is more available to go to the system or to charge the battery. But therein lies the dilemma, you would not have the same load so everything will not be equal.

This compounded by the fact the R/R's doesn't worry about controlling power to the load but rather limiting the voltage at the battery. So if functioning properly and in the voltage range (above drop out), the same battery voltages should be attained and so the same current to the loads. But the one with the larger diode drop will still get hotter. This is one of those times you just have to conclude that there is more charging capacity available because you know the voltage drops are smaller.


Speaking of diode drops. I went back and looked at the data to see if I could figure out the waveforms. Basically when the stator voltage is high, the difference between the stator and the battery is the forward biased upper leg of the diode bridge. When the stator voltage swings down you are looking at the negative of the lower leg forward bias voltage.


I put the attached table together that shows that there is only about 1 volt drop in the upper leg and essentially none in the lower FET side of the FH012AA R/R.


The Honda is more typically 1.7V in both upper and lower legs. So at least in theory that would account for 3.4V of drop (v.s. 1V) in power dissipation which is three times the FH012AA.

While teh OEM seems to have the same type of diode bridge, why the OEM gets so much hotter than the Honda is a mystery although we suspect a case of 'crappy design". I guess I should measure drops at higher load again.


OK where is that scope again.....


Basically the table proves that we are comparing full wave diode bridge rectifier in the Honda and the OEM R/Rs' and a "low voltage drop upper diode with power MOSFET lower diode" bridge.

Pos

[posplayr]

OK to WRAP THIS UP

I got my current amp to work again. Watta you know the 9 volt battery died after just 10 years.

Here is the test setup.



Looking down into the battery compartment; shows the current clamp, and the timing light with RPM readout to control the charging. I moved the current clamp between total current supplied by the RR (about 10-12 Amps) and the battery charging current (smalll current -1 to +2 amps)



The Honda and the OEM regulator modified with fresh crimp connector and soldered to prevent corrosion.




The installed FH0112AA; mounted upside down. I still need to RTV those connectors.

continued.....

[posplayr]

Current Plots

Here are some selected shots of the FH012AA only. This graph is at idle with showing the blue AC current coming from one of the stator legs. The Yellow is the voltage relative to ground of the stator leg. You can see here something just under 15V and accounting for 1 volt drop from the diodes the output voltage was about 13.8V.




Here is the same thing (idling) but looking at the total current coming from the R/R. You get about 10Amps, the RR is not regulating and so you only see the harmonics associated with a 3 phase full wave rectifier. The label says 2 ADC but that is wrong. The scale is 5 amps per major division.



The charging current moving from the R/R to the battery is basically the same but can be adjusted up and down (positive and negative) by the idle speed; it is near zero and I could only get it up to 2 amps at 3K RPM.

Here is where it gets ugly and you see the effect of the "crow bar" control at 3K RPM. They basically look the same except there is about 10 of DC current in the R/R current. The charging current is 2 amps in this graph as measured by my averaging scope.





Not sure this ads alot but it confirms what we already knew. It is interesting that the R/R produces 10 amps total. That has been my guess for a long time .

Pos

[rustybronco]

It is interesting that the R/R produces 10 amps total. That has been my guess for a long time .
Pos

That information is what I have been looking for.

all that I have read, the stock stator is around 275-290 watts output.

out of your FET regulator, you have 14.0v@ 10 amps=140 watts which would make sense, because of the 42%-55% efficiency in the shunt style R/R. the output current is made worse (the greater the difference, the lower the efficiency) because of the very high input to low output voltage.

changing the stator to a larger diameter wire with less turns, may help increase the efficiency of a "crow bar" style R/R, by bringing the input voltage closer to the output voltage. combining a "rewound" stator, with this FET regulator, may show improvement in the available current output.
but it will "probably" sacrifice low speed charging system output. how much it will sacrifice is unknown at this time.

since most of the running of our bikes is in the 4-6,000 range, it just might make a nice trade off.

your thoughts Pos, or am I off base here? (go easy, I'm just a parts guy)

***edit***removed incorrect information. see this post... http://www.thegsresources.com/_forum...0&postcount=85

Actually I measured about 12.5 maximum, so the power delivered to the load is 14.5V@12.5 amps (at 4000 RPM) = 181 watts.

So for the Honda or OEM we should get a total power dissipation of:

P_loss_honda = 3*(2/pi)*15*(1.7/2+1.7/2) = 48.7 watts
P_loss_FH012 = 3*(2/pi)*15*(1/2 + 0/2) = 14.3 watts

[duaneage]

http://www.radio-electronics.com/info/circuits/scr_overvoltage_crowbar/scr_over_voltage_crowbar.php

I thought this would come in handy for those that are wondering what a crowbar circuit looks like. Usually they are used for overvoltage protection, the same principle can be applied to regulation if the sense voltage is applied to the scr trigger input and clamped by external transistors.

[posplayr]

Rusty

That information is what I have been looking for.
all that I have read, the stock stator is around 275-290 watts output.

out of your FET regulator, you have 14.0v@ 10 amps=140 watts which would make sense, because of the 42%-55% efficiency in the shunt style R/R. the output current is made worse (the greater the difference, the lower the efficiency) because of the very high input to low output voltage.

Actually I measured about 12.5 maximum, so the power delivered to the load is 14.5V@12.5 amps (at 4000 RPM) = 181 watts.

The loss of the regulator can be approximated by knowing the stator current and the bridges voltage losses.

The voltage drops are assumed to be relatively constant during a 1/2 cycle of the power signal so the R/R power loss is 3 times the average power loss for both upper and lower diode legs.

P loss = 3 * (2/pi) * I_stator_peak( 50% * VD_upper + 50%* VD_lower)

(loss of the R/R)

The (2/pi) I_stator_peak gives the average current during a 1/2 cycle of the stator current wave form (sine wave)

The diodes are only conducting 1/2 the time so there is an average power calculation

The leading 3 is to account for the three phases.

So for the Honda or OEM we should get a total power dissipation of:

P_loss_honda = 3*(2/pi)*15*(1.7/2+1.7/2) = 48.7 watts
P_loss_FH012 = 3*(2/pi)*15*(1/2 + 0/2) = 14.3 watts

changing the stator to a larger diameter wire with less turns, may help increase the efficiency of a "crow bar" style R/R, by bringing the input voltage closer to the output voltage. combining a "rewound" stator, with this FET regulator, may show improvement in the available current output.
but it will "probably" sacrifice low speed charging system output. how much it will sacrifice is unknown at this time.

since most of the running of our bikes is in the 4-6,000 range, it just might make a nice trade off.

your thoughts Pos, or am I off base here? (go easy, I'm just a parts guy)

***edit***140 watts ain't jack... and add to it a R/R that is less efficient? whoooweeee!

I had a talk with one of the experienced power guys about this topic and his general comment was tread lightly. The crow bar regulation as ugly as it is works because there is substantial source resistance in the stator. Basically the stator resistance limits the current when you short it. If you were to reduce that resistance by 1/2 say, you would increase the current by x2 when crow barring and either the stator or the R/R may not be able to take it. Other than that there are too many factors to generalize in this forum.

Without changing the stator, there is an extra 34 watts avaliable to the load if required so that is about a 20% increase in avaliable output power.

Pos

[duaneage]

I had a talk with one of the experienced power guys about this topic and his general comment was tread lightly. The crow bar regulation as ugly as it is works because there is substantial source resistance in the stator. Basically the stator resistance limits the current when you short it. If you were to reduce that resistance by 1/2 say, you would increase the current by x2 when crow barring and either the stator or the R/R may not be able to take it. Other than that there are too many factors to generalize in this forum.

Without changing the stator, there is an extra 34 watts avaliable to the load if required so that is about a 20% increase in avaliable output power.

Pos

Like I said before, the EE who designed the RRs took into account the stator and it's unique characteristics.The stator impedance interacts with the Q of the circuit represented by resistance, reactance, phase, voltage, current and frequency. One thing to consider is the Honda stators are electrically very similar to the Suzuki stators. Their regulators are similar as well with the significant difference being the regulator reference and an obviously larger heatsink area. Other than that they are quite compatible.

[posplayr]

Like I said before, the EE who designed the RRs took into account the stator and it's unique characteristics.The stator impedance interacts with the Q of the circuit represented by resistance, reactance, phase, voltage, current and frequency. One thing to consider is the Honda stators are electrically very similar to the Suzuki stators. Their regulators are similar as well with the significant difference being the regulator reference and an obviously larger heatsink area. Other than that they are quite compatible.

I think that this comment is trying to suggest that the Honda regulator is a better match for the OEM stator than the FH012AA. It is using my comment for justification.

The point of my comments was that stators for "crow bar" control systems have to maintain a certain about of resistance in the wire. That means thinner wire and more windings. This is necessary because when the crow bar shorts out the stator coils, the stator and the R/R can fry unless there is sufficent resistance in the stator to limit the current.

The specific point being brought out is to not drop the stator impedance too much as if you do and short it's leads then you can fry it. It has little to nothing to do with a matching of the regulator to the stator. A R/R with crow bar control will short the stator whether it is OEM,HONDA or FH012AA.

Most all of the discussion has centered on a R/R that is efficient and can produce regulation in the face of minimum stator voltages. No where has the issue of MATCHING of R/R characteristics to Stator characteristics been discussed or described other than in duaneage prior post and baseless suggestion.

So in Jim's court I find the following statement as a not so subtle form of "fear mongering" to promote running to Honda regulators so that the last 5% of market share can be grabbed. In the interest of GS science I must protest.

the EE who designed the RRs took into account the stator and it's unique characteristics.The stator impedance interacts with the Q of the circuit represented by resistance, reactance, phase, voltage, current and frequency.

There is nothing mis matched about a FH012AA on a GS stator as far as I can tell.


duaneage

I really was hoping you would not post such comments. Maybe you just don't know any better.

Pos

[Matchless]

Jim,
Just another stone in the bush as we say.
I noticed some wiring diagrams for bikes using the FET R/R that have Delta configured stators, cannot recall the models now off the top of my head. I was wondering if the use of these would now be more "efficient and reliable" together with the FET 50A R/R.
Honda did try a model (VFR) or so with delta wound stators a few years ago and I think the charging on that bike was not very reliable and there were many problems with R/R's at the time and all sorts of fixes.
There are also star "heavy duty" stator rewinds and aftermarket ones advertised as giving say an extra 20% and being more robust etc etc.

Some great analysis work you have done here. I am looking forward to the comparison summary table between the two units!!
Keep up the good work.

[Matchless]

One thing to consider is the Honda stators are electrically very similar to the Suzuki stators. Their regulators are similar as well with the significant difference being the regulator reference and an obviously larger heatsink area. Other than that they are quite compatible.

Duanage,
I see what you are saying and I want to agree with you. Most of these "older" SCR R/R's around are mostly made and designed by Shindengin for Suzuki, Honda, Yamaha, Kawasaki etc and over the years some variants and improvements were made, such as Honda did not equip every bike with a sensing type R/R. Honda also used field regulation in the early 80's already. I saw some diagrams where Suzuki also used a reference R/R.
I would not be surprised if some different models have the exact same components inside them. I would also not expect the same manufacturer to specifically manufacture a less robust one for other manufacturers, but rather provide a proven design as per the rated capacity required.
My personal guess here is that Honda demanded a slightly higher current rated type and as such had a larger safety margin built in and maybe used this down to smaller bikes as well.

I have seen sites where people also bemoan the Honda charging problems. The VFR sites are examples.

Please keep on sharing your wealth of experience with us, as your service to the forum is very well respected and will be used by many until you pull the plug one day. The majority of owners just want a part and the info on how to connect it and is not much interested on the detail we are looking at here.
Keep well and keep adding to this thread!

[Larry D]

Duanage,
Please keep on sharing your wealth of experience with us, as your service to the forum is very well respected and will be used by many until you pull the plug one day. The majority of owners just want a part and the info on how to connect it and is not much interested on the detail we are looking at here.
Keep well and keep adding to this thread!

I am one of those people that just want to know which one to use and how to wire it up. I've been reading this thread since it's beginning and understand next to nothing about the techinical aspects portions. The research, testing and time spent are greatly appreciated.

This is what I get out of this...we possibly have a more technically advanced r/r to use on our electrical challanged bikes. We do however, already have a great alternative with the use of the Honda r/r. It's proven it's effectivness over time.
So, over the past few days I've searched around the net for a FH012AA and found only two. Both were around $80-$100. They can be purchased new and the cheapest I found was about $120. If you can find one for less on the net or new...let me know.

My point is this, why would I spend $80-$100 for a part that I can get for $40 ??

Okay you guys can go back to the technical electrical stuff now.......