Rectifier/Regulator replacement list

[posplayr]

FH012 Benefits

While we EE's could debate the relative merits of different regulator approaches till the cows come home, I will try and summarize why this FET technology really represents a break through for the average GS owner.

The topology of the FET approach is significantly different from the other R/R configurations in that it combines two of the primary functions of the classic shunt R/R into a more sophisticated function which switches FET's on and off to achieve the same function. Because of this direct control of the power waveforms, higher efficiency and regulation can be achieved.

The inexpensive power generation approach used on many of these early motorcycles used a simple permanent magnet generator. The faster the motor spins the more power is generated. There is no way to control the magnetic Field in the generator as is done on a more sophisticated generator or alternator. Feild control limits the output power; without it you have to dump generator power somewhere.

So the problem that falls to the R/R is to convert all of this energy into DC and get rid of the excess that the battery and charging system can't use. The approach is rather brute force. The stator output legs are "crow bar"ed. That is essentially directly shorting the power leads coming from the stator. Ouch; as you can imagine things get hot, both the R/R and the stator.

So to summarize why the FET approach is potentially better I'll elaborate on some practical points for the layman below.

[Point #1 and #2 was validated by showing a reduced dropout voltage, increased output voltage]

1.) The FET approach doesn't implement a classic full wave diode bridge rectifier. So if the old regulators are producing 10 amps of current, then there is about 2x2Vx10amps=40 watts being dissipated in the old R/R alone because there are always two diode legs in saturation (e.g. 2V drop).
The FET style drops this in 1/2 because the lower leg of the effective diode bridge is implemented with FET's which essentially have no forward drop in comparison. Both side could have been implemented in the FH012 but they chose to only do one side.

2.) Because the FET regulator is not dropping a full +4 volts across the discrete diode bridge, there is more voltage available to the output and a higher output voltage at minimum RPM can be achieved (say 14.25V v.s. 12.8V for the OEM R/R).

Point #3 is true, but it also appears true for both the Honda unit as well as a CBR linear regulator. The OEM unit appears unique here in that it does crappy closed look regulation

3.) Because the regulation is implemented using FET switches as well, the sensing function is separated from the electrical control and so more precise regulation can occur (no over charging to above 14.5 V like the stock GS which can go to 15.5V as per the manual). The FH012 is specified to 14.5 +/- .3 v max. This is much better than the OEM.

4.) NIX THIS ADVANTAGE IT DOESNT EXIST:
This last element is a bit of conjecture, but in principle at least, the FET based control doesn't have to shunt current to control the maximum output voltage. It is possible to simply disconnect the stator windings from the R/R output. This should dramatically improve the life of the stator windings !!Whereas the OEM R/R shunts, shorting the stator winding producing very high currents in the stator, creating thermal stress and eventually smoking the windings.
In contrast, the FET approach could simply let the stator go open circuit and no current flows.
I'm going to test and see if I can find any evidence this is how the FH012 is implemented.

The FET approach still appears to shunt, it just does it more precisely.

I know it regulates very well

(as does the Honda)

.

Pos

[posplayr]

Ed

I did a search and found this old thread. This might explain why the CBR regulator does a better job. It is a linear regulator. All the links to the schematics were dead.

I suspect that the CBR still has a full wave rectified diode bridge and is still dumping alot of current to ground. This appears to be perhaps more complicated but provides better regulation. The FET approach still offers the benefit of effectively turning off the stator rather (going open circuit) rather than shorting the legs. I need to verify this of course.

Pos


http://www.thegsresources.com/_forum/showthread.php?p=465431&highlight=regulator+schema tic#post465431

Hey gremlin,
Thank's for the pat on the back, though I'm not the only guy out there working on this old bike problem, I do like the positive comments! I don't really have the skill's at electronic's that Cletus does and what would take him 5 minutes to figure out probably would take me a week... but, I have been hard at this thing for quite a while now. Cletus is the real pro, I just tinker and test and hope I don't burn up to many parts! My reg has been in the bike and working perfect for maybe a year now??, I'm not sure really how long it's been, but it's still working fine. Maybe I have it right, that would be shocking! In any case, there are many different way's to tackle this problem, as many as there are opinion's on the subject. And Rick, if you are still following this thread, I would like to hear more about your system! I looked at building a better SCR shunt type reg at the start of this and after getting my hand's on the schematic's for the Kawasaki, Suzuki and Honda R/R's, I started out that way. There are 2 way's they work. Some regulate by triggering the SCR's with the AC and some, like the Honda's trigger the SCR's with a "sense" voltage from the battery or any other B+ from the bike. The SCR shunt type are all VERY simple and compact and have the least amount of part's, they also from a production point of view have the CHEAPEST part's as well. It was after looking at the schematic for a Honda CBX R/R that got me thinking about something maybe better. The CBX reg is totaly different than the other Honda reg's. I don't know why and my buddy that own's 2 CBX's had no clue either, but the reg on them is more like a linear regulator. In fact the way you test a CBX regulator is by hooking up 2- 12 volt battery's in series to get 24 volt's going to it and then you hook up a variable load machine to it and load it up! It was at that point that I went with the linear design. The CBX reg is complicated and has allot of pass transistors in it, the more load the more transistors get switched in to handle the output. Linear reg's are bigger though and they have small resistance's in the circuit's that cause's heating and this means increased heat sink size, BUT, on the plus side they do offer near perfect voltage AND current regulation regardless of load, and they have built in safe guards against damage. Short circuits, fried battery's, overload's and even heat have no effect on them, they just shut down till things are back to normal. This protect's the rectifier's and stator from damage as well. As long as the wires from the reg to the stator don't get shorted out, nothing you can think of is going to hurt the system. Unless maybe your bike is on fire, and then all is lost anyway! Anyway, thank's for the positive input and I will post some schematics in the near future.
Keith

[Gumbo76]

Charging System Troubleshooting

Not sure if anyone has posted anything like this up here, but it was VERY helpful to me when I bought my bike... The guy I bought it from told me it needed a new battery cause the one on it wasn't holding the charge... I drove it to the local Autozone and bought a new battery. To make a long story short, come to find out (as I was riding it home) the lights started dimming. I was losing quite a bit of power. Turn out the douche that sold me the bike hooked up the ground of the R/R to the positive battery terminal... Friend the R/R. I used this chart to help me figure out what part of the bike was causing the battery to not get charged... If you're handy with an Ohm/Voltage Meter, you'll be able to troubleshoot ANY charging problem with this...

Charging System Troubleshooting Guide

Hope this helps!!!

[posplayr]

Charts Charts Charts

[Edited Note]
I have updated the plots for the Honda Regulator to reflect a better test setup. In the second set of plots for the HONDA unit, the connections were improves with better spade connections from the stator, a direct wire was run between the R/R + and the battery (avoiding the fuse box) and the sense voltage lead was alligator clipped to the positive battery post. The plots reflect the strong regulation of the Honda unit after the stator output gets above 15V and the device can start to regulate in the 14.3-14.4V range.
All changes are shown as quotes

OK this is not exhaustive but it makes the point I hope. I'm comparing the regulation performance of the FH012AA against a conventional OEM type R/R. This happens to be the HONDA unit but, any R/R that uses SCR control to crow bar the full wave rectifier will probably do about the same (OEM/HONDA/Electrosport). Ed's CBR R/R which is a linear regulator will not do this.

There are 4 plots showing various combinations of stator output and battery voltage for the FH012AA and the Honda. These are taken at idle and at elevated RPM (3-5K RPM) with the same bike everything identical the only change is to switch out the regulator and attach the HONDA sense reference to the battery.

The voltage scale remains the same for each plot at 5 volts per major division. That is about 1 volt for each tic. The numbered arrow tell you where zero volts is. Zero is color coded an typically 1 major division from the bottom of in the middle.

I'll add a description below:

1st plot if of the FH012AA at idle. You can see the blue battery voltage is just below 14V and the square wave which is one of the stator outputs is just under 15V. Using a Fluke DVM I measure the battery output to be 14.25 so the scope is a little off. I did not record the battery output plot at 6000 RPM but I did previously take a picture. It was about 14.45V with the same DVM.




here we see the increased output voltage of the FH012AA as RPM increases to 3K RPM. The blue is a cleaner wavefrom but neither ever go much over 15V same as idle. If u carefully count the timing divisions, you will see that the duty cycle actually dropped at higher RPM indicating that the higher output voltage of the stator for a shorter period of time provides the near constant battery output voltage.

[EDITED TEXT and new pic]:
The next plot shows the Honda regulator at idle. You can see that the stator outputs are at about 14.0 V. In fact depending upon the initial charge to the battery, it would idle anywhere between 12.6 to 13.3V depending upon RPM (1200-1300), my headlight and blinker settings. More current draw would drop the voltage. The duty cycle here looks to be a little under 50% (equal time down as up)

[EDITED TEXT and New Pic]
Now I rev up the engine and you can see that the output stator voltages have increased to about 14.3 at 3000 RPM and did not go over about 14.4 at 6000 RPM. The Duty cycle has dropped from about 50% to 46.7% at teh higher RPM to compensate for the increased current/voltage coming from the stator.

By spec the FH012AA is rated to generate 14.5 +/- 0.3V output voltage. It seems to acheive this by having a lower drop out voltage that can provide 14.25 volts at idle and then effectively regulate at higher RPMs to maintain the spec voltage limit.

Pos

[Matchless]

Jim,
Some very good info you have there! Have you ever tried measuring the current on any AC phase and comparing between the FET and SCR R/R?
This should prove your theory as well.
Excellent stuff!

[rustybronco]

Now I rev up the engine and you can see that the output stator voltages have increased to about 17V and the voltage across the battery has as well 16V. Eye balling it the duty cycle doesn't appear to have changed at all. That causes the increased average current to the battery and the over charging.




By spec the FH012AA is rated to generate 14.5 +/- 0.3V output voltage. It seems to acheive this by having a lower drop out voltage that can provide 14.25 volts at idle and then effectively regulate at higher RPMs to maintain the spec voltage limit.

Pos

It would seem that the regulator section of your Honda R/R is bad, if it isn't able to keep the voltage regulated within its spec'd output voltage.
the measured battery voltage of my sh530-12k is 14.61-14.75V

http://www.thegsresources.com/_forum/showpost.php?p=341368&postcount=8

[duaneage]

What is the control voltage measured at the sense wire? The regulator references this voltage.

[rustybronco]

It should be at battery voltage.

These are taken at idle and at elevated RPM (3-5K RPM) with the same bike everything identical the only change is to switch out the regulator and attach the HONDA sense reference to the battery.
Pos

[posplayr]

Sense Wire

[EDITED Text and new Pic]
: Post #49 was edited and updated with new plots for the HONDA unit. Also included here is a plot of the OEM regulator output at idle. All changes shown as quotes.

For comparison here is a plot of the OEM regulator at idle with the same stressing conditions for low RPM and high load. The voltage is only 12.3V. The OEM R/R provides the lowest output voltage under these conditions.

Sense wire was clipped to the + battery terminal with an alligator clip. There were times that it was slipping off and so I did go and reverify it again last nite after posting to see if I confirm the changes or lack of changes to the duty cycle using the meters "measurement" function.

I don't have perfect grounds and there is about as much as 0.5V off offset in my positive leg at 5K RPM.

A further complication was that I was measuring voltages at the R/R connector (where it clips into the OEM harness) rather than at the battery directly. As I recall when I clipped a VOM to the battery directly I was getting voltage fold back so. The voltage at the battery was dropping and after hitting a high 13V peak.

I can also do a measurement with the OEM unit as I now found it.

Rusty: That link sure is a blast from the past.

duaneage, I got the Honda R/R from u a while back.

Matchless: I tried my clamp on current sensor, but it seemed to quite working and I'm not sure what it will really add. We should see basically the same current waveform as teh voltage. I'll check it again

I will see about improving the connections tonite and swap out the OEM regulator as well.

Regardless of the sense wire the lower RPM voltage is improved. I will look to see if i can get any closed loop control action from the Honda R/R or the OEM.

See updated post #49 where the strong regulation at 14.3-14.4V was recorderd for the HONDA unit

Pos

[duaneage]

duaneage, I got the Honda R/R from u a while back.

I'll pay return shipping on that Honda RR if you think it has a problem. I've never seen one that didn't work and I'd like to test one that doesn't.

[posplayr]

duaneage

I'm close to a semi productive test setup. I need to reconcile why there doesnt seem to be any control. I will cripm and solder connections to all of the R/R and then plug them into the female conenctor on the harness. If i need to add a diret connection between R/R + and Battery I guess I can do that there are as well but will try and keep things consistent.

To be honest from all the voltage reading I have done with the OEM unit when it is working, I dont expect excellent regulation from the Honda as well. However this one seems stuck at 50% which is wide open.

More later.

Pos

[rustybronco]

Regardless of the sense wire the lower RPM voltage is improved. I will look to see if i can get any closed loop control auction from the Honda R/R and the OEM.
Pos

I can see the improvement in the low rpm voltage output with the FET style R/R and what should be the improvement, in the current output of the stator at low rpms.
The only consideration I would have is, because of the increased efficiency (lower voltage drop) of the FET R/R, would the excess power not used, be given up as an increase in heat in the stator? and can a 30+ year old stator handle it?

ahhhh, for want of a balanced charging system, or a true alternator....

[posplayr]

Mosfet

I can see the improvement in the low rpm voltage output with the FET style R/R and what should be the improvement, in the current output of the stator at low rpms.

Rusty,
I'm an EE but not a power circuits guy. I spoke with a couple of more knowledgeable guys from work yesterday.

There are a couple more twists associated with this FET design. The FETs' are probably actually MOSFET's and they have something called a "Body diode" that is part of the MOSFET operation. So full wave rectification will occur even if the FET is turned off. At low voltages, the output can be increased by shorting the lower body diode when the MOSFET is turned on.

As far as higher power operation, the SCR control is less precise and the MOSFET control is much more absolute. Presumably to get the +/-0.3V tolerance they have a more precise voltage reference than the typical zener/SCR control.

The conclusion was that the FET style should be more precise.

Yes I had considered the increased power drop across the stator coils.

What can I say? "watta u want a boiled battery or a fried stator?

Best to only do this with a stator with fresh insulation .

Pos

[rustybronco]

Rusty,
I'm an EE but not a power circuits guy.

There are a couple more twists associated with this FET design. The FETs' are probably actually MOSFET's and they have something called a "Body diode" that is part of the MOSFET operation. So full wave rectification will occur even if the FET is turned off. At low voltages, the output can be increased by shorting the lower body diode when the MOSFET is turned on.

Pos

You have more than I have. I sell auto parts.
I will now go and refresh my memory on MOSFETs. specifically "body diode".
Thanks for all the information you have provided and I look forward to any additional information you are able to provide.

will you have time to connect an ammeter in your tests?

***EDIT*** I see that you have tried it.***********

Matchless: I tried my clamp on current sensor, but it seemed to quite working and I'm not sure what it will really add. We should see basically the same current waveform as teh voltage. I'll check it again.

what I was looking for is, increased current flow at low rpms. (which there should be)

[posplayr]

Rusty

For your convenience

http://en.wikipedia.org/wiki/Power_MOSFET

Body diode
It can be seen in figure 1 that the source metallization connects both the N+ and P implantations, although the operating principle of the MOSFET only requires the source to be connected to the N+ zone. However, if it were, this would result in a floating P zone between the N-doped source and drain, which is equivalent to a NPN transistor with a non-connected base. Under certain conditions (under high drain current, when the on-state drain to source voltage is in the order of some volts), this parasitic NPN transistor would be triggered, making the MOSFET uncontrollable. The connection of the P implantation to the source metallization shorts the base of the parasitic transistor to its emitter (the source of the MOSFET) and thus prevents spurious latching.
This solution, however, creates a diode between the drain (cathode) and the source (anode) of the MOSFET, making it able to block current in only one direction.

will you have time to connect an ammeter in your tests?
***EDIT*** I see that you have tried it.

what I was looking for is, increased current flow at low rpms.

OK I'll see if it still works; it would not be the first time it was not hooked up right. I was a little rusty on the meter and software last nite.