12v regulator rectifier for motorbike PMA

[scriberman]

First post, so drinks are on me.. Just had a quick scan of the forum and looks like a wealth of information , some of it I can even understand. Electronics is not my first discipline but I know just enough to be dangerous.

Long story.
I have a KTM 250 enduro bike that has a split charging system. Center tapped coil with one side being rectified to charge battery and the other regulated AC for the lights. The common mod is to disconect the center earth tap and have one AC output. Because of this increased output the standard rec/reg unit is overloaded. You can buy a dedicated (trailtech) one but they are £50.
I can rectify the output with a simple £1 bridge rectifier but the problem I'm having is finding a 7-10a regulator. 5a is the max reasonable priced one I can source.
SO,
I have a DC voltage of ~60V and need to charge lead acid battery.
Can I run two LM1084's in parallel?
Is there a better way?
Will a fixed 12v reg charge a battery or will I need to set one at 13-13.8v?
It's being regulated before the battery so does it matter if it's a positive or negative regulator.
Sorry for the long post but you can't have too much information

Thanks in advance
Steve.

 

[Resqueline]

The only diagram I was able to see (without paying for it) was very poor but it showed 6 wires coming out of the stator.
Not having seen the wiring diagram (only pic's of the physical mod) I can't say what's done and how it works.
The principle of having a grounded coil for the charging system is very poor, and lifting the ground sounds like a good idea though.
Using linear regulators is not the way to go. Either use the ordinary SCR regulators, or design a switchmode regulator (to take advantage of the high voltage available).
13.8V is neccessary to charge a battery. 12.0V will not do at all.

 

[duke37]

You cannot charge a 12V battery with 12V, you need 13.8 to 14.4.
A LM317T 1A regulator can be set to such a voltage. It will need a power transistor across it to enable high currents to be passed. It will also need a heat sink to dissipate the exess power.

An LM337T is the negative equivalent and can take a npn booster.

I presume that the bike is negative earth, a positive or negative regulator could be used if the supply is floating. You will need to protect against reverse current.

 

[Resqueline]

Linear regulators are never used in conjunction with Permanent Magnet Alternators, for a number of reasons.
1: A PMA delivers a relatively constant current , the voltage being limited by the RPM's. Thus a very high voltage can be generated if the load is small.
2: The power dissipated in a linear regulator is thus potentially extremely high requiring huge heatsinks.
3: This power has to come from somewhere, namely the engine, robbing it of power and increasing the gas consumption.

Therefore you use a pseudo-switchmode regulator that either shorts or opens the stator winding when neccessary.
The easiest (& preferred) method is using thyristors to short out any half-cycle as needed to keep the voltage down.
The current is constant so nothing burns, and the power dissipated in the rectifier/regulator is kept low, saving hp's and gas.

 

[spankey666]

My first post, and on a subject close to my heart I designed and built racing generators and regulators for all levels upto WSB in my last job
a couple of questions, is your ktm a 2 stroke or 4 ?
Understanding how a motorcycle regulator works is key to solving your problem. most modern ones are 3 phase, and the voltage to the battery is controlled by the regulator dumping any excess energy to earth in the form of heat. this is the only way to combat any back emf issues created by switching . . a simple thyristor switching 2 phases will generally fail especialy at high rpm when peak volts produced can exceed 60v and massive back emf spikes. we spent months of research and development getting it right. my advice to you in this situation, buy the one that works. not very helpful i know, but it will be cheaper than replacing cooked batteries and / or stator windings, which believe me will fry in the twinkling of an eye if you get it slightly wrong.

I'll get my coat

 

[scriberman]

Thanks for the replies, a couple of different views? The bike is a 2 stroke modified to a 2 wire AC output, estimate 100watt total. I don't think the stator will be a problem, it can take Continuous shorting. Any thoughts on going old school and running a zenner diode to dump excess voltage?

 

[Resqueline]

A zener would be the absolutely simplest but worst possible solution. It would require a huge heatsink and I'd consider it unreliable.
I don't see why a (properly dimensioned) thyristor regulator should fail, even at high RPM's. It doesn't dump the "excess energy" as heat either.
The advantages of a shorting (shunt) regulator is good charging at low RPM and the absence of spikes & high voltages, the disadvantage is a slight heating at all times.
The advantages of a disconnecting (series) regulator is no heating when the battery is fully charged, the disadvantage is high voltages & spikes, + poor low-RPM charge.
A modern switch-mode regulator could take advantage of the high voltage possible at high RPM's to produce 500W (for example) worth of charging, with no drawbacks.
Did you connect the two windings in series?
How much (short-circuit) current and (open-circuit) voltage does it produce at different RPM's?
At what RPM do you spend most of the time?

 

[scriberman]

After doing some more research and from previous experience every regulator I've found or worked on shorts the output to regulate the voltage.The old English bikes used a zenner and the Japanese use a zener to trigger a thyristor.
So, Two more questions. Regulate then rectify or rectify then regulate?
Any ideas on the simplest , bulletproof, agricultural way to achieve a DC 13.8 volt output from a 70v AC Input? Thanks in advance,
Steve.

 

[spankey666]

this works

the battery needs to be above 8.5v to make the reg work, the 9v supply (can be 12v battery voltage )should be switched to the battery as there is a 30ma constant drain
the pot is a 1k trimmer that you set to 14.4v output accross the battery at max rpm. you may want to put a 45v varistor accross both scr's main leads to kill any nasty spikes

 

[scriberman]

Cheers Spankey, I'm also work in a racing enviroment (F1)and know the way things don't want to live or work in a supposedly simple situation. Nice to have a proven design.