Motorcycle Voltage Regulator ---- AC regulation

Thread Starter

abuhafss

Joined Aug 17, 2010
205
Hi

I have been working on 4-pin Voltage Regulator (Single Phase) for Honda CG125. After trial & errors I learnt that, that model has dual coils in the stator, both having one end grounded to the chassis. 12-50VAC (depending upon the RPM) from one coil is fed into the Regulator Unit which converts it into 14.4V DC for charging the battery. Similary 12-50VAC from the other coil is also fed into the Regulator Unit but, it is also connected to Head Lamp & Tail Lamp via switch. The Regulator Unit have 2 circuits, one converts to 14.4VDC and the other regulates to 12-14VAC for the Lamps.
4-pin Regulator Wiring.png
20161130_212101.jpg

I ripped a Chinese Regulator unit and found circuit as below:
20161125_175249.jpg
2016-11-27.png
This design does not provides regulated charging voltage to the battery.

I also found a similar circuit on the net as below:
4PinRegulator110cc.jpg

I am unable to simulate either of these circuits.
However, I came across an old Thread on AAC, http://forum.allaboutcircuits.com/threads/12v-ac-motorcycle-regulator.39419/page-8
The circuit posted by Sgt.Wookie at #153 perfectly regulates 12-50VAC to 12-14VAC, but it requires a huge 25A bridge. I also attach the spice file for this circuit.

MY CONCERNS:
1) Why the other two circuits do not simulate? I have used LT Spice and Falstad.
2) Alternatively, I would like some other circuit with minimum number of components.

Thanks in advance.
 

Attachments

AlbertHall

Joined Jun 4, 2014
11,159
The circuit posted by Sgt.Wookie at #153 perfectly regulates 12-50VAC to 12-14VAC, but it requires a huge 25A bridge.
It only needs a 25A bridge if your alternator is capable of producing 25A which I seriously doubt. That circuit also dissipates lots of power in the transistor while the thyristor versions keep heat to a minimum.
 

Alec_t

Joined Sep 17, 2013
11,988
I am unable to simulate either of these circuits.
For LTspice try
1) modelling the generator as an AC voltage source with a finite series resistance (say 2 Ω),
2) adding a load across the SCR (say 6Ω).
3) adding the directive ".options gmin=1e-7".
 

Thread Starter

abuhafss

Joined Aug 17, 2010
205
It only needs a 25A bridge if your alternator is capable of producing 25A which I seriously doubt. That circuit also dissipates lots of power in the transistor while the thyristor versions keep heat to a minimum.
I had tried physically with 4 x 1N4007 (considering that the Chinese circuit used 0.5A bridge). Two of them were blown. So I have to try 5A or 6A but again 4 x 6A diodes would occupy big space. And yes, I had already replaced the transistor with BT151.
 

Thread Starter

abuhafss

Joined Aug 17, 2010
205
For LTspice try
1) modelling the generator as an AC voltage source with a finite series resistance (say 2 Ω),
2) adding a load across the SCR (say 6Ω).
3) adding the directive ".options gmin=1e-7".
No. 1 & 3 worked. No. 2 drops the voltage further.
But the output remains constant, altering the zener value has no effect.
 

Dodgydave

Joined Jun 22, 2012
9,923
The scr with the zener sets the minimum voltage, once the thyristor has fired it will continue to conduct,, until the alternator output is zero, so the battery over-charges, your better using a mosfet and aTl431 variable zener to set the voltage.

Ideally you need to know how your alternator is wired, then you can design one.
 

Thread Starter

abuhafss

Joined Aug 17, 2010
205
The scr with the zener sets the minimum voltage, once the thyristor has fired it will continue to conduct,, until the alternator output is zero, so the battery over-charges, your better using a mosfet and aTl431 variable zener to set the voltage.

Ideally you need to know how your alternator is wired, then you can design one.
The Regulator Unit or the Charger Unit comprises of two parts, as I described in my first post. Please see the first diagram in that post.
The first part (the yellow box) converts the AC voltage from the generator to regulated 14.4VDC to charge the battery.
The second part (the green box) converts the AC voltage (12-50V) to regulated AC voltage (12-14V) for Lamps only.
In this thread, my focus is only on the the second part i.e. regulation of the AC voltage.
 

Thread Starter

abuhafss

Joined Aug 17, 2010
205
Here is the snapshot of the simulation. The maximum regulated voltage for the LAMP OUTPUT should be 14V.

2016-12-02 (1).png
 
Last edited:

Alec_t

Joined Sep 17, 2013
11,988
But the output remains constant, altering the zener value has no effect.
Not so when I tried a sim of the circuit. Part of the negative half-cycle gets shunted to ground. The timing/size of that part depends on the zener voltage.
The maximum voltage for the LAMP OUTPUT should be 14V.
I think the RMS voltage is more important at the sort of rpm which will give a generator output greater than 12V.
Use of a constant current load (I2) is unrealistic and can give strange sim effects.
 
Last edited:

Thread Starter

abuhafss

Joined Aug 17, 2010
205
Not so when I tried a sim of the circuit. Part of the negative half-cycle gets shunted to ground. The timing/size of that part depends on the zener voltage.
I think the RMS voltage is more important at the sort of rpm which will give a generator output greater than 12V.
Use of a constant current load (I2) is unrealistic and can give strange sim effects.
Would you please share the schematic of your sim?
Even if I replace I2 with a load resistor, same results :confused:
 

Thread Starter

abuhafss

Joined Aug 17, 2010
205
Here you go
Hmm, now I got it. But I don't think this regulation is giving any sort of protection to the Lamps; correct me if I am wrong.
Frankly, I was expecting the complete cycle to be regulated as sim by Sgt.Wookie's design which is supplying constant regulated AC voltage.

2016-12-03 (1).png
 

Thread Starter

abuhafss

Joined Aug 17, 2010
205
Modifying the Sgt.Wookie's design, I could get following results:

2016-12-03 (6).png
2016-12-03 (3).png
But I could not get why the power dissipation is increasing after 8sec.
Shall appreciate valued comments for optimizing this design. And suggest me suitable triac in TO-220 package.
 
Last edited:

Thread Starter

abuhafss

Joined Aug 17, 2010
205
I assembled the above circuit using BTA16, 0.5W zener and 120Ω resistor. The triac became hot and the zener became warm in 30 sec. To avoid any damage I power off the bike. Then, using the OEM regulator, I took measurements at point indicated by blue arrow:

AC Regulator.png

Idle = 8VAC, 0.5A
6000 RPM = 12VAC, 2.45A

Idle+Lamp ON = 6.5VAC, 0.5A
6000 RPM + Lamp ON = 12.3VAC, 1.70A

Observing the measurements, any ideas why the BTA16 is getting hot?
 

Dodgydave

Joined Jun 22, 2012
9,923
The Triac is shorting out the AC supply for regulation, so it will get hot, looks to me like a stupid design.
 
Last edited:

Alec_t

Joined Sep 17, 2013
11,988
Any shunt regulator will involve components getting hot, since the shunt action has to dump excess generated power somewhere. You could include a power resistor in series with the triac in your design, so that the dumped power is shared between the resistor and the triac.
 
Top