regulator

[Johnr]

New Member

I came across your site as I am trying to build a regulator for my motorcycle. I have installed an AGM battery that requires a higher voltage to charge up. The regulator I have, puts out about 13 volts but it would seem I need about 14.4 volts. The regulator has a sensing wire which when connected to a 12v source on the bike, feeds about 13 volts to the battery. The sensing wire does not appear to be carrying any current, as when I put an ammeter in the circuit, does not appear to show anything. Can anyone suggest a circuit to help. When the sensing wire is disconnected, the voltage, from the regulator, goes to about 18 volts. Any help would be appreciated.

I have attached a file, from this forum but it is used to provide a lower voltage than I need.

Thanks

John

 

[Gary B]

Personally, I would just connect a diode or two in series with the sensing wire. You will probably have to bypass the diode(s) with a large resister (1K or so) to prevent it from becoming a peak detector. The voltage drop will fool the regulator into thinking the voltage at the battery is lower than it really is.

 

[Johnr]

Hi Gary

Sounds like a good idea, I will try it out. Is there anyway of stabilising the voltage (sensing wire) to the regulator so I get a consistent 14.4 volts out. This would be ideal as the alternator voltage varies according to the rpm of the engine.

Thanks

John

 

[MikeMl]

Lot of useful info here and here

 

[Gary B]

It is already regulated, that is why the diodes. Silicon has a forward voltage drop of .7 volt. 2 diodes + the 13= 14.4 volt charging voltage.

 

[Johnr]

Hi Mike

Some very interesting reading there thanks for that. Gary I will try the two diodes and see what I get. Post the results when I`ve done it

Thanks
John

 

[Johnr]

Hi Gary

I tried the 2 diodes with a resistor and at idle it is about 14.4 volts. As the revs rise however the voltage drops to about 14.15 volts. I will try with 1 diode to see what happens. Is there anyway to regulate the circuit as per my first post attachment, but produce a steady 14.4 volts

thanks john

 

[MikeMl]

Is the existing regulator sealed so you cant get into it? It may have an internal trimpot.

Another idea is to temporarily connect a 5K or 10K pot, ends across the battery, wiper to the sense wire running toward the regulator. Setting the pot with the wiper near the Bat+ end should let you select a running voltage that will charge your AGM.

You cannot make this be a permanent fix, because the current through the pot would discharge the battery when the bike is parked. The trick would be to find a source of battery voltage which is switched via the key (ignition) switch.

 

[Johnr]

Hi Mike

I have just put a 5k variable resistor between the battery and the regulator. The regulator is sealed, so I can`t get access to it. I took the feed wire from a relay, so that when the engine is off the power supply to the sensing wire is disconnected. The voltage is still variable going from the rectifier to the battery. The range seems to be about 14.1 to 14.5 volts. I will find the setting, voltage that needs to be supplied to the sensing wire, to produce an output of 14.4volts. I would then need a method of producing this voltage from an unstable 18 volt supply that the rectifier/ regulator would like to produce, sensing wire disconnected

Thanks

John

 

[MikeMl]

I was thinking that you just need to feed the regulator k*(Battery Voltage), where k is like 0.9; hence the voltage divider. The regulator effectively multiplies it's output by 1/k, or 1.11, which is what you want.

You do not want an external voltage reference like you are proposing. The existing regulator already has a built in voltage reference.

 

[Gary B]

That is weird. I would have expected it to go the other way. I told you to bypass the diodes with a resistor. I suspect the problem is the bypass resistor is the wrong value. Also, to raise the voltage at the battery, you would add another diode rather than remove one. If you don’t idle it much, you can get by with a slightly high charging voltage at idle and not worry about damaging the battery.

One other thought, I don’t know the rating of the diodes you used, but I would suggest 1N4001 to 1N4007 as you may be able to find. They are rated one amp and there may be more current in that sensor wire than expected. It could be the diodes you are using are changing their forward voltage drop with the change in current.

 

[Johnr]

Hi Gary / Mike

Thanks for the reply. The regulator with sensing wire disconnected puts out somewhere in the region of 18 volts. This will attempt to fry the battery and electronics. With the resistors connected, at 4k rpm I can adjust to get 14.4 volts but at idle this rises to 15 volts. Seems that what ever voltage you supply to the regulator it will put out about this voltage. If you connect the battery directly to the regulator it puts out 14.35 at idle but only 14.1 at 4K, too low to charge. By adding resistors to the 18 volts I can get the 14.4 but only at specific rpm levels. From 18 volts unregulated I would like a constant 14.4, in an ideal world that is. There does not appear to be much if any current flowing through the sensing wire, couldn`t get an amperage reading with a digital meter. Open to any suggestions

Thanks

John

 

[Gary B]

The diodes are the important part; they act as a low voltage zener. That is where the regulation comes from.

 

[MikeMl]

Look at it like this: originally, this regulator was designed to hold the voltage across the battery at 13.8 to 14.2V. If it now goes to 18V; something inside it has failed.

Connecting a diode or resistor externally is not going to fix the failed regulator. You either have to replace it or discard the failed one and homebrew a new one.

 

[Gary B]

He got the 18V when he disconnected the sensor wire.

 

[MikeMl]

He got the 18V when he disconnected the sensor wire.

I forgot that. So any technique that makes the regulator see a lower voltage on the sense wire should trick the regulator into outputting a higher voltage.

Ok, since he measured the current in the sense wire and it was very small, temporarily connect a 1.5V AA cell in series with the sense wire, plus pole toward the battery. That would subtract 1.5V from the actual battery voltage, causing the regulator to output 1.5V more than it normally does.

 

[patatas64]

Hi, just want to ask how can i directly sense voltage and its possible schematic. im building a load controller and i need to sense the voltage from the generator and feed it to my controller as feedback. thanks...

 

[MikeMl]

Hi, just want to ask how can i directly sense voltage and its possible schematic. im building a load controller and i need to sense the voltage from the generator and feed it to my controller as feedback. thanks...

Two resistors connected as a voltage divider? For example, if the regulator were to be based on a TL431/LM431, the internal voltage reference is 2.495V, so if you wanted to set your output voltage to 14.25V, you would make a voltage divider that turns 14.25 into 2.495V. If the output voltage is too low, say 14.00V, the output of the divider would be low by the same ratio, i.e. 14.000*2.495/14.250 = 2.451V, which due to the gain of the control loop cause the output to increase the output.