Been thinking about this one. When a car with a mid life battery starts up, it's voltage can go up to 14.4 volts. But when cold, after sitting over night, the battery voltage may be at 12.4 volts. Using the scenario in post #9, at 14.4 volts the output would be 3.323 volts. But before starting the engine the voltage (12.4V) will have an output of 2.862 volts. A better way to set a voltage at a given setting and hold it there regardless of the battery voltage would be to use something like a TL431 voltage reference. I'm looking at how to construct such a device now.

[edit] a "Mid-life battery" is an older battery, about half its useful life. A new battery tends to hold 12.6 volts over night and when charging it rarely goes up over 14V. At least that's what I observe daily with my battery and battery voltage monitor.
[end edit]

Depending on the application the voltage reference may or may not be necessary. I would say try without one first.

 

Let's hear what the TS wants to drive and go from there.

Here in Thailand the police have started cracking down on modified exhaust systems on motorbikes. The kids have been putting modified equipment in place of the normal exhaust systems and been racing around towns nosily at nights keeping people awake. Every couple of weeks there is something in the news about police departments nabbing dozens and dozens of these in sweeps and making the kids destroy the modified parts with sledge hammers before being released.

 

Let's hear what the TS wants to drive and go from there.

I agree. I merely interjected an assumption that the TS wants to power a GDO Remote without having to be concerned about the coin cell battery that so many of them use. But it IS just an assumption.

 

To provide the voltage for the door opener remote transmitter the simple and reliable and durable circuit will be a shunt diode voltage divider with the correct number of series diodes as the stable portion. If the coin cell delivers about 3 volts at no load then put four 1N4000 diodes in series and feed it the supply through a resistor to run about 100mA through the diode string. That will give between 2.8 and three volts across the diodes to power the device. If it only needs 1 1/2 volts use just two of the diodes in series. Better than the opamp and since regulators of that voltage are in small packages, much easier to work with.

 

To provide the voltage for the door opener remote transmitter the simple and reliable and durable circuit will be a shunt diode voltage divider with the correct number of series diodes as the stable portion. If the coin cell delivers about 3 volts at no load then put four 1N4000 diodes in series and feed it the supply through a resistor to run about 100mA through the diode string. That will give between 2.8 and three volts across the diodes to power the device. If it only needs 1 1/2 volts use just two of the diodes in series. Better than the opamp and since regulators of that voltage are in small packages, much easier to work with.

Not true.

For small loads, an SMD op amp and resistors could be used and made to fit on the tip of a fingernail. Even the standard-sized ones are only about an inch long and probably half as wide, so not much in the way of bulk there.

The op amp solution also consumes MUCH less power (and in fact the consumption can be driven down to essentially zero with big enough resistors). Even the circuit that I posted earlier uses more than a hundred times less than the diode string you've suggested.

Plus the op amp circuit is arguably just as stable as a diode reference. (Don't believe me? Check it yourself.)

 

Have seen penny-pincher keep a car battery long after it should be replaced. My neighbor regularly uses a battery with a cold charge of 11.7 volts.

 

Been thinking about this one. When a car with a mid life battery starts up, it's voltage can go up to 14.4 volts. But when cold, after sitting over night, the battery voltage may be at 12.4 volts. Using the scenario in post #9, at 14.4 volts the output would be 3.323 volts. But before starting the engine the voltage (12.4V) will have an output of 2.862 volts. A better way to set a voltage at a given setting and hold it there regardless of the battery voltage would be to use something like a TL431 voltage reference. I'm looking at how to construct such a device now.

[edit] a "Mid-life battery" is an older battery, about half its useful life. A new battery tends to hold 12.6 volts over night and when charging it rarely goes up over 14V. At least that's what I observe daily with my battery and battery voltage monitor.
[end edit]

The high and low voltage in your scenario would be acceptable but the load on the motorcycle circuit isn't. I'm inexperienced when it comes to electronics but I know the difference between a dead short and a glow plug. That stated I expected heat from the resistance but I had not considered the amperage that it would draw. I ordered voltage regulators which is what I started with but they were defective.

 

The efficiency does not matter a whole lot because the circuit would only be active while the horn is sounding. But the mechanical durability of the circuit matters a great deal, and being in the right range without any calibration adjustment is also important. And the shunt diode regulation system is overload proof., so a momentary short to ground will do no damage at all. And the 2% tolerance as the supply varies from 8 volts up to 16 volts will have no effect at all, while the divider and opamp is ONLY a stable buffered divider with no regulation ability at all. Those were the other reasons for my suggestion.

 

The efficiency does not matter a whole lot because the circuit would only be active while the horn is sounding. But the mechanical durability of the circuit matters a great deal, and being in the right range without any calibration adjustment is also important. And the shunt diode regulation system is overload proof., so a momentary short to ground will do no damage at all. And the 2% tolerance as the supply varies from 8 volts up to 16 volts will have no effect at all, while the divider and opamp is ONLY a stable buffered divider with no regulation ability at all. Those were the other reasons for my suggestion.

Fair points. Voltage levels in automotive systems can be fairly unstable.

 

Garage door openers built into motorcycles is actually a very popular add-on. Moreso for convenience than anything else. I know I wouldn't want a god awful bulky garage door opener clipped to my bike somewhere and even if it were a little keyfob it's just something else I shouldnt be fumbling around with while riding. Most often they use the brights circuit as a trigger, when flashed a certain number of times the garage door opener is activated. Turn signals would work too I guess as long as you don't use them pulling into your driveway, unless the turn signal you always use pulling in the driveway corresponds to the door you want to open that is.

 

What fun! A thread discussing all kind of ways to supply a not very accurate 3V when a single TO92 part will give a far better result for pennies.

Okay, you might need a capacitor with it, though I have used them without, and never seen a problem

Bob

 

What fun! A thread discussing all kind of ways to supply a not very accurate 3V when a single TO92 part will give a far better result for pennies.

Okay, you might need a capacitor with it, though I have used them without, and never seen a problem

Bob

You seem to be talking about a simple transistor-based regulation scheme. Care to elaborate?

 

What fun! A thread discussing all kind of ways to supply a not very accurate 3V when a single TO92 part will give a far better result for pennies.

Okay, you might need a capacitor with it, though I have used them without, and never seen a problem

Bob

On a motorcycle mechanical durability and electrical ruggedness matter a great deal. Probably that little cheap TO92 package is not short circuit proof. AND this application does not need precise regulation.

 

If I didn't mention it before, I tried my hand at such an application in my truck. Took a 5V adapter that plugs into the lighter socket and modified the output to 3 volts. The modification still works but the fob I used to transmit the open and close command has burned out. Originally the fob used a CR2032 battery, which is why I went with a 3V modification on the adapter. However, and I don't yet know this for a fact, I think the fob may have been susceptible to an over-current situation. Maybe after the weather cools a bit I'll pull it out and investigate it more fully. But like an LED, if you provide too much current it will burn out. Even if you supply more than you should but not enough to burn it out - it could still burn out after a short while. LED's will survive at 30mA (typically) but not last as long as if you supply 20mA. 40mA for a moment will also not burn out an LED, but within a few minutes at that current level it will burn out. I'm thinking my fob may have suffered the same fate.

Fobs are cheap. I can make another attempt at it. MY reason for building a GDO into the automotive circuit is because when you have a fob sitting in plain sight someone may break in and steal it. Along with automotive documents kept in the glove box. Then they have an address and a way into the house. Something built in isn't so quickly recognized or stolen. And keeping the fob out of sight is an inconvenience when coming home because you have to stop and dig it out. So maybe I'll go after my issue in the next few days. They're saying Monday and beyond is looking like temperatures that do not reach into the 90's. So we'll see.

The question remains: Are you (the TS) building into your bike a GDO that you can operate without having to dig into a pocket or some other location? Knowing exactly what you're doing will help us provide the circuitry help you seek.