Building a simple DC voltage regulator

[EMCLARK]

Hello, I need a voltage regulator that can take the variable input from a car (12-15 volts) and turn it into a constant 10v that can handle a draw up to 20 amps. I have a basic understanding of this stuff, but math has never been my strong point, these formulas and such easily frustrate me.

Looking around, with that kind of amp rating, it would seem I need to build one myself. Any assistance in this would be much appreciated.. as I said, I have a basic understanding of this stuff, if someone were to provide a simple schematic I could take it from there.

Again, any help would be greatly appreciated.

 

[skimask87]

7810 regulator with a series pass transistor.
Problem you might run into is at 12v input, you'll be running really close to the drop out voltage of the 7810.
So, get an LDO with the same basic spec's and usage.

 

[ronsimpson]

If you use a linear regulator like the 7810+Transistor there will be 2 to 5 volts drop in the regulator. 5V X 20A = 100 watts You will need a large heat sink!

Here is a switching power supply. Sorry it is not simple. It should be 96% efficient so there will be only 8 watts of loss.

 

[EMCLARK]

A step in the right direction

This is what I'm working with (don't judge the application..)

This is a musical air horn system, the compressor is 12v - the problem is, even with the car not running, it's getting more than 12v, which makes it faster than it should be.

In this application the regulator I need would only be working less than 10 seconds at a time. Does this change the solutions I was given above being it's a momentary application?

If I use a regulator where the the activation input is 13v, that'd be fine, I can accept not using the horn if the car isn't running.

I found this while searching:

To me this seems as simple as can be.. but the voltages are wrong.. can this be modified for 13-15v input and a 10v output?

Please forgive my ignorance.

 

[skimask87]

Change the 7812 for a 7810.
See what changed there?
See what the difference is?
Have you looked at the datasheet for the 78xx yet?

 

[EMCLARK]

*lightbulb*

Thanks!

 

[JimB]

As you are driving a compressor with this power supply, be aware that there will be a BIG current surge at switch-on time.

When you say "up to 20 amps" as the current requirement, is that the normal running current or does it include the surge during start-up?

If you do not allow for the surge, at best the initial voltage will be low until the compressor is up to speed, at worst it will destroy the pass transistors in the power supply.

JimB

 

[EMCLARK]

Would it be safe to assume this includes the start up surge? It's an electric maintenance free compressor, with a 20amp fuse on the source wire. I haven't actually gotten a tester onto it myself, but I've found a few tidbits here and there that say once the power is supplied, it's a constant 14-15 amp draw beginning to end.

 

[JimB]

Would it be safe to assume this includes the start up surge?

No, not really.

Although it may be fused at 20 amp, that does not mean that the surge does not exceed 20amp for a short while, (less than 1 second).
Fuses can run for a long tome on a small overload, but transistors were invented to protect the fuse by blowing first.

It could be that I am being over cautious here, but to me this is one of those things that needs to be verified if you want something which is reliable.

JimB

 

[EMCLARK]

Would upping this to 20amps be as simple as adding a 4th 2955?

At this point I'm just trying to figure out if this is possible.. I intend on testing the draw before building anything to be sure it's correct.

 

[EMCLARK]

I successfully built this using a prototyping board, I used strands of wire for the tracks. I added a 4th 2955 to make it 20amps, and replaced the 7812 with a 7810 for the desired voltage. As the entire circuit is controlled by a relay, the board only receives power when the horn button is pushed. It works exactly as I hoped regardless of the car being on or off (the issue being a 12v motor receiving more than 12v from the alternator and spinning faster than it should. The reason I put to it to 10v is I felt it was still too fast even at spec voltage, a manufacturing error perhaps.) Heat isn't an issue, no components feel remotely warm after normal use (which lasts about 4 seconds).

Thanks for the assistance

 

[tcmtech]

Just stack some high current diodes in series to get what ever voltage drop you need. at a fairly consistent .5 - .6 volts per diode dropping the voltage a few volts will not be much of a challenge.

 

[EMCLARK]

The challenge was having a variable input and a constant output. I needed something that would take the 12.5 to 14.5v, depending on if the car was on or off, and have a constant 10v out.

The amp draw was an issue to, it needed to handle up to 20amps.

I don't know if there was an easier way to to do this, but it's working and it was a lot of fun to do, I've never built a circuit before.. I tried that FM mic bug from radio shack about 20 years ago when I was a kid, and it didn't work.. so this was good

 

[tcmtech]

What part of this is going to fast?
The musical tune or just the compressor motor and why do you feel its going too fast?

 

[EMCLARK]

The tune was playing too quickly, it sounded as if the horns were being cut short. Compared to youtube, etc. examples of the same horn I found, mine seemed faster, which was also what I found others saying (too fast/too slow).

There was also a clear difference in speed car off versus on with the higher alternator output. Since the manufacturer says it's a 12v compressor, it seemed apparent that, even with the car off, it was still getting more than 12v (battery is healthy sitting at 12.5v). I took a chance on 10v working, but seeing the speed difference from 12.5 to 14.5, I thought by dropping it 2v it would be right where I wanted it.. thankfully I was correct.

 

[MrAl]

Hi,

Sounds interesting yes.

The difference between running the compressor continuously and running it for only 4 or 5 seconds is the key to this design working without a huge heatsink. The transistors take a few seconds to heat up even without a heatsink so the max temperature after 5 seconds will be less than after say 500 seconds. That fact will allow for a much smaller heatsink.

The caution however is that the transistors also take a while to cool down, and how long they take to cool down depends on the specific heat capacity of the transistor bodies plus the specific heat capacity of the heat sink. It is entirely possible that they hold the heat for a good 60 seconds or more. This means that if they were fired once every 30 seconds that they may eventually reach a temperature which destroys them.
This in turn means that there is going to be a certain minimum 'off' time to allow them to cool before firing them again. It could be 30 seconds, 60 seconds, 2 minutes, or 5 minutes. The only way to be sure is to measure the temperature before the first turn on and then measure it after the turn off and see how long it takes to get back down close to ambient. If you dont fire them up too often it will be ok.

 

[EMCLARK]

I don't use it very often, although that is very good to know. Knowing it would only be used 4-5 seconds at a time, I still used heatsinks as the schematic called for. I used the correct to-3 heatsinks for the transistors, as well as one on the 7810 regulator itself. I know in other applications, amps, etc., the heatsinks the transistors rest on are massive. When testing the circuit, I played it about 6 or 7 times within a few minutes, which is way more than I would normally, all the components were still cool to the touch.

At the end of the day it's just a cheesy musical air horn, lol, I just wanted it to work and sound as I thought it should.