Battery, Alternator or its cutout?

[danhr]

Ok I am not an electronics guy but 7812 is also drawing 0.28A.

What I did is constructed the regulator part and tested it directly off the 24V battery. Output is 12.4V. Now I removed the -ve wire and introduced a mulitmeter to check the ampere draw of regulator. It was 0.28A. Thats bad aint it?

 

[Ron R]

Yes it is.
For the regulator to function properly the current to flow through it should be at least 5 mA. So if a device would need no more than 5 mA this regulator should not be used.
But for your alarm and memory the current will be more than that, so no problem in that regard.
I understand from your posting you tested the regulator without load. This can produce odd results.
12.4 V is at the high side but still within specs which says ot should be min 11.5V - max 12.5V
Try the following to find out what the regulator consumes:
Measure the draw for the alarm when connected to 12 V.
Now use the regulater to feed the alarm from a 24 V source. Compare the two values. The latter will be about twice the first.
see also: http://www.fairchildsemi.com/ds/LM/LM7812.pdf.
This is the datasheet for the various LM-regulators.

 

[danhr]

Ok let me get this right... You saying that with a load such as alarm and CD memory the draw would be reduced?

Yes I did not had any load on the 12V side. I just checked the draw of this homemade regulator and on the 24V side it was 0.282 ampere.

The IC I am using is L7812CV. I am going to put some load and update.

 

[Ron R]

Danhr,
There must be something wrong in your setup. The amount of current needed for the regulator itself should be in the 4 to 8 mA range. Defenately not 200+ range.

Did you connect the IC the right way? Like it is shown in your photo, input (+V) is the leftmost pin, output is the pin at the right and ground (-) is centre.
So if you connect a source (24V batt) to the left pin (+) and the centre pin (-) than you should find a voltage of ~12V when you connect your multimeter to the centre and right pin.
Best is to put a resistor (56 Ohm) between the centre and right pin and then measure the voltage. Then there also should be a current flowing of 0.21 A (=210 mA). To find this, disconnect the resistor from one of the two pins and put the multimeter in there (one lead to the pin of the IC and the other one to the resistor), set to current.
If you don't find this value (~~210 mA) then there is something wrong.
Now measure the current at the input side. This should also be about 210 mA.

 

[danhr]

Well something is wrong here. I still get 0.282 amps. With load I get 0.27amps (On the 24Volt side i.e. L7812CV is drawing 0.27amps with load of CD memory and antitheft).

Connections seem fine and its replica of your circuit drawing.

Checked the output and its 12V. Good news is that anti-theft alarm worked well. Bad news, the problem remains intact, as the draw is even more than the 15A Argus Switching Regulator.

Other thing I observed was L7812CV was getting hot very quickly. Even though this is not a problem (as soon as I touched it to metal body, it cooled down). But the heat dissipation also suggest a heavy draw.

I even took a fresh L7812CV and without diode, cap (at input/output) tested it. Draw is 0.282amps. It is pretty disappointing.

All now seems that either all of the regulator IC that I have are fake rip-off and thus are consuming a lot of power or this tweak is not going to work at all.

I am going to check if I can get ones from Fairchild or National Semiconductors. If not, I will either stick to current setup and recharge the battery every week or so or use a supercap/small rechargeable 12V AUX battery.

 

[Ron R]

Well something is wrong here. I still get 0.282 amps. With load I get 0.27amps (On the 24Volt side i.e. L7812CV is drawing 0.27amps with load of CD memory and antitheft).

Connections seem fine and its replica of your circuit drawing.

Checked the output and its 12V. Good news is that anti-theft alarm worked well. Bad news, the problem remains intact, as the draw is even more than the 15A Argus Switching Regulator.

This is bad.

Other thing I observed was L7812CV was getting hot very quickly. Even though this is not a problem (as soon as I touched it to metal body, it cooled down).

Yes this behaviour is very normal. That's why I use the aluminum box as a heat sink.

But the heat dissipation also suggest a heavy draw.
I even took a fresh L7812CV and without diode, cap (at input/output) tested it. Draw is 0.282amps. It is pretty disappointing.

This indeed is dissapointing. Something is wrong.

All now seems that either all of the regulator IC that I have are fake rip-off and thus are consuming a lot of power or this tweak is not going to work at all.

Why would it work for me and not for you ? Maybe indeed fake IC's

I am going to check if I can get ones from Fairchild or National Semiconductors.

Yes give it a try. Another option is the new kind of regulator that I described. Look at this page: http://www.recom-international.com/press/The-Best-just-got-Better.html

I just discovered they now also produce 1A versions for input voltages higher than 18 V. See the R-78Bxx-1.0 halfway down the page. (these were not available about two months ago AFAIK)
These are switching regulators, but fully integrated. I use two of them to feed my PDA (7805.0, the 5V version)and my GPS( the 3.3
V-version. They work fine and no or virtual no heat. This because they have a efficiency of up to 95%.
The regulator is pin-compatible to the linear regulators like the LM78xx.
The housing is just fatter. Whereis the LM 78xx is about 5mm thick, these are about 10 mm thick and a bit wider. Hight is more or less the same. They don't have the metal heatsink....not needed!

If not, I will either stick to current setup and recharge the battery every week or so or use a supercap/small rechargeable 12V AUX battery.

If it won't work that's all there is left.

 

[danhr]

Let’s see what’s locally available to me. I hope substitute works.

RECOM R-78B12-1.0 seems awesome.

Extremely sorry for the bad pics. But I just don’t know what’s wrong with my camera.

 

[Ron R]

Looks nifty, little hard to make out how you connected everything.
But what is important for the capacitors to work properly is that they have to be as close to the regulator as possible.
To be honest, for my LM7812 setup I have not used any board.
I just bolted the regulator to one of the walls of the alu-box, bend the pins a bit away from that wall and soldered everything directly on the pins. Just wrap the leads coming from the C's around the pins (in the right way, the electrolytic capacitor -the large one- has a minus and a plus connection!). Do the same with the diode (ring away from the IC) and connect the leads. +24 V(input) to the left pin, +12 (output)to the other end of the diode.
Make sure the soldering rod is the right temp and solder has raisin core. Solder each point quickly, let the IC cool down in between each soldering point.
Middle pin = same as metal back = mass (ground) = connected to body/chassis therefore you don't necessarely need aq minus lead.

I made sure the leads where fixed so no strain can be excerted on the IC-pins.

 

[danhr]

Good news and bad news.

Good news, I have pin pointed the components that 'draw' amps.

1. Clock
2. 15amp Argus converter currently supplying CD player, antitheft and CB.


Something is wrong with the clock since when I plug it in, it draws about 0.23amps. Currently I have removed its 'connecting jack' and draw dropped to 0.1 then after 10 mins to 0.070 then after fifteen to 0.050amps and its dropping. I will fix that later.

For the converter, the bad news. I cannot find LM7812 (quality ones) locally. Now I have asked someone to get them for me hopefully I will get them soon.

Ron R:

I have thought of slight modification to your circuit design. Along with use of diodes, I will use relays as well. That would ensure nothing 'wrong' gets happen. Furthermore, if draw increases suddenly (say, you somehow managed to turn on the CD player with homemade regulator supplying power), it would turn on the 'big' converter to compensate for draw.

And of course, fuse protection.
I will try to keep it as simple as possible.

 

[Ron R]

Danhr,
good you found that problem with the clock.

The idea to use relays is not bad at all and the way you want to use the system (switching on the inverter when a you have a big draw) is clever.
How are you going to detect the current-level for the activation?

sorry for responding after several days, but my computer died a sudden death.
Had to get a new one and despite having backups of everything I had a bad time restoring everthing because my previous system was XP and this one is Vista.....Thanks Bill....

 

[danhr]

Ron R:

Amps can be found out using a Hall sensor effect.

Generally speaking, components may involve:

a) A hall sensor (to sense amp draw)
b) A Microcontroller to sense and report/do changes. A crude way would be to use a Comparator, instead.

Now a since we do not want great accuracy and are not interested in readings, we will use a comparator. Using comparator, we can set a threshold to activate a relay. My plans are at a halt since I still didn’t get LM7812...HUH!

I am now trying to make adjustable regulator using LM317, I had them laying around. Hope that works (has a lower draw). Plus point with LM317 is that one can pump up voltage to 13.8 or 14.4 using a potentiometer.

Also, need to know that the 24-12V regulator that I have has Green, Yellow, Red, Black, Purple and Orange wires.

RED 24V-input
Black Ground
Purple 24V-input : is for ACC but I linked it to 24V direct to turn converter ON.

Now orange supplies 14.4V (same as output) but when converters power button if OFF, it supplies something like 9V.

I have no idea what green and yellow wires are for. Can someone help?

p.s. the other option is an ammeter with a contact switch where the needle will push a 'button' to activate a relay. Not very practical but might work.

 

[danhr]

Ohh and forgot to mention, with LM7812 if you put a couple of diodes to -ve forward biased, so you can pump up 0.7volts per diode.

See the pic. Its LM317T (adjustable 12V regulator). You can see a variable resistor, I used it to increase the voltage to 13.8V. Even with LM317T, the amp draw is 0.271 with or without load. See the pic, I used a small 12V fan as a load.

Extremely disappointing.

 

[Ron R]

Ron R:

Amps can be found out using a Hall sensor effect.

Generally speaking, components may involve:

a) A hall sensor (to sense amp draw)
b) A Microcontroller to sense and report/do changes. A crude way would be to use a Comparator, instead.

Now a since we do not want great accuracy and are not interested in readings, we will use a comparator. Using comparator, we can set a threshold to activate a relay.

You said you were going to keep it simple

My plans are at a halt since I still didn’t get LM7812...HUH!

I am now trying to make adjustable regulator using LM317, I had them laying around. Hope that works (has a lower draw). Plus point with LM317 is that one can pump up voltage to 13.8 or 14.4 using a potentiometer.

Adjustible is fine. I used an 7812 because I wanted it's voltage to be clearly lower than that of the inverter, otherwise the invertor might have difficulties keeping it's voltage as I found out. But in my setup both (the 7812 and the invertor were connected the same time. As far as I understand your setup differs here. You want to use relays to separate.

Also, need to know that the 24-12V regulator that I have has Green, Yellow, Red, Black, Purple and Orange wires.

RED 24V-input
Black Ground
Purple 24V-input : is for ACC but I linked it to 24V direct to turn converter ON.

Now orange supplies 14.4V (same as output) but when converters power button if OFF, it supplies something like 9V.

I have no idea what green and yellow wires are for. Can someone help?

I assume you have a special inverter. When not activated (purple not connected to 24V) it delivers a small current for standby-equipment (radio-memory, alarms).

But I don't know this specific inverter. Can't you find info about it on the internet?

p.s. the other option is an ammeter with a contact switch where the needle will push a 'button' to activate a relay. Not very practical but might work.

And not reliable. If so I would go for a so called 'reed' relais.
It's a small glass tube with a switch in it.
This switch closes when the tube is in a (elektro-) magnetic field.

if you run the feed through a coil that sits around this tube you can decide at what current the switch closes by altering the coil dimensions. The bigger the coil the stronger the magnetich field for a given current.

 

[Ron R]

Ohh and forgot to mention, with LM7812 if you put a couple of diodes to -ve forward biased, so you can pump up 0.7volts per diode.

See the pic. Its LM317T (adjustable 12V regulator). You can see a variable resistor, I used it to increase the voltage to 13.8V. Even with LM317T, the amp draw is 0.271 with or without load. See the pic, I used a small 12V fan as a load.

Extremely disappointing.

Well, you have extra losses through the resitors. If you want to go this way (using an LM316, I mean), use a combination of resitors with the highest values that are usable.
You are using a 180 Ohm resistor with an 2Kohm (?) variable resistor here. Try something like 1k5 resistor with 20K variable. This might improve (reduce) the draw. The current for the adjustpin should be in the 50 - 100 micro(!)Amps. Hardly measurable with an ordinary multimeter.

 

[danhr]

Ok now I know you are an electronic-wiz.

270-ohm resistor with 5k variable resistor. OK I will increase both and see what happens.

I used LM317 primarily since I had them lying around. It is strange but I am not able to get quality stuff.

A hall sensor is fairly easy to use. Comparator is something you can salvage from an old non-functional PSU. Yes, it is a little time consuming (takes a few hours to get the job done), nevertheless it’s doable. I do not know much about reed relays. But sound like it would be the simplest of all the methods.

For the 24-12V converter, I cannot find any manual. It’s a Japanese make, brand name ARGUS and its 15A. Another one with similar wires is made by CellStar rated 24A. I could not find any manual for any.

 

[danhr]

Hankook the Worst battery manufacturer

Ok now see this… I was thinking of an explanation to the ampere draw by such a small regulator. Then it hit me, it could be that these Hankook batteries cannot handle any parasitic load at all. This I observed since which ever converter I use 6amp/10amp/15amp or 24amp the draw is 0.27amp for each. Moreover, if I use two converters, the amp draw increases by 0.0xx.

Same is true for LM317 and L7812. So I took a PC PSU and checked it. The PSU is suppling 12.25-12.45Volts without load. I initially checked to see if it could re-produce enough voltage to turn on the fan (small load). In each case, either passing through LM317 or L7812 the voltage was between 10-11Volts. Enough to run the fan.

Therefore, I checked the draw. LM317 with load was 0.134, without 0.133amps. L7812 showed 0.132 with and without load. I noticed that the draw is high but it is still LOWER than 24Volt Hankook batteries, which are stated to handle 70amp of draw for each battery.

Furthermore, the PSU I had can handle 2amps of load on a single rail (at max). I bet if the PSU were made stronger/produces higher voltage like say 24Volts, the amp would become a fraction of what it is now.

Now if all the above is true, Hankook is the worst battery manufacturer that I have ever seen both support wise and product quality wise.

 

[danhr]

Ron R:

Thanks again for the link. As I said, I have placed order for RECOM switching 12Volt 1amp regulators.

These recom ICs will not be able to handle heavy load and that’s an issue. LM7812 used to get pretty hot even when idle (no alarm). What do you propose for that, Ron?

You are right about the transistor. If I use a transistor with RECOM IC, it will be a linear regulator.

 

[danhr]

Ok till date Cutter.com has not sent me the parts (Don’t know why).

Strange thing happened today. I had removed the battery terminal to recharge the batteries. Once done, I connected the terminal and observed a spark. After hours of working, I located the problem to door lock. The door lock spring was rusted and it didn’t turn back the lock to its position resulting in the solenoid and its relay to be activated. Mind you, I recharged the batteries because they drained out within three days. So this is another problem that comes over time i.e. rusty lock spring=battery drain. The drain was about 0.7amps.

Furthermore, just for update, Clock drain is 0.2mA (miliamps). The converter drain is about 0.232A (amps) with our without parasite load. Hopefully when I get RECOM ICs, this issue will be resolved.