Cap & diode in DC environment

[bluedevil]

Car environment, 12VDC. There is a sort of computer in the car that displays various things, among them date & time. When
weather is cold, and after a standstill of 2 days or more, when cranking the said computer obviously loses power for a sec and
the date/time reset. I got tired of adjusting it, so by an advice I bought a 35V-10000uF cap and a 45V-7.5A schottky diode. I wired the diode on the + wire that goes to the computer, and after it I put the cap's + pole. Cap's - pole is on the - wire for the
computer. I tried it - works: even after 4-5 days of standstill it would keep the computer from resetting. But, as the weather
grew even colder (under freezing point of water), and after a standstill of 2 days or more, the resetting thing came back.
So I got 2 more caps, 35V-10000uF each, and I wired them in series with the existing one (I belive, at least: all thre + poles of
caps are together, as are all three - poles). But it didn't help.

What am I doing wrong? Do I need two more diodes, one per each cap? Or bigger diode? Or what? I don't belive the said computer draws much current, because it only displays inf it picks up from sensors. Only thing it remembers are dat & time, all the other info it reads from sensors. Please advise what to do?

Thanx!

 

[prateek_k_chd]

" I wired them in series with the existing one"


Capacitor are different from resistors,Mister devil.

You wire them in series - you lower their capacitance.
And the decrease in temperature has already lowered their capacitance.So you need to INCREASE the capacitance.

For that you have to wire them in PARALLEL.

try it and let us know how it goes.

 

[ark5230]

There is one more possibility
These days USB, SD MP3 players in the cars keep date and time for long time.
The technique there is that power of clock is direct from battery.
The power to the rest of the power consuming circuit comes from ignition.
Some similar input may be present with the device DEVIL is using. He can try searching the wiring diagram for the specific equipment he is using. It will certainly have provision for such a power supply connection to retain date and time over loner periods from practical use.

 

[Pjdd]

From his description, the OP seems to have connected the additional caps correctly -

all thre + poles of caps are together, as are all three - poles

that is, the caps are in parallel. And since the first capacitor did the job, it appears that the wiring was done correctly too.

@bluedevil: When the battery is at its normal voltage level, it keeps the caps charged via the diode. When the battery voltage drops, the stored charge of the caps keeps the computer thingie running, but the diode prevents it from flowing back to the battery and whatever else might be drawing power from the battery. Two possibilities come to mind as to why it no longer works that way:

1. The diode got shorted and no longer works as a one-way passage. Can you check it?

2. The caps are not rated for low temperatures and have much reduced capacitances at your current temps. Caps are rated, among other things, according to their working temperature range which is often printed on the can. Some caps are rated for 0 to +70 deg C, some for 0 to +85, -20 to +125, etc.

 

[bluedevil]

1st, my mistake: I wired caps correctly but I wrote wrong. So they are wired in parallel, as they should be, to increase capacitance.
So, I'm gonna check out if the diode is still working and is not shorted, and then I'm gonna check the temp rating on the caps too.
I didn't know temp could cause such a big difference in capacitance :/

So, for the time being, thanx for the replys and I'm gonna check those two things out...

 

[BradtheRad]

Cranking voltage can drop below 9 volts on a cold day, and the battery has been sitting for a few days.

You are correct to install the diode. And notice the diode must carry a large inrush current to recharge 30,000 uF or more of capacitors. And it happens very quickly if there's little resistance.

Therefore if you find the diode needs to be replaced, consider installing a resistor inline with the new diode. The purpose is to limit surge. Between 5 and 50 ohms should be okay.

 

[bluedevil]

Uh... Now...
How do I test the diode? Since it is so small, it didn't have any marks as to which pole is where, so I was told to just plug it in
(without caps) and see: if the computer thingie works, it's the right way. If not, it should be reversed. So, how do I check if
the diode is working OK? I have a multimeter.

Also, I like the idea of "protecting from surge" even if it's not necessary. It can't hurt, can it? So I'd get a resistor, since I'm
already wiring & rewiring stuff and just plug it in there. There are no + and - poles to the resistor, right? I just get a 50 ohm one, wire it first, after it the diode, and finally cap(s), and that should be it, right?

 

[wa1kij]

Multimeters often have a diode testing function. Put the switch on the diode schematic symbol. Failing that, one of the resistance ranges may work. If the diode is good, you'll get a resistance reading in one direction, infinity in the other.

---------- Post added at 05:10 ---------- Previous post was at 05:07 ----------

...and it doesn't matter which side of the diode you put the resistor on.

 

[bluedevil]

Great! So I'll check the diode, I'll replace it if necessary, and I'll put the resistor in. Plus check on the cap's temp rating.

 

[bluedevil]

OK, tomorrow morning I'm tearing this apart to check everything. Just one question: since I have five 10Ω-½W-resistors at hand, would one of those be enough? Or two? Three? And if so, if I need to put 2 or 3 (or 5) together, do I put them in series with eachother or in parallel?

Thanx!

 

[Pjdd]

Uh... Now...
How do I test the diode? Since it is so small, it didn't have any marks as to which pole is where,

A 7.5A diode shouldn't be too small to have any markings.

so I was told to just plug it in
(without caps) and see: if the computer thingie works, it's the right way. If not, it should be reversed.

That's probably still the most practicable way for you. If it works one way but not the other, then the diode is OK. If it works both ways, then it's shorted.

 

[bluedevil]

I found a 8A-45V diode, so I'm gonna go ahead and replace it anyways. This one (new one) is cylindrical, black, and has a white ring on one end - I guess that's the (+) pole, that goes towards the battery?

The other diode was (is) square, black, and no markings at all apart from the manufacturers's code. Just two "legs" and that's it.

 

[Pjdd]

No, the ringed side is the cathode and goes to the load. The other end goes to the battery.

 

[bluedevil]

Uh, good thing I asked You just saved me some nerves & work tomorrow
So, ring-side towards the load (computer-thingie & caps) and no-ring-side towards the battery. Thanx!!!

 

[RetroTechie]

3x 10000uF caps to keep a realtime clock from resetting? That's crazy... 8-O

I'd have a look at insides of that onboard computer - chances are there's a realtime clock component somewhere, with place for a backup battery (like the CR2032 you find on PC motherboards), and that battery is either missing or gone bad. If it's there & you can locate it, just replace (and forget about that big capacitor nonsense, capacitors are no good as battery backup).

 

[bluedevil]

I did open it some time ago because I was suspecting some bad soldering points - no battery there. When you disconnect car's batt, even for a second, clock & date reset. I'm guessing since it's cold weather and standstill the car's battery get's low on charge, and when cranking it uses all the batt's power for a moment, so the clock resets. And the caps were the way to prevent the cranking from stealing all the power, keeping that small amount of power just for those 2-3 seconds.

Maybe caps lose much capacitance when it's cold so the 10000uF which did the job well when it was a bit warmer now can't store enough? But my best guess is that diode got shot so the current passes from caps towards the batt when cranking... Since there is no backup batt & I have all the parts ready, I'll give it another go with 2x10000uF caps (2 are idenctical, the 3rd cap is a bit different although identically rated, so I'll leave it out for the moment) and a new diode.

 

[RetroTechie]

Can you identify exact board type that's used in the onboard computer? That might help to determine what component keeps the time & how that is powered.

A small backup battery (or perhaps a small 0.1-1F supercapacitor) for that still makes more sense than adding big capacitors to keep the whole computer powered while the car is started. For example it would also keep the time if you replace the car battery at some point.

 

[bluedevil]

I guess I can take it apart again and take pics of stuff that's there, maybe then somebody could identify them... I'll try to do just that...

 

[Raza]

blue,
Batter put ONE diode each capacitor, as the 30,000µF capacity may damage the single diode while short circuiting it. Or put a bigger capacity diode in circuit. An other suggestion, IF POSSIBLE BUT A BETTER ONE, just simply crank the engine every day for 5 minutes and put it off even if the car is not driven. It will keep the battery to a better charged condition and the capacitors will fresh their charge also. Or if you put an INRUSH CURRENT CONTROL resistor in series with the diode it will help also. For connecting diode put the RING SIDE to the load and the other side to positive terminal of the battery. Do not know whether it helps.:!:

 

[bluedevil]

@ Raza: yeah, it helps. I will take pics of components inside the computer-thingy if I manage to open it because IIRC it was a real PITA with all the brittle plastic tabs 'n stuff. As for the caps, I do have 2 diodes and will use them both with 2 caps, to try that one also. I'm just not sure how to wire it? Like this maybe:


(-) BATTERY (+)
----------------
|...........|
+-II--<I-+
|...........|
+-II--<I-+
|...........|
----------------
(-) LOAD (+)

---------- Post added at 16:18 ---------- Previous post was at 16:15 ----------

Because so far I have it like this:
(-) BATTERY (+)
----------------
|........|
|........V
|........|
+--II--+
|........|
+--II--+
|........|
----------------
(-) LOAD (+)