Cap & diode in DC environment

Use an SMPS power supply before your device. while cranking, the battery voltage sharply fall below the normal level. An SMPS can tolerate this surge..

And what is that, this SMPS power supply?

---------- Post added at 16:29 ---------- Previous post was at 15:31 ----------

OK, I Googled it Waaaay to much for me. I understood that it should be a device that would accept input power of 5-15VDC, and give constant, clean, spikeless 12VDC output. But... I can't find it online to see size, price, etc. All I could find were some PC power supplies of 90+ watts, big as a cereal box (or more) which I'm sure is too much.

Use LM2576T- ADJ DC-DC Converter.. its quite simple to build.. i thnink u have sufficient knowledge in electronics..

Refer this Link for more detail.. its a low cost (you can built it in around 1.5USD)

**broken link removed**

Yeah, I think your problem is that the radio is grounded to the engine, but the engine isn't grounded very well to the body. When you crank the starter, all that current raises the voltage of the engine block with respect to ground. Connecting to that bolt ought to do it.

Great info! Gonna try the good grounding first, and then build the gadget Sreejith posted...

When cranking: if I turn the key to on, clock is ok. Then I crank, the engine comes on, and the clock resets.

Click to expand...

Although a grounding problem is as likely a culprit as any, here's another possibility. There's a chance that it's not a low voltage condition that resets the unit.

Consider that it might be a high voltage spike.

When you stop cranking, the starter cuts out. The solenoid cuts out. Current flow stops suddenly. These inductive components can raise an inductive kick. Namely a high voltage spike going throughout the electrical system of your vehicle.

It does not show up on a meter or LED display. It takes an oscilloscope to show it.

The spike is most severe when your starter has been drawing the most current. It just so happens the starter draws the highest current when battery voltage is low. Say, on a cold day when the battery has sat for a few days. As you describe.

But there are those massive electrolytic capacitors. You might ask, why don't they absorb the voltage spike? It's for the reason that both a small and large capacitor are recommended across the power leads to a sensitive circuit. The large capacitor does not respond quite so well at absorbing a spike. The small capacitor is better at it.

The usual value is 0.1 uF. Inexpensive. It should be rated for 100 or 200 volts.

bluedevil,
To me till we all are trying to surround the problem could not decide till now the trouble is on the sudden voltage drop on the battery while cranking or some other culprit making us fool. If still you are getting the clock reset even after connecting 30,000µF, I do not believe it is a cause of sudden voltage drop on battery.. How long the engine takes to come up after the cranking? 30 seconds, 60 seconds or less or more? I do not agree to diagnose while connecting a Bulb or a LED for monitoring the drop, as the resetting of clock takes just a blink like time and can not be detected without an Oscilloscope. First of all you must know the reset voltage supply of the computer thingy and then you can try to monitor if the battery voltage is the cause. Once you have connected the capacitor in parallel to the battery, the battery voltage drop will not affect the clock rest but the drop on the Capacitors. It seems there is some Ground or Supply connection problem on LOOSENED contact on the Computer and while you crank the vibrations (may be some time only) make that connection disconnected and the computer resets. This is a bit time consuming but a better finding on this assumption, remove the computer keep it on the SEAT while the connections are there and start the engine and see if it helps.

^ will try that, on saturday morning 'cause I work 8-4:30 and don't have any daylight nor garage, so weekends is the only option. Unfortunately, due to two big connectors that go to the computer, I can't move it far away from its post - just 3-4 inches maybe, not enough to reach the seat. I will, however, use a wire to a good ground and splice it into 3 existing GND wires on the computer, to assure there's a good GND.

Why I don't belive it's a lose contact is that it never resets (nor blinks, for that matter) while driving. And the roads here are bad to say the least, a lot of holes, speed bumps, irregularities, not to mention when I go to unpaved roads, everything shakes & vibrates, but it never ever did anything to the computer thingie. Just the cranking.

As for the cranking duration, after I give it contact, I hear fuel pump priming, then when it stops I turn the key to crank, and it turns around for 1 or 2 seconds, never more. Just 1 or 2 secs, and it's on. I do, however, have a little instrument in the cluster which tells me the batt voltage when I give it a contact (and tells charging voltage when the engine is running): after a standstill it shows 9V, or a tad over it, like 9,6V but not more. When I crank, of course, all goes down to 0, interior lights dim a bit etc - like in any other car.

Ok agreed as you said for the loose connection. An other question, does it same thing in Summer ? Or just in winter and specially in low temperatures ? It in winter only have you any idea it starts about at which lower temperature ? In fact I now want to segregate the problem situation, as if it is defined the remedy will be easy. As there are Blocking diodes connected in between the main battery connection and the capacitors there is no way that when battery goes to low voltage the capacitors also discharge that quickly. The caps are meant to feed the computer only. Can you reply early for some other ideas to come forward.

It didn't happen in summertime, not once. It began with autumn temps, when they started being under 12C (53F) the whole day - meaning lower in the morning, of course, and up to 12 C (53F) in the coldest part of the day.

---------- Post added at 11:49 ---------- Previous post was at 11:48 ----------

(give or take a degree or two (in celsius), of course)

As you said the cranking takes just 1 or 2 or may be even 4 seconds, I do not think it will cause the computer to reset while the caps are connected. Just monitor the voltage on caps terminals while cranking. The volts here should not go down more that 11 volts or so and this will not let the computer reset. Simply put the multimeter probes on both the terminals of caps and start the engine. Another experiment, easy to do. Do you have any 12 volts supply like an AC adapter or any other spare battery may be small size. If yes, just connect that to the caps terminals and remove the battery connection to the caps and start the engine, if the problem is battery volts drop, the computer in this situation should not reset.

I drove last on moday evening, clock resetting & everything. But no other problems. Today (thursday) around 6 pm I went to doctor's only to find I can't crank her - fast clicking from starter and all lamps dimming. So I guess batt is shot, and maybe all my problems were with the batt afterall.

Only question I now have is: could the 30000uF cap + 3 diodes have drained the batt? I thought the strain on the batt would be almost inexistant since caps would fill up and that's it - no other draining would occur. Is it so or am I wrong?

Check the battery connections. They may be loose.

As tight and as clean as they can be. Tha was the first thing I did back in late september when the clock-resetting started.

The only other thing I'd check is how well the negative cable is attached to the body & the engine block. The capacitors wouldn't drain the battery unless there was something wrong with them, or the polarity reversed.

bluedevil said:

Today (thursday) around 6 pm I went to doctor's only to find I can't crank her - fast clicking from starter and all lamps dimming. So I guess batt is shot, and maybe all my problems were with the batt afterall.

Click to expand...

Yes, a weak and aging battery will account for this. Especially as cold weather begins.

Check out whether your starter is drawing overmuch current.

The fast clicking may be your solenoid. See my earlier post about high voltage spikes. Touch your hands across the positive post and a ground. Ask someone to start the engine. At the same time you hear the rapid clicking, expect to feel high voltage spikes go through you.

Only question I now have is: could the 30000uF cap + 3 diodes have drained the batt? I thought the strain on the batt would be almost inexistant since caps would fill up and that's it - no other draining would occur. Is it so or am I wrong?

Click to expand...

You are correct as long as no cap is leaking more than a fraction of an amp.

A dome light (or trunk light, etc.) is known to drain a battery overnight. These little bulbs can draw as much as an amp. Especially if the battery is old. It's happened to us and people I know.

That's the reason it's a good idea to find out how much current your computerized unit draws.

bluedevil,
How old is your battery, a year or more? If the same battery was in your car last winter, too, was there any problem like this ? If there was no problem since the battery was new, then forget every thing and get a new battery and get rid of every thing but if the same problem was last winter, too, then you have to think right left. The ageing batteries in cold season do drop their efficiency and at the same time engine takes bit long (comparatively) to get started due to low temperatures.

Batt was bought 11.29.2010. and disn't have problems untilthis september. Now I jumped it to start, went to the shop (20 minute drive) and it didn't recharge to crank it again. So They tested it and it's shot. Got anew one and now I'm gonna replCe the old one under warranty & sell it. They also checked for proper alternator functions and parasitic load when shut down, and all is ok. Connections too. We'll see in a few days what gives...

Did your battery ever get fully discharged, and then freezing temps came through the area? At any time, even just once? That can ruin a battery.

Fully charged battery acid can stay liquid way below 32 deg. F.

However when a battery gets discharged, the acid turns to water (mostly). It can freeze at a higher temperature. Lead plates can warp. Internal connections can disconnect. The damage remains even if you charge the battery.

It did discharge once, in July, when I left interior lights on for 2 days. But in summer. Maybe that killed it. Anyhow, I got a good quality Varta now, and a new one for the broken, in warranty. It's got 2 years of warranty, so it's in the ads maybe I'll manage to sell it since it's brand new.