Car key/fob remote repair

[ro0ter]

Hello everyone,

I`m struggling with two keys belonging to same car. Since I`m good with electronics, all friends come by asking me to fix this and that.. So I try to help everyone, but lately my time is limited (1 x 2-years-old + 1 x 2-weeks-old are more than enough )

Pardon me for asking, but right now I am in a dead end.. A friend of mine (obviously) has a car (I`m not sure whether I should post the car`s name here... not to be considered advertising or something.. but whatever, maybe there is a common problem with this brand name..): 2001 skoda octavia mk1 wagon... The problem is that both keys drain batteries in days.

I noticed however that the crystal oscillator had a very bad connection to one of its legs... The copper pad below the respective leg was totally missing (corroded or smth), but no trace of anything. The key circuit was not removed from its case (I had to heat the melted plastic pins holding the board against the buttons half). Still, someone would have had access to the PCB side holding all the chips (buttons are on the other side, original ones).

1. Batteries drain in days (no matter that it`s Duracel, Varta or dolalr-store type).
2. You can`t reverse-polarity the battery, the battery holders will simply connect to the same side.
3. Circuit does not have any traces of corosion/deposits.. Anyway, washed with isopropyl alcohol (99.9%) and no success.. I even removed and reseated the chips, same problem..
4. Connected a 300ma power source, saturated board with isopropyl and the isopropyl alcohol on both chips starts to evaporate... not a good sign...
5. 1 corroded oscillator pad, no traces of anything..

Note: The transponder is a glass-sealed capsule, not part of this circuit. This circuit only servers locking/unlocking doors (no trunk, only 2 buttons).

Do I stand any chance of making this car remote work? How about the second one?...

[Rerouter]

If its draining power in a few days, that means its pulling a fair amount of current (multiple mA) still tiny, but its measurable,

Now to start with, you have the IC's off, well hook up the battery via a multimeter measuring mA, if no reading try uA, your looking for a PCB defect causing a power draw,

If nothing, you insert 1 chip back on and see if it changed, and so on

The other one is try desoldering the buttons, and see if its consumption levels off, too many times I have seen semi conductive switches due to water migration.

[mbless]

The problem is that both keys drain batteries in days.

Do I stand any chance of making this car remote work? How about the second one?...

The first sentence implies the FOB is working, the second one says it does not work. So is the FOB working or not?

Regardless I would do what Rerouter said and find what is drawing the current. If the chips are drawing a lot of current, evaporating IPA faster than rest of the board, and are not working, then it would appear the chips are dead.

[ro0ter]

Both FOBs work. Push any button, they light up the red led and strongly emit. Both FOBs have the same problem: drain a CR2032 (Varta/Duracel/Auchan/no-name) in 2 days.

I only have 1 FOB for diagnosing/fixing (hopefully not autopsy), the other one is used by owner... PCB does not draw any power while connected with no chips.

I`ll try desoldering one chip at a time, see if only one is the clurpit. Providing 3.1v (max current 300mA) makes both chips heat (maybe internal voltage regulator of one of the chips floods the second chip). Did not allow them to heat more than 30-40*C. Sprayed a good amount of isopropyl alcohol and connected power source, the heating is obvious in both chips, starts where I added the red dots (I`ll upload a gif next time).

I will try desoldering the buttons first, the "wet button" hypothesis is plausible, yet I saw no water damage or trace and the remote pcb was still held to the buttons half case with the original plastic-melted pins.

What intrigues me is the "consumed" oscillator pad... I`ll have to confirm again whether it was consumed or broken. It might be possible that someone squeezed the remotes so hard (when they normally ran out of battery after a couple of years) that they broke one of the oscillator pads, and if they provided a new battery afterwards - I am curios how would the chips behave when there`s no oscillator, but only a 3+v battery..

Thank you all for your contribution! Will keep you posted.

[Rerouter]

without a working clock, the chip should remain in a low power reset state, with all pins set as inputs with something like a 100K pullup,

however the TX may involve a more powerful switching, so there may be some leaky mosfet or similar for that (cannot see rear side of PCB)

[ro0ter]

Should be something like this: (image from internet)

[ro0ter]

I guess I have a clue on what might have happened.. The crystal oscillator pad was not "corroded" but rather it was broken by pressing too hard on the button exactly on the other side, directly below the oscillator.. The oscillator might as well be broken, the key no longer transmits.. Now, while the oscillator is not connected I guess there is no clock for whatever standby or watchdog and the battery simply drains through some mosfet.. or maybe it requires the clock for generating the on-chip reference voltage.. hell knows how these beasts work...

I`ll replace it with some 13.56mhz crystal and see where I get

[ro0ter]

***SOLVED***

The crystal oscillator was the culprit. When the oscillator was not connected, it was drawing up to 400mA from the PSU... When pushing the button, the broken pad would momentarily connect with the crystal and the key would be alive, thus emitting a valid signal.

I`ll have to see whether the second key has the same problem, but I bet it`s the same...

The PCB of this key is quite thin and pushing a button makes it curve a little (e.g. ~0.5mm for a normal button press, I bet it`s more when the battery is dead and someone squeezes the remote). I added a hot glue-molded section between the pcb (chips side) and the battery, therefore when the key is pressed, the pcb no longer bends but it transfers the force to the battery and to the other half of the case (therefore to the counter-finger).

Thank you for all your help and ideas!

[uy2618Mc01]

Nice analysis and thanks for posting the diagnosis. Too often the fora get presented with the problem but not the confirmed solution.

[ro0ter]

without a working clock, the chip should remain in a low power reset state, with all pins set as inputs with something like a 100K pullup,

In this case it seems it is drawing A LOT if current... very lol, inded.... much fun...

[G7PSK]

If the problem was caused by pushing the buttons are they making poor contact and need cleaning or replacing or is the owner just heavy handed with such things.

[ro0ter]

If the problem was caused by pushing the buttons are they making poor contact and need cleaning or replacing or is the owner just heavy handed with such things.

The buttons work like a charm, they are mint-condition... on a 17-years-old-car... These are "premium", never-failing unless-you-tamper-them-with-a-screw-driver-buttons. They require moderate pressing force, but there are humans on this earth who press squeeze the buttons like they want to squash the remote once its battery is low... I don`t know where that comes from, but that is the very same that happens with tv remotes, causing the conductive rubber on the other side to fail.. The PCB is however very thin and when a beefy finger presses the remote button very hard (either low range due to low battery, battery installed wrongly (both battery contacts receive the +), no battery installed in remote, depleted battery, dog chasing human or simply "forgot how to human"), the pcb bends.. That is not healthy for parts on the other side (expecially for the oscillator.

NOTE: Second remote had the very same and exact problem - one beautiful solder joint crack on the same side of the crystal oscillator (as the first remote).

Attached pictures with "hot glue" remedy. sits between battery and pcb, I should probably glue in a single spot to the pcb..