Hello Everyone,

I wonder if anybody could help me?!
I recently fitted a battery holder for a CR123 3v lithium battery for the memory back up. In the space of a couple of months one battery has already gone flat and another seems to be losing around 0.1V per 24 hours. My question is, is this because the Duracell CR123 battery isn't somehow as stealthy as the Panasonic original? or is there some problem in the backup circuit, i.e a capacitor acting as a reservoir to slow the flow of voltage? or something completely different?

Any help would be appreciated!

Cheers

Paul

That does not sound right, the battery draining that fast. My Xpander
has had a plain CR2032 backup battery (installed in a hacky way) for
at least 7 years now. That suggests you really don't need some mega
beefy battery there.

The battery power line ("+5M") is only used for a handful of things
according to the schematic. You can try gradually replacing parts
involved with it. I think I would start with D4 and C14.


Op za 13 mei 2023 om 12:15 schreef <technogeist@...>:

Hello Everyone,

I wonder if anybody could help me?!
I recently fitted a battery holder for a CR123 3v lithium battery for the memory back up. In the space of a couple of months one battery has already gone flat and another seems to be losing around 0.1V per 24 hours. My question is, is this because the Duracell CR123 battery isn't somehow as stealthy as the Panasonic original? or is there some problem in the backup circuit, i.e a capacitor acting as a reservoir to slow the flow of voltage? or something completely different?

Any help would be appreciated!

Cheers

Paul

On 13 May 2023, at 21:15, Mondovermona <ipc.de@...> wrote:

?Hi Paul,

Sorry to hear.? This is not a problem with the battery, it is with the backup circuit or one or more of the memory chips itself.? When ?CMOS memory chips are coming into age, they start draining more current in their memory backup.?

To resolve, you might want to measure the backup current flowing from the battery.? It should be around 1uA (0.001 mAmps).? In any case if the current draw would drop significantly with the memory chips removed then at least one of them will be faulty.? If the fault is not wirh the memory chips, then you are left with a faulty on board memory backup voltage circuit. Oftentimes it is one of the transistors therein.?

Good luck.?

In any case, before hampering with your valuable device, you should feel confident that your electronic and micromechanic skills are sufficiently developed.? Else, keep your fingers off. ?

Mondo?

--

?_

Dr. Hartmut Schwahn

72070 Tübingen?

Telefon: +49 163 1631495

Thank you for the information Mondo.

I’ve been studying the battery backup circuit and the three 6264 memory chips, but I’m unsure which transistors you’re referring to? I can’t see any? If you could point me in the right direction, that’d be great.

I shall measure the current draw of the battery tomorrow.

Cheers

Paul




On 13 May 2023, at 21:15, Mondovermona <ipc.de@...> wrote:

?Hi Paul,
Sorry to hear. This is not a problem with the battery, it is with the backup circuit or one or more of the memory chips itself. When CMOS memory chips are coming into age, they start draining more current in their memory backup.
To resolve, you might want to measure the backup current flowing from the battery. It should be around 1uA (0.001 mAmps). In any case if the current draw would drop significantly with the memory chips removed then at least one of them will be faulty. If the fault is not wirh the memory chips, then you are left with a faulty on board memory backup voltage circuit. Oftentimes it is one of the transistors therein.
Good luck.

In any case, before hampering with your valuable device, you should feel confident that your electronic and micromechanic skills are sufficiently developed. Else, keep your fingers off. ?

Mondo
--

_
Dr. Hartmut Schwahn
Lange Gasse 14
72070 Tübingen
Telefon: +49 163 1631495

On 14 May 2023, at 18:37, Mondovermona <ipc.de@...> wrote:

?

Yes, you're right no transistors in this circuitry. (different though in the OB-Xa and OB-X)

I agree with Jacob: after you have checked the CMOS memory chips, check D4, C14, C13, C2, C3, ?and also U14.

I would start with removing each of the CMOS memory chips and then CMOS IC U14 because this IC is also a socketed and thus easy to check. If the high current goes away with a party removed you have identified the faulty part.?

make sure that you can reliably measure the low current flowing from the battery during the procedure. Do not use anything else but a battery to check this. Do not use any other voltage source as a replacement for the battery. Use a battery operated current meter. ?You will otherwise obtain erroneous readings as ?low currents are involved. ?

Good luck
Mondo

On 14 May 2023, at 21:54, Mondovermona <ipc.de@...> wrote:

?

Paul, instead of current, you might want to measure the voltage drop across R30. You do not neet exact readings, just enough accuracy to tell the difference between ok. situation and high current draw situation.?

Mondo

[Edited Message Follows]

Hi.?

?

Have you finally checked the "good" chip on all three sockets? ?The problem might arise from a latched CE signal or the like in two of the three sockets. ?

Although the CMOS memory chips are prone to show this hi-drain behavior when faulty, I would not let the CMOS controller chips U13 U14 off the hook. ?Did you test with U13 U14 removed? ?

?

?In any case, its good to have some spare memory chips. ?

?

Look for the most recent issues of this type or, if you want to stick with vintage stuff, look after the "LP" versions, as there were two versions and only th LP version was meant for battery buffering. ?The "normal" version would drain your battery even faster.?

good luck?
Mondo

?

[Edited Message Follows]

Hi.?

?

Have you finally checked the "good" chip on all three sockets? ?The problem might arise from a latched CE signal or the like in two of the three sockets. ?

Although the CMOS memory chips are prone to show this hi-drain behavior when faulty, I would not let the CMOS controller chips U13 U14 off the hook. ?Did you test with U13 U14 removed? ?

?

?In any case, its good to have some spare memory chips. ?

?

Look for the most recent issues of this type or, if you want to stick with vintage stuff, look after the "LP" versions, as there were two versions and only th LP version was meant for battery buffering. ?The "normal" version would drain your battery even faster.?

good luck?
Mondo

?

Update:

Placing new RAM chips in causes the battery to drain at 0.001v per second!

With U13/14 removed and the old RAM chips placed back in, this also drains at the same 0.001v per second.

So, I'll assume U13-14 are the problem?

The new RAM chips were LP versions as Mondo recommended.

Any recommendations on U13-14 chip type?

Cheers

Paul


On 17/05/2023 15:13, Mondovermona wrote:

[Edited Message Follows]

Hi.

Have you finally checked the "good" chip on all three sockets? The problem might arise from a latched CE signal or the like in two of the three sockets.

Although the CMOS memory chips are prone to show this hi-drain behavior when faulty, I would not let the CMOS controller chips U13 U14 off the hook. Did you test with U13 U14 removed?

In any case, its good to have some spare memory chips.

Look for the most recent issues of this type or, if you want to stick with vintage stuff, look after the "LP" versions, as there were two versions and only th LP version was meant for battery buffering. The "normal" version would drain your battery even faster.

good luck
Mondo

Draining at 0.001V/s sounds like a measurement error. In your original
email you said you were losing 0.1V/24h. Why would this rate have gone
up by a factor 1000x?


Op do 18 mei 2023 om 13:44 schreef [ P J D ] <technogeist@...>:

Update:

Placing new RAM chips in causes the battery to drain at 0.001v per second!

With U13/14 removed and the old RAM chips placed back in, this also drains at the same 0.001v per second.

So, I'll assume U13-14 are the problem?

The new RAM chips were LP versions as Mondo recommended.

Any recommendations on U13-14 chip type?

Cheers

Paul


On 17/05/2023 15:13, Mondovermona wrote:

[Edited Message Follows]

Hi.

Have you finally checked the "good" chip on all three sockets? The problem might arise from a latched CE signal or the like in two of the three sockets.

Although the CMOS memory chips are prone to show this hi-drain behavior when faulty, I would not let the CMOS controller chips U13 U14 off the hook. Did you test with U13 U14 removed?

In any case, its good to have some spare memory chips.

Look for the most recent issues of this type or, if you want to stick with vintage stuff, look after the "LP" versions, as there were two versions and only th LP version was meant for battery buffering. The "normal" version would drain your battery even faster.

good luck
Mondo

I wish it was a measurement error, but it is not.

The original 0.1v 24 hour loss was with the old RAM. The new measurement
was using the new RAM.

Note that U13-14 were removed for the old RAM to read the same as the
new RAM with U13-14 in place.

Cheers

Paul


On 18/05/2023 13:48, Jacob V wrote:

Draining at 0.001V/s sounds like a measurement error. In your original
email you said you were losing 0.1V/24h. Why would this rate have gone
up by a factor 1000x?


Op do 18 mei 2023 om 13:44 schreef [ P J D ] <technogeist@...>:

Update:

Placing new RAM chips in causes the battery to drain at 0.001v per second!

With U13/14 removed and the old RAM chips placed back in, this also drains at the same 0.001v per second.

So, I'll assume U13-14 are the problem?

The new RAM chips were LP versions as Mondo recommended.

Any recommendations on U13-14 chip type?

Cheers

Paul


On 17/05/2023 15:13, Mondovermona wrote:

[Edited Message Follows]

Hi.

Have you finally checked the "good" chip on all three sockets? The problem might arise from a latched CE signal or the like in two of the three sockets.

Although the CMOS memory chips are prone to show this hi-drain behavior when faulty, I would not let the CMOS controller chips U13 U14 off the hook. Did you test with U13 U14 removed?

In any case, its good to have some spare memory chips.

Look for the most recent issues of this type or, if you want to stick with vintage stuff, look after the "LP" versions, as there were two versions and only th LP version was meant for battery buffering. The "normal" version would drain your battery even faster.

good luck
Mondo

Out of curiosity, what are the markings on your new RAM ICs?

Either way it sounds like you should stick with the old RAM IC's and
look for the problem elsewhere.

I would try replacing C14 next because it's an electrolytic capacitor.

Op do 18 mei 2023 om 14:54 schreef [ P J D ] <technogeist@...>:

I wish it was a measurement error, but it is not.

The original 0.1v 24 hour loss was with the old RAM. The new measurement
was using the new RAM.

Note that U13-14 were removed for the old RAM to read the same as the
new RAM with U13-14 in place.

Cheers

Paul


On 18/05/2023 13:48, Jacob V wrote:

Draining at 0.001V/s sounds like a measurement error. In your original
email you said you were losing 0.1V/24h. Why would this rate have gone
up by a factor 1000x?


Op do 18 mei 2023 om 13:44 schreef [ P J D ] <technogeist@...>:

Update:

Placing new RAM chips in causes the battery to drain at 0.001v per second!

With U13/14 removed and the old RAM chips placed back in, this also drains at the same 0.001v per second.

So, I'll assume U13-14 are the problem?

The new RAM chips were LP versions as Mondo recommended.

Any recommendations on U13-14 chip type?

Cheers

Paul


On 17/05/2023 15:13, Mondovermona wrote:

[Edited Message Follows]

Hi.

Have you finally checked the "good" chip on all three sockets?  The problem might arise from a latched CE signal or the like in two of the three sockets.

Although the CMOS memory chips are prone to show this hi-drain behavior when faulty, I would not let the CMOS controller chips U13 U14 off the hook.  Did you test with U13 U14 removed?

  In any case, its good to have some spare memory chips.

Look for the most recent issues of this type or, if you want to stick with vintage stuff, look after the "LP" versions, as there were two versions and only th LP version was meant for battery buffering.  The "normal" version would drain your battery even faster.

good luck
Mondo