Hello all,
I have a couple Timemaster HO cards with the original Data Sentry branded NiCad batteries sodlered on. One of the PCBs appears to have fared pretty well, with just a little corrosion on the solder pads. The other though appears to have some trace damage perhaps from the battery leaking.
Do any of you have preferred methods for cleaning a board which suffered NiCad leakage after removing the dead battery? Internet results are a mixed bag and I trust you guys more. I just want to be sure I safely remove the old batteries and any corrosion risks, before installing a modern replacement.
Cheers,
Alex
Can you post some pics? I generally either clip the leads or desolder. Cleaning I would normally do with concentrated alcohol. 90% Iso or better, or something like Clear Spring or Everclear. Gently with a cotton swab. When I replace a battery I usually prefer to use some kind of battery holder rather than soldering the new one directly on.
Nickel-cadmium cells' electrolyte is potassium hydroxide, a strong base, so it is effective to clean using (2–4%) acetic acid. Cleaning with an acid guarantees that you will react any KOH left on the board, ensuring no future corrosion risk. The acetic acid will evaporate on its own, but you can rinse it using an alcohol-water mixture to flush away as much as possible of the salt contaminants that are a product of acid-base reactions. You can also rinse using water spray from a sink faucet or use an ultrasonic cleaner if you have that available. Cleaning corrosion damage is often done using multiple steps like these to leave the cleanest board possible for the following repair steps.
Unfortunately, much of the leaked KOH will have already reacted with copper and formed black copper hydroxide, an insulator, which will have made circuit traces inoperative. In some cases you may be able to scrape away the black oxide, re-tin with solder, and call it a day. Otherwise you are looking at replacing the lost copper using wire, copper foil, or other reconstruction techniques.
Thanks guys, pictures follow.
I indeed want to neutralize any KOH, if needed. I noticed something on the trace that concerns me most; it looks basically delaminated from the substrate, and I have seen this on several 5.25 Apple Disk II controllers as well. I have put a picture of that last.
No worries on having to replace traces with wire or other repair if they are actually damaged, that I can almost certainly handle :)
20230103_151152.jpg
20230103_151042.jpg
The trace with the blue hand writing on it is the one that I am curious about, here is the Disk II card I have that looks similar (I have several that look like this). I have to wonder if it was an issue with heat in manufacturing or something that delaminated the trace, but I am certainly open to being wrong on that.
20221230_130507.jpg
That "wrinkled" effect seems to be caused by solder flowing under the mask layer during wave soldering. At least that is what it looks like going from images, perhaps up close and personal it is obviously different from that. If the trace moves (even slightly) when probed it could instead be a sign of corrosion, but in my experience corrosion from battery leakage is patently visible on the surface and not hidden.
Your first pic looks like the trace coming from pin #38 of the PIA is cracked? If so I would use a toothpick to put a tiny drop of epoxy under the lifted side, clamp it down somehow, wait 24h for it to cure, scrape back the mask over both sides, and tin the exposed traces, then use a 0-ohm SMD resistor or short section of wire (or small section of cut copper foil) to jump over the sides. The easier method is to just glob on solder so it sticks to both sides, but that can be risky because of the extended time you might mess around heating the board with your iron.
Repairing a cracked or scratched trace was one of my first successful repairs, on a 30-pin memory SIMM. I was amazed that the crude tools I had available at the time (no flux even) could do that, but it worked.
Thank you, I appreciate the info!
I have done a lot of different trace repair for arcade games, I'm just very careful with old batteries in particular. I don't want to leave behind a corrosive time bomb.
I'll desolder the old batteries and check any suspect traces and go from there. I opted not to do holders for now because I found proper batteries with the same PCB footprint. They should look pretty original other than any possible repairs.
You should use vinegar, (probably a glass fiber brush for mechanical cleaning), IPA (isopropyl alcohol) and afterwards destilled water and drying to remove any corrosion.
https://www.eevblog.com/forum/repair/cleaning-and-repairing-battery-corrosion-on-a-pcb/
the wrinkling is of no concern. my TMHO has been like that since I bought it in 1984.
wanna sell one?
The wrinkles seem solid on the Disk II cards when prodded. It would make sense as a possible side effect of wave soldering on these two cards... today I learned :)
I am not looking to sell these, I have machines in mind for them, but if I get them both working I'll keep you in mind. These came from another member here, so if I did send one to you I would not ask much. I'm also always up for interesting trades.
I've seen it on ZX Spectrum motherboards, Commodore 64 motherboards and all sorts of peripherals of the era.
Nothing to worry about as long as it's not discoloured. Discoloured usually means something corroding under the soldermask.
Chesh
+1 to this... spot on and neutralize it first. I 'd also agree with the later comment about the large trace "look". Probe it, it could be prefectly fine thick trace. Things don't always look great, but as long as it's got solid form I'd bet excess solder and that shouldn't be a problem.
If you're talking about what I think you are, I thought that was a white hair. But super easy enough to check. =)
That all makes sense. It sounds like this is the plan:
-Spot clean what I can with acetic acid 1st, then iso (mainly just the corroded looking solder pads at this step, though I'll be removing that corroded solder with a vacuum desoldering iron)
-Desolder the old battery / remove dead solder
-Clean the battery end of the PCB with acetic acid and alcohol to be safe / complete any reaction
-Inspect/verify/probe traces, repair if needed
-Install new battery
I'll give this a go on each card and report back. Much appreciated as always.
Are you planning on replacing the NiCad?
Yes but not with NiCad, I'll be replacing them with a new NiMH of the same footprint and voltage. The brand is a four letter word around here but this is the one:
https://www.amazon.com/dp/B079VQP6X8
skate323k137 wrote:
Yes but not with NiCad,
The charging characteristics of NiMH are quite different than NiCd so take that into account.
As others have already said - no worries because of the wrinkling. I have a bunch of Apple III cards which are all like this - and always have been, since the day there were new. That's 1980s PCB manufacturing for you. Often mainly affects the wide traces. But as these Apple III examples show, sometimes all traces were made like this. Hasn't been an issue for the last 42 years. They are all solid and will easily outlive any of the semiconductors... ;-)
WaveSoldering.jpg
Nice job, I was asking exactly for this reason!+1 on the other Jeff's comment, NiCad charge profile is different and not good for NiMH. Unfortunatly that's not an easy thing to work around. Will either have to figure out a different charge method, or stick with the chemistry the board was desiged to support.
Normally I would completely agree.... but this battery is advertised as a replacement for the data sentry NiCad. Not that I trust youtube (usually don't) but I did find a video of someone installing this same battery on one of these cards. I'll do a little more reading though to be safe.
You have nothing to worry about with charging profiles. The charging circuit for the NiCad was probably just a crude - overvoltage to get charging to occur and then (most likely) shutoff of the charge voltage for the remainder of the power up period.
You won't really do any harm to the NiMH battery. I've charged hundreds of NiMH AA and AAA cells using old NiCAD chargers and they often charge better than with the chargers the batteries came with.
I can guarantee you that the "charging circuit" on this board is just a single resistor to maintain a constant float charge on the battery. I only mentioned this because many people replace the NiCd with a Li based primary battery. These can power the board for up to 10 years. And in this case, you'd want to remove the charging resistor.
I have done that with several Neo Geo motherboards [remove the charging resistor and just install a different non-rechargeable battery]. I think I'll be OK here with these replacements, but that is certainly on the books as a backup plan.
NiCd is special because it can be "trickle charged" indefinitely without being damaged. A common rate for trickle charging is C/50, in other words, the rate that would take 50 hours to run the same number of coloumbs as the capacity rating through the battery. [It is equal to the capacity in amp-hours divided by 50 hours.] However, since NiCd has very low charging efficiency, it normally takes 100 hours to go from empty to full State-of-Charge this way. But sometimes higher rates like C/20 were used. As jeffmazur noted, designing this charging circuit only requires a single resistor between the Vcc line and the battery's + terminal to appropriately limit the charge current.
NiCd also has fairly high self-discharge characteristics which make trickle charging desirable, as well as simple and cheap to implement. The continuous trickle charging compensates for the self-discharge, and the excess energy is simply lost as heat.
NiMH has much better charging efficiency, lower self-discharge, and doesn't tolerate being trickle charged that well. Normally a NiMH charger requires a charge cut-off based on time, temperature, and continually measuring the open-circuit cell voltage. A NiCd rapid charger also needs to implement much of that, and probably works better on NiMH cells than a trickle charger.
I appreciate the abundance of caution. Here is the part number of the battery I purchased, and it's datasheet:
Varta 55615-703-012
https://static6.arrow.com/aropdfconversion/d4e730ae615f737912312eadc91f66baa1e54df7/pi_55615703012.pdf
The data sheet says "Direct replacement for Ni-Cd," and indicates the main usage of this part is RTC/Memory backup. So in this case I believe the manufacturer intended these to work for this application.
What I learned (years ago) is that it's the NiMH battery doesn't allow the charge circuit to sense a complete charge, unless the charge is based on a timer, there's a good chance of over charging.
One other difference, like you were discussing is that NiCd can easily handle charge rates well above 1C. Since both NiCd and NiMH charge under only CC, there's a good chance there is no trickle and would be full charge after fully charged until stopped. I'm not sure about the use of a rapid charger, but I would expect the charge circuit on the expansion card to be fairly simple. Designers are always asked to cost resuce to the bare minimum to meet the needs.I'd probabily hook up a data logger and monitor the charger output to understand the charge cycle used by the card.
It's possible the Varta includes some passive components in the pack to mimic NiCd. It's also possible that they mean it's a dimentionally and connector compatible battery. We're also tlaking about 2020 batteries being used in 1980's system. This is not the most common application for these batteries. I'd likely make a 5 minute call to Varta's tech support to confirm the battery does what you think and is compatible with NiCd charger circuit from the 80s. The agent may be surprised by the question, but should have some useful feedback.
That actually sounds like a fun and useful phone call. I like that idea :)
I have not heard back from Varta yet, however, I went ahead and removed the old batteries and cleaned up the cards. The Rev 5 suffered less corrosion, the Rev 4 needed a good cleaning under the battery after removal. Both boards were cleaned with 3% acetic acid, and then alcohol swabs.
I went back and checked the one youtube video I had found (https://www.youtube.com/watch?v=eywEZl5MZY8) and the guy leaves it for a month at the end before rechecking it. I certianly know this does not satisfy all the questions regarding charging characteristics but I will follow up if/when I hear back from Varta technical support. However that was enough for me as someone who just uses the machines casually to feel OK installing these for now.
Both cards seem to work fine with the new batteries. I will re-check them in several days, but I think they will be OK. The most corroded trace on both boards actually appears to just run to a test point, but I did appropriate repair on pads and checked trace continuity as I went along. Overall not a bad repair at all, I just wanted to be safe cleaning the NiCad, so I appreciate the help there.
tkiL2B2.jpeg
VqAbLSk.jpeg
Looks great! I'm not sure if it wasn't clear but there should not be an issue with the bulk of the chage, but rather what happens when the computer is running longer than a charge time and if it continues to supply current into the battery. If that happens, I don't know how the battery will react other than expecting some sort of damage or decrease in usefulness. I'd expect that clock ot run a long time before draining that battery, can't draw much current.
I have some of the AE RAM keepers from the GS if I get a chance I'll take a look at the circuit and maybe characterize exactly what the charge cycle is.
Gotta love the 4 character year!!
Guessing he had to change the century digits in the Basic program...
Also, the clock chip doesn't really "drain" the battery per se with NiCds. The self-discharge rate actually determines the battery life. At roughly 20 uA, even the NiMH battery would last for over a year.
For the curious, this is the reply I got from Varta:
"The datasheet does not indicate any additional components in the 3/V 150H NiMH battery pack, just three (3) cells in series to create 3.6V.
I do not see any passives such as PTC or temperature monitoring thermistors or similar.
Our attached Handbook contains charging information and usage suggestions."
As not to re-host the handbook, this is the documentation I was provided. https://www.varta-ag.com/fileadmin/varta_microbattery/downloads/service/battery-documentation/nickel-metal-hydride/Sales-Literature-201810_HANDBOOK_Rechargeable_Button_Cells_NiMH_en.pdf
If this is indeed the case, why do you guys even bother with rechargeable batteries?? The NiMH ones will not last 10 years for sure.
On a board where the footprint is already for a coin cell, I will usually disable a charging circuit and install a normal coin cell or holder. In this case, the battery was this unique size/shape/footprint and this offered an opportunity to keep them looking reasnoably original. If they don't last several years then surely I will consider other alternatives.