Power Function Rechargeable Battery 8878

[kage28]

Hi all,

So I left my 8878 untouched for way too long, thus totally drained and it refuses to charge after numerous attempts, some attempts going into days (just flashes orange, and unable to turn on after unplugging). Looks like the only way out is to attempt a change of the internal cells, and I've went around looking at potential replacements. Since I can read Chinese, I went scouting around Taobao and found a few potential candidates (mainly based on voltage, capacity and dimensions). I was wondering if any of the 3 cells pictured would be suitable replacements? My electronics/ soldering are basically non-existant, and I would probably engage a 3rd party to do the actual replacing.

Thank you very much

 

A.) 2 wires output, no connector

B.)2 wires output, no connector

 

C: 3 wires output (3rd wire for NTC protection? is it the same as the 3rd wire in 8878?), no connector

 

To refresh, this is the internal circuit of 8878:

 

Edited January 11, 2022 by kage28
Added 8878 circuit picture

[Toastie]

10 hours ago, kage28 said:

just flashes orange, and unable to turn on after unplugging

Hello there,

couple of questions before diving deep;)

1. What type of charger do you use, the original LEGO charger or a third party device?
2. Did the 8878 turn on with the charger connected to it?
3. Did you already open your 8878? This is really straight forward, and then you can check for the type of LiPo they used on that batch. If I am not mistaken, at least 2 of my 8878's have two pouches, each 3.6V. Maybe it is better to go with that, but I do not remember the wiring.
   EDIT: Sorry, first read then babble ... you have already selected the 2S type.

@Philo's pages regarding tech stuff are superb, this one covers LiPo's for PoweredUp, so not exactly what you want, but is a very nice recource.

Let's talk about 1-3 first and then let's see what's next ...

Best
Thorsten

Edited January 11, 2022 by Toastie
First read than think then write ...

[kage28]

1 hour ago, Toastie said:

Hello there,

couple of questions before diving deep;)

1. What type of charger do you use, the original LEGO charger or a third party device?
2. Did the 8878 turn on with the charger connected to it?
3. Did you already open your 8878? This is really straight forward, and then you can check for the type of LiPo they used on that batch. If I am not mistaken, at least 2 of my 8878's have two pouches, each 3.6V. Maybe it is better to go with that, but I do not remember the wiring.
   EDIT: Sorry, first read then babble ... you have already selected the 2S type.

@Philo's pages regarding tech stuff are superb, this one covers LiPo's for PoweredUp, so not exactly what you want, but is a very nice recource.

Let's talk about 1-3 first and then let's see what's next ...

Best
Thorsten

Hi Thorsten,

The answers to your questions:

1. I used a 12V power source, after reading your research that using adapters between 9v to less than 20v are fine.

2. Yes, during “charging” it turns on, but goes dead as soon as the plug is out or wall power source switched off.

3. Nope I have not opened it yet, don’t have a drill yet to remove the plastic covering the screws. 

Also, the truth is… embarrassing.  I have bought this battery way back around 2011, and totally forgot about it and keep it in storage until…. Recently. Yes… 10 years unopened. In other words, no, I have never actually used it ( I know I know……what an idiot!) and since I have read that people have only left it in storage for months before encountering a similar predicament as I have, I conclude the cells are way beyond salvageable. So thought I’ll do this somewhat as a learning project and see if I can rescue this from the ditch and learn a thing or 2, plus, can’t hurt to have another usable battery pack for my growing train collection!

My main concern is actually the wiring, since LEGO uses a 2S cell with 3 wires. Not sure what exactly the 3rd wire does exactly in conjunction to the PCB and how critical it is to safety or even proper functioning. Also, so far I have not come across any resource or examples of anyone changing the cells. A example of that would really be educational methinks, and I can at the very least to let my appointed solder-er know some potential pitfalls to avoid, or be mindful of certain things.

Thank you very much.

[Toastie]

OK, next test ;) When you have your 8878 on the charger (12V is perfectly fine!), does it function properly, let's say with a PF motor attached, using that orange dial? Does a PF receiver work properly as well? That means everything is fine with the output electronics.

18 minutes ago, kage28 said:

Not sure what exactly the 3rd wire does exactly in conjunction to the PCB and how critical it is to safety or even proper functioning.

The 3rd wire in plain vanilla 2S LiPo packs (= nothing additional fancy built into the pack of two) is for balancing the charging voltage. Individual LiPos (each "3.6/7V" cell) exhibit individual aging and charging behavior.  A fully charged LiPo cell reaches about 4.2 no-load voltage. Charging it with higher voltage than that will most probably damage it. When you use just two wires, you'd assume that the charging voltage - let's take for simplicity 8.4V - is equally distributed between the two cells connected in series. However, as current flows, the voltage drops across the two cells may be different, and may lead to a situation, where one cell sees 5V and the other the remaining 3.4V - which means one cell will certainly suffer. Upon monitoring the voltage at the point, where the cells are connected, the charger can "balance" the different charging behavior of the two cells.

So the three wire cell(s) would be my favorite.

I'll open one of my 8878 tonight and let you know what's inside, OK? I also bet, that others have done that already and may very well know what to do!

Best,
Thorsten   

[kage28]

Thanks very much for the information!

Regarding the “load test”  I’ll have to test it come morning (it is 1:30am where I am)

Thanks again!

[Toastie]

UPDATE ;)

Opened up one good working 8878 from back then.

The LiPo pack is of the type 2S, with 3 wires exposed, black (negative terminal), red (positive terminal) and white. The third (white) wire is indeed going an NTC (thermistor = temperature sensitive resistor with the behavior the warmer, the lower the resistance) wired to the negative terminal of the cells. Quickly checked: The resistance between white and black is about 8.8 kOhm at room temperature and swiftly drops to 8.4 kOhm when in my hand - so this seems to be a plain vanilla 10k NTC, commonly used in LiPo packs. In other words: The electronics of the 8878 is monitoring the temperature of the cells for not (over)stressing them.

The balancing of the two cells (I cannot imagine that TLG would go without, it would be against all the safety rules, even in 2010 ;)) thus seems to be done with a small protection board sandwiched between the two flat cells. I cut the enveloping shrink type plastic away and yes, there >seems< to be something between the cells other than the NTC, which is there as well. Can't get any further without risking of damaging the cell pouches, they seem to be "glued" together. Damaging would be bad (cf. Ghostbusters: "I am fuzzy in all good bad things, what do you mean: bad?" - as they may go off in flames).

I'd thus say what you need is a 3-wire 2S LiPo pack, where the third wire is an NTC output and not the balancing lead. Plus the LiPo pack needs to have the balancing electronics on board. I believe - but I would only take this route when replacing the LiPo pack.

Here is a two photos:

Best
Thorsten

 

Edited January 11, 2022 by Toastie

[kage28]

Thank you for the very informative guide and explanation Thorsten, very much appreciated! Looks like if everything else was working ok, this replacement cell would be the one to get, and a swap would pretty much be enough:

 

 

Quote

OK, next test ;) When you have your 8878 on the charger (12V is perfectly fine!), does it function properly, let's say with a PF motor attached, using that orange dial? Does a PF receiver work properly as well? That means everything is fine with the output electronics.

However......the bad news...  I did the testing for this:

1. Direct to train motor: No response at all when turning the dial either way

2. Via the IR receiver and then to train motor: no response, and the LED on the receiver doesn't light up as well.

Set up the above to the standard AAA battery box (88000), all was working as expected, so it is clear 8878 is the problem.

Looks like I have to get my hands on a hand drill pronto, to see whats really going on inside the battery pack, and maybe a multimeter as well.

 

 

[Toastie]

6 hours ago, kage28 said:

this replacement cell would be the one to get, and a swap would pretty much be enough:

Yes, this LiPo pack looks very good indeed. Any chance to have a link to the company selling it?

Sorry to hear that your 8878 does not behave on the output - but maybe there is some protection circuit telling the output electronics not to do anything, when the LiPo is "bad" - who knows. After all, LiPo's want and need to be treated carefully, dead or not.

Yes, having a look inside may be good: I used a 4.5 mm sharp drill on a cordless electric screwdriver, but it can be done manually as well, takes a little longer. The very old 8878 have a tiny plastic cap covering the channel leading to the screws, the more recent models have slightly strengthened "caps" (this is what I remember from opening one 8878 back in 2013 - the one I opened yesterday seemed to having thicker caps). Once the caps are removed, you need a "thin"crosshead screwdriver, as the screws sit "deeply buried" in the plastic channel connecting the bottom cover to the top panel.

Good luck!!! And be careful not to physically harm the LiPo pack.

Best
Thorsten

[kage28]

6 hours ago, Toastie said:

Yes, this LiPo pack looks very good indeed. Any chance to have a link to the company selling it?

Yes, this is the LINK to the family of products on Taobao. Note, site is in Chinese, and you may need a friend within China to help to ship it internationally.There might be similar ones in Aliexpress, l'll probably try to look for it later.

6 hours ago, Toastie said:

Yes, having a look inside may be good: I used a 4.5 mm sharp drill on a cordless electric screwdriver, but it can be done manually as well, takes a little longer. The very old 8878 have a tiny plastic cap covering the channel leading to the screws, the more recent models have slightly strengthened "caps" (this is what I remember from opening one 8878 back in 2013 - the one I opened yesterday seemed to having thicker caps). Once the caps are removed, you need a "thin"crosshead screwdriver, as the screws sit "deeply buried" in the plastic channel connecting the bottom cover to the top panel.

Thanks for the tips... I "drilled" using a screw driver and  pull out the plugs limiting access to the screws. Sucessfully opened.... Behold! The cells in all their glory..... or not:

Looks like really bad news.... and it seems to me that I better cut them loose and properly dispose of them as soon as I can! Given my lack of tools, can I simply just snip off the 3 cables one at a time?

Lets take a look at the PCB:

Looks ok to me.... nothing looks fried.... and it shouldnt, since they have pretty much never been used.

Pardon the small pictures... but EB limits me to only slightly over 100KB overall uploads per post.

 

 

Edited January 12, 2022 by kage28

[Toastie]

4 hours ago, kage28 said:

one at a time?

Wow! One looks like gone bye bye since long. 

This: Yes, cut one wire at a time. First ground (black) then VCC (red), then white.

The pouches are chemical waste - I believe though that the blown up is also chemically dead, the lower one that is. Be careful with the upper pouch. Seems to be intact. As said, this is hazardous waste - don't throw into the household waste bin.

Now, as you are researching this to a level I believe no one has reported here on EB: Is there some sort of electronic circuity in between the pouches? And please use cheap water-resistant gloves for doing this. After work is done, throw them away.

With regard to picture upload: EB is not a hosting site; use whatever hosting site is available to you (Flickr and the like) and copy/paste the links to the picture into your post. As far as I am concerned: I believe this is excellent information you are providing here.

PCB looks good, indeed.

Really exciting! Thank you very much for the a) the link and b) all your efforts!!!

Best wishes,
Thorsten

[kage28]

5 hours ago, Toastie said:

Is there some sort of electronic circuity in between the pouches?

Happy to to that, though I have to ask: How do I do that? With a multimeter? I have one arriving tomorrow. Where do I probe? I assume I can do that even after I detach the wires/ cells from the main PCB. 

Edit: Misunderstood the question on 1st pass. No prob, I’ll separate the pouches and see what kind of circuitry/ PCB is between them. For science! 

5 hours ago, Toastie said:

With regard to picture upload: EB is not a hosting site; use whatever hosting site is available to you (Flickr and the like) and copy/paste the links to the picture into your post.

Got it, thanks!

5 hours ago, Toastie said:

Thank you very much for the a) the link and b) all your efforts!!!

Not at all, thank you for guiding me instead!

Edited January 13, 2022 by kage28

[kage28]

Pictures of the connections between the cells in 8878

8878 7.4v battery cells by Kj Toh, on Flickr

8878 7.4v battery cells by Kj Toh, on Flickr

8878 7.4v battery cells by Kj Toh, on Flickr

8878 7.4v battery cells by Kj Toh, on Flickr

QR Code/ link to the 703048 1S cell page (many products on the same page)

Edited January 14, 2022 by kage28
added qr code link to the store's cell page as the earlier link can't seem to load

[kage28]

Hi all, to all who might be interested, especially @Toastie , who had been a great help, an update on the 2S-cell replacement for 8878.

After a couple of months since Jan 2022, with the help of friends in China, I have finally received my replacement cells to start my project to overhaul the 8878 that I have never got to use.

I've received the cells on the 22nd March 2022. 2 days later, I found some time, and gathered all the necesary stuff and headed to the place I've scouted beforehand to do the necessary de-soldering of the old wire stumps and to solder on the new cells.

Unfortunately, I underestimated the actual size of the cells and the measurements are a little too big (length by about 1.5- 2 mm or so, width just a hair wider than 30mm, and some trimming of the plastic will be necessary to fit the new cells into the compartment.

New cells by Kj Toh, on Flickr

New cells by Kj Toh, on Flickr

Af course, before that, got to charge and test the battery:

Prior to charging, I wanted to see it it powers on at all. No problems there.

New cells by Kj Toh, on Flickr

 

 

 

Using a 5V-12V converter with min 1amp current, I've started charging it. Red lights blinking....which is normal (even with the spoilt old cells).... the real test is whether it stops blinking and become solid red, which means fully charged.

After about 20 mins, solid red...... SUCCESS!

Charging by Kj Toh, on Flickr

After disconnecting it from the charging cable, I've connected it directly to a train motor to test if it can run (previously was a no-go, even with the battery directly connected to the charging mains)....I turn the battery on, and then the dial......SUCCESS!

 

Now, I haven't really done any form of heavy testing (I only bring out trains from storage for my boys... and myself   at the end of the year), but I have little doubt that it will perform like new, or at the very least, close to it. Now, the only thing left to do is to fit it into the battery compartment.... and discharge it slightly for storage til the later part of the year.

 

 

Edited March 25, 2022 by kage28
Typos mainly

[Toastie]

8 hours ago, kage28 said:

I turn the battery on, and then the dial......SUCCESS!

CONGRATULATIONS!!!

This is so nice to read. Now you have brought a beautiful piece of engineering back to life! It feels good, doesn't it?

Really, really great. 8878 is such an expensive (but nicely equipped) piece of electronics - that damned chemistry box causing the trouble is now, as a result of your carefully executed and very nicely documented work, is now replaceable!

Thank you very much again!

Best regards,
Thorsten 

[kage28]

Quote

CONGRATULATIONS!!! 

Thank you, I couldn’t and wouldn’t have have gone ahead without your valuable input on this, so my thanks instead!

Quote

It feels good, doesn't it?

Absolutely! Especially considering the fact that I’ve never got to use it at all during its original life cycle!

13 hours ago, Toastie said:

8878 is such an expensive …… piece of electronics - that damned chemistry box causing the trouble ……… is now replaceable!

Honestly, given the God-tier levels of expertise on these boards regarding what had been done modifying existing battery packs for LEGO,  I couldn’t believe the fact that I could not find any video or article about anyone replacing the cells for 8878 as of Jan 2022. I’ve always wondered if it was mostly because the God-tier customisers already made their own solutions for battery units, eschewing 8878 altogether?  Or maybe the ones who have the skills are also the ones who have taken better care of their batteries, hence no need to perform such a “surgery”.
 

Either way, for the people who have the same problem as I had (or will have, in the future), your 8878 CAN ( previously only theoretical, but now with empirical proof) be given a new lease of life after the stock cells expire, as they will, inevitably.

Just err on the side of smaller dimensioned cells.