Here's a 3 page PDF showing diagrams of how the cradle is wired. The cradle is a defective charger. It's meant to make it easy to switch between 1S/2S/3S/4S balance charging configurations.
Only the PCB is shown. The red/black circles at the bottom go to the positive and negative contacts of the battery bays.
I staggered where the wires exit the case to reduce the possibility of alligator clips touching each other.
Traces on the PCB are cut to electrically isolate bays 1&3 and 2&4, but not illustrated on my diagram. I stripped the board, but only to make room for my wires. A hot air soldering station or heat gun will greatly speed up this part of the mod.
It might be hard to see, but I used solid copper wire for the posts that the alligator clips will attach to. To attach them to the wires inside, I hammered one side flat, bent it into a hook, inserted the bare ends of the wires and hammered the hook shut. After taking these pictures I hot glued the posts down from the interior because it was too easy for them to fall inside the case. Don't forget to lube the sliders. It makes a huge difference.
The figure shown looks right to me (though I’m no expert). I’m not familiar with these traces you mention though?? If the circuit is as shown then it looks correct to me. What charger will you be using to log data? Iirc you just bought a reacktor??? The main leads from the charger will need to connect to the + of the first cell and - of the last. I’m assuming the alligator clips will run to a balance plug and into the charger.
Thanks for double checking for me. Yes, you understand the way I mean to attach the charger power and balance leads. I mentioned traces because it's a Nitecore i4, and the bays are connected 1/3 and 2/4. Surprisingly, I had some connectivity between the cathodes of bays 1 & 2 too, but maybe that shouldn't be a surprise since this i4 was defective.
I did just order the Reaktor. I originally wanted an iCharger, but ended up with the AC680 first, but now I have the iCharger too, at least a clone of it anyway. I'll use both chargers to log data, but I'll use the Reaktor when I want internal resistance data too.
Edit: I expended the sharing permissions further. Let me know if there are access problems.
Nice.
Does the thunder displays the individual cell statistics also on it’s I build display?
I can’t see Pictures of the cradle?
I always wanted to make one of these but my i4 works too good to kill it.
In my hobbycharger there is no need to use different balance leads for different cell counts as they are connected inside anyway.
So a 4S balance lead would work for charging 1-4 cells. So the cradle just needs one 4s balance plug one hard soldered positive/negative lead and 4 bananaplugs to put in the other positive/negative lead.
You mean the Reaktor? It should, especially since the AC680 does. I'll get some pictures this evening. Were you able to open the pdf? Here's the first diagram from the pdf.
The other diagrams show 2S, 3S and 4S configurations.
I think Werner is implying you only need one balance plug, just fill the cells into the cradle from one side to the other, then move the main charge lead to the last cell. The balance plug can stay in the largest port.
Oh, for sure. I'll use a 6S cable most of the time. It'll have alligator clips on the ends, so I'll only use what I need, whether that be 2S or 6S, or just not attach the clips to the cradle for 1S.
There aren't any additional balance wires inside the cradle. All the wiring does is move the contacts away from the sliders. At first I was using strips of solder wick attached to the contact points of the sliders, but that was tearing up the wrappers of my batteries, and it was way too difficult to keep clips apart. This takes care of both problems.
Now I'll see about taking some pictures. I wanted to take it apart anyway to add plastidip and hot glue to a couple spots.
To save the hassle of so many aligator clips and keep it simple you could just solder the wires from the cradle to the balance plug. Then you would only need to worry about the aligator clips on the main leads.
I'll add pictures of the modified i4 to the top post in a moment. They're uploading now.
That would have worked, but I only have 2S and 3S balance cables. I have more in the mail, but I didn't want to wait. Another reason to use clips is that I could use this with other cradles to do 6S balancing.
It took looking at the discharge graph for me to realize that doing a discharge test with cells in series with balance cables wouldn't produce the same data as doing it one at a time because the balance wires are bleeding off power occasionally, which lengthens the charge time. It's still useful for finding the capacity as long as it's not critical to know the current or time though. If I'm wrong, and I hope I'm wrong, please let me know.
This mod is still worth it for me, but now I don't see the need for a second one unless I get something like the Supbeam X60M that uses six cells in series.
Could a lot of resistance in the wiring cause some horrific test results? None of the cells that I've charged with the new cradle have performed well. The voltage sagged so much with the cell below that I thought it was dead, then somehow it kept chugging all the way to 2139 mAh. I switched to HKJ's method of using a c-clamp, and my first cell off the charger with the c-clamp holder had a much better looking discharge charge and another 300+ mAh.
That connection would almost certainly have to be in the charger since I was simultaneously using bays 1 & 4 with separate hobby chargers. I'm not reading any resistance with my multimeter. I'll try getting one of the "bad" cells tested again. Unfortunately it'll take about ~7 hours.
It's only 8 minutes into the new charge, but it's already looking much better. Something must be wrong with the i4 cradle.
Here's that same cell retested in the c-clamp holder. The charger was in li-on discharge mode instead of li-po discharge, but still discharging to 3V. The graph looks better, and there was another 240 mAh.
Why is that looking so much better than when it's charged in the i4 cradle?
Both of these tests were run with the AC680 monitoring software, which is buggy, but I was getting the same type of surprisingly bad results with the Reaktor while using Logview for logging.
I would think the capacity difference is due to the cradle set up reading voltage droop. An idea to check that would be discharge a cell to 3v using the cradle then remove it from the cradle and check the voltage.
