Cherry67
Member
- Joined
- May 13, 2018
- Messages
- 518
Hi Guys, as anounced in my powerwall thread i will describe a few issues of my current, nearly finished, 2nd harvesting campaign.
I am intersted in any comments you might give, since i do several things a bit different than the most of you.
I will talk about
- my harvesting style (done)
- coming result of my harvesting campaign on real bad packs
- Imax B6AC usage (DON'T use the clones)
- Opus usage (new for me) (Care for the heat dissipation) (done)
- Resistance measurement (want to have)
I will not be able to describe that in one run, so stay tuned and show by occasionally if you are interested.
Harvesting style
My first campaign went mostly using my B6ACs, at that time 2 clones.
At that time i had only normal use in mind, the idea of a powerwall issue came later
I had some long-term island supply in mind, for party and weekend use, so I had no peak power on the cells. I guessed i will stay at 1/10 C max loading, and 1/5 C of max charging current. Thus i can be more genereous in using supoptimal cells, but of course nothing known risky.
The imax are nice to process the cells in bulks, means it can charge, and especially balance, arrangements of up to 6s, and to even increase the amount of cells in operation i went up to 6p simultaneously.
I must mention that i leave the usual pair of cells together. I guess that these cells have the same history and should behave quite similarly so my guess was there is little advantage in processing them separately. Additionally, that leaves the originally welded connector stripes on the cells, and with a little foresight i where to cut them each cell gets a nice Solder point on the stripe, which eases soldering and reduces the heat transferred to the cell. Disadvantage is that the soldered package looks gruesome, at best. (Even the triples wil be reworked to be pairs, i separate one cells, this are processs as single and re-paired).
In general i seem to use lower currents than recommended here. Since i process usually 3-4s 6-8p, i have 18 to 28 cells "on the board", which are processed together, i can easily take the time for a 1/10 C load. Discharging was to slow with the B6AC, it has a 2 Amp/5 W limit.
I built a small extender, basicaly a current resistor, operational amplifier, power transistor and vented cooling, which measures the (discharging) current and surges another 9 times the current of the imax off the battery itself.
So the imax is unmodified, everthing is external, but i can easily consume 50 Watts, if i like. Now, its only in steps of 1,2,3 4... Amps, but alas, this way the current and Ah display is just the tenth of the real. When the imax is accurate at all, which it is not. I will talk about that later.
Then came the powerwall idea, together with PV usage start, and i hade usage of my about 100 cells to build my fisrt two modules, as mentioned in the other thread. Since i will supply only basic load (no full islanding), the low power request on the cells ist the same.
-------
2nd harvesting campaign
Now to the second, my current, harvesting campaign. I got hands on a bunch of expectable bad cells, about 50 Laptop packs, and wanted to recheck and improve my processing style.
(my third campaign is already planned, i will buy cells from a board member here)
My imax stock has grown to 4 units, because 2 chrashed ( my fault), i ordered new ones, but was able to repair the broken ones.
I had to improve the connection of the cells to the Imaxes, i built "Boards", about 22*30 cm (8 by 12 inches) which got a balancing connector, plus the main plug cables, to a connector stripe. The cells are are arranged mostly as 3s 2P as described, and i put 3 or four of these on the board. wires and balancing wires are soldered to the connector stripe. In the end i had about 5 boards in operation, which gave me quite a capacity for cells in process.
The cells are harvested from the packs, separated to the pairs, and Voltage measured. A pair with a voltage over 3 V goes to a board, cells below are slowly fed up with laboraty power supplies to about 3,3 V. I did this in the past already, before i read the same experiences here on the board.
On soldering them together on the board i look a little bit to voltage differences, i care not to to connect a good cell on 80 % SOC to a 3,3 V cell.
Then they get (usually) 3 Cycles on the boards, with 1/10 C to 1/5 C parameters. I have experienced that the capacity increases on the second cycle, and a bit on the 3rd cycle, so something happens there. This takes a complete day per board per cycle, but this processes the mentioned about 20 or so cells to gether, and that is not so bad.
All capacities on discharging are noted on the boards, together with the current used. I decide on the cycle numbers on the capacity results, if nothing happens with the capacity i load them fully, separate them from the boards for sending them to stock for 14 days, to the self discharge check procedure.
I do not so much care for the absolute capacity of a board or the cells, usually the are not paired very good, means they are not equal in capacity, and one cell level is empty first, when others gabe still 3,6 V, as example. Since i mainly want cycling them, that is good enough for me.
The last step, after discharge check, has to determine the capacity. Since that is not so good with the imax, when they are different, my newly improvement is an Opus CT 3100, just for discharge capacity measurement. The cells still go in there as pairs, that was the idea. The bad news was that the opus not only gets HOT especially on discharging in all four slots, the poor fan is just an excuse. The temp in the Opus is one thing, but I measured the cell(s) on the right side get heated up to about 50 degress C (122 F) just from the Opus heat.
I dont want that, not just so from the point the cells cannot stand that, but merely of two considerations: The capacity result maybe different, AND, the cell, as a "bad one" MAY not like the temp. IF the cells are ready to go better not to give them too much excuses to go....
So, i built a small stand with a bigger fan, plywood of 6 mm (3/10 inch), in the most simple way.
The opus stands just in there (with removed bottom plate), nothing is attached or screwed to the opus.
The back side is left without a piece to support the Opus, this opening is for air exhaust of the fan to get air streaming out.
The new fan.....
....is plugged in to the opus instead of the own one (its a 12 V original fan, about 65 mA). Sure my new one was checked to match that. I cared not to overload the small switching transistor in the Opus.
Now the same right cell gets to about 30 C, on 26 C ambient, for a 1 Amp discharge. That is better.
Btw, i read about the current/power supply issue of the opus in an older thread here, i will care for this if i want to use the loading features. Old PC PS are in stock here, of course.
I am intersted in any comments you might give, since i do several things a bit different than the most of you.
I will talk about
- my harvesting style (done)
- coming result of my harvesting campaign on real bad packs
- Imax B6AC usage (DON'T use the clones)
- Opus usage (new for me) (Care for the heat dissipation) (done)
- Resistance measurement (want to have)
I will not be able to describe that in one run, so stay tuned and show by occasionally if you are interested.
Harvesting style
My first campaign went mostly using my B6ACs, at that time 2 clones.
At that time i had only normal use in mind, the idea of a powerwall issue came later
I had some long-term island supply in mind, for party and weekend use, so I had no peak power on the cells. I guessed i will stay at 1/10 C max loading, and 1/5 C of max charging current. Thus i can be more genereous in using supoptimal cells, but of course nothing known risky.
The imax are nice to process the cells in bulks, means it can charge, and especially balance, arrangements of up to 6s, and to even increase the amount of cells in operation i went up to 6p simultaneously.
I must mention that i leave the usual pair of cells together. I guess that these cells have the same history and should behave quite similarly so my guess was there is little advantage in processing them separately. Additionally, that leaves the originally welded connector stripes on the cells, and with a little foresight i where to cut them each cell gets a nice Solder point on the stripe, which eases soldering and reduces the heat transferred to the cell. Disadvantage is that the soldered package looks gruesome, at best. (Even the triples wil be reworked to be pairs, i separate one cells, this are processs as single and re-paired).
In general i seem to use lower currents than recommended here. Since i process usually 3-4s 6-8p, i have 18 to 28 cells "on the board", which are processed together, i can easily take the time for a 1/10 C load. Discharging was to slow with the B6AC, it has a 2 Amp/5 W limit.
I built a small extender, basicaly a current resistor, operational amplifier, power transistor and vented cooling, which measures the (discharging) current and surges another 9 times the current of the imax off the battery itself.
So the imax is unmodified, everthing is external, but i can easily consume 50 Watts, if i like. Now, its only in steps of 1,2,3 4... Amps, but alas, this way the current and Ah display is just the tenth of the real. When the imax is accurate at all, which it is not. I will talk about that later.
Then came the powerwall idea, together with PV usage start, and i hade usage of my about 100 cells to build my fisrt two modules, as mentioned in the other thread. Since i will supply only basic load (no full islanding), the low power request on the cells ist the same.
-------
2nd harvesting campaign
Now to the second, my current, harvesting campaign. I got hands on a bunch of expectable bad cells, about 50 Laptop packs, and wanted to recheck and improve my processing style.
(my third campaign is already planned, i will buy cells from a board member here)
My imax stock has grown to 4 units, because 2 chrashed ( my fault), i ordered new ones, but was able to repair the broken ones.
I had to improve the connection of the cells to the Imaxes, i built "Boards", about 22*30 cm (8 by 12 inches) which got a balancing connector, plus the main plug cables, to a connector stripe. The cells are are arranged mostly as 3s 2P as described, and i put 3 or four of these on the board. wires and balancing wires are soldered to the connector stripe. In the end i had about 5 boards in operation, which gave me quite a capacity for cells in process.
The cells are harvested from the packs, separated to the pairs, and Voltage measured. A pair with a voltage over 3 V goes to a board, cells below are slowly fed up with laboraty power supplies to about 3,3 V. I did this in the past already, before i read the same experiences here on the board.
On soldering them together on the board i look a little bit to voltage differences, i care not to to connect a good cell on 80 % SOC to a 3,3 V cell.
Then they get (usually) 3 Cycles on the boards, with 1/10 C to 1/5 C parameters. I have experienced that the capacity increases on the second cycle, and a bit on the 3rd cycle, so something happens there. This takes a complete day per board per cycle, but this processes the mentioned about 20 or so cells to gether, and that is not so bad.
All capacities on discharging are noted on the boards, together with the current used. I decide on the cycle numbers on the capacity results, if nothing happens with the capacity i load them fully, separate them from the boards for sending them to stock for 14 days, to the self discharge check procedure.
I do not so much care for the absolute capacity of a board or the cells, usually the are not paired very good, means they are not equal in capacity, and one cell level is empty first, when others gabe still 3,6 V, as example. Since i mainly want cycling them, that is good enough for me.
The last step, after discharge check, has to determine the capacity. Since that is not so good with the imax, when they are different, my newly improvement is an Opus CT 3100, just for discharge capacity measurement. The cells still go in there as pairs, that was the idea. The bad news was that the opus not only gets HOT especially on discharging in all four slots, the poor fan is just an excuse. The temp in the Opus is one thing, but I measured the cell(s) on the right side get heated up to about 50 degress C (122 F) just from the Opus heat.
I dont want that, not just so from the point the cells cannot stand that, but merely of two considerations: The capacity result maybe different, AND, the cell, as a "bad one" MAY not like the temp. IF the cells are ready to go better not to give them too much excuses to go....
So, i built a small stand with a bigger fan, plywood of 6 mm (3/10 inch), in the most simple way.
The opus stands just in there (with removed bottom plate), nothing is attached or screwed to the opus.
The back side is left without a piece to support the Opus, this opening is for air exhaust of the fan to get air streaming out.
The new fan.....
....is plugged in to the opus instead of the own one (its a 12 V original fan, about 65 mA). Sure my new one was checked to match that. I cared not to overload the small switching transistor in the Opus.
Now the same right cell gets to about 30 C, on 26 C ambient, for a 1 Amp discharge. That is better.
Btw, i read about the current/power supply issue of the opus in an older thread here, i will care for this if i want to use the loading features. Old PC PS are in stock here, of course.