Honda Insight Forum banner

161 - 173 of 173 Posts

·
Administrator
Joined
·
11,459 Posts
Greentec have two of my BMS OBDIIC&C LTO gadgets so can get BMS ID's and cell voltages on every block. ;)

If a cell voltage was way off they would probably not ship the block.
 

·
Super Moderator
Joined
·
7,989 Posts
Greentec have two of my BMS OBDIIC&C LTO gadgets so can get BMS ID's and cell voltages on every block. ;)
Yes, and they recorded the ID's on my most recent shipment. I kinda doubt they are doing cell level capacity, since that would be a tremendous amount of work. Cell level voltage - maybe, but even that requires removing the covers, which is dangerous and a fair amount of work. If they were doing that why would they not write it down like the other stuff? Opinions may differ;)
 

·
Registered
Joined
·
6,660 Posts
...Cell level voltage - maybe, but even that requires removing the covers, which is dangerous and a fair amount of work.
Yeah but, I thought Peter just said they use one of his devices to read the cell voltages - so they don't have to remove the covers...

The one thing Greentec said that didn't make sense to me, was that they check the capacity of one module for each pack - but they also said they didn't charge them: How do you check the capacity of a module without discharging it, and then, how do you discharge it without re-charging it?
 

·
Premium Member
Joined
·
2,856 Posts
Discussion Starter #164
The one thing Greentec said that didn't make sense to me, was that they check the capacity of one module for each pack - but they also said they didn't charge them: How do you check the capacity of a module without discharging it, and then, how do you discharge it without re-charging it?
Although possible .. I think such a thing would have a rather large margin for error in the testing resolution.
soo , maybe only reliable to see if there is a very large difference between them.

I see two approaches:
#1> dV under load.
All else being equal .. a larger capacity cell would be expected to have less dV under load than a smaller capacity battery.

ie .. if everything else were the same .. and the only difference is cell A is 15Ah vs cell B is 20Ah .. it would be reasonable to expect cell A to have a higher dV under load than cell B .. because at the same say 100A load cell A sees that as a much higher 'C-Rate'.

#2> dV per Ah or Wh extracted.
again .. if everything else is the same .. two cells starting at the same point .. and both discharged the same rate for the same amount of Ah or Wh .. it would be reasonable to expect the one with a larger drop in rested terminal voltage should be the lower capacity battery.
 

·
Super Moderator
Joined
·
7,989 Posts
Yeah but, I thought Peter just said they use one of his devices to read the cell voltages - so they don't have to remove the covers...
Yeah, that is true. With the interrogator, which they have, they can probably check the cell voltages.

They still can't check cell level capacities far as I can see. They might have test equipment to simultaneously cycle 12 cells, but it would still require them to remove the covers and hook up a bunch of sense wires - I think. I haven't been told about any such testing.

The good news is that in both anecdotal accounts and my testing, there have apparently been only 2 bad cells - maybe fewer. Pretty darn good record.
 

·
Super Moderator
Joined
·
7,989 Posts
#2> dV per Ah or Wh extracted.
again .. if everything else is the same .. two cells starting at the same point .. and both discharged the same rate for the same amount of Ah or Wh .. it would be reasonable to expect the one with a larger drop in rested terminal voltage should be the lower capacity battery.
This is basically what the PL8 does, in 8 cell segments. Slight variations in cell capacities account for the divergence of the segment "tails" below 2V. On discharge, the weakest cell reaches 1.5V first, at which point the highest tail is the strongest cell.

So, bottom line in my mind is that Greentec isn't doing cell, segment or pack level capacity MEASUREMENT. If you want the data, you have to measure it yourself. The good news, to me, is that I haven't seen a bad capacity pack in the 17 that I've tested.
 

·
Super Moderator
Joined
·
7,989 Posts
Continuing testing, finding that accurate internal resistance measurements are a lot more difficult than I had first expected, so perhaps the range on good packs is smaller than I had thought. When hooking up the PL8 "tap" wires, it appears that any slight variations can make a sizable difference, perhaps .2 milli-ohm on a .5-.7 milli- ohm measurement. Apparent any slight oil from hands, and any variations in screw torque can make a difference. There is apparently slightly increased IR caused by the strap between the two packs.

Toshiba had used surprisingly high torque on all these connections, or perhaps some sort of conductive compound. The screws have a "snap" release, like they are slightly seized.

I'm wondering if there is some sore of conductive compound which improve the contact resistance with these connections?
 

·
Registered
Joined
·
6,660 Posts
I'm wondering if there is some sore of conductive compound which improve the contact resistance with these connections?
I spent quite a bit of time several months ago trying to find/researching a 'conductive compound' - like an anti-oxidizer that was actually conductive. My conclusion was that they don't exist. From what I gathered, products of this ilk can improve conductivity merely because they reduce oxidation - they help preserve an air-tight/gas-tight seal. When you smear the stuff on a joint, such as between a terminal and a busbar, the busbar and terminal will make good metal-to-metal contact when tightened down, the compound will fill in any voids and squish-out around the joint, making it gas tight (and water-tight). But the compound itself won't be conductive...

edit: Here's a blurb from one (supposedly good) product's description. I think claims of this style are the closest you'll find to a compound being 'conductive':

"Lowest Electrical Resistance
Alnox contains extremely hard metallic particles with sharp and irregular shapes of carefully controlled grain size. This forms many metal to metal contact points for low resistance connections."

I recall reading a thread or two, though, where engineering/material sciences-types scoffed at such claims, saying something like the metal particles don't do much, it was mainly about the compound preserving that gas-tight seal...

Not that this is at all what you're looking for, but I figured I'd attach a decent report I came across that discusses various 'electrical connection' standards and practices. There's a short review on page 5 of a handful of compounds...
 

Attachments

·
Premium Member
Joined
·
2,856 Posts
Discussion Starter #169
I think ~0.0001 ohms (the range described above) .. is very very tiny .. there are soo many (too many) things that can cause tiny variations on such small scales .. you are welcome to try .. but I think there are diminishing returns trying to get smaller and smaller tinier and tinier margins of error .. especially on a compound (aggregate of many many things) property like ohms are.

FYI Note .. PL8 displays Ohms based on battery AC impedance .. which is not identical to ohms of DC resistance .. even if both are measured in ohms.

----
Personally .. the way I approach that concept/aspect of battery performance .. I use the PC export funciton of the PL8 .. out to a spread sheet .. and then let the spread sheet give me the nominal voltage under those test conditions .. 2 cells at the same start ( temperature / pressure SoC / etc) .. discharged out the same Ah , the one with more resistance will have more voltage drop .. thus less net overall Wh , and a lower net overall nominal voltage for those conditions .. I mainly look to +/- 1/100th of a Vnominal (0.01) .. I take the mV and smaller parts calculated with a grain of salt (maybe for how close it seemed to be to rounding up or down).
 

·
Super Moderator
Joined
·
7,989 Posts
Thanks for the report. I haven't read it all yet, but it has lots of though provoking information. It seems that much of the information is slanted toward the "utility" problem of aluminum conductors. Those aren't the metals at stake here. The contacts on the Toshiba taps appear to be stainless steel. The balance/tap wires on the PL-8 appear to be tinned steel.

The "strap" between the 12cell halves of the pack also look like tinned steel. These straps definitely develop surface dullness, or corrosion. The bending load seems like steel.

The "puck" spacers which pass through the covers are some sort of metal I've never seen before. It seems even lighter than aluminum, though now for sure. It also nevers seems to develop any surface corrosion at all.

Interesting materials. I'm sure there must be some voltaic effects with all these different metals, plus the tinned copper cable terminals we mostly use.
 

·
Super Moderator
Joined
·
7,989 Posts
I think ~0.0001 ohms (the range described above) .. is very very tiny .. there are soo many (too many) things that can cause tiny variations on such small scales .. you are welcome to try .. but I think there are diminishing returns trying to get smaller and smaller tinier and tinier margins of error .. especially on a compound (aggregate of many many things) property like ohms are.
Yeah, I have limited patience for this sort of thing myself. It only gets my attention when I see IR at 1 milli-ohm or more. I'm finding that there are usually reasons, like slightly higher contact resistance. The two PL8 balance/tap wires which bridge the inter-pack strap frequently tend to show a bit higher IR. This pretty easy to spot.
 

·
Registered
Joined
·
6,660 Posts
....It seems that much of the information is slanted toward the "utility" problem of aluminum conductors. Those aren't the metals at stake here. The contacts on the Toshiba taps appear to be stainless steel. The balance/tap wires on the PL-8 appear to be tinned steel.
What are the "contacts on the Toshiba taps?" I recall Insightbuyer thought the busbars were nickel-plated something, I guess it could be copper though that would seem slightly strange since they're welded to pure aluminum terminals... I wouldn't think the 'taps' are steel, what I imagine the taps to be, that is. I can't imagine that a hobby charger would have steel cell port terminals ("balance/tap wires"?) - I'd guess they're zinc-plated copper - I think that's what most pins and terminals of this 'genre' are made of... You rarely see steel in any of these uses, steel's not a good conductor and as I recall there's some other issues with steel or any ferrous metal in uses that roughly involve electricity/electronics...
 

·
Super Moderator
Joined
·
7,989 Posts
Most of the materials are a guessing game, unless perhaps there are some references in some of the Toshiba data sheets. The balance/tap wires on the charger may have coated copper connector pins, I don't know. Revolectronics sells the balance harness without terminals, so I crimp and solder ordinary harware terminals onto the harness. I found this necessary to keep the resistance of the connection as low as possible. Even then, the tap harness plug to the PL8 is kinda flimsy, and they have a tendency to loosen up with wear.

I agree that steel is a relatively poor conductor, but the force required to bend one of the interpack straps certainly makes it seem to be steel. It certainly isn't anything as soft as copper. Maybe some special alloy of steel???

It is kinda interesting that the current Orion II BMS uses gold flashed contacts in their harness to BMS connector. I guess they found that increasing resistance over time was an issue.
 
161 - 173 of 173 Posts
Top