SkyRC — IFA 2014 — MC3000 charger-analyzer

[aznsx]

New firmware 1.21 is available. I noticed the following changes compared to v1.18:

• modified IR calculation (higher values than before; could be an interesting change as IR calculation was considered broken for firmwares released after v1.15)

Thanks for posting this, as it might be significant.

If possible, could you try to quantify the change you observed in DCIR measurement after you did this update? Any rough estimate of the approximate magnitude of the change would likely be helpful in understanding this issue.

The info I've seen on the problem of concern is very limited, but below is the best example I've seen of a quote of a user-reported description of the problem which was introduced in the past FW update. As you can see, the error was not small, it was huge, with measured values dropping to ~20% of their previous, correct, same-cell measurements, rendering them effectively invalid.

Based on the info below, do you think the previously-introduced defect / bug has likely been resolved / corrected in this latest version, or perhaps just changed marginally, but that the measurement is still likely incorrect?

Thanks!

Users have encountered serious issues w/ 'internal resistance' measurements using later version(s), and I've heard nothing regarding resolution of that problem to date.

Here's one quote:

***********************************************************************
Just something I noticed after updating my MC3000 from v1.15 to v1.18:

the DC IR resistance reading becomes too low.

for instance, I had a high-drain 18650 battery that registers around 28-31mOhms back in firmware v1.15
but after updating the v1.18 firmware, the DC IR reading is just 6-7mOhms.

another button-top 14500 battery (button-top will usually have much added resistance),
in firmware v1.15, it was reading around 100-110mOhms
but in firmware v1.18, it was just reading around 20mOhms

since I had previously been jotting down the DC IR values of many of my batteries (using the v1.15 values), I decided to just flash back to the v1.15 firmware, which appeared to restore to the old DC IR values
******************************************************************************************

 

[Lipti]

If possible, could you try to quantify the change you observed in DCIR measurement after you did this update? Any rough estimate of the approximate magnitude of the change would likely be helpful in understanding this issue.

I've got only very rough estimates as I have upgraded from 1.15 via 1.17 via 1.18 to 1.21 and haven't had time to do scientific tests. But maybe even those results may help you:

Note: I only tested with an old AA Eneloop from 2016 as my primary goal is to match lots of NiMH batteries.

v1.15: I quickly inserted the battery one time in each slot, contact could be bad and was different between the slots: 323, 463, 315 and 409 mOhm (slot 1 to 4)
v1.17: Same quick insertion, only one time in each slot: 56, 46, 49 and 52 mOhm
v1.18: Same quick insertion, only one time in each slot: 58, 56, 48 and 56 mOhm
v1.21: As I noticed slot 4 with a complete different value than the remaining, I exercised the spring mechanism, took care about battery contact and tried several times for lower values: 167, 169, 172, 173 mOhm

As the method was different between testing of v1.15 and v1.21, I can't say they reverted the calculation to v1.15; What i can say is, that I had a value at about 300 with the bad contact. So either they reverted, but I assume it's a different calculation than in v1.15 (with IRs between v1.15 and v1.17/1.18 but more to the v1.15).

 

[aznsx]

@Lipti

Thank you very much for your very thoughtful and considered reply. It is greatly appreciated;-)

In the absence of the kind of conclusive data we might ideally like to have, I consider almost any data, such as yours, to be of potential value in understanding this issue. I'll give this some more thought in light of your info, but at this point I believe my initial thoughts would be much like the working assumption in your last sentence.

For the moment, I'll conclude that it remains best for me to keep my units(2) on V1.15 (which produces measurements I am highly confident are indeed approximately correct), but will continue to consider the issue, keep an open mind, and watch for any further info which may come to light.

Thanx again!

 

[Lipti]

@aznsx

You're welcome. I wanted to inform about the new version of the firmware as soon as possible without testing it for days/weeks, as the existence of a newer version than 1.18 could be interesting to more people. Furtermore, I didn't want to hold back the changes I noticed. Unfortunately, I couldn't test them in-depth in such a short time. And there's always the compromise between being academic in things occuring randomly (like the new firmware) and doing the work which I intended to do before I noticed the new firmware.

Maybe I can test the IR values the next days more in-depth. I'm more interested how the new values are related to the values of other test equipment (like my three Vapcell YR-1030). I can't judge if v1.15 was correct, it just seems that v1.17/1.18 were way off. If v1.21 would be in line with my YR-1030 devices, it would be perfect - for me. Other might want the same values as in v1.15.

 

[symphonystars]

Upgraded the firmware to v1.21 on my MC3000 from v1.15 and after some testing i can confirm the IR values it gives are pretty much the same as with the v1.15 fw(+- few mOhm). I guess they simply put back the old algorithm

 

[aznsx]

Upgraded the firmware to v1.21 on my MC3000 from v1.15 and after some testing i can confirm the IR values it gives are pretty much the same as with the v1.15 fw(+- few mOhm). I guess they simply put back the old algorithm

That information doesn't appear to be consistent with this information below from post 5742. The way I read that post (5742), the V1.21 measurements appear to be roughly/approximately 50% of those measured w/ V1.15 (which would leave them still well outside of the 'valid' range IMO, and certainly not "pretty much the same"). Perhaps someone can reconcile this apparent contradiction...or am I missing something here? Please advise if I am, as especially with only 2 data points available, I do not want to get this wrong.

There could be procedural issues involved. When making mOhm range resistance measurements, all connection points must be clean / free of oxidation at time of testing. Furthermore, the MC3000 runs/uses 2 different DCIR measurements; one abbreviated / less accurate (per the manual) is done as a rough check whenever a new cell is inserted and an operation is initiated, and the more extensive and more accurate test that is done on specific command (using the keys). All of the measurements being considered here must be made using the latter, complete DCIR test to be comparable.

FWIW, I just now stopped and tested a protected, button-top 16340 cell using both the MC3000 (FW V1.15), and an Opus BT-C3100 (FW V2.2), and the 3 measurements each from the MC3000 were 175, 168, and 172, while the measurements from the Opus were 174, 160, and 149. Aside from the better consistency / repeatability of the MC3000 (which has been documented), the 2 units are in fairly close agreement, within acceptable tolerance, and is also consistent with the accuracy of both devices being considered to be good/very good when HKJ did his initial testing. That should serve as a relative 'sanity check' on the absolute accuracy of the MC3000 using V1.15, just in case anyone doubts its absolute accuracy using V1.15.

Thanks.

v1.15: I quickly inserted the battery one time in each slot, contact could be bad and was different between the slots: 323, 463, 315 and 409 mOhm (slot 1 to 4)

v1.21: As I noticed slot 4 with a complete different value than the remaining, I exercised the spring mechanism, took care about battery contact and tried several times for lower values: 167, 169, 172, 173 mOhm

 

[symphonystars]

not 100% sure why lipti's test has such big gap in IR values but possibly just a matter of bad contacts(they mentioned making sure about the contact quality only on the v1.21 test)

either way i repeated the test with 4 ikea ladda 1900mAh and the IR results are consistent and very similar on both firmware versions

the values are from the test pressing both up and down keys(the wire from the negative contact goes directly to the pcb and was installed cause i had contact issues in the slider assemblies some time ago)

v1.15

v1.21