OffGridInTheCity Build

Sucker I got the same alert and was contemplating it. They are among my favorite cells. Perfect for Powerwalls
Best of luck.

Wolf

Wolf said:

Sucker I got the same alert and was contemplating it. They are among my favorite cells. Perfect for Powerwalls
Best of luck.

Wolf

Click to expand...

Me too. Takes less cells per pack because of the 2900-3000mah specs and the ones I've done are among the most stable packs. For example, the pink Samsung ICR18650-26F from Battery Clearing House modem packs had the highest quantity of 'slow self-discharging' I've ever seen. I had to toss at least 20 cells of otherwise good test metrics after 6 weeks of sitting - I use <4.110v as my criteria. Usually it's more like 5 for the same quantity of cells.

Sweet cells although the way they say to open them makes me nervous especially since Panasonic cells of that type have particularly thin wall metal.

cak said:

Sweet cells although the way they say to open them makes me nervous especially since Panasonic cells of that type have particularly thin wall metal.

Click to expand...

Agree about especially thin (easy to dent) wall. We'll see - I'll report back.

Just got some of the 'never cycled' NCR18650A cells from BatteryHookup - https://batteryhookup.com/products/new-case-of-108-panasonic-ncr18650a-2900mah-cells

I have to say the cells look in good shape but the 1st 27 all have date codes "0611" which gives 2010. The BatteryHookup description says the packs are "new" but they don't say the age of the cells in the description. The devil can be in the details sometimes. But these feel like they're in really good shape.


Not that bad to open - here's a pictorial sequence of what I did....
1) Start by prying an edge from the end - less likely to dent the cells

2) Then slide the screw driver down the side - it will split pretty easily / very flimsy case...

3) Back to the end and pry up (on the BMS inside) to start a split down the other side. **Doesn't seem to be a problem with shorts or anything like that but of course be careful

4) Repeat step 2 at least 1/2 way down on the other side

5) Pry / rip off the top. The point of screwdriver is pressed on the BMS circuit board, avoiding damage to the cells. It's flimsy and was able to jut rip it off with my hands - e.g. pulling doesn't dent the cells.

6) Top off - BMS on top - pry gently on the positive end as this is strongest part of cell / won't dent but will let you start separating the cells from the bottom of the case. Flimsy / easy as I finished it by hand!

7) Cell units are loose and no dents and no shorts

8) I prefer to try to keep/clean the wrapper BUT in this case, for me, it's looking much easier to slice the the wrapper and pull out the cells. The glue is soft but very sticky!



9) OPUS Quick Test. I know it's not accurate but based on my experience, these are *very low* IR readings - as low as any I've ever had with these OPUSs.


10) Starting the 1st capacity test.... Looks like they are 2.5v/2.6v'ish out of the package....


Stay tuned for future results

That's a bargain indeed Why don't you rewrap them before using them, that negative case is so dangerous

italianuser said:

That's a bargain indeed Why don't you rewrap them before using them, that negative case is so dangerous

Click to expand...

Yes, I'll rewrap but couldn't wait to test the 1st 12....

And the first discharge numbers are 100% - https://secondlifestorage.com/index.php?threads/panasonic-ncr18650a-cell-specifications.1742/


I'm excited but of course need to process several hundred to get a true picture. One unique thing about these so far is the box/cardboard packaging is a lot compared to the actual cell packs

To correct an earlier comment - with a little patience the cells clean up OK and don't need to be sliced out of the original wrappers. To scrape off the glue takes a light touch rather than being aggressive. Running ~40min / medical pack to get 9 clean cells.

Wanted to get the word out before they're all gone

Finished testing the 1st 50cells from 3 random boxes and all are at 100% (3050-3150mah per OPUS). They seem so good not sure I need to test them.... but you know, foolish not to as even new cells can have a self-discharger or 2 in a 1,000 cells.

Best cells/price I've ever purchased and 100% the same type of cell - all are NCR18650A - which lowers the guess work and minimized the testing to get a 'battery' of all the same cell type. For example, the pink cells I just did for Battery #8 - I had to buy/process 3,000 to get 1,500 good ones.- e.g. double the test work. As another example, the RING packs were fantastic but 60% / 40% different types of cells.

Maint Update....

One of my Progressive Dynamics (240v@50a) ATSs - https://www.amazon.com/gp/product/B003VAWNVK/ref=ppx_yo_dt_b_asin_title_o00_s00 - started 'sticking' - e.g. would not finish the relay open/close cycle unless I banged the box with my fist. Since I use several of these as my way to automatically consume off-grid solar power production - it's of concern to my setup.

The one at the main (house) panel is the one that went bad - relays shown in the pic below. This thing has always HUMMED (loudly) when grid only.. but the other 3 are quiet. It could be the broken bracket shown below that wasn't securing it tightly so it could hum.

The relays in the pic were approaching 3,000cycles (grid -> inverter) or (inverter -> grid) over the last 4 years. Since I have 4 of them - I'm curious if they wear out at a few thousand cycles or this is a malfunction or the DC coil model is less reliable or does it really matter that I mounted it upside down from the advised instruction sheet....

I made an info request of Progressive Dynamics support asking about expected cycle life - we'll see what they say. I have 4 of them around the house - all are slightly different because they were purchased at different times - but if they're only good for 3000 cycles then it will become an ongoing maint issue.

The good news - it wasn't hard to swap out the relays - they're held in the box with 4 screws/plastic hold-down pieces. This is good since the metal box at the main panel is sheet-rocked flush with the wall and I don't want to tear open the wall

Misc: They also have DIN rail mounting capability once I had it out and could see the back.

Progressive Dynamics replied.....
----------------------------
From: PDI Service [mailto:service@progressivedyn.com]
Sent: Tuesday, September 6, 2022 7:38 AM
Subject: Relays starting to 'stick' – any remedy or do I need to replace the unit

I would replace the unit. if that relay is sticking it may not be putting proper pressure on the contacts. This could cause it to wear out quickly as they deteriorate. Do not apply lubricants .

I estimate the lifespan to be around 7 years in an RV situation. Probably not switching as often as you are. The life span will be reduced if the switching is done under load.

David

Service Department
Progressive Dynamics
-----------------------------


Sigh....
It has been nearly 4 years of trouble free operation, so I'm not unhappy with the product, but I was hoping for more. Not sure how to define reasonable life-span/expectations on something like this. Time will tell, but looks like I need to lay in replacement relays and expect failure every X years/cycles OR eventually move to different equipment consume off-grid w/grid assist.

This is my first maintenance issue and it renews my apprehension about the lifespan of the Midnite Classic controllers, AIMs inverters, and Batrium. The DIY powerwall lifespan may outlast the equipment

Update: More panels and Shed

Panels: Hooked up an additional 3s (3 x 285w = 855w) on the ground array. This puts Midnite Classic 2 at 3s6p - a bit over paneled. Overall PV goes from 12.825 -> 13.68kw.



Shed: Outside of the shed is complete / ready for winter! Part of the shed will house a propane driven generator for winter battery charging. Was a challenge to build this (get the roofing on) underneath the ground array

June...

July (additional panels are up - stabilized by the roof)

Sept - outside done!


Next is to add gutters on shed + ground array for rain harvest.

Maint Issue.... (Just starting my 4th year of operation)

Yesterday I started getting inverters on/off every 5 mins - not good!. The inverters on/off is controlled by the AUX1 relay on Midnite Classic #1 based on battery voltage. Rebooted the classics hoped for the best.

Today the problem returned! Went to solar room and all the Midnites are blaring GND FAULT!. Smelled that something had burned. Cycled power and noticed that Midnite #1 was showing a strange voltage as in 60v on a 52v battery. Power cycled it and still wildly off. Assumed I burned up the Midnite Classic somehow.

Broke out a new Midnite Classic and replaced the old one. It wouldn't even boot, just went into a loop at startup. Noticed that the battery voltage at the new Midnite went down from 52v to 40v or 44v or 36v as soon as I energized the Midnite. Yikes - did I just destroy a 2nd one?

Sent off a support request to Midnite Solar asking about startup loop. Got a prompt answer in 60minutes! (Talk about service!!!!! )
---------------------------- email from Midnite ---------------------
From: Ryan [mailto:ryan@midnitesolar.com]
Sent: Wednesday, October 12, 2022 2:13 PM
To: Information; xxxx
Subject: Re: General Information - MidNite

Ken,
You have a bad electrical connection some where between the battery and that classic. it is high resistance so when the classic tries to start it drops the voltage and resets
--------------------------------------------------------------------------

Started wiggling wires. I mean - it's just the wire from the battery bus to the Midnite!. What could go wrong in a 15ft stretch of wire?!? Bypassed the 80a breaker, crimped on a new lug.... Couldn't figure it out.

Took the Mindites to my office and hooked them directly to a 48v battery at my desk - they BOTH worked! (Yea!) but that proves it's the wire in the control room.

Finally remembered a junction box put in by the original electrician and look at this! He butt-joined the wire from the battery bus to the Midnites in this box and the butt-joint burned up. Explains GND FAULT and voltage change in the wire when Midnite energized as there was barely a connection. This was originally wrapped in electrical tape but arc'ed thru / burned / made contact with bonded junction box. Explains the burnt smell.


Replace the butt-joint with bolted lugs


Here's a pic showing the connection box relative to the control room with battery bus.



-----------------------------
All is working!

Takeaways
1) Ground Fault worked!
2) Proper circuit breakers in the control room greatly helped work on this safely.
3) Midnite support was FANTASTIC - can't believe it.
4) I think I'll add a 'max voltage diff' comparisons between the 3 Midnites and Batrium - and do a software alert if they get wildly out of sync. That might have helped bring attention to the problem sooner.
5) Head in sand is not good!

On with the journey....

OffGridInTheCity said:

Maint Issue.... (Just starting my 4th year of operation)

Yesterday I started getting inverters on/off every 5 mins - not good!. The inverters on/off is controlled by the AUX1 relay on Midnite Classic #1 based on battery voltage. Rebooted the classics hoped for the best.

...(snip)...

Click to expand...

Well done OGITC!
I think you did a good job with diagnosis

Sheesh, what was that electrician thinking when putting that butt splice in??! Especially being in a metal box without protection. Wow!

Glad you found the issue and was a cheap/easy fix. Also super glad it was a very uneventful failure. This could have gone the other way where things heated up and caused a fire

OffGridInTheCity said:

4) I think I'll add a 'max voltage diff' comparisons between the 3 Midnites and Batrium - and do a software alert if they get wildly out of sync. That might have helped bring attention to the problem sooner.
5) Head in sand is not good!

Click to expand...

Yikes Just curious how that butt connectors set screws got so rusty? Wow looks like it got pretty hot too.
Good Idea on the V comparison and alarm. Also I would check any and all butt connectors if you have additional ones and change them to lug and bolt type like what your repair looks like. Much more positive connections. Additionally the lack of ferrules on the wire ends going into the butt connectors. I believe "code" requires that.

Korishan said:

Sheesh, what was that electrician thinking when putting that butt splice in??! Especially being in a metal box without protection. Wow!

Click to expand...

It was wrapped initially. Nevertheless a butt connection esp. without ferrules?

OffGridInTheCity said:

This was originally wrapped in electrical tape but arc'ed thru / burned / made contact with bonded junction box. Explains the burnt smell.

Wolf

Click to expand...

OffGridInTheCity said:

originally wrapped in electrical tape but arc'ed thru / burned / made contact with bonded junction box

Click to expand...

What do you think is the root cause was?
Poorly wound electrical tape? Not enough loops, came loose over time?
Or bad butt-joint overheating and cooking the tape crisp? Stranded wire and screw terminals have been proven again and again to become unreliable over time.

ajw22 said:

What do you think is the root cause was?
Poorly wound electrical tape? Not enough loops, came loose over time?
Or bad butt-joint overheating and cooking the tape crisp? Stranded wire and screw terminals have been proven again and again to become unreliable over time.

Click to expand...

Yea, this is the scary part. These 4awg wires carry up to 80a from Midnite -> Battery at max PV.
I assume the 'rust' that @Wolf pointed out was due to the heat? or maybe it was rusty to begin with - e.g. bottom of the barrel from electrician's bag? It was pressing against the junction box - did this cause the electrical tape to melt?

This box routinely get's up to 50C/122F range in spring/summer - but that's not hot enough to melt electrical tape. Maybe heat/cool over the years caused bad connection? In this thread, I put the temp probes in Feb 2020 - https://secondlifestorage.com/index.php?threads/how-hot-is-too-hot-for-pv-array-wire.8736/ This is pic of the box back then.


From that thread above I started tracking max temps / day and added that to my dashboard. It was right in front of me on the dashboard with daily max going from routine 50.6C to 81.4C.... Need to change these numbers to RED >60C or something to draw my attention!


The temp probe was in the box monitoring the incoming PV wire - about 7 inches from the burned up butt-joint. It never get's above 55C at max PV on a hot day.... but look at this. Between 10am and 1:30pm Oct 11th it went up to 81C = 178F. Butt-joint really hot - maybe over 200F! Divide the shunt amps / 3 Midnites + the load = ~60a range as this was burning up.


and then on the 12th... up to 67C / 152F.... maybe final burn-thru...



Clues all around me....

Korishan said:

Sheesh, what was that electrician thinking when putting that butt splice in??! Especially being in a metal box without protection. Wow!

Glad you found the issue and was a cheap/easy fix. Also super glad it was a very uneventful failure. This could have gone the other way where things heated up and caused a fire

Click to expand...

Yes sir.. and to me, shows there's value in electrical hookup standards/codes - e.g. connections in a metal box in this case. If these connections were just 'loose' I might have had a fire in the sub-floor of the house.