Portuguese Powerwall in Poland

Joined
Mar 5, 2017
Messages
101
Ok, let's start document this journey:

so my current status is that I mounted a bunch of sonoff powsaround the house to measure several stuff, and a SDM120 for the overall usage of the house, and with a small python script insert it into influx and visualize it on grafana:


image_wxzoyx.jpg


Since I have 2 tariffs (night + weekends cheaper), I needed a bit of scripting to calculate according to the time and active tariff what is the cost per measured appliance and the total spent on the house.
So now I know exactly how much per month each of my loadscosts me. That's is quite nice feeling.

So, monthly the biggest consumption comes from the fridge (lodowka in polish :), the water heater panels (solar), and computers+router+projector (media). And since these are all small loads, I want in the first step offload them into the powerwall with a small but efficientinverter.

I've been collecting laptop batteries for a few months and here's how my stats look so far:

I consider good cells above 1800mah (opus measured), BUT also if the capacity / original catalogue capacity is bigger than 80%. I hope by doing this I don't put into the powerwall cells which are near their end of life.


image_ioeonj.jpg


As you can see, it doesn't look too good.

I contacted every laptop service around, but the price is rapidly going up, so it's becoming more difficult geting good cells from used laptop batteries, at least here in Poland.
Either they want too much for them, or the cells are not good enought for the powerwall.

I still have around 70 laptop packs to open, hope the stats get a bit better.

I decied to buy directly from a ebike guy100x INR18650-29E unused(original capacity)for 1.42EUR/cell, to help me get to my first goal of 560cells for the 14s40p pack.

I recently bought:

3x260w solar panels (still not mounted, just testing laying in the grass).
pcm60x

And made my first hook up yesterday.
What an incredible feeling, seeing the juice flowing :)


image_ovwtzr.jpg


I made a small 4s5p pack mainly to practise soldering and bus bars contruction and tested it:


image_sodagi.jpg


But the pcm60x has a 15v limit, so it's time to go to the next step, which is build a 7s15p pack from my <1800mah cells and fromcells with life capacity < 80%. This will help me further practising soldering and building for the ultimate goal of the 14s packs.

Let's see how that goes. I'll try to give more updates as it goes allong.
Please comment, specially if it's advice or critiques, insults are also tolerable :)

I'de like to special thanks Peter for sharing his journey, and Daniel (daromer) for his incredible willto keep commenting and helping others. You guys are great!

Have sun
 
image_igkpuv.jpg
Hey mate and welcome to the community,

Awesome to see another pack going up and how global this trend has become.

A few soldering tips:

Use an adjustable heat soldering unit. They are well worth the extra money! When soldering to the cell and to thickcopper, turn the heat up to its max. This allows you to heat the point of contactquickly and reduces therisk damage due tooverheating.Also use thin solder that has a flux built into the core. The correct tools for the job will save you hours of frustration later.

I use a Hakko FX-888D soldering station and MG Chemicals Sn60/Pb40, 0.025", with 2.2% flux solder. With the heather to 850 the solder melts away into beautiful little pools of shiny liquid in just seconds. You will spend hours and hours and hours using this soldering station, so it's a worthy investment.

Stoked to watch as you progress and happy building mate.

Cheers,

Chris
 
[ATTACH said:
9739[/ATTACH]
Inverted18650 pid='8108' dateline='1499527533']image_igkpuv.jpg
Hey mate and welcome to the community,

Awesome to see another pack going up and how global this trend has become.

A few soldering tips:

Use an adjustable heat soldering unit. They are well worth the extra money! When soldering to the cell and to thickcopper, turn the heat up to its max. This allows you to heat the point of contactquickly and reduces therisk damage due tooverheating.Also use thin solder that has a flux built into the core. The correct tools for the job will save you hours of frustration later.

I use a Hakko FX-888D soldering station and MG Chemicals Sn60/Pb40, 0.025", with 2.2% flux solder. With the heather to 850 the solder melts away into beautiful little pools of shiny liquid in just seconds. You will spend hours and hours and hours using this soldering station, so it's a worthy investment.

Stoked to watch as you progress and happy building mate.

Cheers,

Chris

Hello, thanks for the advice.

I followed daromer's advice and got mysef a 200watt soldering iron which has a thick mass.


image_rbnrvc.jpg


I also got some flux pens similar to what has seen on average joe:

https://www.aliexpress.com/item/Kes...32481864294.html?spm=2114.13010608.0.0.AfGPbp

and so far is has been a surprising pleasure. It' really easy, as daromer explained in his video (thank you very much Daniel).

The only problem I seem to be having is that after soldering 20 cells the tip is already in this condition:


image_qkkgaa.jpg


Would really appreciate some advice on how to not allow this to happen, at least not so quickly. Maybe too much flux?
 

Attachments

  • image_kgixkr.jpg
    image_kgixkr.jpg
    582.8 KB · Views: 300
Its because you got your self an iron with pure copper. You need a tip that is plated with iron for instance for it to hold :)

Dont worry when it comes to above. When you have wore it out you just exchange the tip for a new copper piece :)
 
" The only problem I seem to be having is that after soldering 20 cells the tip is already in this condition:

Would really appreciate some advice on how to not allow this to happen, at least not so quickly. Maybe too much flux? "

daromer said:
Its because you got your self an iron with pure copper. You need a tip that is plated with iron for instance for it to hold :)

Dont worry when it comes to above. When you have wore it out you just exchange the tip for a new copper piece :)

I had the exact same issue after my test build. Since the tip is busted I took a Sanding machine too it and sanded off the first 3mm to get to a normal looking tip again but now it doesn't seem to get as hot. How come?

Daromer: So I can just put in a piece of copper with the same thickness and use that?
 
You can put in any copper yes and use it.

Note that there are 2 different tips. The ones with iron coating will last almost forever. As long as you dont file it! Its very important.
Copper tips on the other hand will wear out very very fast. and even faster if you dont threat it right. Like make sure its wet and use solder with less flux. I use 60/40 lead based and it seem to work pretty good. Not sure if non lead based will be more aggressive.

You can always put in your own copper tip so if you can get hand on some use that. My bigger iron is just normal copper tip to and it wears out but i have soldered 1000s of cells and even more stuff without problem. yes it digs out the tip fast but on the other hand the tip doesnt cost me much.
 
So I completed the 7s15p test pack just in time to catch the last sun rays today and load it a little bit, this was at around 19h, the sun was almost gone in the horizontand panel is rated 260w.


image_hmrctc.jpgimage_opzqco.jpg

image_xvhbuk.jpg


This7s pack is still quite unbalanced, I'm leaving itconnected overnightto my icharger for top balance.


I used repacker for calculating each series capacity to match, so let's see if this helps keep it balance without BMS.

I will then try to use it on an old UPS APC smart 600 and try to test its capacity on some AC loads.


Next step I guess will be to mount the panels on my terrace roof.
Since I want to have a variable tilt system for my south side panels, in order to easilychange the angle 4 times per year, so it will require a bit of thinking, planing something simplebut solid for strong winds.

For my location here's the optimal angles:
winter: 76
summer: 23
autumn and spring: 50

Also next in order is to build a shed for my 14s packs.

I just noticed the batrium discount code that peter got us.
I've been undecidedbeetween the REC BMS and Batrium, so now have to decide until august 6 wether to order the batrium withfree shiping.
 
So, a bit of an update.

Good news is I put up the 3 panels on my terrace roof. I still didn't come up with an solution for a variable angle system, so I just mount them fixed (13 degrees for now), but I guess until winter I'll have a bit of time to think of something that is solid.


image_pdganv.jpg

image_uwzgzc.jpg


I only connected 2 of them for now, since I need mor MC4 connectors for extension cable.

Looking at these first solar panels on roof gives me a very big smile on face :)



image_ntfmuj.jpg


Next I connected an old used UPS APC smart 600 to my 7s15p pack and it works, I've been pulling almost 300W and powering my router, laptop and projector, and this is an incredible feeling. Also it helps convince the wife to approvebuying a good inverter for the 48v powerwall to feedthe all house :)


image_lqztmt.jpg

image_qrudpj.jpg



On the bad news side, I blew up a buck converter from aliexpress:

Adjustable power supply300w


image_flxror.jpg


I don't understand it fully, since it should hold up to 40v on the input, and since I wasn't pulling anything from the pack. Maybe I just shorted it out, being uncarefully.
I unsoldered the caps, and will try to bring it back to life by puting new ones, since I'de like to use it to power my laptop directly without DC/AC/DC wasting.
If anyone has a good source for buck converters for 7s or 14s and around 10amps, please let me know.

Also on the bad side, my stats from salveging laptop cells took a big turn for the worst:


image_tiqezi.jpg


I've now tested around 800 cells, and these stats are notveryencouraging, since the price of these laptop packs are getting higher and higher.

Still a bit of cells and packs to open and test, so let's hope and see what that will bring:


image_arilbi.jpg

image_jfiuzr.jpg


Next step is to build a shed for the packs, and decide on the batrium discount thing....what do you guys think? are you also considering takeing up on this opportunity??
 
Great on the additions
Sorry on the Buck and Cell return currently. Keep chuggin away at it. I'm sure your return on the cells will go up, though.

And yes, the wife must be happy and convinced of the practicality of the units that are needed to be involved ;) Besides, keeping the wife in the loop of decisions helps for a happy marriage.
 
I just realized that this old used APC smart UPS 600 which I got for 7 EUR (!!) produces quite a good sine wave which can be used for fridge and electric motors appliances.

So this is quite a nice surprise, and since this morning I've been using almost nothingfrom grid, since all my basic uses are quite below 400W.
Only very occasionally do Ineed to use stuff which is above 100-300Watt, and very briefly.

I justmowned mylawn with a 250Watt trimmer from sun energy.
I'm a happy camper :)

I also was able to connect the PCM60x to my raspberry pi, and get the stats to influx/grafana:


image_kleily.jpg


Was waiting for Daniel to get me the pcm60.js out of that ISO he made, since I didn't want/have a SD card to burn the ISO, but started from the original node.js script from this threadhttp://secondlifestorage.com/t-pip4048-raspberry-pi-influxdb-grafana and finally modified it to serve my needs.
If anyone needs it, just let me know.

Nice.
These small successes are very motivating, to keep working towards the big 14s goal and a good inverter.
 
As i said its the same script as is on mu ISO :)
 
I ordered the MPP PIP 5048HS, let's hope it will work as I expect.

I'm now approaching the 560 good cells mark, and want to start building the 14s packs, so for now it would be 40p.

So I'm thinking about a few ways to group the cells, and would love some opinions from more experienced users.

I think someone in here or in some video said that they're grouping by chemistry, so basically trying to build packs that have same kind of cells.....do you guys think this is good idea?

So let's say that you're aiming for example for 100ah per pack, and depending on what kind of cells and how many, then each pack would have different amount of cells, but all with similar chemistry and all with same capacity.So if you build on 4x 4x5 cell holders, would still have some space left for increasing capacity later.

The other option is to distribute all the cells you have into the 14 packs,trying to get same amount of ah, either by randomly or repack calculated. Each pack having the same amount of cells. Thisapproach seems to be the most common among the builds here shown in the forum.

Either way, I'm thinking about doing something I haven't seen anyone here doing, which is to build 80p packs but only filled with 40 cells for now (cause that's all I got :-/ ), and later will just insert more cells into empty spaces.

I guess Simone was doing something similar, but he kept empty spaces inside the pack. I'm planning on adding 40 cells more at the end having already bus bar ready for soldering it.

Something like this:

image_nrxytz.jpg


What do you guys think about it?
 
Yes, if you're aiming for a certain capacity per pack, then you will have different number of cells per pack depending on what the average cell capacity is.

Mixing chemistries is more for LiCo's with LiFePo's with LA's or another chemistry. Most, if not all, your cells will probably be LiCo's (if they all came out of laptop batteries). So, you should be fine mixing Reds with Blues with Greens with Purples with..... You get the idea.

If you're using the plastic cell holders, it'll be hard to add cells later; unless you break the plastic ears that hold cells inside. That's one reason why I'm not going to use these style of holders. Hard to change pack size; or replace a bad cell.
 
Yay a Portuguese hacker :) I'm a Vieira from Madeira but born in Australia .. hehe.. just to confuse things.
I'm thinking of making (just for fun and learning curve) a BMS system using Pi zeros and remote influx and grafana, which I think you beat me to it somewhat.
 
Just ordered the Batrium starter kit.
And the PIP 5048HS should be delivered tomorrow.
I feel like a kid on christmas day. Yey! :)

On the otherhand, my 18650 stats still are quite bad, after 1212 cells processed.
Feels kind of frustrating to get only 33% good cells for thepowerwall, when most people I read about they havethe opposite: ~30-40%
bad cells...


image_ffsubi.jpg



I tried all laptop repair shops in town, and afew electronics recycling facilities, but their prices are going up with demand.
Probably the ebike DIYmovementis also hiting hard on buying used laptop packs.
I didn't try schools or hospitals, since I don't believe that in such relative poor country as Poland, they'll have unused equipmentto throw away or sell cheaply.

I still have around 300-400 cells to process, and when buying my last laptop packs I got more picky, trying to choose from the pilethe packs with6cells and capacity>=55wh, so maybe the lastcells will improve a bit my stats.

I have until now payed around 165 EUR for all thelaptop packs (didn't keep track of how many), and so far thecalculatedAh on the good cells gives me 0.2 EUR/Ah (this is after taking 10% off the Ah, since it's Opus measured).

Considering I can buy unused cells for0.47 EUR /Ah, I will seriously consider in the next phase (next 5kwh)if all the time spent unpacking and testing the cells is worth doing this further....or instead just buy the cells, and use the time saved on something else.
 
"I didn't try schools or hospitals, since I don't believe that in such relative poor country as Poland, they'll have unused equipment to throw away or sell cheaply."

You won't know till you ask. You may be surprised. After all, Hospitals have to have battery operated equipment and they must replace it after a certain amount of time, not uses. So it's quite possible to get some crash cart batteries that were only used a few times in several years.
Schools on the other hand, they are a bit more abusive with the batteries as they let them drain till they won't even power on to tell you to plug the device in. But still worth a shot.
 
a bit of an update:

- PIP 5048HS is here, waiting to be mounted

- Batrium arrived and I've mounted it on the 7s15p test mini powerwall that has been running my low wattage stuff in the house (lights, fridge, computers) for the last 2 months.

It works great:


image_fxpvrt.jpg


Now I feel a bit more relaxed about burning down the garage, and don't have to run every couple of hours and measure each cells group voltage.

- I finished processing all the laptop battery cells that I had and here's how the stats look:


image_rzvdfk.jpg


These plus 200 cells unusedthat I bought cheaply, will allow me to build a 14s60p powerwall.

I've been soldering the cells using the glass fuse wires of 1.5A (according to Average Joe tests it should blow at something like 5amps),
and I hope in the next few weeks to finish it and put it up on the wall connected to the PIP and Batrium.

- Meanwhile thesd card from my raspberry got corrupt and of course I forgot to make any backup of the scripts I had, so had to start from zero.
So I decided to give a try to Daniel's ISO, and it works quite nice.
Of course one has to tweak it to own's needs.

Here how it looks right now:


image_nydurc.jpg


This shows:
-SDM120 meter connected to grid incoming wires (1 phase).
- PCM60x solar watts
- Batrium SHUNT watts goind out from OR intobattery


This shows realtime wattages going around, I'm trying to get something similar to what tesla app shows for its powerwall 2 (check it out:
)

But I'mstill thinking about how to calculate daily and monthlykWh bars for solar,battery and grid to get an idea of what are my usages of those periods of time.


The SDM120 from grid has this ready to use, since it allows to get the total kwh.
But the batrium SHUNT and PCM do not, so probably have to use the x samples per minute that we get from them and then somehow calculate the kwhs.
You guys have any advices on how to do this the right way?
 
Back
Top