diy solar

diy solar

Easy simple DIY LFP battery tester build instructions, supply list, and diagram

Maast

Compulsive Tinkerer
Joined
Oct 31, 2019
Messages
773
Location
Washington State
This is the first and easiest DIY battery tester for lifepo4 battery cells, this is for a SINGLE 3.2V cell ONLY. If you attempt to use a 12V battery you'll melt the load, set your dog on fire and make your kids drop out of school. You have been warned.

Overall cost is about $130 US which is dirt cheap compared to buying a 60A tester.

In this example diagram the cell is a 120AH LFP cell, the charging power supply is a 60A Meanwell HRP-300-3.3, and a 60A load. There is a reason for these values:

Cell amp-hour capacity is usually measured at .5C with C being the capacity in amps. .5C means we'll be pulling .5 of its capacity which in this case is 60 amps and the battery will be exhausted in 2 hours. .25C would be 30 amps, etc. You can use a smaller charger but it makes the test that much longer. As it is a full test cycle takes 4 hours or so.

The power supply is a constant current limited power supply which means it'll reduce its voltage to avoid over-amping itself. If you do not use a current limited power supply it'll just shut itself off, start turning on and off until the overload is cleared, or just permanently die. This Meanwell will actually put out about 70 amps initially but because the cell voltage rises while charging and will drop below 60A fairly quickly. In addition the PS has a range of 2.8-3.8V and will need to be adjusted to 3.65V out of circuit (no load or battery) using the little phillips screw head near the power light. There are lots of meanwell clones out there if you hunt around. Whatever you use it has to be current limited.

The HRP line goes from HRP-75 all the way up to HRP-1000 (75 to 1000 watts) depending on how high you want your charging amps to be and how much you want to spend.

The best price I've found new is on Ebay for $75 including shipping, you can often find them used for a lot less.

The 12V source can be another power supply, or a 12V battery on a trickle charger. The meter only draws 2 watts so you dont need much. The Amp-Hour meter needs it's own separate power supply because LFP cells do not provide enough voltage for it to operate.

Tester Diagram.jpg
This is all assembled on a 24"x36" sheet of 1/2" plywood.

The meter is a PZEM-015 with 300A shunt can be found on Amazon for about $20 US. It comes with some wire but they tend to be a bit short, you'll probably need some extra stranded 18 gauge wire.
The amp-hour meter also displays voltage, current amperage, time watt hours and resistance. The only thing we're concerned with is the voltage, amperage and amp-hours. I prefer to use the shunt vs the hall effect amp hour meters because they're more consistant and accurate.

The 150A Fuse/Switches can again be found on Amazon for $16 but less expensive fuse/switches can be substituted, I just liked the convenience of the connections. Keep in mind these types of fuses' amperage are often badly overrated, in my tester I'm using the 300A versions.

All the main current carrying cables are multi-stranded 4 gauge "battery cables". The cheapest source on Amazon is 4 gauge jumper cables that you snip the ends off. $20 for 20 feet but you only need a few feet here. Shop around.

The load will need two 9 inch pieces of 3/4 copper pipe, a spool of .032 stainless safety wire and 24 small #6 stainless screws, the best I've found so far are the small screws that are used for aluminum gutters. You'll also need four 3 inch #10 stainless screws, they have to be stainless because it doesnt conduct heat like straight steel or especially copper.

This tester does not have any over discharge alarms or protection, you'll need to watch it like a hawk when you get near the end of the discharge. At 2.5V the cell is dead, remove it immediately or you risk damage to the cell. A better target is 2.7 or 2.8 volts.

Upgrades will be posted later for a low voltage alarm and to automate this with an automatic low voltage disconnect from the load. If there is enough interest I'll add onto that with how to automatically switch over from discharging to charging and toggle cycle counter using time delay relays. Somewhat complicated to do though.

We'll get into building the load and tester assembly in the next post.
 
Last edited:
The load is made from two pieces of 3/4" copper pipe spaced exactly 15 inches apart on center and raised 1 1/2" above the plywood by the 3 inch screws.

Solder the 4 AWG cable directly onto the pipe, I've found a easy way to do it is by sanding the last inch and a half of the pipe to bright metal, clamping down the copper cable with a screw down pipe clamp and using a propane torch. The solder doesnt stick to the stainless. Don't squish the pipe end onto the cable and solder it because we want both ends to be open to help cooling.

Avoiding your soldered cable, predrill two small holes half the size of the shaft of the long stainless all the way through the pipe on both ends. While you're at it drill 12 holes half the diameter of the small screws in-line with the holes you just did along what will be the top, space them about 3/8" apart.

Now do it again with the other pipe, keep the small holes directly across from each other so everything lines up parallel. Now sand off any ridges that formed while you were drilling and get the strip of copper where the small holes are down to bright metal. 80 grit is perfect for this.

Mount your pipes to the plywood, take a 6 inch piece of 2"x4" and, avoiding your larger predrilled holes, lay the 2x4 chunk where you want the load to go and lay the pipe on top of it, now screw down through the pipe into the plywood using the 2x4 as a spacer.

Be VERY careful to not over tighten the screws in the copper pipe. We need the screw threads to hold the pipe up off the plywood, then do the other pipe making sure the pipes are absolutely parallel. If you do strip the copper you can use a stack of stainless nuts as a spacer or even a glob of JB weld. Doesnt have to be elegant, just work.

Now screw down the small screws halfway in, then string the stainless safety wire from pipe-to-pipe one strand per screw pair. It doesnt have to be 'tight' but it does need to be straight, pre-straightening the wire will help a lot. When your done stringing screw everything down. Be careful you dont bend over your long screws, I might help to leave the 2x4 chunks under it to support the elevated pipe while stringing and remove it after you're done.

It's strongly recommended you lay down a film of a zinc based antioxidant where the sanded down strip is, it'll help immensely to keep the copper from oxidizing and interfering with the copper/wire contact. Noalox can be found anywhere in the US.

It's about 5 amps of load per strand, you can adjust your load up or down depending on what you need to hit your .5C target. In this example we're shooting for 60A of load. If you adjust your load above 100A I'd recommend mounting your load onto a piece of concrete backer board. I dont know for sure if it's necessary since the wires dont actually get all that hot.

You can also add more strands by using both sides of the screws and doing loops with your stainless wire. FYI .032 stainless wire is about 6 ohms a foot.

If your load gets up above 150A you should switch to 2 gauge battery cable and upsize the switches. Reference table say you can go up to 200 amps on 4 AWG at these short distances but it'd be better to upsize your wire.

Under load the stainless wire does get hot enough that you don't want to touch it, but not smoking hot. If the pipes were closer together each strand would draw more amperage and get hotter.

Tester assembly: working back from your load start assembling like the diagram shows. When you get to the 3.65V power supply you will need THREE strands of 12 gauge wire of both positive and negative to connect, one stand 12 AWG can only handle 20A. The small screwdown terminals on the power supply only allow up to 12 AWG stranded wire, solid core didn't fit. You can do bare wire w/ Noalox or crimp on a U terminal to the wire and then connect it.
Terminal is better but I admit I couldnt find one small enough that would also take 12 AWG wire so I just put greased bare wire ends in and squished it down.
 
Last edited:
Tester with Alarm
Tester w alarm.jpg
The alarm is a lipo low voltage alarm that's adjustable from 2.7v - 3.8v or off with a small button between the speakers. They're 110db wake-the-dead loud so it'll get your attention if you're anywhere nearby. Its actually so loud I stuffed bits of earplug foam in the speakers to quiet it down enough that I could endure being around it when it was alarming.

The alarm only works if it has a 2S to 8S connection AND it won't reliably power up with just the power from a LFP cell so you have to tie in your 12V power source to power it up. This also means you have to have a common negative between the 12V source and the negative of the LFP cell. It'll give a weird reading for pin3 but it's usually above 5v so we don't care. It'll trigger on the lowest voltage it senses.

Pin 1 is the combined common negative
Pin 2 is the 3.2V LFP positive
Pin 3 is the 12V positive

The alarm can be connected with a common 2S balance lead. It's voltage resolution is usually off, sometimes by as much as .2v so it might behoove you to buy a few and find the most accurate one, or just take it in to account and offset your trigger voltages.

Still working on the alarm with disconnect, it's tricky making a intelligible diagram because there is a lot more wiring involved.
 
Last edited:
Sorry, not seeing the pics
I'm doing the diagram for the low voltage alarm and assembling supporting photos for the load build and next post. After editing the above two posts at least 40 times I'm doing it all offline and polishing it before I post it.
 
In this example diagram the cell is a 120AH LFP cell, the charging power supply is a 60A Meanwell HRP-300-3.3, and a 60A load. There is a reason for these values:

Awesome job. Thanks for sharing this. Do you find the Meanwell holds the voltage setpoint pretty tight? I plan parallel top balance my packs because I will avoid ever going below 20% SOC, but will occasionally push > 90% SOC. I will use a Daly BMS to shut down if a cell falls off the cliff. I was going to purchase an adjustable benchtop PS for balancing, but for about 1/4 the coin I can pick up the HRP-300-3.3. If it holds the setpoint V and tapers the current to 0, I should be able to use this for an unattended top balance. I'd use a battmon and OV cutoff as accidental protection, but I'm planning to try this. The Meanwell is in the mail.

--mark
 
The Meanwells will hold the voltage rock solid steady and the amperage will taper to 0. In addition you can stack them both at the exact same voltage so if you have a 120A (600W power supply) and a 60A (300W) you'll get 180A if your starting voltage is 0, as the charge rises the amperage tapers off by the time you hit 2.5V you're at about 140A of the 180A available.
 
I know tthese are "knockoff" Meanwells but this is what I was sent by Shunbin for my 24V pack. Haven't used them as yet. Got absolutely no documentation of any kind so I'm not sure what the wiring setup is.... Maybe Maast or someone else may know where I can get docs for these... something / anything ?
LFP-Charger.jpg
 
I know tthese are "knockoff" Meanwells but this is what I was sent by Shunbin for my 24V pack. Haven't used them as yet. Got absolutely no documentation of any kind so I'm not sure what the wiring setup is.... Maybe Maast or someone else may know where I can get docs for these... something / anything ?
I found specs on the fingali site, evidently they're also sold on ebay and aliexpress for about $40 a pop. Looking at the picture I can tell you they're wired in series - which means the positive output of one is fed to the negative input of the next giving you a total of 24V.
 
Thanks for sharing your tester design, I am using my version (75A load) to test 200Ah AL shell cells purchased on Aliexpress - so far so good (yes, I should have gotten it up to 100A but ran out of pipe).

One thing that was tripping me up at the beginning was the meter (PZEM-015 with 300A shunt) was initial set for the 100A shunt, so all my readings were 1/3 of what they should have been. I initially thought the batteries were junk, then realized the readings were oddly more or more exactly 1/3 of what they should have been.

I have had trouble getting the alarm to work, keeps alarming when it cycles to higher cells that aren't connected - so I have to watch the discharge carefully near the end.

Power supply will also be great for top balancing the batteries.

Cheers!
 
@Maast ..... Wondering what you used for the low voltage disconnect?

Thanks
I attached the output of the low voltage alarm (with a 100uf capacitor to smooth the signal) to the input of a time delay relay control board which switched power to a 400A Normally Closed contactor on the discharge side then a 200A normally open on the charge side - switching it from discharging to charging. The time delay board was set for long enough to charge the cell plus about 20 mins of absorption time.
 
I attached the output of the low voltage alarm (with a 100uf capacitor to smooth the signal) to the input of a time delay relay control board which switched power to a 400A Normally Closed contactor on the discharge side then a 200A normally open on the charge side - switching it from discharging to charging. The time delay board was set for long enough to charge the cell plus about 20 mins of absorption time.
(y)
 
Thanks for sharing your tester design, I am using my version (75A load) to test 200Ah AL shell cells purchased on Aliexpress - so far so good (yes, I should have gotten it up to 100A but ran out of pipe).

One thing that was tripping me up at the beginning was the meter (PZEM-015 with 300A shunt) was initial set for the 100A shunt, so all my readings were 1/3 of what they should have been. I initially thought the batteries were junk, then realized the readings were oddly more or more exactly 1/3 of what they should have been.

I have had trouble getting the alarm to work, keeps alarming when it cycles to higher cells that aren't connected - so I have to watch the discharge carefully near the end.

Power supply will also be great for top balancing the batteries.

Cheers!
How did you reset the 300A shunt meter so that it is set to properly read a 300A shunt? I have the same problem and thought the meter was bad.
 
How did you reset the 300A shunt meter so that it is set to properly read a 300A shunt? I have the same problem and thought the meter was bad.
IMG_20200123_121853~01.jpg
  1. In the normal display interface, long press the button till the display changes to "Set Voltage"
    1. Release the button
    2. Short press the button
      1. The display will change to
      2. Set A current.
    3. Longer press the button until the display shows the shunt rating in use.
      1. Release the button
    4. Short press the button to cycle through the 4 options.
  2. Long press the button to save the setting.
 
Last edited:
  • Like
Reactions: HMR
View attachment 5732
  1. In the normal display interface, long press the button till the display changes to "Set Voltage"
    1. Release the button
    2. Short press the button
      1. The display will change to
      2. Set A current.
    3. Longer press the button until the display shows the shunt rating in use.
      1. Release the button
    4. Short press the button to cycle through the 4 options.
  2. Long press the button to save the setting.
Thanks so very much! I finally found this section in the tiny print manual. See my next message in the batteryhookup thread to correct my findings.
 
I recently revisited this thread and thought I'd update it with something I had to do recently for a 12V pack: I made a simple rectangle of schedule 80 PVC pipe and put both copper pipes on one leg of the rectangle on opposite sides using 3 inch stainless screws. When I strung the wire I ran it from one pipe, to the other leg of the pvc through little stainless eye hooks and back. It'd get WAY too hot for open air and melt the PVC so I dunked it in a plastic 30 gallon barrel and it worked like a charm. I had to drill some strategic holes in the PVC to get it to sink and I left the copper pipes long enough to stick out from the water. Something similar with more back-and-forth legs of the wire would work for higher voltages.
 
Lots of good information here. Thank you. The Juntec 1300 I was trying to use was not up to the task for this application. I rigged it up to a cell and powered it up...it was counting current @ 787 amps and accumulating with no load applied. Spent way too much time trying to re-calibrate it with no luck. Voltage was easy to calibrate to my multi meter, but when disconnected from 12 volt power it went back to factory defaults. Oh well, live and learn. Sometimes ya just live. Sure would like to see some pics of your work.
 
Thank you! Well here's a quick photo of my current tester. It's rigged up for the 24V BYD module thats sitting stripped-down next to it. The test load is a 11 year old 3000W victron inverter/charger I picked up on craigslist for cheap. Each of the terminal blocks are wired to one of the cells and also have the various displays, low voltage alarms, and active balancers attached. The small power supply is just a simple 12V to power the relays and amphour meter and the big one is a Meanwell PSP-600-27 that I'm using as a 20A charger that I picked up on ebay for 30 bucks.
Its not pretty, but it works and prevents overdischarging during a test cycle. Other than the $20 a pop active balancers I think I have less than $100 into it. When I'm done with the testing the balancers will go on the big bank.
tester3.jpg
 
That's cool. It's funny, DIY batteries and no name solar modules will be the least expensive components in the system unless you Hunt Craig's List. I spent 20 years in the panhandle of Idaho. The wife wanted a Florida zip code. I suggested a post office box.... I live in Florida now.
 
Back
Top