DIY portable, useful LiFePO4

[markopolo]

Some notes on lead acid & lithium pack in semi-permanent parallel via Schottky diodes.

Start of test:
LFP Pack 13.155V
Lead Acid 13.385 (not long off float charge)

20 hours elapsed:
LFP Pack 13.151V
Lead Acid 12.989V

31 hours elapsed:
LFP Pack 13.148V
Lead Acid 12.977V

9 amp load applied to lead acid battery. Current was shared more equally at first but quickly lead to LFP pack supplying most of the current. Voltage difference was 0.4V under load.

44 hours elapsed:
LFP Pack 13.145V
Lead Acid 12.968V

PD converter/charger attached to lead acid battery which started in 13.6V Normal mode.

45 hours elapsed:
LFP Pack 13.147V

70 hours elapsed:
LFP Pack 13.152V
Lead Acid 13.599V

The results are interesting. Just under 0.2V difference with no load and 0.4V difference with a load.

There is some reverse leakage through the Schottky diodes evidenced by the increasing voltage of the LFP pack while the PD charger is on.

I think can work with those results and will probably proceed with the plan outlined here: https://www.classbforum.com/forums/f...tml#post106918 with the update posted here: https://www.classbforum.com/forums/f...tml#post107227

I have three different charger control relays to experiment with.

[markopolo]

The lead acid battery has been in parallel with the lithium battery for 120 hours. The PD converter/charger has been on for 76 hours.

120 hours elapsed:
LFP Pack 13.151V
Lead Acid 14.401V (happened to catch it during a 15 minute Boost mode event)

It's clear that a lead acid battery can be maintained at 100% SOC (they like that) and keep the lithium battery near whatever SOC you are comfortable with (50% or less is a common recommendation) when not in active use.

Here's a video on paralleling Schottky diodes for anyone interested in doing this:

https://www.youtube.com/watch?v=IFJk1KEeokQ

Edit: Just adding that if the Schottky diodes weren't there the lithium battery would have been at 100% SOC for days now. Maintaining a lithium battery at 100% SOC is generally thought to be not a good idea for battery longevity.

[markopolo]

Just another project note about parallel lead acid and lithium batteries.

The old wet cell and the new lithium battery were both fully charged 7 days ago. Right after individually charging them they were directly paralleled, no diodes used this time.

After 7 days the paralleled batteries are at 13.295V.

Approximately 1Ah per day from the lithium battery was used to maintain the lead acid battery at 100% SOC over the week.

Suitable fusing between the batteries is necessary.

[markopolo]

Another app works with my battery packs - Polinovel 2.0

https://play.google.com/store/apps/d...olinovel&hl=en

Play Store image:



You can't change settings but if you just want to check SOC, voltage, current and pack temperature then it's an alternative to the other apps.

Interesting video from Polinovel on cylindrical cell manufacturing:

https://www.youtube.com/watch?v=mCf0cxd-tJQ

[markopolo]

Still on the subject of paralleling lead acid and LiFePo4 ( lithium ) batteries via Schottky diodes.

Victron offers a Schottky diode combiner - https://www.victronenergy.com/batter...tery-combiners - consider it as being the opposite of an isolator.

I was tempted to purchase one but the cost including delivery to me in Canada was high, like over $90 delivered. You can probably get the 80A version in the US for $50 or less. The relatively low current ratings of 40A or 80A likely won't be enough for some folks with lithium batteries if running an air conditioner for example. I would think that they can be paralleled to increase the current handling ability. Also, I didn't really need the combining feature. I just needed an easy to implement Schottky diode package.

I've opted to use 242NQ030 Schottky diodes. Datasheet: International Rectifier 242nq030 datasheet pdf


Schottky diodes.JPG

The $6 per piece cost including delivery combined with the 30V 240A rating seems perfect for my needs. It's pretty easy to use them singly or create a combiner or even an isolator by using more than one with a copper busbar on the output or input ends depending on which device you are making.

If anyone sees anything of concern in the datasheet please let me know.

[@Michael]

Based on your original diagram, I think that if I were to adopt something like this, I'd have to make sure that the path through the diode would be blocked any time the path through the B2B was not blocked.


In my particular case, my concern is that when the B2B is charging, the lead acid side of the circuit will see a voltage drop that would create a >.3V different between the two batteries.

My case might be unique though. I have a 40A B2B, 210AH of AGM's, and only 6ga wire, so unless either the alternator or converter are putting out close to 40A, I end up with significant sag on the AGM side when the B2B is charging at 40A and (for example) the solar is only charging at 10A.

[Luv2Go]

Looks like a nice part (of course I'm biased as I worked for IR). As I'm sure you know, this diode is only good for 240A when suitably heatsinked. Also, since the cathode is the base the heatsink needs to be isolated. There are heatsink sizing guides on the web and figure 6 (the DC line) will give you a good idea of the power dissipated by the device at the current you are planning on using.

[markopolo]

Originally Posted by @Michael

Based on your original diagram, I think that if I were to adopt something like this, I'd have to make sure that the path through the diode would be blocked any time the path through the B2B was not blocked.

In my particular case, my concern is that when the B2B is charging, the lead acid side of the circuit will see a voltage drop that would create a >.3V different between the two batteries.

My case might be unique though. I have a 40A B2B, 210AH of AGM's, and only 6ga wire, so unless either the alternator or converter are putting out close to 40A, I end up with significant sag on the AGM side when the B2B is charging at 40A and (for example) the solar is only charging at 10A.

Thanks for the info. I can see how that could also happen with my setup.

It's not clear to me if the sag is normal to your system or something that can happen but you avoid it.

I'm guessing it's because the Redarc functions like a step-up or boost converter. The energy to output the 40A has to come from somewhere and with solar only supplying 10A then the rest has to come from the AGM's in the absence of other charge sources.

When only on solar do you need to use your Redarc (and consequently pull down the AGM's)?

To avoid it, I have control of both current and voltage on my DIY DC to DC charge controller so I could reduce the output or stop the output to make sure the AGM side gets priority and the lithium only gets the surplus. Additionally, the phone app for the BMS lets me turn off the lithium battery so there is an unseen on/off switch in place between the lithium battery and the diode.

I haven't thought about the design lately. I'm leaning toward simplifying it a bit if the loss from the lithium side to the AGM is as minimal as my testing indicated.

[markopolo]

Originally Posted by Luv2Go

Looks like a nice part (of course I'm biased as I worked for IR). As I'm sure you know, this diode is only good for 240A when suitably heatsinked. Also, since the cathode is the base the heatsink needs to be isolated. There are heatsink sizing guides on the web and figure 6 (the DC line) will give you a good idea of the power dissipated by the device at the current you are planning on using.

Thank you, very good points that have to be considered and managed. The nicely heat-sinked and isolated Victron was very tempting to buy. They've dealt with all of that.

[@Michael]

Originally Posted by markopolo

It's not clear to me if the sag is normal to your system or something that can happen but you avoid it.

I'm guessing it's because the Redarc functions like a step-up or boost converter. The energy to output the 40A has to come from somewhere and with solar only supplying 10A then the rest has to come from the AGM's in the absence of other charge sources.

Yep - the Redarc is a B2B that can step-up/boost, and it's either 40A or 0A, nothing in between.

Originally Posted by markopolo

When only on solar do you need to use your Redarc (and consequently pull down the AGM's)?

I'm re-thinking the solar side. It was set up to only charge the AGM's so to charge the lithium from solar I had to use the Redarc.. My plan is either charge the lithium directly from solar or A/B the solar charger.

After reading your various posts on interfacing a lead acid with lithium, I bought and temporarily installed a Victron 80A battery combiner. I'll experiment a bit and see if it makes sense to keep in place. So far (one afternoon) as expected the AGM voltage tracks the lithium voltage less the 0.3v drop from the diode. Current draw is a few amps from lithium to AGM. Because I have a voltage sensitive bidirectional battery relay between the coach AGM and the chassis AGM, the lithium will charge the chassis battery too, providing the chassis is below ≈12.6V and the AGM is above ≈13.0V.

Here's what I have as of today, minus the solar A/B.



I need to make sure that the SSR that front-ends the Redarc B2B is off anytime the mechanical On/Off that connects the Victron Combiner is on. Other than that caveat, I think it's going to work, even if it's a bit fiddly.

I think that I can run the coach from either lithium or AGM, charge AGM from lithium using the combiner, alternator or converter; charge lithium via alternator/converter/AGM through the Redarc; and charge either using Solar.

I expect to run from AGM most of the time, with the combiner keeping the AGM charged as long as possible. I expect that by the time the low voltage disconnect hits it's 12.2 threshold, the lithium will be at 12.5v, the AGM will be at 12.2v, and both batteries will need to be recharged.

--Mike

[booster]

Just a note of caution on the solar switching. As we have discussed in the past, you have to be careful about having panel output to the controller and a battery connected. Without the battery, the controller can throw a spike, especially if the panels are running up near max with maybe 21.+ volts for many "12v" panels.


A/B switches may be fast enough to complete the changeover before the voltage to the control drops to far, but that might be risky.

[hbn7hj]

Originally Posted by booster

Just a note of caution on the solar switching. As we have discussed in the past, you have to be careful about having panel output to the controller and a battery connected. Without the battery, the controller can throw a spike, especially if the panels are running up near max with maybe 21.+ volts for many "12v" panels.


A/B switches may be fast enough to complete the changeover before the voltage to the control drops to far, but that might be risky.

Just switch them through the both setting.

[booster]

Originally Posted by hbn7hj

Just switch them through the both setting.

I think that may or may not work, as it will all depend on the make/break thing. I don't know if the ones with a both position do a break when going in and out of it or if there is a break, how long it is.

[booster]

Originally Posted by hbn7hj

Just switch them through the both setting.

I think that may or may not work, as it will all depend on the make/break thing. I don't know if the ones with a both position do a break when going in and out of it or if there is a break, how long it is.


Simplest is to just have another switch right there to turn off the panels while you switch.

[hbn7hj]

Originally Posted by booster

I think that may or may not work, as it will all depend on the make/break thing. I don't know if the ones with a both position do a break when going in and out of it or if there is a break, how long it is.


Simplest is to just have another switch right there to turn off the panels while you switch.

I switch them all the time. There is no break going in.

[booster]

Originally Posted by hbn7hj

I switch them all the time. There is no break going in.

I just dug a bit deeper into the Blue Sea literature. Some of the switches, like the E series, specifically say make before break so no interruption in power for battery select situations. Others don't have that statement, so I think that they would have break before make and could be an issue, but they don't state that they are that style. Switches of most kinds are break before make in most other industries.


So I think my statement would have be to be sure to test any switch you are going to use in a spot that could throw spikes as it probably could be either way.


They all will break if you accidentally go through the off position when changing banks.

[markopolo]

There's a new version of the Xiaoxiang app available on JBD's site ( jiabaida.com ). It's version 3.1.1020 or more precisely, version 3.1.1020.24.

I managed to get two screen captures but can't get the app to work without freezing or causing greater issues. Looks like there will be more info available right on the Dashboard:

xiaoxiang 3.1.1020.jpg

xiaoxiang 3.1.1020_.jpg



I tried three phones but couldn't get it to work. With at least one of my phones the Bluetooth modules on the battery packs actually get locked in an always on state and have to be physically disconnected to release them.

I suggest just viewing the screen captures as a preview of what's coming and stick with using Xiaoxiang 3.1.1015 ( 3.1.1015.23.apk ) for now. It's full featured and very stable on my phone.

[markopolo]

The full access app version 3.1.1015 is now available in the Google Play store from Carplounge Tackle -> https://www.carplounge.de/en/loungebox-outdoor-power

Google Play: https://play.google.com/store/apps/d...unge.bms&hl=en

The LoungeBox looks ideal for multi-use portable lithium power. Bluetooth model capacity ranges from 75Ah to 150Ah

https://www.youtube.com/watch?v=T1_RR-QfH4g

I hadn't thought of mounting Anderson connectors like that. I like that idea.

Summary of most current app versions:

App Name: CarploungeBMS
Version: 3.1.1015 (23)
Minimal Android: 5.0 (Lollipop)
Target Android: 10.0 (Q)
Maximal Android: 10.0 (Q)
Compile SDK: Android 6.0 (Marshmallow)

App Name: xiaoxiang
Version: 3.1.1015 (23)
Minimal Android: 5.0 (Lollipop)
Target Android: 9.0 (Pie)
Maximal Android:
Compile SDK: Android 9.0 (Pie)

App Name: xiaoxiang
Version: 3.1.1020 (24)
Minimal Android: 5.0 (Lollipop)
Target Android: 10.0 (Q)
Maximal Android:
Compile SDK: Android 10.0 (Q)

Looks like at least 4 commercial uses of JBD Bluetooth BMS with apps; Liontron, Solmax / Solarfam, Polinovel & Carplounge

Carplounge App:

Carplounge 1.png

Carplounge 2.png

Carplounge 3.png

Carplounge 4.png

[rvsprinterguy]

Can I ask for a summary of what you've built, its specs, and its cost?

[markopolo]

Originally Posted by rvsprinterguy

Can I ask for a summary of what you've built, its specs, and its cost?

Costs in US $
$378 for 135Ah LiFePO4 at 12.8V nominal
$235 for three 150A BMS boards (450A total)
$90 for my DIY 20A DC-DC charge controller
$60 copper buss (have some left)
$50 for three cases
$45 wire (have some left)
$20 for active balancer
$13 for terminals
then there's miscellaneous stuff like solder, kapton tape, bolts, nut, washers etc. (have some left)

I think these BMS boards are a great value and are what sets this apart from typical drop-in batteries. I have full access to 58 settings in the BMS plus an accurate state of charge meter all accessed through an app on my phone via Bluetooth connection. I can view pack temperature, state of cell balance, voltage, capacity remaining and even turn off discharging. I choose the high/low temperature & voltage setting and balancing voltage and precision etc. It's exactly what I wanted.

If the packs are connected in parallel the combined capacity is 135Ah. If discharged at 3C they could power a short duration 5,000 Watt load.

Pack 1: 58Ah 18 lbs
Pack 2: 58Ah 18 lbs
Pack 3: 19Ah 8 lbs

The DC-DC charge controller is also controlled by app on my phone through a Bluetooth connection. I really like that I can set both voltage and current. The latest app version lets me recall 9 different programed presets. The missing feature is automatic charge termination. I have to be within Bluetooth range to stop charging or I can set the voltage lower if away or just turn it off. I will try using macro apps on Android to automatically click the Off button after elapsed time. If that works then I'll try a more ambitious macro that recalls a preset based on elapsed time.

I have no defined use for these batteries and have various options. Each pack with an inverter could be placed in a different areas of the house during a power outage. I could place one or more packs in the van to supplement the AGM's.

I've learned so much during the entire process and that's basically priceless.