Regenerative Charger / Discharger

Hi All,

On this forum some very good designs can be found for chargers / dischargers and capacity testers.
Most of them discharge the cells trough load resistors.
Which results in that you are actually building an expensive heater :s

So, the cells can be harvested but in the process energy is wasted.
Would it be possible to also use this energy?
For example in a multibay setup, one cell can theoretically be discharged while other cells are charged with that energy.
Increasing overall energy efficiency by connecting them on a common power rail.



Has anyone played around with this concept?

There are some cheap single cell battery modules on the market which include a boost convertor.
Has anyone played around with these? Would it be possible to switch them in parallel on a common bus?

Any thoughts are appreciated

The iMax charger does this, I believe. There's a few custom chargers as well. I think rev0's charger uses this method too

The icharger line of Charger dischargers has regenerative discharge mode on there chargers, Not sure about the IMax charger as the one I have most likely is a bad clone.

later Floyd

Korishan said:

The iMax charger does this, I believe. There's a few custom chargers as well. I think rev0's charger uses this method too

Click to expand...

Thanks,
I forgot to mention in the startpost that I am looking for a self designed version.
As for my application I'd like to build a charger which can interface with another system and possible bretty's DB.

I've looked at Rev0's charger. That one seems to closely resemble what I'd like to achieve.
Don't know if I understand that one fully. At first sight his design looks rather complicated
Am I correct that he dumps the charge not in the second cell, but in a seperate buffer pack?
I can't find figure that out from the youtube vid's/wiki.
Does that charger need to be connected to the PC to start charging?

Propably should ask rev0 direcly

I don't have a lot experience yet with these kind of application.
Don't know if it is possible. But an dedicated charger / discharger chip on I2c sollution is preferred.
As that would overcome the limits of the analog I/o on the Arduino designs here on this forum. Without getting into complicated multiplexer setups.

Ideally, my tester bay setup would consist of a common DC voltage bus and a I2c line.
Connected to an arduino controller.
Resulting in a simple, modular and highly customizable design.

Challenge is, that designing simple things most of the time is complicated :angel:

Re: iChargers: beware that there are Turnigy Reaktor clones of the Junsi iChargers, and while the early versions have almost the same features, later versions have very different firmware and lack many features. So check to be sure ifthe clonesinclude regenerative (dis)charging

Regenerative (dis)charging is discussed at length in most major RC forums, e.g. RCGroups and Endless-Sphere, so search there for much further info. In particular you need to be careful using this feature or you can damage the charger.

EVtinkerer said:

I've looked at Rev0's charger. That one seems to closely resemble what I'd like to achieve.
Don't know if I understand that one fully. At first sight his design looks rather complicated
Am I correct that he dumps the charge not in the second cell, but in a seperate buffer pack?
I can't find figure that out from the youtube vid's/wiki.
Does that charger need to be connected to the PC to start charging?

Propably should ask rev0 direcly

Click to expand...

I'm not entirely sure, really. But I've asked him to make an appearance here

EVtinkerer said:

I've looked at Rev0's charger. That one seems to closely resemble what I'd like to achieve.
Don't know if I understand that one fully. At first sight his design looks rather complicated
Am I correct that he dumps the charge not in the second cell, but in a seperate buffer pack?
I can't find figure that out from the youtube vid's/wiki.
Does that charger need to be connected to the PC to start charging?

Propably should ask rev0 direcly

Click to expand...

I can say this much yes it needs to be connected to your PC.
You cancontrol it with putty and see the output.
You can alsoinitiate a charging/discharging etc. python script that will write to sqlite or thingspeak.

The rest IDK yet.

Wolf

EVtinkerer said:

Thanks,
I forgot to mention in the startpost that I am looking for a self designed version.
As for my application I'd like to build a charger which can interface with another system and possible bretty's DB.

I've looked at Rev0's charger. That one seems to closely resemble what I'd like to achieve.
Don't know if I understand that one fully. At first sight his design looks rather complicated
Am I correct that he dumps the charge not in the second cell, but in a seperate buffer pack?

Click to expand...

So it will do both, but you need a buffer pack that can contain 2x highest cell capacity in WH to be safe (e.g. when both cells aren't synchronized and both decide to discharge at the same time). If one cell is charging and the other is discharging, the discharging cell will naturally provide power to charge the second cell. There is about a 54% round trip efficiency (from discharging cell, boosted to 3s voltage, then bucked down to charging cell). Not the best, but it wasn't exactly my focus while designing this. Certainly more energy saved than 100% resistive

EVtinkerer said:

I can't find figure that out from the youtube vid's/wiki.
Does that charger need to be connected to the PC to start charging?

Propably should ask rev0 direcly

I don't have a lot experience yet with these kind of application.
Don't know if it is possible. But an dedicated charger / discharger chip on I2c sollution is preferred.
As that would overcome the limits of the analog I/o on the Arduino designs here on this forum. Without getting into complicated multiplexer setups.

Click to expand...

Yes, right now it requires a PC. Ideally I would like to add a way to connect each cycler to an ESP8266 to upload data directly to a server/cloud, but alas, I am a hardware engineer that knows just enough programming to be dangerous. If you want, you can see how bad a programmer I am by looking at my CCR source code, all 4727 lines of non object oriented mess: https://github.com/jkenny23/CCR_v2.0/blob/ccr_v2p4/stm32_cycler_v2p4.ino

EVtinkerer said:

Ideally, my tester bay setup would consist of a common DC voltage bus and a I2c line.
Connected to an arduino controller.
Resulting in a simple, modular and highly customizable design.

Challenge is, that designing simple things most of the time is complicated :angel:

Click to expand...

Sounds like a good idea. Indeed things get more complicated than you would think. Now you need to define both directions I2C protocol, packet formats, etc., and fit this asynchronous communication into your charger code without bunging up any of the time critical closed loop control functions of managing a high power supply. So I have not done this I had to abandon the idea of using a display because my code is on the bleeding edge of functional, with how much CPU time it takes to run all the floating point calculations 2000 times a second. Displays use way way more CPU time, guess that should be obvious. Mostly because I couldn't figure out DMA; if you can shove all your data into a DMA FIFO and have it dump that to the I2C bus (in your case) or display SPI/I2C without any blocking, then it would be doable.

Welcome to the rabbit hole! I encourage you to use as much useful info you can get from my design, my github is there and has all the hardware and software design, my project is fully open source.

gauss163 said:

Regenerative (dis)charging is discussed at length in most major RC forums, e.g. RCGroups and Endless-Sphere, so search there for much further info. In particular you need to be careful using this feature or you can damage the charger.

Click to expand...

Thanks, I'm currently browsing trough these forums.
Most topics are about commercial products like Ichargers. Not applicably to the diy I'm looking for.
Will share here if I find some relevant topics.

Oh and regarding your question "When expanding your design to a multibay setup. Would it be possible to have a common voltage bus. Which each single bay will draw current from or supply to?", yes, you can just dump energy into a common bus, and my charger is basically designed for that. The input power line to each CCR is this common bus. You put a 3s buffer pack here for regenerative mode, but theoretically if you have enough CCRs in various modes, they'll act regeneratively, pulling power from other CCRs in discharge. I'm not sure how big of a pack you would need for a given number of chargers. Clearly statistics will kick in eventually and you won't end up in a situation where all 5, 10, whatever CCRs hit discharge at the same time and want to dump 30AH into the buffer pack simultaneously. That's why the latest version combines both regen and resistive, with each CCR able to dump up to 25W or so of resistive power into a FET/heatsink, clamping the common bus to no higher than 12.5V or so.

Well, Ive caught myself in a nice learning experience.
In my original post I was describing a PT5000 as a charger.
Yet, it seems that there are dedicated charging ICs which very closely resemble what I like to achieve.

For educational purposes lets stick to the 5V bus.
I came across a whole bunch of ICs which support USB OTG.
Which basically means the charger can go into reverse operation.
When providing power on the USB connection these chips can utilize that power to charge a Li-ion cell. Some ICs even support 3A charging current.

When there is no power provided at the USB connection they can go into OTG mode. Which turns the IC in a boost convertor, draining the battery and supplying power to the USB connection.



(image is just an example)

There are so many ICs and so many manufacturers, its hard to know where to start. Some have very low discharge current. (500mA for USB 2.0 support) Yet, some offer up to 5A discharge current.

Then there are different power paths.
https://www.monolithicpower.com/how-to-select-lithium-ion-battery-charge-management-ic



In this use case, it would be a pass-trough or NVDC power path.
Meaning that the IC is in control whether it is in Charging, or reverse operation.

However, I still wonder if multiple of these ICs can be used in parallel.
And how to actually ensure that the cells discharge when you want them to.

---

rev0 said:

Welcome to the rabbit hole! I encourage you to use as much useful info you can get from my design, my github is there and has all the hardware and software design, my project is fully open source.

Click to expand...

Hahaa, Well a rabbit whole it is!
Im alternating between, Simple -> Complicated :huh: -> Simple -> Complicated :huh:

Im more of an Hardware engineer myself. Yet more in the area of PLC and relays.
As a person, most of the time I am looking for existing solutions to expand upon an connect in a creative way.

Well, my IC path, might be more in your area of expertise.
Im very curious what you think of my thoughts on the usage (or mis-usage) of USB OTG devices.
(I'm at the point that I want to cut some corners to keep progress)

rev0 said:

Yes, right now it requires a PC. Ideally I would like to add a way to connect each cycler to an ESP8266 to upload data directly to a server/cloud, but alas, I am a hardware engineer that knows just enough programming to be dangerous. If you want, you can see how bad a programmer I am by looking at my CCR source code, all 4727 lines of non object oriented mess: https://github.com/jkenny23/CCR_v2.0/blob/ccr_v2p4/stm32_cycler_v2p4.ino

Click to expand...

That is quite some impressive code. But to be honest TL;DR for know...
Actually designing a whole charging / discharging system from scratch, wasnt my objective.
Actually it is a very small part of my project. But it got so interesting, so spent already quite some time on it.

It seems that you are doing stuff in your code, which I try to find an dedicated IC for.
Its quite a wish list:
I2c communication
Power path control
Temp protection
Reverse Polarity protection
Enhanced charge mode (pre charge, constant current, relative quick 2A)
Discharge / Reverse / boost mode
IR measurement
Undervoltage protection
Fuel guage
And so on...

Already Quite impressed that there are a lot of these devices to choose from. But it's hard to find the right tree in the forest. :blush:
Specifically in an new area of expertise, I get lost quickly.

Maybe my idea of finding the right IC, stick it on a break-out board (or let jlcpcb do that) connect it to an Arduino and re-use bretty wattys code is way too optimistic.