Thread Rating:
  • 0 Vote(s) - 0 Average
  • 1
  • 2
  • 3
  • 4
  • 5
Modified power management (LiOn instead of NiMH)
#1
Hi Jan,

very interesting project I found today in the German "Heise-online" article and congratulations for this great idea and the complete instructionss and more...

Since I´m also doing some Arduino programming and 3D printing I have read everything to understand in detail what is done here. :-)
Currently I do not have abike but waiting for my dream bike is back in production. Seems that a lot of parts from China are missing and production in Germany is on hold.

However I developed an Arduino Mini based anti-thief alarm circuit with 105dB sirene based on a acceleration sensor.

And here comes my idea for an improvement of your electronic design. I used a LiOn single cell 18650 (3450mA), a DC StepUp converter (3A) and a LiOn charge / protect module.

The cell has a an output of 3.7 - 4.2V nominale and can be discharged until 2.9 safely. The output voltage can be adjusted via the DC Buck-converter to 6V for the Arduino and servo. So there is no need for 2 batteries. Additionlly I would not use NiMH cells due there is still a little memory effect but also I would say that your charging concept with 2 cells packets in row with different voltages and different discharge current is not very healthy for them. All cells placed in row should have the same dischrage level before charging them.

Here I have attached some links to cells and the modules. maybe you think about this for a future design.

https://www.conrad.de/de/p/samsung-inr18...99572.html

https://www.amazon.de/Lademodul-Schutz-A...B07FB6PLKM

https://www.amazon.de/LAOMAO-Step-up-Con...MDQQ6DHNE8

https://www.conrad.de/de/p/panasonic-ncr...36402.html
#2
Thanks for your thoughts on the battery arrangement. Indeed, my original design is not really TopNotch... I settled for that because of the economics - LiOn chargers are usually a lot more expensive than NiMh. But with that module you posted here, that's another story... Are you sure that single cell battery will perform well under 3A discharge peaks?
Quote:2 cells packets in row with different voltages and different discharge current is not very healthy for them
Everything has been tested for a lot of kilometers before releasing. The batteries are still doing fine. I always use the discharge function on the charger before I begin to charge.
#3
Well, I did not measure the voltage under such high current but I recommend to perform a practical test on your side. The components are cheap (the 2 moduls an da single cell) should not cost more than 12-15Euro/$.

I searched for the maximum current of a servo and found only a 13kg servo with 1.4A at 13kg maximum at 6V. So I think a 17kg or 20kg servo will stay below 2.5A. However here is a 4A module:
https://www.ebay.de/itm/DC-DC-Wandler-St...Swy2tdUYaW

Regarding the NiMH cells... ok if the praxis shows it is working, it´s fine but theoretically you will stress the 2 cell packets since they always have different capacity when starting to recharge them.
#4
Another power delivery option I'm testing is a USB Powerbank, powering both the Arduino and the servo with 5V. Advantages: cheap, self-protecting, easy chargeable, and you can carry a spare one while on the road. Disadvantages: USB cable connection may be vulnerable while mounted on a bike, would require an extra housing or pouch to fix it to the bike. Good results so far on the workbench test-setup, using a 7500mAh powerbank
#5
I'm gonna give this a try.
Two questions for you (thanks for co-thinking about a better power management)
- you quoted two batteries, I guess it is a choice between the two? Panasonic or the Samsung?
- I'm planning to charge one battery with my dynamo fronthub (delivering 6v/2,4W) but I'll need a good power delivery to the charge module. Should I go for a simple 7805/condensor combo, or do you have suggestions for a separate board to regulate the voltage&current?
#6
(10-01-2020, 10:06 AM)Jan Oelbrandt Wrote: simple 7805/condensor combo

no and no and always no....

Every solution ist better like a 7805 solution!
#7
I'll add my experience from powering an automatic shifter off a front hub dynamo for the past few years. I’m not recommending you do it this way, may not be safe so do so at your own risk. I should also make it clear that in my case the dynamo has never made it 100% self sufficient so little risk of overcharging the battery. Depending on the shift resistance and type of riding it can do anything between a few hundred and several thousand miles between charges.

The original shifter was powered off a 3 cell lipo (11.1V rated, 12.6V fully charged) from a powerbank. The Arduino directly connected via the VIN pin and the servo with a servo specific step down converter. The battery pack did have a battery protection board which I kept. When I added a front dynamo I connected it via nothing more than a full bridge rectifier.

I shortly after upgraded the circuity and changed the battery to 2 x 18650s ( 7.4V rated, 8.4V fully charged, around 1.5Ah capacity left) savaged from old laptop batteries. Making sure to use a 2 cell battery protection board (under/ over volt protection only) but again with dynamo hub connected directly via only a full bridge rectifier.

On my latest auto shift road bike I am using a high voltage servo rated 8.4V which can be connected directly to the 2 cells. Simplifies things doing away with the regulator, less to go wrong.

I am aware of circuit ideas to boost the output of a dynamo hubs but keeping things simple it brings up an interesting question, what is the optimum battery voltage to charge?

The dynamo I have puts out approximately 1V per mph. A battery won’t start charging until the dynamo voltage is greater than it plus any rectifier voltage drop. Bicycle dynamos have very good current limiting, usually around 300 – 400mA.
Will give a couple of examples, a 12V battery will only charge at a speed over around 13mph but will if going fast enough (> approx. 17 – 18 mph) charge with a max power of around (12 x 0.4) 4.8 watts.
A single 18650 on the other hand will begin charging at about 5mph but with only a max power at (> approx. 10mph) of (4 x 0.4) 1.6W.
Obviously most people will spend more time riding at over 5mph than over 13mph but the power available with a higher voltage battery is far greater if going fast enough. What I’m getting at is there must be an optimum, yet to do the research myself on this one. I’ve settled with 2 cell li ion (7.4V) at the moment for convenience and reckon it's not far off optimum.

I have to add that lithium cells do need to be treated carefully but I’ve yet to have one get angry at me. I don’t use any balance charging with the dynamo and its not been an issue. I do recommend making sure the cells have similar capacity and have been in the same pack together if reusing old. As I do have to charge them occasionally with a proper charger they get checked for balance.
NiMH does seem like a far safer option and don’t see why they won’t charge just as well from a dynamo hub via a simple rectifier.
#8
@elythomaslumber: I've tried a single 18650 cell, coupled to a DC-DC boost module, I can't figure out why but it wasn't able to power the servo. Thinking of what Ed_R said regarding simplicity and safety, I'm trying this setup now as attached. Thus powering both the Arduino and servo from the 6V NiMh battery pack. The dotted arrow in the schematic means the dynamo hub is not hard-wired but attachable using a RCA plug. I did this because, measuring the voltage output of my dynamo hub, it went way up above 8V. Could be harmful to the max 6.6V rated servo. So I'm leaving it disconnected until I get a low-battery warning, then plug it in to charge the battery while riding. I'll let you know how it turns out in the long run (done 100kms to this day with this setup)


Attached Files Thumbnail(s)
   
#9
I'm wondering if the DC - DC boost module is perhaps not capable of supplying the peak current demand of the servo. On my original auto shift bike I’m using a servo specific buck module called a UBEC rated 3A max 5A, servo is rated 20kg-cm http://www.hobbywing.com/goods.php?id=376

When did you measure the voltage output of the dynamo? Open circuit or with the battery connected? My experience of batteries is that they are very good at clamping the voltage unless they have unusually high internal resistance. Obviously lithium based batteries most never be over charged but NiMh is quite tolerant of continuing to receive current once charged. Especially the low current of a dynamo, it will just be converted to heat and from my experience still keep the voltage down.
#10
(11-22-2020, 10:37 PM)Ed_R Wrote: I'm wondering if the DC - DC boost module is perhaps not capable of supplying the peak current demand of the servo. On my original auto shift bike I’m using a servo specific buck module called a UBEC rated
Yes, that's my guess as well. The one I've got says 3A input current, so I thought it would deliver at least 2A output current. Maybe I'm too optimistic ;-)

(11-22-2020, 10:37 PM)Ed_R Wrote: When did you measure the voltage output of the dynamo? Open circuit or with the battery connected?
With battery connected.
(11-22-2020, 10:37 PM)Ed_R Wrote: My experience of batteries is that they are very good at clamping the voltage unless they have unusually high internal resistance. Obviously lithium based batteries most never be over charged but NiMh is quite tolerant of continuing to receive current once charged. Especially the low current of a dynamo, it will just be converted to heat and from my experience still keep the voltage down.
As posted previously, I'm running the 6V NiMh battery pack straight to the Arduino & servo. The simplicity of the setup, the safety aspect, less electronics that can fail during trips,... make me to still prefer NiMh over Lithium. I carry a connector to charge-while-riding from the dynamo hub. Doing great so far (+300 testmiles since mod)


Forum Jump:


Users browsing this thread: 1 Guest(s)