Adaptto DIY 70A charging coil

[Doctorbass]

Adaptto diy 70A charging coil

not an easy project finally!! :lol:

It require a parallel capacitor and a serie inductor between the power supply and the controller charging connections. I already have the right quality capacitor but the inductor is something a little bit more complicated….

After i did many test with various inductor that I modified and various core that I have wound to meet the suggested inductance value, I finally decided to order the real recommended core “E” (The Epcos E 70/33/32) and to wound my own inductance.

According to the Adaptto manual the recommended inductance shall be between 15 to 30uH.

Unfortunately my Meterman XR37 DVM multimeter can normally measure inductance but it had a poor result at such low inductance value so I have ordered one specific inductance meter on ebay for about 25$ that work like a charm and is very accurate for low inductance. In fact it also measure the capacitance very well!

So I have measured many high current inductors that i already had from various salvaged power supply. And in some cases I removed some turn to get between 15-30uH.

Finally what happened:
All of them worked and was able to boost the current to 91-92V for charging but their core was becoming VERY hot !!! and would have melted the wires if I would have not cut the power in time!. That is something strange…in fact the wire temp was OK but the core was very hot!.. So I suspected that the core was saturating and was probably too small for that amount of power. I was testing from 250 to 1200W with meanwell RSP 1000-48. I have tried many combinaisons of toroidal and E shape core with multiple parallel small strands from switching power supply I salvaged without positive success.

Yesterday I received the Epcos big ”E” core and add a air gap between both E core parts to lower the inductance as it is recommended ( about 1.5 to 2mm) and also installed 12 gauge turnigy silicon wire for about 10 turns and I got around about 35uH. Then connected it to the power supply and capacitor and controller input and I was able to push the power up to 1100W and the big E core remained ice cold !!! success!!

So I can conclude that the core volume and cross sectional area is very important in regard of the power amount and also frequency that is used. In the case of the Max-E with lastest firmware Rc9b it’s 18khz.

Today at the lunch I have discharged my 22s battery to 50% soc and wanted to push more power to it with the new working coil but this time strangely the 100v 680uF capacitor was buzzing seriously and heating… that never happened before…. And laso the current the power supply was giving was higher than the charge current measured by the controller… finally it heated so much that the negative leg melted !

Next plan are to investigate why with lower SOC battery the capacitor was having a problem and I was not able to push power to the battery.

Then I will replace the turnigy 12 gauge wire by a parallel bundle of 22 gauge wires. I will use probably 36 to 40 strands in parallel but isolated by varnish 9 only soldered at their end)

Apparently there is also another parameter about inductor that some people are using witch is a different way to take all the strands together. It,s called Litz wire witch consist of twisting multiple pair together and then to twist all these pair together but in opposite twisting direction… this apparently reduce the skin effect. Witch is something I did not tested yet…

Well inductor are not as easy as I expected as well. Some of you would have probably preferred to order directly the full harness coil assembled with the caps and connector and plug and play!

But I want to understand how it work and I already have some parts so why not!

I have learned a lot about inductance here: https://www.youtube.com/watch?v=3b8Wol-X4tE



LCR meter from ebay:

http://www.ebay.com/itm/High-Precis...2ba26&pid=100011&rk=2&rkt=10&sd=221630445296

The Epcos ferrite that I bought ( same as Adaptto big 70A E coil)

http://www.kosmodrom.com.ua/pdf/e_70_33_32.pdf

Litz wire thread:

http://endless-sphere.com/forums/viewtopic.php?f=2&t=66615&p=1001898&hilit=litz+wire#p1001898



some advices form andreym:

more inductance you have = easier for controller, But I`d suggest to fill maximum by adding strands in parallel, and you can add more air gap between the to E core parts, this will decrease the inductance as well.

Electrotransport.ru thread about that:

http://electrotransport.ru/ussr/index.php?topic=25546.0


Inductor charging through the controller on Electrotransport.ru
http://www.microsofttranslator.com/...ssr/index.php?topic=18930.msg327156#msg327156


Doc

 

[liveforphysics]

Nice work Doc! A compact 70A charger is an awesome thing to have on a bike!

 

[Allex]

Older adaptto coils used the chunky E shaped cores on both 30A and 70A coils.
I think that they use 2 round ferrite toroids together on the new 70A type coil and one ferrite ring on the smaller 30A coil.

from top: Old style 30A, New 30A, New 70A

 

[Doctorbass]

Thanks Allex, That,s interesting!

Right now I'm still searching for what type of winding the multistrand wire bundle is made of. I still have not answer for the famous question: Is the wire bundle twisted in the LITZ method or not? and if so, how many twist per inch and how many sub bundle to make the final bundle etc...

From my understanding Litz reduce skin effect. And skin effect is bad because the current flow more on the external diameter of each strand instead of the hole cross sectional ( full volume) witch I guess increase resistance (impedense) and dissipate more heat?

So twisting the multiple strands together in Litz method should reduce skin effect and btw reduce resistance and finally heat dissipation ?

This litz method seem very technical degree and I see also a lot of Hi end hi-Fi purist use that method to make their own RCA and speaker cables to get better performances.. as well as many switching power supply transfo as well as table top induction oven....

Otherwise can I just wound all the parallel strands ( not twisted) and wound them normally like our hub motor are wound?


Litz wire references:

Wiki
http://en.wikipedia.org/wiki/Litz_wire

http://www.newenglandwire.com/products/litz-wire-and-formed-cables/types-and-constructions.aspx

Doc

 

[Luge_a_foin]

Hello,
I've lost my charging coil (stolen in my office?) and I'd like to make several (for my office & home). Basically 5A and 19V or 38V should be enough (labtop psu) so I should be able to make it smaller. Does anyone have any ideas of the 2 components I need for that?
Thanks.

 

[steveo]

Hey Everyone,

I'm looking at building my own charge coil just like doc did, however i'm aiming at building a coil that is compact and as light weight as possible.

I need the coil to support a maximum charge rate of around 1200W

My trike uses a max battery voltage of 50.4v (12s lipo)

My hubmonster ebike uses a max battery voltage of 63v (15s lipo) -- (might upgrade to 16s lipo soon)

I want to also be able to charge with 12v+ of course which should be possbile.

How do i determine what stands of copper wire are best to use? do i simply go by inductance which as Doc mentioned should be between 15 to 30uH? how about length of wire? and capacitor values?

The core options, i know there is the "E" shape core (this seems to bulky and heavy).. and the "o" shape core just like adapttos 30amp version.

i'm gonna keep doing some reading, but any light anyone can shed would be great.

Edit*

Some sold info here for anyone to see, from docs research.

hopefully it translates from russian to english for whoever tries to view it.

http://electrotransport.ru/ussr/index.php?topic=25546.0

thanks
-steveo

 

[Allex]

Here is an important post by inventor if do your own.
http://electrotransport.ru/ussr/index.php?topic=18930.msg492946#msg492946

 

[steveo]

Has anyone here beside Doc successfully made their own charge coil?

I want to make a smaller version of the coil that Doc did, I would say 30-40amp range would be sufficient.

I did attempt to build my own with a E shaped core which only opens at the top of the "E|" << -- like that.

I ran only 200w of power through it and it is already getting warm, i wrapped it with about 1 meter of 12 gauge wire, same experience as Doc, core getting hot.. wires staying cool.

I used 4 caps that are identical to the ones on the 30a charge coil.

SOC of battery is about 20%

so is my problem the core style i'm using? or more to do with the inductance of the coil i did?

if i have to i will follow docs footsteps how he build his.. but its just hudge, and heavy coil, I want to keep this guy in my ebike.

I also prefer the "E" shaped style coil and not the ferrite core...

I do have a farrite coire also that is approx 4 inches wide & about 1 inch thick..

the "E" shape core is about 4 inches wide by 4 inchs tall.. about 1 inch wide.

thanks
-steveo

 

[steveo]

wow .. no reply yet here .. hmm.. i guess nobody out there has built there own beside Doc...

I ended up ordering the core & same wire type doc used.. I made 2 different coils and experienced same hot core problems like Doc.

I hope the new core will work well.. and dam.. the thing has to weight like 5lbs at least!

oh well.. at least i can charge super fast!

i wonder if it will fit in the tiny space left in my frame..

hmm

-steveo

 

[Okami]

Hi there! Heads up! Unfortunately I cannot give any tips, though wanted to give thumbs up for trying to do this at home!

Were looking through some ebike stuff then stumbled upon this! Glad I found it, since I could not understand what the odd looking coils were for, now I at least know that it can be handy in the future to boost up the voltage way higher than the regular small scale boosters can

 

[ElectricGod]

It would be uber cool to make a stand alone high current step up converter. Clearly components inside the adapto and the charging coil get used together to make the step up converter. Does anyone have a schematic of what is happening inside and outside of the adapto to make this work? I have probably 200 4110 mosfets so building something that can deliver the current isn't really a problem to me. It would be cool if it was adjustable so that I could turn a pot and set the output to 48 volts and then turn it again and set it to 82 volts or 120 volts or whatever.

 

[steveo]

Here is mine..

Ignore some of the wiring as it was wrong, I was going to setup my bike with charge coil inside it, but it turns out you can't keep adaptto charge coil connected when controller on, i thought you could

anyway the coil works good.. 18 winds of 10 gauge audio amp wiring.

I've charged at 1-1.1kw with it getting slightly warm soc from 25% - 100% .. i will push up the power to 1.5kw soon to see what it can handle. 1.5Kw is good enough for now till i have 240v in my garage

thanks
-Steveo