July 7, 2009
The 2008 Chopper ( Here ) with an X5305 and 48v 12ah prismatic cells was ok last year but the 12ah cells are fizzing out and they are dying one at a time making the chopper a sitting duck, so it's chopper upgrade time !
Since I have a stock of PSI cells on hand, I decided to put them to use on the chopper, but this time with a crystalyte X5304 ( higher RPM motor than the 5305 ). And to compliment that, a controller modified to push 80 amps !
1st step was to build the battery pack.
When I purchased the PSI cells from Taiwan they came as loose cells without any materials to assemble them into packs other than copper bars to link them and a " Voltage Monitoring System " in a seperate box. A few months later I was advised that there were now assembly blocks ie: lego's available to make pack assembly much easier.
An email and payment to :
I got my Lego blocks !
I've decided to build the pack in 12v blocks, given my history of relocating bits and pieces to different projects as i burn/blow things up
The blocks are incredibly well made, sturdy, strong, low flex, and they lock together very tightly
After lining up the notches, I used a block of nylon and a hammer to set them in place..
I dug thru my stash of carboard and found the template i made last year when building the 12ah pack.. as you can see, the 6 x 12v packs will fit in the frame perfectly with not much room to spare once i add sides to the packs..
One layer of electrical tape and the cells wedge right into the blocks..
I started the electrical tape between the white barcode sticker and the cell's shrink wrap stamp, this way the stickers and tape seams all point in the middle and out of sight ! ( Obsesive Ccompulsive Disorder ?? )
Up to this point, I have not been using any BMS's on my lithium packs, all the packs i've purchased so far that came with them have been problematic, causing charging problems and cutting out when they should not.. instead of doing what they are supposed to do.
This version of the " VMS " only monitors the voltage of all the cells, it does not have an Active Cut-Off to shut down the power if a condition occurs, but it does have allowance for a 5v signal that you can use to activate your own method of cut-off.
Each of the 6 circuit boards monitors 4 cells, they in turn connect to each other with ribbon cables, and all that goes into a main Data Board..
Now, imagine 24 cells, and a wire going to each connection, it gets messy and fast, managing all those wires is a must, making it look good is darn near impossible...
A few experiments later, I found that a dremel tool and a router bit on one corner of the block would allow me to run the wires all to one side of the 12v pack without having to pinch them on the outer edges of the blocks.
If i was going to build the whole 72v pack permanently into the chopper, i'd hook up all the blocks with the tabs, but i'm building 12v packs.. so off with their heads !
The hardest part in al this is making up your mind what to cut, where to run the wires and all the fine details, the rest is just a matter of copying the process over and over again, strangely, I like production work !
July 14, 2009
All 6 packs assembled, and stacked in a way I can bench test them. That's ALOT of wires !!!!
Back side, nice and clean !
Ok.. so here we go.. time to connect all the 12v into 72v.. I checked and rechecked my polarities a few more times and hooked it all up !
Next step.. connect the VMS and start cycling the pack using my lightbulbs...
Soldered all the wires to each 6pin connector, all ready to go.
Well, what's worth doing is worth doing in excess huh..
Got my hands on a full 4x8 sheet of 1/8" thick polycarbonate, the acrylic was cheaper at 80$ for a 3x8 ft sheet, the lexan was 200$ +tax, .. I like having extras, this stuff will be usefull in the future, no doubt..
I"ve had good sucess with heat guns and angle iron strips clamped in a bench vice, I am ordering a strip heater right now !!! seen some videos on youtube using those.. makes life alot .. ALOT easier..
It's a tight squeeze ! and i had to make sure none of the boards short out against each other.. tore strips of foam from the cell boxes and taped over them as insulators.. BMS condoms ?
I Individually charged each cell with single cell chargers to 100%, ready to hook up VMS for the 1st time !!! Each of the 6 boards has a tiny dip switch you turn on one at a time... and Voila !
Then we start testing the pack !
BRIGHT in here ! I get 5 amps from 4 bulbs, 6 more i can add later for aprox 12 amps !
And above, the 3 Soneil 24v chargers, charging at 5 amps on 10ah cells means I go from empty to full in 2 hours !
( Charging at 5 amps = 417 watts in !! You can't do that with cheap batteries ! )
July 19, 2009
When i plug in the charger, the LED"s wake up, but not always on all boards.. I know that if a cell or group of cells have below 8v( 4S ) the BMS will not wake up, but all cells are well above 3v so they should. Somestimes they do.. sometimes they dont.
I have not yet gotten the buzzer to kick in on Overvolt or Undervolt, I plan to drain a deep DOD tomorrow and confrm that it will trigger..
I'm not sold on this VMS thing yet, cell voltages are not ballanced with variance way wider than No BMS that I have been doing on a 48v pack.. but it's early still.. a bunch more cycles on the bench and I will either start to see a pattern or dump this VMS and bolt up a Gary and Richard unit !
July 21, 2009
On Sunday night I fully charged and let the pack sit overnight, last night I went for a deep discharge to confirm the buzzer works .. and it does !.. The VMS bleeds off a significant amount of juice, brings all cells to 3.33x volts. Got 8.8ah before i shut down the test
Well.. at this point in time.. i'm seriously considering my options
- Get 2 x 12 (edit: channel not V hehe) LVC boards from tppacks.com and charge via Single Cell chargers and the 5amp Soneil's ..
- Get Gary and Richard BMS and proper power supply
- Stick with VMS, tho after a 2nd deep discharge and a full recharge, the cell voltages are all over the place.... got the OVP ( Over Voltage Protection ) beep as one cell was a 3.9v and the lowest one was a 3.45v.. i'm not all that impressed with this thing right now.. needs more testing... but time for bed !
July 22, 2009
A Video I made to explain some of this :
Video : VMS in action
July 31, 2009
Ok.. so after thinking about this whole VMS thing, ALOT, i'm ordering a 24 channel board from AndyH ( from the Endless Sphere forum ) the vms is not going on the chopper.. i'm not giving up on it but I want to get the chopper on the road NOW... and don't want to waste too much more time fiddling with the unknowns of the vms.
This weekend I start cutting lexan !
October 24, 2009 !
Ok, so the chopper sat for a while, the RC bike parts arrived and it took over my spare time.. but oct 24 and i'm back on the chopper project.
Cut panels for the 12v PSI packs, even tho this will be a 24S pack I want each 4 cells seperatly bundled.
Table saws rule, but Daaaaamn dangerous ! I'm past the " afraid " level, i'm now " extreemly cautious " .. had one piece nudge the side of the blade, go flying in the air, land back onto the blade and shoot across the garage catching the end of my middle finger, some nice blood spatter !
The white template is the opening inside the chopper frame, so far so good, the pack will be wide but since the chopper has no pedals or chain.. it will be just fine !
Some of the lexan panels installed, i'm leaving the middle vertical one's out so i can use the brass tabs between the packs 1/2 and 3/4/5, those 2 sections will be joined with big thick copper straps.
Brand new shiny BMS built by AndyH !!! THANKS ANDY !!!! looks great !
I plan to install anderson powerpoles on every wire and have them hidden behind the BMS..
Nov 16, 2009
Finished soldering wires and cut a piece of G10 fiberglass board to protect the wires from the back of the circuit board.. Hot glued the wires flat nice and neat..
This is serious stuff, when you work with this many batteries don't screw around, clean work space is a must and don't connect anything until you absolutely have to. By keeping each 12v block seperate I reduce the chance of a MAJOR short, a single cell can create enough spark to really scare the crap out of you so 24 cells would really light up a room !
Nov 23, 2009
Today I finally made my bus bars to hook up the 12v packs for 72v.
I had some copper bars left over from my original PSI order , they don't fit the lego cubes so i made use of them.. Double thick copper on one end, double 12g wire on the other side..
Then crossed my fingers and plugged it all in ! No smoke, No Fire ! Woot !
Some observations, the 3x soneil 24v 5amp chargers I have plugged in series provide 83.7v peak ( / 24 = ~3.48v per cell ) just enough to make one channel blink orange for a split second before the current tappers off to nothing. The main LED stays red and never goes to green.
I added a single cell 2 amp charger in series with that and then I get a whole lot of blinking going on, very cool to look at ! and I get 87.6v, that gives 3.6x v per cell and almost all channels blink until 87.6v the current stops again and all the channels turn off with the main LED still on red.
Bottom line is that the pack is basically fully charged, and does ballance while the BMS prevents any single from going over 3.7v nicely..
So far so good !
Video : BMS in action!!!
Nov 26, 2009
Over the weekend ( after painting the whole house.. man I hate painting!!! ) I cut the side panels... looks blurry because one of the peel away layers is still on there, removed the white side of the polycarbonate to see the lego block holes and pre-drill them.
Spray painted metalic black on the inside of the Right side panel to hide all the BMS wires, the Left side panel will be left transparent on the clean side.. Numbered my BMS wires with gold stickers between a piece of clear tape, once the panels go on it's impossible to know what wire came from where.. I've done this in the past, spend hours building the pack and completely forget to label my wires.... get to the end and realize...aww crap...
Doubled up on the 10 gauge wires and an extra + and - 14 gauge leads for the charger, soldered them to a copper tab, heat shrink and ready to go.
Sandy giving me that " are you done yet ? " look..
Fits in the frame perfectly !
Found a spot for the upgraded controller, bend a small piece of lexan and bolted it to the bike. Used a hose clamp up top with a piece of copper to secure it to the seat tube.
PSI Chopper !
Dec 28 , 2009
Ok.. so the day after my test ride we got 20 cm of snow..
For the last while i've been testing the battery/BMS indoors. On Christmas morning i made myself a cup of coffee and put the pack on charge, i did notice that cell no.24 was not lighting up the LED's and the pack kept trying to ballance.. i checked with my volt meter at the BMS and was showing 3.6xx volts so figured the cell was full but just not quite up to tripping the green light.. fine.. disconnected and plugged in the lighbulbs.
Went next door to plug in the neighbor's new Wii and came back home to 7ah showing on the CA. 70% DOD.. good. put the charger back on. and well... that last LED just would not light up.... then i decided to dig deeper.. checking the ballance wires directly something was obviously wrong... ie: cell 24 at 0.004 volts !!!
Thank god this is the top cell on pack 6 and easy to get at.. pulled the lexan panel off the top and sides, loosened up the nuts on the left side, pulled the leggo blocks off the non-hot glue side of the pack, and dropped in a new cell.
Looking at the dead cell i can smell electrolyte and a small ooze of liquid near the - terminal.
I can't yet explain what caused this, but am investigating...
January 17 , 2010
Alrighty... done some more testing and have info to report.
Since replacing the dead cell, cycling the entire pack a dozen times to 25~50% dod..
- If i leave the BMS plugged in with the ballance leads and the + and - leads hooked up to the pack, cell no.24 gets drained significantly ( like 1 ah over 48 hrs )
- If i leave the ballance leads plugged in but not the + and - everything is ok.
Something strange i've observed is that chanel 24 will start to flicker faintly when the pack is just starting to ballance but cell 24 is still at the 3.4xx volts, the other chanels only start to flicker when cells get to 3.6xx volts
I checked resistance of all chanels ( BMS not connected to pack ) and get 250k on all 24 ( This is version 2.3 of the board btw )
When charging with the BMS i am using a combination of a 48v power supply and a 36v LFP charger in series, the lfp charger limits current to 2 amps and total voltage of them in series gives 96v, if i boost cell 24 with a single cell charger the BMS will go on green and shut down as expected.
Janurary 21 2010
I tested the BMS with a potentiometer ( variable resistance knob thing, just like a volume knob on a stereo ) but not having the right bits this test did not work out.. live and learn lol..
April 11 , 2010
1st.. i just got back from an awsome ride !!!!...
Anyone who's driven an X5 at 72v and 80 amps in a 20" wheel knows exactly what i'm talking about, but for the rest of you who have not.. take the most fun you can have with you pants on and have 2 healthy servings of it .. then add 2 more bonus points for Joy and insanity.. you get pretty much what this is like.... oh man.. THIS IS TOO MUCH FUN ! if such a thing is possible.
I have to calibrate the meter but i get roughly 60 km/h and she gets there in short order ! and in complete silence, only a nice howl off the line as the power comes on strong... i frickin love it !.. LOVE IT>....
But... But.. ... But...
There is also bad news..
I sent the BMS back to AndyH for repair, he did his thing and sent it back !! Thanks again Andy !!..
So i quadruple checked my wires to make absolutely sure it's all correct, wired it all up, setup a DC/DC 12v setup to power a big honking Kilovac EV200 relay, all works great, wired up the e-brake wires ( this time checking polarity. ) and after the ride i took out my new to me Zivan NG1 charger..
I know the Zivan will need caps to filter out the BMS pwm at the end of charge during ballancing, so i have caps on hand, but before getting into that i wanted to try the charger as it is.. so i plugged the zivan into the BMS and powered it up.. beep beep.. and it starts charging... but .. i did not yet have the " sw " wire connected to power the BMS's charge circuit.. blinky blinky orange lights on the channels and the main LED blinks red/green..
Sooo.. i get my soneil chargers out, wired them up and nadda.. green lights.. If i bypass the BMS and charge the pack directly with the soneils they work as expected...
i think i'm going to cry..
I let the soneil chargers do their thing, pack charged and channel no.24 lit up briefly at the end, checked voltage and indeed 3.75v and the rest of the pack is just about perfectly already ballanced so no big deal.
I fired up the motor for a minute to drain a few watts, plugged in the soneil direct at the cell leads and it powers up as expected ( shows a big 10 amp dump on startup then goes to 2 amps and ramps up slowly ) at wich point 20 of the channels lit up indicating cells over 3.65v so i unplugged the zivan
Good news is the voltage profile of the zivan is right.. if the BMS was working as i want it to.
I'm on vacation next week, so tomorrow morning i go for a ride and will try this again on a fresh head.. and hopefully some feedback from the forum.
Picture of the other side, looks much cleaner.. once i get all the details worked out i will make some shrouds for all the wires and tidy things up some more.
Oh, and also on the list for this week, REAR BRAKE CALIPER MOUNT !!!!.. like whoa this thing realllllly needs back brakes !!!