Powersoccer chair that really kicks

[JoeC]

That's great news about the motors! I know a few people who will be pretty excited about that.

Looking at the current sensor build pictures- are these shunt resistors, or something like a Hall effect?

[Burgerman]

Yep, eBay, quite cheap. 100A DC reversible. Hall.

http://www.allegromicro.com/en/Products ... r-ICs.aspx

5V & 0V supply from Roboteq and signal output from 0 to 5V (2.5V centre at no current) feeds 2 analog input pins. Script takes this and controls motor compensation values. You just add your mOhm figure.

http://www.panucatt.com/Current_Sensor_ ... s-100a.htm

[JoeC]

That's outstanding!

I hate to ride on the hard work you all have done, but I might have an R-net system or two that I'm willing to part with now.

[Burgerman]

Seriously - that's what this forum is for. Its hard work setting everything up, getting the correct parts and making looms professionally etc. Harder to get it all correct than you think. But its worth it.

[sepos]

Finally today we switched power into Roboteq and download the script. All seem good so far. Thanks for good insights!

Next thing would be connecting the joystick. I have QR-7LT from Quickie, but cannot find any user manual to figure out what to do with wiring. Red & black should be obvious, but what about yellow and violet cables?

Or should i just test which is which? On Roboteq side AIn4 (Throt), AIn5 (Sp-Rot) and Ain6 (Steer) would be the right counter parties?

I suppose some of the cables bring data from controller to joystick's direction? Would be interesting to get battery level indicator working too.

[LROBBINS]

You cannot connect the Roboteq to the wires in the stock controller-to-joystick cable - those carry serial digital messages. What you need is the analog signal directly from the joystick can inside the module. It would probably be simpler to buy a joystick and skip hacking into the one you have. An example would be the Apem 3000 series, for example RS product number 6932310 or Farnell number 1829789.

Ciao,
Lenny

[Burgerman]

Read the roboteq manual carefully and slowly.

[sepos]

there is digital input also for Roboteq.
Is possible to use RS232 (RX data, DX data, GND and V) with standard joystick, or am i lost?

[Burgerman]

Seriously. Read the manual.
You are lost!

The powerchair controller uses some custom serial protocol. You will be using:

ROBOTEQ INPUTS:
Analog: voltage swing LEFT/RIGHT
Analog: voltage swing BACK/FORWARDS
Analog: CURRENT SENSOR LEFT MOTOR
Analog: CURRENT SENSOR RIGHT MOTOR
Analog: SPEED POTENTIOMETER
Pulse: RADIO CONTROL LEFT/RIGHT
Pulse: RADIO CONTROL BACK/FORWARDS
Serial: SCRIPT
Serial: USB

ROBOTEQ OUTPUTS:
Digital: BOTH MOTORS via SSR
Digital: STATUS LED
Digital: BATTERY STATE
Digital: MOSFET FAIL to RELAY to cut power if fault

ROBOTEQ POWER
Pos + neg 5V pins: RADIO CONTROL
Pos + neg 5V pins: SSR to control MAIN POWER LATCHED RELAY
Pos + neg 5V pins: JOYSTICK
Pos + neg 5V pins: STATUS LED
Pos + neg 5V pins: POWER ON
Pos + neg 5V pins: 2X CURRENT SENSORS

[LROBBINS]

As John implied, the stock controller is not just plain RS232 serial. It is usually, though not always, some variant of CANbus which is a differential serial with a very tightly defined message packet structure (each frame contains a message ID, up to 8 data bytes, a CRC check sum, and so on) and a lot of features that ensure reliable transmission and receipt of messages. It also operates at a much lower voltage than RS232. The particular CAN message dictionary used by each manufacturer is a closely-guarded secret. I'm sure it could be hacked, and the Roboteq's raw CAN could be programmed to interpret it, but it would be an enormous amount of work. Short of that, or making your own CAN system as I am, you have to use the analog signals directly.
Ciao,
Lenny

[sepos]

Hi, i could not locate any advice how to load these batteries.

Trying to load one half (36 cells) with 12V (like in this picture http://www.wheelchairdriver.com/convert ... ithium.htm)

I am using cellpro powerlab8. It needs few parameters, which i am not sure:
1) number of parallel packs: isnt it 9 in this case?
2) charge rate: it is proposing A1.0C
3) discharge rate: 1.00A

Powerlab and my battery setup dont seem to match as it is saying "pack not present" when i am trying to start monitoring.
What else should i consider?

[Burgerman]

read the powerlab manual. Twice. http://revolectrix.com/support_docs/item_1377.pdf

Download the PC software.
Update the firmware.
Set balance nodes to XH/EH wiring
Set all max input Amps and input power supply voltages.

Set ALL these following parameters correctly. viewtopic.php?f=2&t=4988#p66617
Or use the attached preset (with at least 40 settings set up correctly already). Remove the .zip from the end of the file name:

Import to software/charger. You need a PC.

Watch it charge on screen - open graph...

Tell me what it does!

[Burgerman]

I am using cellpro powerlab8. It needs few parameters, which i am not sure:
1) number of parallel packs: isnt it 9 in this case?
2) charge rate: it is proposing A1.0C
3) discharge rate: 1.00A

Parallel packs none! Your BATTERY has 9 in parallel, but that's not important. its still one pack. A 108Ah pack. That's for charging 5 or 6 hobby packs at once. It multiplies all the settings by x amount... Don't touch it.
Charge rate - see previous post - You can charge at up to 40A if your power supply is man enough. it needs a LOT of things setting up. And the correct pre-set. And no auto C rate charging... The preset will do all of this. You cant really configure this charger without the PC software... There's a lot of tech stuff and limits.

[sepos]

Thanks again.

One parameter was missing
Balance Start V
Error message tole that it should be lower than Charge Voltage, which is 3.50V so i set this to 3.40V.
Or should it be smaller?

Now as i start to charge it gives either message "No Pack" or "Pack Removed (rise 8%)"
Any ideas for this? Could it be the balance wiring issues?

[Burgerman]

Did you use my file? You should.

I suspect you are using the wrong pre-set. Or have not updated the firmware.
Probably a lipo one... That wont allow proper balancing for a dozen reasons.

Charge voltage should be 3.600v OR 3.650 volts initially. First few charges.
Balance should be set to balance only at CV for now which greys out the balance voltage start settings. Until a new update that I have sorted out with them in firmware allows for a 3.4v balance to start with a LiFePO4 preset. And that screen should not ALLOW 3.4v to be set if you are using the correct pre-set.

There will be at least another dozen important setting you have wrong if you are attempting to do this on the charger without a pc connected. You need a PC and a interface - Nobody listens!

Now as i start to charge it gives either message "No Pack" or "Pack Removed (rise 8%)"
Any ideas for this? Could it be the balance wiring issues?

Yes it may be. You did set wiring mode to XH/EH instead of FMA didn't you? I suspect not...

[Burgerman]

viewtopic.php?f=2&t=4988#p66617

READ all of this. Same problem...

YOU NEED HIGH POWER A123 PRESET. THEN:
And you must set ALL of these.
For which you NEED a PC...

The PC software is about 90 percent of the charger. Realistically you need a modern PC or laptop...

You can use it without but will need to set up a mountain of settings manually with a tiny screen and you probably don't have the newest firmware installed either...
Get a PC (OR access to one!)

1. You need to choose the HIGH POWER A123 PRESET. The others are 10A limited.
2. You then need to set all of the following OR CHARGE MAY BE INCORRECT, not complete/ balanced, or cut short.
Trying to do this on the charger may not be possible or easy and some of these settings might be inaccessible or missing it looks so difficult I never even looked. So:

DETECTION
Low Voltage (L.V.) restore: 2.55V
L.V. restore time: 3 mins
L.V. restore Amps: 500mA
Parallel packs: NONE
Cells: Auto detect

CHARGE
Charge Amps: 30.00 (whatever amps you want based on your battery capacity, or power supply. The charger can do 40, most power supplies cannot.)
Increment: 5mA: NO
Charge Voltage: 3.600V (3.50 to 3.55 will take longer to charge, longer to balance but healthier for the battery. 3.65 volts is faster to balance, less good for the battery. All end up full! if time allows)
Conv Frequency: 125khz
Cold Weather T: NONE (reduces charge voltage in winter for safety, mostly for lipo batteries we don't need it)
F. Average past 1min: NO

DISCHARGE
Disch. Mode: Internal Regenerative
Discharge Profile: Constant Current
Discharge Amps: 5.00A
Discharge volts: 2.50V
CCCV Disch. : NO
Cycles: 1
End Cycle: Charge
Cool Down: 5 min
Store or discharge: NO

BALANCE
Bal Mode: PROPORTIONAL
Bal Speed: 16X
Bal Set point: -1mV
Bal entire charge: NO
Balance Disch.: NO
Balance deadband: 2mV
Balance limit CSET/10: NO
Balance Current: 1000mA

TERMINATION
Term Mode: MANUAL
Manual A setting: (70mA Larger figure like 140mA ends faster. Smaller figure like 40mA ends in a more complete charge. But depending in pack, age etc it may never end. Use whatever works best).
Require Balance Done: YES
Require all charge voltage: YES
C.V. Time out: 8 hours

SCREENS
Ignore all settings

FUEL
Pack capacity 0mAh
Use fuel table: NO
Ignore the rest...

SAFETY
Safety charge stop: NO
Charge timeout: 16 hours
Discharge timeout: 4 days
Err. Amps change: NO

Then make sure your power supply and other settings are correct. Set max Amps at or just below your supplies max. Set minimum volts to half the usual volts. Set power supply volts to your supply. Again its easy to do all this with a PC and the software.

SOME of these may be absent and only adjusted via PC... You need a PC!
Make sure you update to latest firmware 3.32 or newer - check often... Also needs a PC.

[sepos]

PC i have.
Firmware updated to 3.3.2
Now using directly your presets.
EX/EH i missed earlier, but now corrected.
supply: 12V min (supply does 24V), 54A (supply promises 55A)
All should be there, but now
"invalid cell count" is the reply, when i try to start monitoring. Need check the balance cabling.

[Burgerman]

yES.

THIS IS 8 CELL 24V? 9 WIRES?

This should be as per the diagram in the manual I linked to.

As you look at the charger, it goes

(neg end) 123456789 (plus end)

1 goes to same place as fat neg charge wore on end of pack.
9 goes to same place as fat positive charge wire on other end of pack.

2345678 go to each cell "join"... between.

Then it will work.

[sepos]

this is like in your picture
http://www.wheelchairdriver.com/convert ... ithium.htm
trying to load on set at time

so only 5 wires.

or should i combine both sets together?
then there would be 9 wires (number 5 and number 1 from both sets combined)

[Burgerman]

Its an 8 cell charger. Why would you charge two 4 cell packs separately?

Nothing stops you doing that, but it will take double the time.

If you are using 4 cells, 5 wires only, then how do you have the connectors to the charger connected?

Should be:

Neg goes to pin 1
cell1+ to pin 2
cell2+ to pin 3
cell3+ to pin 4
cell4+ to pin 5 (or to pin 9 if you choose FMA wiring)

[sepos]

now combined packs together (i misunderstood from some where that it would be better to charge them separately).

in my balance cable there is red wire on left side, 7 white wires in the middle and black wire on right side

watching the picture (http://www.wheelchairdriver.com/convert ... ithium.htm)
i connected

first pack
-red (pos)(the same where thick red wire starts) goes to Pin1
- following wires 2,3,4, to pins 2,3,4
- leaving black wire out
second pack
- red to pin 5
- following white to pins 6,7,8
- black to pin 9

, or am i misunderstanding the picture and combination of two packs?

[Burgerman]

Neg goes to pin 1 --- PIN ONE IS THE ONE CLOSEST TO THE LEFT the negative side of battery, MARked with - on the balance port.
cell1+ to pin 2
cell2+ to pin 3
cell3+ to pin 4
cell4+ to pin 5
cell5+ to pin 6
cell6+ to pin 7
cell7+ to pin 8
cell8+ to pin 9 -- this is the plus side of the battery ON THE RIGHT Marked +

Excuse caps, battery is flat on keyboard...

CHARGER below!
O NEG- O POS+ NEG- 000000000 +POS


http://scriptasylum.com/rc_speed/lipo.html

[woodygb]

Have you joined the two packs ... the last cell of the first pack joined to the first cell of the second pack?

1-2-3-4-5-6-7-8

[Burgerman]

AND page 62!

http://revolectrix.com/support_docs/item_1377.pdf

Voltages are for LiPo so ignore those...

[sepos]

Have you joined the two packs ... the last cell of the first pack joined to the first cell of the second pack?

1-2-3-4-5-6-7-8

Yes i have.

[sepos]

can balance cable be connected to first +cell of each group or is it necessary that cable is in middle of the blade connecting 2 groups?

[Burgerman]

Its the same thing?
It is electrically connected together?


1-2-3-4-5-6-7-8

Can go to 4 or 5, (or both...)

[sepos]

Almost gave up. There was one bad connection in balance wires, which was causing the trouble. Now first loading cycle on going!

[Burgerman]

Loading? charging... 1st time will take a long time.

Watch the cell volts graph on screen.

Save the graph data. email it to me.
Then, the 2nd charge you do all cells should balance in just a couple of mins.

[sepos]

graph data in your mail.
that round of charging was almost 13 hours.
more Amps/time needed or what to balance fully?