The charger is 24v 3A
I’m planning on going up there Tuesday night just to test the motors.
Robomo Club Robot Project
MadTooler
#22
Before you go ordering, it may be good to look into Lithium Iron Phosphate Batteries (LiFePo4). Higher capacity, longer run, significantly better number of cycles, but of coarse, more expensive and a bit more effort to find.
A quick google search came up with some details here:
robomo
#23
I think the AGM batteries will be great. The LiFe batteries are better but might be Overkill. The max range is 15 miles already. We don’t want to be showy.
robomo
#24
The robot moves! The charger is working as well. I have a video but can’t upload it here. Now I’m looking at the controller.
frankjoshua
#25
It’s got a 10k pot for speed control. There are 3 wires from the board going to the base. Power, ground and signal. 8 wires from the joystick into the PCB.
frankjoshua
#26
If anyone works on the robot. The brakes must be engaged to control the robot.
I am leaving the smaller lead acid batteries in the base for testing.
frankjoshua
#27
How should we interface with the base? Emulate the joystick? or Send signals through the same lines that the joystick controller uses?
I probably should have looked inside the motor controller case. Maybe it’s a known controller with an established serial protocol.
frankjoshua
#28
I can’t find much info on how to interface with the joystick or on board motor controller. Most of the projects I’ve found online have just replaced the motor controller with something like this https://www.amazon.com/Sabertooth-Dual-25A-Motor-Driver/dp/B008OMQUXC/ref=sr_1_12?keywords=Dimension+Engineering&qid=1571456539&sr=8-12
I’m not sure if 25A is large enough. The have a 60A version for $60 more if needed.
MadTooler
#29
Did you identify the driver circuits? FETS, driver chip, etc.?
I would stick with the drivers on the scooter and spend the effort figuring out how to make them work. A lot of issues can be avoided being re-invented/engineered that way. Some component identification, a few readings with a meter, and a few probes with the scope should get a lot of the basic bits spelled out. I am not available right now to do it, but there are several others in the club that I bet are very capable of getting it done with even better working knowledge.
The task, along with what Craig was mentioning about his RV gauge, could make for some practical application examples for a monthly club talk.
Not knowing the joystick circuit, here is a conversation about similar where an analog output was needed. Maybe an external ADC.? Again, this approach would be a worst case scenario if the direct to motor driver is too much of a pain.
frankjoshua
#30
This article talks about using PWM to emulate the Joysrick. There were some issues that are over my head but it seemed to be possible. I’m not sure on how to proceed.
@MadTooler I haven’t identified anything. I posted the pictures from the open controller but I don’t know what’s what.
frankjoshua
#32
No sorry. I really should have opened that also. I didn’t think about it until later. I’ll try to get up there soon to get pictures. I still don’t have a key so I’ll have to go when there’s an event happening.
DH1952
#34
A few thoughts about the motor driver and the directional control;;;
One way to make the control wireless and still use the existing motor driver and joystick control is to create a “transparent” RF link. This is different from a Bluetooth or similar wireless transmitter/receiver pair in that the data that is input to the transmitter is not stored as bytes or packets of bytes to then be encoded and sent to a receiver that decodes the packets and sends the information on.
Instead, binary “1” state is simply transmitted as a frequency and the “0” state is transmitted as a different frequency (or possibly as just “off”). The result is that the timing and protocol of the baseband signalling is preserved. The idea is to make the transmitter/receiver pair look like just a wire.
This method works best with UART-type serial or with SPI. It would be very hard to implement with I2C.
DH1952
#35
Incidentally, I tried to find more information about the motor driver module, model ELEASMB6991. Although there were lots of internet hits for this, they were either people wanting to sell the module or they were the owner’s manual for the Jazzy wheelchair. I was unable to find any schematic or other detailed description.
One thing to note - the big IC on the PCB picture that Josh posted looks like it’s an MCP51, which the internet claims is a NVIDIA audio codec. That’s really weird. I can’t imagine what that would be doing in there. Everything else looks like transistors and voltage regulators.
DH1952
#36
One more thing - the transmitter and receivers that are available for the “transparent” transmission mostly top out at about 9600 baud.
frankjoshua
#37
Mark was able to get a clear signal and it looks like Serial data to us. Although, after talking about it for a while it still seems like a major project to simulate the serial connection. We are leaning to using the Sabertooth motor driver. It’s just much less complicated.
DH1952
#38
Okay. One question - what was the time interval of the narrowest pulses? The reason that this is important is that if the baud rate is no greater than 9600 (100usec for the narrowest pulse) then it’s a good possibility that the “transparent” RF link will work, which means existing motor driver and existing joystick controller can be used without modification.
DH1952
#39
I’m starting to think that the Sabertooth motor controller would be the way to go. The Sabertooth can take R/C servo inputs so it should be easy to connect that to a radio-controlled receiver.
frankjoshua
#40
The time setting on the oscilloscope was 200us.
If we go with the Sabertooth controller, I was thinking about using relays to switch between the old controller and new. That way we could still use the joystick to move the robot around. I don’t think the wiring would be that difficult.