Dyson vacuum motor drive

[najrao]

Much has been done in this forum to explain the frightful battery bms in Dyson vacs.

I am trying to get the motor to run on a bench supply. Proving to be very tricky, and I have failed so far.

What is with the two-pin extra connection between battery and motor? At the motor end, both feed into small caps, which measure 2.7uF in situ. Beyond that, it gets complicated to trace the feed. I would have thunk that Dyson somehow used the one power mosfet on the battery board itself as a main motor switch.

In the event, simply feeding 24V to the motor plug causes a draw of some 40mA, after the dc bus has charged. Even the display at the back of the motor does not light up (backlight seems to glow, seen in the dark).

I believe my motor is from a V11, can't be sure.

What more should I do get this going, or is it a stupid endeavor?

Surely, others on this forum would have attempted something like this?

[davidmpye]

Hi,

I can give you some information, but probably not what you want to hear.

The 2 pin connection is a 115200 baud serial link between the battery and the motor unit.
Red wire is battery->motor
Black wire is motor->battery

This appears to serve 2 purposes as far as I can see: 

'proof' to the motor unit that the battery is genuine
reports state of charge back to the motor, so on the V11 you can see predicted runtime.

I'm about to create a separate thread asking for any protocol gurus to see if they can help me figure out the serial traffic.  But in short - without the data flowing, the motor won't power up.

David

[najrao]

Thank you, davidmpye. I had given up all hope of getting response.

Indeed, the battery puts out a signal on the red wire on pressing the trigger. Sighted on a scope. This lasts some milliseconds, and does not repeat till the next trigger press.

Unfortunately, I have only a red-blinking battery*, and this does not turn on the battery power mosfet at all. So there can be no response from the motor.

The bldc driver is not enabled unless there  is a successful handshake, evidently. Bench supply does nothing.

My only hope is locating a node which enables/disables the inverter function.

I look forward (with bated breath) to your new thread, even as my software education was sadly neglected -- from 50 years ago.

najrao.

*a good battery costs much more than what I paid for the three V7-11 motors.