Dyson DC34 AnimalPro Repair

My friend saved this one from the trashcan. Because it was defective I offered to repair it for her.

Defects

1. The filter retainment clip on top is broken.
2. The vacuum works as intended but shuts off after a short time. When pressing the trigger again the vacuum again works for a short duration. The on-time decreases after consecutive trigger pulls.

Issue 1

Scanning

The broken clip is scanned with an EinScan 3D scanner. This scan is used as a reference in CAD to model a replacement part.

3D scan of the broken clip

Printing

The part is printed on an FDM 3D-printer in PETG filament matching the original colour.
The orientation of the part during printing should maximize the strength of the pins holding the clip in place. To improve the appearance of the part special care should be taken by placing the seams on the underside of the finished part. For this, Cura offers special settings for seam placement, the “smart hiding” option was used in combination with manual seam orientation. The seam position is marked as white dots on the sliced part in the picture below. To download this model, please visit the GrabCAD page.

Issue 2

Working theory: The battery cells are old and as a consequence of that their voltage drops fast when discharging until it triggers the undervoltage protection of the BMS (battery management system). Once the output is shut off the cell voltage recovers and the protection mechanism disengages only to trigger again, but quicker the next time.

Battery disassembly

The battery shell is held together with clips in the same location as the other Dyson batteries.

The BMS

The best feature of this battery is that it doesn’t deactivate itself like the V6 battery BMS does when the cell voltages get unbalanced, which can happen very easily if you leave the vacuum lying around for a couple of months!

Anyways, the most negative cell voltage is connected to P4 and the most positive cell voltage is connected to P2. I was not able to identify the microcontroller on the front.

Close-ups of the backside.

Cells

The battery is comprised of 6 Sony US18650VT cells with the following stats:

The cells were removed from the battery pack and charged/discharged to measure their remaining capacity, which turned out to be just 700 mAh @ 1 A / 28 mAh @ 8 A discharging current. Obviously this is not sufficient and confirms the thesis. The cells were replaced with OEM cells with a capacity of 3000 mAh from a salvaged e-bike battery.

Tests

To confirm that the new batteries could provide the required current the new battery pack was investigated with a current probe attached to an oscilloscope. The first screenshot shows the current draw (100 mV/A) of the vacuum in normal mode, it first ramps up to 6.5 A and then drops to a constant current draw of 5 A. In MAX mode the ramp can be observed at the start but then the current increases in two steps, first to 8.5 A and finally to 11.2 A. At first I tried different cells (the light blue one displayed in the thumbnail) which could not handle the full 11.2 A, as a result the vacuum managed to draw the 8.5 A but shut down shortly after the second step, probably because the voltage drop tripped the undervoltage protection, but I cannot confirm.