MKIII low power mode Q&A UPDATE: I took one for the team



:thinking:

Just a btw, 20uA is impossible without a switching device; according to the engineers who designed the FOC Unity, electrolytic capacitor leakage current is much higher than this, on the orders of hundreds of microamperes.

However, it is theoretically possible to get an antispark with an operating current this low, I just can’t be bothered to design one because it would be ridiculously expensive.

Parts needed are LM5163, LTC7000 or LTC7003, and some ultra-low power MCU, like STM32L0 with half-decent firmware, if someone else wants to take a stab at it.

20 Likes

@Trampa I would love to know your thoughts on this.

3 Likes

We meassure the current during hibernation. The outcome is 20uA. Datasheets are one thing, reality is another thing.
The current you mark is the operating supply current, non switching. That is different from the sleep mode.

2 Likes

Pics or it didn’t happen?

6 Likes

:man_facepalming:

The buck converter is never disabled.
Even then, the rest of the DRV8301 consumes at least 20uA in standby mode, anyway.

This is also not counting the electrolytic capacitor leakage current.

2 Likes

As there doesn’t seem to be a VESC 6+ anymore, someone eventually will buy the MKIII and it should be easy then to measure it once and for all.

3 Likes

Hmm, thinking of taking one for the team and ordering one to do a teardown.

I would most likely end up bodge wiring and reprogramming it to run a custom trapezoidal control algorithm though.

5 Likes

That would be interesting!

Why trapezoidal and not sinusoidal?

Going to put it in a belt drive scooter, FOC offers minimal benefit, and considering the International Rectifier garbage thats in these things, would just make it heat up excessively.
Also a custom FOC algorithm would take me a lot longer to implement.

1 Like

FOC has proven to be better. We have done tests at the very limit of the devices capabilities. Faster, less heat, less noise, more torque.

5 Likes

FOCBOX would like to know your location

Well I mean, probably helps that BLDC is not implemented as optimally as it can be. It works but could be better.

2 Likes

But 524€ for two MKIII Vescs are high. Sheesh.

3 Likes

I understood zero of this conversation and I am both an electrician and an engineer , I think I need to go back to college

23 Likes

But does it make my board go zoom?

5 Likes

Deffo. 100A per vesc is a reaper

2 Likes

How is the switch wired?

Is the switch necessary for normal operation? Or Is a jumper needed

1 Like
  • For normal operation nothing is needed.
  • If you want roll to start >> Jumper
  • If you want a power switch, you wire in a simple power switch (normally closed).
  • If you want a LED, you wire that to 5V or 3.3V
  • A single switch can be wired to two or more VESCs if all VESCs are wired to the same battery.
4 Likes

what type of switch, momentary or latching?

3 Likes

momentary

1 Like

Alright guys, I took one for the team and ordered a tramp vesc 6.

Going to test it in my scooter and subsequent (non-ddestructive) teardown and analysis, can’t wait to report to you guys what I find. I expect the actual sleep current to be around 1-2mA.

Interested to see how it performs compared to Flipsky’s 6.6 with enclosure, since that one uses much better MOSFETs, but has a worse QC roll.

Expect the teardown to have me rant about the MOSFETs for a bit.

Expect in the next few weeks for a <$200 trampa vesc 6 being sold on the forums.

30 Likes