Little FOCer! 84V 5kW VESC-Based Controller

[shaman]

No drawbacks that I’m aware of. The goal surely isn’t to make anything worse

[ed_co]

Any plans of improving it then?

[shaman]

@shaman

Since I have rev up even for the most minor change, then I might as well add stuff to improve the overall performance of the Little FOCer. That being said, v3.1 is being prototyped. I added some stuff that helps when pushing the Little FOCer to its limits.

yes

[ed_co]

Any chance of:

• Handling lights and buzzer
• Improve BT signal
• Change to a more adequate pull down resistors

Previous latest two, as per seen in some videos

[shaman]

@ed_co

Handling lights and buzzer

Big LED strips draw too much power. Buzzers are fine and people like @surfdado have made code that makes use of the spare GPIO pins to drive a buzzer.

@ed_co

Improve BT signal

v3.x has much improved BT signal vs v1 and v2. I’m further improving it in v3.1 by switching to a ceramic antenna and reducing the copper content in the lid.

@ed_co

Change to a more adequate pull down resistors

This is entirely subjective as to what “adequate” is. 10k is a generic value that is added. If you want stronger pulldown on any ADC channel, then you have to add it yourself.

[ed_co]

@shaman

Big LED strips draw too much power. Buzzers are fine and people like @surfdado have made code that makes use of the spare GPIO pins to drive a buzzer.

There are VESCs which can handle led strips. I guess maybe having a buck converter for that? Not sure.

@shaman

This is entirely subjective as to what “adequate” is. 10k is a generic value that is added. If you want stronger pulldown on any ADC channel, then you have to add it yourself.

Allegedly 1k was more addecuate than 10k for less persistancy and more immediate while stopping.

[shaman]

@ed_co

There are VESCs which can handle led strips. I guess maybe having a buck converter for that? Not sure.

This is dependent on the power of the on-board buck converter. The buck converters of the lower voltage vescs can supply more current. the tradeoff of having a buck converter a capable of accepting higher input voltages (20S) means it can’t provide as much current as a lower voltage counterpart.

@ed_co

Allegedly 1k was more addecuate than 10k for less persistancy and more immediate while stopping.

For what application though? OW? If so, then that’s a particular application where 1k is better. But that doesn’t mean 1k is better for all applications of which the little FOCer is used for

[ed_co]

@shaman

For what application though? OW? If so, then that’s a particular application where 1k is better. But that doesn’t mean 1k is better for all applications of which the little FOCer is used for

I was just talking about the FM version of LFOC which I guess was just created for OW. So maybe is worth it.
Thanks for all your quick replies.

[surfdado]

@shaman

educing the copper content in t

There is only one LFOC - no FM versions. You might be confusing that with bundles?

[ed_co]

Yes, there is (there was, there is no stock for anything). At least in Europe.

[firepower]

When more stock of little focers expected?

[shaman]

I’m working on it as fast as I can. Stay tuned for an actual eta probably this week

[shaman]

Bad news: New buck converter is crap. Can’t handle the fluctuating battery voltage from sag and regen. I have found stock of the old buck converter (tried and true) and I’m throwing that back into the design. I need more time to shop for a decent substitute.

Good news: Everything else is working fine. The new IMU is proving to perform well in testing so far.

[Battery_Mooch]

@shaman

Bad news: New buck converter is crap. Can’t handle the fluctuating battery voltage from sag and regen.

Would you be comfortable sharing which converter this is?

[shaman]

http://www.xlsemi.com/datasheet/XL7005A%20datasheet-English.pdf

Kind of embarrassed that I even tried it. The ample stock was attractive though. But I wont knowingly sacrifice quality.

[Battery_Mooch]

Not much data in their datasheet but certainly nothing there to indicate it would give you the problems you had. I would have tried them too. Sorry it didn’t work out but glad you found stock of the other converter!

[mr.shiteside]

Spitballing here but would a Schottky before the input cap in the buck circuit help you out? I haven’t seen your schematic but it would save the input cap from backfeeding and effectively dealing with huge ripple currents. If used as per datasheet<5W output, let’s say 6-7W input, 33uF input, ~40V input, you could pretend it’s a ~175mA input current or ~225R input impedance, giving RC of around 7ms. Without bulking up input capacitance to buffer it out you’d at least get over the high frequency stuff by isolating the buck from the ESC and could significantly reduce input sleep rate

This is a student guess though, this could already be in place or have a big hole in the logic

[shaman]

Yeah there’s both a reverse bias protection schottky diode and a TVS protection diode along with a small current limiting resistor and some input capacitance. It’s got every bit of help I can toss it with the space I have to work with

[mr.shiteside]

Yeah never mind then, wasn’t sure how standard that stuff is because I haven’t seen them in many app notes but I guess they’re just assumed you’ll use it if required like this

[ed_co]

Looks like customwheel.shop, just removed the prices. I am feeling for the worse. I think €350 was quite a bit already. I am so looking forward to buy one but prices are getting crazy lately in the DIY Scene in general