DIY

@cw_o Welcome :)

If the price is okay I'd take the risk. Getting actual chips from companies like Mouser, Digikey and LCSC can be pricey once you get to the shipping screen. I've gotten chips from eBay and AliExpress (likely the same source) without many issues for things like IMUs, Mosfets and a handful of logic ICs.
Get multiple though just in case you need to redo it. No more than 5 should be fine. IMU's are fragile so you may need to try a few times.

@cw_o

• If they have a reflow/hot air station then that's the tool you need :) Air for the chip and an iron to touch up the pads.
• Before removing the chip make note of it's orientation for when you replace it.
• Use some isopropyl alcohol and a cotton swab to clean the area and remove some of the conformal coating. It'll get in the way otherwise and possibly compromise a good solder job.
• 180c-200c on the hot air station should be fine to get it off without putting too much heat in the pcb. If you find after a minute it's still not showing signs of moving you can carefully increase the temp but not by much.
• Airflow setting somewhere in the middle. Not so much it flings off parts but not so slow it doesn't put heat in fast enough to fight the power planes heat dissipation.
• Don't hold the heat in one spot as you can popcorn a PCB. Rotate it around the chip so that you get a more even distribution of heat.
• Flux is your friend and will help get under the chip to get the solder molten.
• Don't pry it off, wait for it to be loose and lift up gently with tweezers.

• Once off clean up the pads gently with some solder wick and then use some cotton swabs to take up some flux before cleaning and visual inspection to make sure all pads are good.
• When wicking it's best to cut off a small section and press the iron onto the wick. You can move gently over the top with the iron or if easier use tweezers on the wick.
• Don't pull too hard as the pads can momentarily solder to the wick and tug which can result in a torn pad. If you find the wick has become soldered just apply flux and hold the iron there for a few seconds then try to move it sideways gently while still applying heat. Never pull the wick up.
• As the pads for this chip are under it you'll need to prepare the pads with solder before placing the chip. Go over with a soldering iron and pre-tin them so that they are slightly raised enough that they'll contact and flow onto chip. Not too much that they might squish and bridge. Bit of an art to knowing how much is right. Make sure this iron doesn't get too hot. The pads are delicate and too much heat from an iron can cause them to de-laminate.
• Apply some flus in the middle of the pads where the chip will sit and place it on top as centered as you can get it.
• Bring the hot air back over. The flux will melt and pool around the chip. To stop it shooting off I prefer to try hit it with heat top down so the air doesn't try to blow it away wince we're not holding the chip.
• Eventually you should see it wiggle into place as the solder and flux pull it into alignment. You can give it a gentle tap with the tweezers to make sure it's actually soldered because it'll snap back when slightly poked.
• You can press down a little on top to squish the solder and make sure all pads flowed to the chip. Don't do it with force, just a little to press it down and remove the heat. Hold that for 20-30 seconds to ensure the solder ahs set and you can let go of the chip.
• Use another cotton swab to cleanup the flux and then isopropyl to get the rest of it.

Before powering on I recommend grabbing a multimeter and checking the voltage rails and some of the capacitors for shorts to ground. The caps should have 1 to ground and the other to data so if any are bad then you may have a bridged solder joint under the chip and will have to repeat the process with less solder.

Below is a reference for the traces in the area if it helps at all.

@cw_o Did the seller say what hardware revision it was. It looks like it's pre-serialisation but the 4210-4213 models complain about mismatched BMS's. It can be worked around with additional modding but it's a factor.

4206-4209 are fine so if it's one of those it'll be a simple swap.

@cheesej Great work getting that fixed! Surprised that damage was enough to break the inductor.
Thanks for sharing your process :)

@cw_o Is this a stock board or a VESC?

@cheesej Hiya :) That's a 270uH power inductor. If there's actual damage to the windings on it that could cause issues.

The only time I've seen similar when the battery was fine is dodgy caps on the balance IC.
Cell voltages can report incorrectly if any of the caps around the balance chip at the other end of the pcb are slightly detached or faulty. Common to do when unplugging the XT60 and bumping one.

Might be worth giving this post a lookover to see if your board is doing similar if the inductor doesn't solve the issue.
https://owforum.co.uk/topic/403/xr-overcharged-cell-that-isn-t-after-battery-upgrade-repaired

Board is holding up well. Thought I would run through some things I've discovered and came up with while working on this board, and tinkering with it.

First- absolutely do not put off any maintenance or work on the board longer than needed. Will get back to this later.

I don't have any pictures, but I managed to get access to a makerspace at my new college, which is amazing. The parents aren't as happy that I've now got a place to go work on my projects that has A/C, access to 3D printers, CNC, waterjet and laser cutters. I've signed up to get training on the CNC, waterjet and laser cutters so I can gain some new skills and expand on what I can do next. I have a few projects I want to embark on once I get settled in and see how this semester goes. (Stepper-Motor powered guitar anyone?(Learned that Bambu A1 printers generate their boops and beeps via motor vibration, I'm now obsessed with playing music out of my printer. Roommates might hate me now))

I got the chance to put on some new heatshrink over the cable harness that runs from the battery to the controller. The hard part was running the new heat shrink over all of the cables. I started out with the largest wires, then pulled through the smaller ones using a length of string tied to the ends of a set of wires through the heat shrink. Also discovered that I had somehow lost length on the XT-60 even though I didn't cut it (black magic I guess). Gonna make a two inch extension using some length of wire and a female/male set of XT-60s so the main battery power leads aren't any tension. Remember to put the cable gland on before you put your XT-60 on or else it'll be trouble.

Did discover that if you leave neopixels powered but without data plugged into an ESP32, interference from the motor will cause the LEDs to flip out. Plugging in the data from the LEDs into the ESP32 negated this, and there is no reported issues so far. Will let the smart people figure this one out.

Old wiring harness heatshink wasn't made for the length of BTG, so I barely had any cable gland on the meat of the heat shrink, making for some very poor sealing. Waiting on the new MK8 Ennoid to come in, which has 120V FETs to allow for 20-22S shennanigans once I finally figure out how to properly route and manage wires. Thinking of using some zip-ties and hot glue to work this, and some pre-planning.

Ordered stainless steel m3 screws to replace the ones in my battery box which have finally stripped out. (I hate tiny torx for a reason. I hate tiny screws for a reason.)
Need to look at sealing the front and battery box, though I want to do this after I'm absolutely sure my board is 100 percent wroking. Found a couple trails near the college that have some trecherous terrain I could totally throw the board down to stress test it.

Thinking back on the Hall sensor swap and VESC Tools Hall Analysis tool, I think I may have a reason as to why hall analysis was showing very wacky hall readings. It could be due to the fact that BLE isn't stable and may cause data to be intermitently dropped, leading to data being lost during trasmission. Could be the Vesc software itself on my iPad or computer being generally pretty jank and having some issues that cannot be diagnosed. It could be the VESC itself is old and is just showing it's age. Yes, I'm blaming my tools for a problem I can't solve, but it's the only thing I could think of outside of user error.

Off-topic time:
I went to San Japan over Labor Day Weekend. It's not my normal crowd, but it was fun. Sucked going alone, but it was genuine fun to see people and get to meet people who seemed to be important. Used my board to get to and from the convention between the convention center and my hotel. Got caught in the rain saturday night on the way back, and that soured my mood for a moment.

Anyways, they had a car meetup thing in one of the halls where they had a bunch of crazy cars with anime/waifu wraps, similar to how Warstik was back in 2024, but better. It made me want to do it again, but Craft and Ride is dead and even if they weren't, I don't wanna drop the money on rail gaurds that wouldn't be removable or easy to protect. Came up with the idea there to have removable rail guards using the screw holes found in stock XR rails for those rail clips. BTG WTF rails don't have these, but they can be added with the help of a drill and tap set. Could have a set of screw holes, two on each half of the rail, one on top and one on bottom. You'd have a bit of the rail guard wrap over the rail and would be bolted in place to hold them on. Can't really prototype without proper training on a laser cutter or CNC, but it doesn't mean I can't make ideas. Now I just have to find a way to print on acrylic on the cheap.

@onedangts-xl The tire size increase was a must on the GT side. I mean the frame itself looks so much better now with the larger tire.

Speaking of, how is the tire?

She's alive.

Board runs great, and I've already given it hell. The new campus I transfered too is very hilly and has some pretty steep inclines, so I've been pushing my board up the hills to try and stress it out.

I got the Ubox 100v installed, though I had to design my own mount for it and rip it out of the metal housing it came in. Found out the NRF data port and ADC 2 is dead, so that's fun. I have a new MK8 coming in with 120V FETS. Apparently running 150V FETs with a 75.6V battery makes the controller run HOT. Plus, switching to 120V means I can get insane amperage (About 198.6A for Soft max, over 230 abs. max!)

Motor feels great, and I did the hall sensor replacement on my own. Ended up doing all three, took pictures and then made a pev.dev post about how to do it. It's poorly written, but hopefully it's enough.

One of the riders at 1Wheelparts hooked me up with their quickies. I modified the pubpad to fit specifically the quickie, and I'm going to modify the foot hooks from VOW to fit the quickies specifically.

My new footpad runs a V4 Stoked Stock sensor with the first 5mm printed out of thick PETG, then rest out of TPU

Looking into adding PWM control for my COB LED lightbar, but that will come later if I learn PCB design.

@lia I saw it, can confirm was awesome 😊