Jolt.js charging circuit damaged

[ZacharyLightman]

I was not careful and I plugged a lipo in the the jst connector that was the opposite polarity. the jolt still functions when powered any other way but I would like to repair it. Is it possible to get the schematic? I also have another jolt.js that failed, but I don't know how or when, I had been powering it through the terminals and left a battery attached, It is possible that one of my kids was messing with it and unintentionally shorted the board through the exposed pads on the bottom of the board(it was attached to a metal robot chassis). It would be helpful to have a schematic for troubleshooting. for now I will use my mostly working jolt and I have ordered another.

[gfwilliams]

Hi,

I'm sorry about that - and for the lack of schematics too - I'd put them online but not linked them from the Jolt.js page! Check out: https://github.com/espruino/EspruinoBoard/tree/master/Jolt.js

There are PDF but you should also be able to use Eagle CAD (or KiCAD can import I believe) to view the original schematic and board file.

If the LiPo charger has gone it's super simple - just an LGS4084HB5 that I think you'll need to replace (but ideally you need a hot plate or air gun for that as it's a bit tight). Does the Jolt not work when the battery is plugged in? Or is it that the LiPo charger gets hot?

For the other board failing, that's a bit of a concern. I'd hoped that apart from the small chance of shorting one of the row of small circular contacts on the back, it should be pretty safe. The first thing I'd try is to get a 3v power source and try and power the microcontroller, and see if the red LED blinks once and it's connectable. Just use GND on the terminal block (or the screw one) and then VCC on Q0 or Q1 (or the little gold circular contact nearest the middle of the board).

If that's the case then it'd be a voltage regulator issue I guess - although the only time I've seen those die is when the voltage spikes above 18v.

[ZacharyLightman]

I was able to inject 3v into the "dead" Jolt and It came back to life , so it is most likely just the regulator that was damaged, as you predicted. I decided to order a handful of them from Digikey. I also ordered a bunch of LGS4084HB5s from LCSC as they are inexpensive and could come in handy for other projects. I still don't know what caused the regulator to fail because I only had a 6v battery wired up, but it should be easy to fix now. thanks for the help.

[gfwilliams]

Great! Glad that's likely to fix it.

I've got to say though I'm very confused about the dead regulator if you only had 6v on it. It should be temperature protected so even if it just sat there with 3v shorted out all the time it shouldn't have blown it up. The Jolt is powered via a diode from the battery connector, so I guess something like that could have blown if a lot of power was drawn?

[ZacharyLightman]

I guess I'll know when the new components arrive, if replacing the regulator doesn't fix the issue, I might be replacing the diode too. Luckily I have access to a lot of passive components in 402/602 sizes in my work place(I work for a company that produces automotive electronics) so , I should be able to replace that without ordering something. I will let you know if I find anything else damaged on the PCB, perhaps what could help determine what happened. I was running a small robot but the currents were under an amp, but with inductive things like motors anything is possible. I have participated in some discussions at work about our actuator designs and our simple BLDC motor design can still produce a lot of voltage spikes from parasitic capacitance and who knows what else.

It is funny that the Jolt with the failed regulator can still charge batteries, so I'm using that jolt to charge batteries for the jolt that has an operational regulator but with a failed charger.

[ZacharyLightman]

I was able to replace the charger on one JOLT and the regulator on the other. both boards are now fully operational. thank you for the help.

[gfwilliams]

That's awesome, thank you! I'm still very surprised the regulator went at 6v though - I guess it could have been a spike of back-emf as you say, but there are a bunch of capacitors that'd have to charge up first.

Having said that, by default Jolt.js is powered from a LiPo battery via a diode (to stop voltage from the terminals over-charging the battery). I guess that means that any extra voltage put in via the diodes in the motor drivers won't actually go back to the LiPo battery (although it'd somehow have to get from 3.7v up to 20v!).

You could swap the jumpers on the back around to fix that: https://www.espruino.com/Jolt.js#solder-jumpers - So short LiPo Pwr and open up USB Pwr so you can still charge by USB, but the battery is connected directly to the terminal block and motor drivers.

But it sounds like when you were powering the Jolt off of 6v that'd have been via the terminal block so wouldn't have been an issue