Torqueboards customer service?

That really confirms my hypothesis. Tho an 8A charger should have a fuse or two internally that might have saved the important stuff.

However, I do have a 42V2A charger that is essentially useless… so I mean for science…

2 Likes

Those aren’t useless; I can use those :smirk:

1 Like

It did have a fuse…however that fuse won’t save your charger, it will save your house. In the time it takes to blow that fuse, all the electronic components in your charger will have fried 10 times over.

7 Likes

I got 42V5A chargers for the 10S crowd. 2A is like a leaky faucet filling up a cup lol

2 Likes

I love the 1.8A and 2A chargers.

When you have 3 skates already charged, there’s pretty much never a reason to use a higher charge current. It’s just slightly reducing your battery lifetime for little benefit.

If I don’t have any charged, or need one charged fast, sure I will put it on a 5A charger or on two 2A chargers.

5 Likes

I’m leaning towards this theory as well.

OP said:

Ya mine was full on combustion, like a pack of fireworks and it all started between the BMS and the battery pack . I know the whole diy thing but the BMS and battery pack I had zero to do with.

1 Like

You would hope that the BMS would stop the short-circuit current quickly though. Has anyone tested that TB model?

3 Likes

Good question, although if a random dude on Reddit built the board and most likely wired the charge port the other way round, he could have maybe even bypassed the BMS all together lol

2 Likes

The TB bms is a rebranded and maybe modified Daly.

1 Like

I was wondering about that, it sure looks like a Daly!

3 Likes

This strikes me as an indication he’s isn’t placing any blame on the esc’s, not that it was literally started between the bms and battery. Just that those two items were most likely the initial problem. He seemingly has very little knowledge on the details of diy and isn’t posting much info.

Based on that first pic the largest visible failure is the L4 group which all exploded and only the caps seem to be there or the L1 group which is completely gone, who know why. With most of the right side fine, I think it has to have start with that L4 group. Maybe it was from the bms traces breaking or something of that nature. But I really can’t imagine a little Daly not going poof before cells would hit thermal run away.

Also I’m told you can get a charge port with battery from TB. If that’s the case, and the fuse was not included, I personally can’t lay blame for that theory on the user. I don’t ship packs without fused charge ports, because that’s not something someone buying a build pack should have to install themselves IMO.

4 Likes

As a builder I think it should be standard practice to include a fuse unless otherwise specified, and to test the pack with a known good charger (like 1 minute to verify it works). A little effort goes a long way, I’d much rather spend an additional 10 minutes to reduce the risk of one of my packs going up in flames

9 Likes

I just want to share this with you guys.

I work for the government, we had a Dewalt charger cause a huge fire in our tool warehouse. Caused seriously millions of dollars in damages. Sprinklers flooded an entire concourse of an airport. When these things catch on fire, there is nothing that will put them out

15 Likes

from previous experience, the old version of TB 12s4p pack don’t have a fuse, but it is wired as BMS discharge iirc

honestly, i would prefer the fuse to be installed no matter what, it is there to save lifes, if user decide on something else, they can take the risk of removing it or whatever, just my 2c

3 Likes

I can tell you from experience that breaking the traces won’t cause this: on the BMS I have tested thus far, a balance connection breaking = charge input disconnects in most cases.

OP isn’t really giving much info on reddit, but agreed that it looks like pgroups actually ruptured.

Things that would help narrow down potential root causes:

  1. Has it been charged before successfully or was this first attempt?
  2. has the battery been used at all? Was it ever overdischarged?

I would lean towards a cell failure while charging, which can be caused by a number of things (damaged/bad/fake cells, bad series connections, etc).

4 Likes

I am fairly certain TB does a charge and discharge cycle on all their batteries before they ship em.

4 Likes

Despite designing the PCB used in that battery, I’ve not actually held/inspected a TB production pack yet. Pretty sure they’re being built locally though. I’ll ping Dexter on this one, I’m curious.

7 Likes

@torqueboards A customer with a Metr Pro CAN wanted to connect it to his TB ESC and found out that it has a fluctuating 3.7-4.2V output on its 5V labeled CAN port pin. Is there also something placed between the “real” 5V and the 5V labeled pin on the TB ESC like on the Stormcore?

2 Likes

I have also issues with the customer service of @torqueboards. No answers from them for 4 months. I bought a single motor kit and a motor and nothing. Never received them…

1 Like

Credit card chargeback if you can, asap. Provide all details and communication attempts.