Sofirn SP36 spewed toxic chemicals throughout my entire house

[quote=Lightbringer]
Reiterating all this…

I thought that was because you couldn’t make it work with your flip phone. ; I think everybody needs to have some sort of charger or preferably USB tester that shows them what’s going on when charging phones,flashlights or loose batteries. Specifically being able to watch volts and amps at various stages is crucial to understanding what’s going on while charging any device. And a USB tester could help identify what might have been the problem for this OP because he still has the same charger and cable and another presumably identical flashlight. The one recently tested by HKJ would be suitable.

Yep.

It will look like one of these.

Which one, as specifically as you can?
Who put together the bundle of flashlight and charger?
Yeah, it’s possible Sofirn bought and bundled something bad. Or someone else did.

Cheap/counterfeit chargers have killed people.
You can look this stuff up:

https://www.google.com/search?client=firefox-b-1-d&q=counterfeit+USB+charger+fatality

Amazon sells counterfeits irresponsibly.

Look up whatever identifying info you can find on the charger, post pictures of that, and
look to see if HKJ has reviewed that model.

https://lygte-info.dk/info/indexBatteriesAndChargers%20UK.html

https://lygte-info.dk/info/indexUSB%20UK.html

There’s a specific safe distance inside on the circuit board required between the AC current conductors and the DC output, and that proper gap is often not done right.

You could easily get five successful charges followed by an overvoltage failure from a bad charger.

Just saying, don’t throw blame until you investigate possible causes. And don’t trust Amazon.

PS: yes, they don’t do QC/QA on their instruction page text, but you can tell what they’re telling you.
It appears to me they only give a cable, not a hardware USB power supply. Is that right? Where’d yours come from and what is it?

I'm sure Sofirn will do whatever they can to investigate the problem and to compensate the customer by refunding him the broken device as well as any shipping costs (please keep the post office receipt ;-) ) that will arise. Apart from that I'm wondering what more Sofirn should compensate in this very situation. If I was Sofirn I'd probably send Jake some kind words of apology for the unpleasant incident (which I could image has happened already via PM) and maybe - as a gesture of goodwill - a voucher to get another Sofirn flashlight of his choice free of charge. That's it.

Everything else has been addressed by now to pinpoint and rectify this problem as soon as possible. I asked Barry to talk to Sofirn's engineering tomorrow. Maybe they can simulate a surge test with 20V input voltage on another SP36S sample to see what happens and if there's anything else that could make the charging circuit go nuclear. The most important thing to do right now is to get ahold of the defective unit for proper forensic analysis.

As far as I can assess the situation, nobody got hurt and nothing else is kaput except for some scared souls and some unpleasant odor on the USB-connnector and inside the room which hopefully will soon disappear. I always get the creeps when someone is asking to file a law suit against someone else. If this becomes the usual way how to handle things....good lord....we can kiss a lot of flashlight manufacturers or at least their contribution and collaboration on BLF goodbye for good. I now feel set back to the Emisar D4 V2 "muggle-mode gate" Hank and ToyKeeper have been dealing with. :-(

Sofirn’s hands-on approach is better than any other flashlight manufacturer I’ve seen yet and Barry is often “walking among us”.

Of course it’s Jake’s call but I think it’s in Jake’s, Sofirn’s and our interest too to figure out what caused this.

My money’s on the charger too, somehow giving too much voltage to the unit for some bizarre reason. But if it is in the flashlight’s charging interface that could affect many other products including the LT1 possibly which is currently in production.

Full disclosure, I have an SP36 and LT1 on the way so, of course, I’m a little rattled! :stuck_out_tongue:

Where is the link to the pictures? I have had experience with various flashlight driver fires and tested various battery types overheating.

If I got the OP right, it was an SP36S light that malfunctioned. I have only used a USB-A to USB-C cable along with some USB-chargers that come with a USB-A socket. On the picture below you see a Blitzwolf BW-S5 QC3.0 (up to 22V) charger charging the SP36S with the USB-A to USB-C cable supplied with the flashlight. The light is being charged at its maximum rated specs, i.e. 12V @ 1.67A (= QC 2.0).

This is an excerpt from the Sofirn SP36S manual that you can also download from my sharepoint here:

https://drive.google.com/drive/folders/1oZvw8Pybww_O198ynTl8yNXkUjUG9ohw

Sofirn never officially stated the maximum input voltage but I was told by them it is rated up to 12V @ 1.67A. Even though my USB charger can deliver up to 22V output voltage (if the right protocol was being used) my sample did not go beyond 12V. AFAIK, Sofirn never bundles USB-chargers along with their lights. We need Jake to provide the exact brand and model number of the USB-charger and cable he used on the SP36S. If it was a charger with a USB-C socket it's possible the PowerDelivery protocol (20V 5A !) was being used instead of the QC 2.0 protocol.

Dunno, wasn’t there.

Could’ve charged fine during the day and then gotten unplugged shortly after finishing the first few times, vs the last time when it was left to cook overnight.

Dunno if any electrolytic caps are there on the board, but when they “blow” from stress (vs just drying out over time), they can leak their electrolyte. No idea what that smells like.

Like I said, without a proper post mortem, it’s all speculation. I’m just saying what it sounds like to me.

Seriously — the power to damage the flashlight probably came through the wires from the charger.
Which charger? Look it up at HKJ’s page:

https://postimg.cc/gallery/2bbhm4vz4/

This is a picture of the original SP36 driver (not SP36S driver!) I found via quick Google search. I made a circle around the components that I think that could be located right underneath the burn mark.

Following. This is more gripping than a Hollywood thriller in ‘whodunnit’ fashion.

The µC can be ruled out (that thing with ATMEL written on it), as the light works. The other IC looks like a constant current driver.

ah ok. can definitely see the overheated charred spot on the driver board in the middle. Looks like an overheated/cooked chip, and the cells look fine. Just curious of that the voltage and amps of that charger being used was when it happened?

I’m not familiar with the USB-C standard. So, is USB-C charging something that we need to be concerned about? If I use a legitimate USB charging brick that supports whatever-the-latest-and-greatest USB-C fast-charging standard, is any USB-C flashlight I plug into it going to be okay? Isn’t the flashlight supposed to negotiate its charging voltage? Or is there a mis-match of charger and device that could cause a melt-down?

Looks to me like that chip or whatever under the word ‘RUBBER’ is the culprit.
If you look at the red/black wires(in the burn’t photo) and reverse it (turned over) in the next photo that seems to be where it burnt. Luckily both photos are together, making it easy to see.

Here're some more, I put some arrows to it to point to the location that could be the origin of the burn mark (not necessarily the source of the defect!). Please note this is NOT the SP36S driver but I presume the layout will not look very different. Left picture is from Wieselflinkpro and right picture from Djozz.

The USB-C specification looks like a minefield for circuit builders. Apparently the circit board in the device being charged can elicit up to 20V from the charger.
And it looks like the charged device can ask for more voltage than it’s designed to handle, and get it.
Ouch.

https://www.google.com/search?client=firefox-b-1-d&q="USB-C"+charging+voltage+negotiation

Wellp, the green-arrowed thing is a 7135, so I doubt that’s the problem.

Yup.

The 7135 is used in Anduril for regulating moonlight and low modes and is not part of the charging circuit. OP reported that the light still works so that’s probably not the problem.

So, this could mean either a different driver layout or some other component burned up and all the heat on the other side could only leave traces on a very distinct spot (around these small holes in the driver's surface).