2 Cell Driver failure. Cant understand why.

22 posts / 0 new
Last post
cornz
Offline
Last seen: 4 days 12 hours ago
Joined: 10/20/2019 - 05:46
Posts: 54
Location: NW UK
2 Cell Driver failure. Cant understand why.

Ok, here’s the background.
Made a lovely UV torch from:
Convoy c8 host
BD39 17mm 2.2A 3V – 9V 1 cell or 2 cells driver
5W LG UV 365nm Ultraviolet UV LED Emitter 20mm MCPCB

Now, the key there is the BD39 driver.
So, the torch works great with one 18650, but the plan was always to make it a 2 cell version hence the BD39 driver.
So, the other day my c8 extension tube arrived, so i screwd it on, put in another 18650 and then watched as the driver and / or led popped the instant i turned it on.

My question is simple. Why? I thought that the driver would simply output a constant voltage and current as long as the input was between 3 and 9 volts (2×18650=8.4v max)
but this doesn’t seem to be the case.
Have i made some fundamental error (happly admit I’m new to this) or is that driver not what i thought it was?

Mike C
Mike C's picture
Offline
Last seen: 2 hours 31 min ago
Joined: 01/22/2014 - 08:03
Posts: 2401
Location: Sweden

This driver? http://kaidomain.com/BD39-17mm-2_2A-3V-9V-1-cell-or-2-cells-2-Groups-4-t...

From what I can see in the description it is a buck/step down driver, it should work fine with 2 cells. It’s dead when you revert to single cell setup?

cornz
Offline
Last seen: 4 days 12 hours ago
Joined: 10/20/2019 - 05:46
Posts: 54
Location: NW UK

Yeah, except its the 1 mode, 1500mAh variant. Just plain on/off.

http://kaidomain.com/BD39-17mm-1_5A-3V-9V-1-cell-or-2-cells-1-Mode-Buck-...

Dead as. However, it could be the LED that’s fried and NOT the driver but it certainly doesn’t work.
You’re thinking like me, its a step down, constant voltage / current driver. I’ve sent a message to kaidomain to ask them for info.
I was on election duty all yesterday so haven’t had much chance to play but when i metered the led, it was dead short but that could be the driver.
I’ll try and get it de-soldered and test the led out of circuit.

moderator007
moderator007's picture
Offline
Last seen: 18 hours 41 min ago
Joined: 12/23/2012 - 04:47
Posts: 3237
Location: North Carolina

That setup should have worked fine. You’ll need individual testing to determine where the problem is or was.
In my experience UV leds turn black on the die when overheated. Its easily seen by eye.
Another possibility is that it had a bad reflow. Sometimes when reflowed there is only a tiny bit of solder that connects the led to the mcpcb. It will work fine until it gets hot or enough current is applied to pop the fuse (the thin connection of solder) so to speak. Sometimes they will even work fine at low currents but at high currents it just shuts off, which I think is more heat related in a bad solder joint between the led and mcpcb. Just another possibility to check.
Once you get it tore down let us know what you find.

cornz
Offline
Last seen: 4 days 12 hours ago
Joined: 10/20/2019 - 05:46
Posts: 54
Location: NW UK

Digging in now. Led looks fine visually but wont light with diode tester although its still in circuit.
Right, both are FUBAR.
The driver has the aroma of fried electronics.

Led is open circuit. Even directly on led contacts.
The driver, well, i dont know enough about their actual make up but, between the spring contact and the + led lead is about .3ohms.
between the spring contact and the led – lead is an immediate resistance of about 230k which starts to increase probably due to a cap or the inductor on the driver.
Between the led outputs is a multi mega ohm resistance which drops rapidly.
Sounds to me like the driver has flunked under 8.4v and dumped the lot into the led.
Either way, its £15 down the toilet through (it appears) no fault of my own.

cornz
Offline
Last seen: 4 days 12 hours ago
Joined: 10/20/2019 - 05:46
Posts: 54
Location: NW UK

Had the driver under my magnifying glass. No apparent damage but it definately smells of toasted electronics.
I have the bigger 2.2amp version. Going to run some simple side by side tests.

new driver:
Meter + to spring, meter – to LED – is open circuit* something my meter cant measure is happening but on the blown driver, the difference is obvious.

cornz
Offline
Last seen: 4 days 12 hours ago
Joined: 10/20/2019 - 05:46
Posts: 54
Location: NW UK

Ok, so i sent a message to kaidomain.
“Hello. I have been running these now for a few weeks and all was well. Yesterday, i fitted my extension tube onto my Convoy C8 so i could run my UV lamp from 2 cells. The instant i switched on, the led flashed, there was a small "pop" and the lamp died. Upon testing, the driver has blown and also killed my UV led. Given that the driver is designed for 2 cells, what has happened”?

So, had a response from Kaidomain.

Hi Cornz.

It seems the LED had burnt by over-current.

UV LED will turn dark when it is over-heated.

Do you get any situation like this?

We don’t install the UV LED by using the BD39 driver.

I will pass the case to our product support and see what is the matter.

Sorry for any inconvenience cause to you.

Lexel
Lexel's picture
Offline
Last seen: 1 hour 22 min ago
Joined: 11/01/2016 - 08:00
Posts: 5582
Location: Germany

That UV LED can handle 1.5A fine

should not burn before 3A or so as someone here tested them that high cant remember possibly djozz

In any way that UV LED can’t cause a driver meltdown as any other LED won’t do so, it simply acted as a fuse at some point of the driver error

moderator007
moderator007's picture
Offline
Last seen: 18 hours 41 min ago
Joined: 12/23/2012 - 04:47
Posts: 3237
Location: North Carolina

From your description in your first post I would guess a short between led wires and reflector or pill. But you said it worked fine with one cell and you just added a tube and another cell and poof, no other disassemble or modification was done other than adding a tube, right? If that’s true, that leaves about nothing to blame but a bad driver. The setup should have worked fine. If you where using a copper mcpcb, 2.2 amps would have been about perfect for the LG, right in its wheel house.

BanL
Offline
Last seen: 5 days 11 hours ago
Joined: 04/10/2015 - 12:30
Posts: 356
Location: China

Hi

Yes, i had taken this case.

I will make a try and see what is the matter with the driver.

If the driver had burnt, do you feel some bad smell from the driver or do you get a low current in now?

Also, do you have any multi-meter to test the LED can light up or not?

Would you mind to let me know what is the battery are you using?

I want to simulate your configuration.

Regards

Ban

kaidomain.com KD - LED Flashlights, DIY Parts and Accessories

Inquiry for KD, please feel free to contact me.

cornz
Offline
Last seen: 4 days 12 hours ago
Joined: 10/20/2019 - 05:46
Posts: 54
Location: NW UK

Hello Banl.
There is the familiar smell of burnt electronics but there is no visible damage on the BD39 driver. nor the led. i have metered the led, it is open circuit.
I had the DB39 with the 5w UV LG led, with one samsung 18650 and it works fine.
All i did was connect the extension tube and another identical samsung 18650, the INSTANT i turned it on there was a small “pop”, the led flashed for a fraction of a second and then nothing.
The LED and driver do not function at all, there is no short in the convoy host, again the only thing i did was connect an extension tube.. I will attempt to post images on here or on a cloud drive.

Banl, the batteries are Samsung INR 18650 25r (green) although i have also had it working perfectly on Samsung INR 18650 30q batteries (purple).

cornz
Offline
Last seen: 4 days 12 hours ago
Joined: 10/20/2019 - 05:46
Posts: 54
Location: NW UK

No, not a thing.
Just unscrewed the tailcap, attached the extension, inserted battery, reattached tailcap. Click-poof.

moderator007
moderator007's picture
Offline
Last seen: 18 hours 41 min ago
Joined: 12/23/2012 - 04:47
Posts: 3237
Location: North Carolina

cornz wrote:
Hello Banl.
There is the familiar smell of burnt electronics but there is no visible damage on the BD39 driver. nor the led. i have metered the led, it is open circuit.
I had the DB39 with the 5w UV LG led, with one samsung 18650 and it works fine.
All i did was connect the extension tube and another identical samsung 18650, the INSTANT i turned it on there was a small “pop”, the led flashed for a fraction of a second and then nothing.
The LED and driver do not function at all, there is no short in the convoy host, again the only thing i did was connect an extension tube.. I will attempt to post images on here or on a cloud drive.

That sounds exactly like a over current situation for the components on the driver and the led.
cornz
Offline
Last seen: 4 days 12 hours ago
Joined: 10/20/2019 - 05:46
Posts: 54
Location: NW UK
cornz
Offline
Last seen: 4 days 12 hours ago
Joined: 10/20/2019 - 05:46
Posts: 54
Location: NW UK

But my understanding / thinking that the driver should output a constant current whether on one or two cells is correct, yes?
In other words, there is no obvious reason why that driver would fail and blow the led on two cells.

KawiBoy1428
KawiBoy1428's picture
Offline
Last seen: 55 min 41 sec ago
Joined: 04/11/2014 - 18:05
Posts: 3955
Location: The Motor City

KB1428 “Live Life WOT

BanL
Offline
Last seen: 5 days 11 hours ago
Joined: 04/10/2015 - 12:30
Posts: 356
Location: China

Hi cornz

Yes, this driver can be used on 1 or 2 Li-ion cell. The input voltage is supported up to 9V.

I am sorry i don’t try to use this driver on the UV LED.

We are installing our UV drop-in module by using another version with a lower current 1000mA.

Therefore, i will simulate all the situation and see what is the matter with the driver and the LED.

Please don’t worry.

I will reply to you asap.

Sorry for any inconvenience cause to you.

Ban

kaidomain.com KD - LED Flashlights, DIY Parts and Accessories

Inquiry for KD, please feel free to contact me.

cornz
Offline
Last seen: 4 days 12 hours ago
Joined: 10/20/2019 - 05:46
Posts: 54
Location: NW UK

BanL wrote:
Hi cornz

Yes, this driver can be used on 1 or 2 Li-ion cell. The input voltage is supported up to 9V.

I am sorry i don’t try to use this driver on the UV LED.

We are installing our UV drop-in module by using another version with a lower current 1000mA.

Therefore, i will simulate all the situation and see what is the matter with the driver and the LED.

Please don’t worry.

I will reply to you asap.

Sorry for any inconvenience cause to you.

Ban

Hello Ban.
Do you have any news for me?

cornz
Offline
Last seen: 4 days 12 hours ago
Joined: 10/20/2019 - 05:46
Posts: 54
Location: NW UK

So, it all went quiet, i sent KaiDomain a message saying i had heard nothing from them and wanted to know what was happening with my “complaint”.

me:
Hello, i am still waiting to hear the outcome of the complaint i put in about the BD39 driver and the LG led.
BanL has taken over but i have heard nothing.
Please advise what you’re going to do about this as the driver and led are both dead and should have worked.
I would like the LED and driver replacing please.
regards.

KD
Hi Andy
I am Ban
Are you meaning the LED has problem too?
This is smaller chance the LED had burnt out.
Do you have any digital multimeter to test the LED is fine or not.
Sorry for any inconvenience caused to you.

ME:
Yes Ban as i have already explained the bad driver destroyed the LED. I have already explained ALL of this to you here and on the forum.
As you have confirmed that the driver was suitable then it was obviously faulty and blew my expensive 5watt LG uv LED. So i would like a replacement driver and LED sending free of charge please.
Regards.
Andy
PS. The led was on a 20mm pcb.

KD
Hi Andy
I am worried it will have the problem again.
Will you consider to get with lower current 1000mA version?
Regards

ME
But Ban’ that LED is rated at more than 1500mAh and putting batteries in series only increases the voltage. I just want it sorting. Please send a bd39 1500mAh driver and 5watt led. I shall revert to one cell which works perfectly well.
Regards. Andy.

KD
I had tested the driver (input voltage: 8.2V) with the LG 5V LED.
It works fine and keeps running for around 20 minutes.
There are many situations that make the build has a problem.
I am worried it will have the problem again and make you wasting your time.
Are you still want to get the replacement?
Sorry for any inconvenience caused to you.
Thank you

ME:
Hello Ban, then the first driver was faulty!! ALL i did was attach a convoy branded extension tube, added a branded 18650 and thats all. So i dont see what other problems could arise. As a single cell flashlight, it had worked faultlessly.
So, please forward me the following.
1×5w LG 365nm led on a 20mm mcpcb.
KD part number [S028430]

1xBD39 Product Code : CDM BD39 1500mA1mode driver.
.

Thank you for all your help.
Kind regards.
Andy.

To be continued.

pirateo40
pirateo40's picture
Offline
Last seen: 1 week 4 days ago
Joined: 02/25/2017 - 16:13
Posts: 96
Location: United States

FWIW – I have several C8s running with this exact configuration – 2.2A, no issues. Same driver, same LED.
Looking at the pic of your MCPCB the solder job seems kind of sloppy (sorry, just trying to help diagnose the problem). In my experience an LED blown by overcurrent shows some sign of stress visually. I gotta wonder if there’s not a solder short that happened to develop coincidentally when you installed the 2nd battery (but I usually don’t trust coincidences in electronics)
Just my 2c

Mark

MinerShop.com

 

cornz
Offline
Last seen: 4 days 12 hours ago
Joined: 10/20/2019 - 05:46
Posts: 54
Location: NW UK
pirateo40 wrote:
FWIW – I have several C8s running with this exact configuration – 2.2A, no issues. Same driver, same LED. Looking at the pic of your MCPCB the solder job seems kind of sloppy (sorry, just trying to help diagnose the problem). In my experience an LED blown by overcurrent shows some sign of stress visually. I gotta wonder if there’s not a solder short that happened to develop coincidentally when you installed the 2nd battery (but I usually don’t trust coincidences in electronics) Just my 2c

Dont see how. Same switch, identical (stored in tubes) batteries. No indication that ANYTHING was wrong. It had been running single cell for weeks with zero issues.
Connected extension tube, click-pop-dead.
That same extension and host works fine on 2 cells with Simon’s recommended driver. No issues at all.
It WAS a duff driver. No other logical explanation.
For what its worth i did strip it down and man, some of the soldering on these drivers is appaling. But then anything with eutectic / lead free solder is crap anyway. Whiskers and cracks all over the place.

cornz
Offline
Last seen: 4 days 12 hours ago
Joined: 10/20/2019 - 05:46
Posts: 54
Location: NW UK

So, after much tooing and froing, Kai domin have sent my replacement parts. The LED is on a 16mm pcb, not a 20mm and is therefore, useless. They knew, i told them it was a 20mm on my original order.
Sigh. I never have these problems with Simon. Stick to him folks, you get what you order and more, everytime.
Sorry Kaidomain but your aftercare service sucks balls.
My last orders with you have been placed. There wont be anymore.