LED strip tester will it work

[DarkZero]

i was wondering if my circuit for led testing will work the idea was to get delayed startup and current limiting but that the delay will start when device under test is connected

[jmelson]

320 V DC power source?  Are you sure this is the right value?  What is the purpose of Q1?  Seems like it will just be always on.
The other two transistors seem to be a current source, but the MJE13007 Darlington power transistors are optimized for saturated switching, and this looks like a linear design.  I think those are a bad choice for linear operation, but look up the Safe Operating Area curves to be sure.

Jon

OOPS, I thought it was a Darlington, but the MJE13007 is NOT a Darlington, but a very low-gain plain bipolar, which is going to really not work well with the high resistor values in the original circuit.

Jon

[DarkZero]

Q1 is supposed to ramp up voltage from 0 to 320v so the led doesn't burn before regulation and is supposed to do so when DUT is connected, other 2 transistors are supposed to limit current to ~ 10mA and ill power it from a 240 10VA isolation transformer.

[Zero999]

Just one question: why?

Most LED strip is designed for 12V or 24V, not mains, which I presume is where the 320VDC is coming from. Whilst this circuit will limit the current and not damage the LED strip, it's still dangerous, if the LED strip isn't adequately insulated for mains usage.

Use a lower voltage <60VDC, as it's much safer.

[spec]

Hi DarkZero

I can only emphasize what the other members have said about the dangers of your circuit, and if your LED strings do not need such a high voltage, it would be highly advisable to lower the voltage as much as possible.

That being said, I have had a look at your circuit and find it quite clever, but there seems to be two issues.

The constant current generator appears to be indeterminate because of the connection of R1, and there is liable to be a significant voltage drop across R3, bearing in mind the low current gain of Q1 (high voltage transistors tend suffer from low current gain, especially at low VCE). I had a go at fixing these issues and came up with the circuit shown in the attached schematic for you to have a look at.

[DarkZero]

to clarify why 320 its because of this led tester

https://www.amazon.com/0-320V-Output-Multipurpose-Backlight-Application/dp/B01A1479ZS

i get that 320v is kinda high and i don't know why it was picked for the tester i mean i don't know how can it be safe if u touch probes and it somehow manages to push 20mA trough the body

spec that circuit looks much better

i have one question wont voltage drop on the device being tested to whatever value it needs while in current limiting ? if it does wont they be at their working voltage meeting the isolation requirement ?

[spec]

to clarify why 320 its because of this led tester

https://www.amazon.com/0-320V-Output-Multipurpose-Backlight-Application/dp/B01A1479ZS I get that 320v is kinda high and i don't know why it was picked for the tester i mean i don't know how can it be safe if u touch probes and it somehow manages to push 20mA trough the body

Hmm, it is strange that the Amazon LED tester is safe.

spec that circuit looks much better

Thanks 

I have one question wont voltage drop on the device being tested to whatever value it needs while in current limiting ? if it does wont they be at their working voltage meeting the isolation requirement ?

Yes, that is correct and is the very nature of a constant current generator, which is what your circuit is.

Because the UUT will have it's operating current forced through it, it will by definition also have its correct working voltage across it. So, yes, that will test the UUT for insulation. If the insulation were faulty the UUT would either not work (LEDs would not illuminate) or you would see it arcing over.

[DarkZero]

well ty

i tought this to be very basic circuit but it turned out to have a lot of basics i didn't consider but i'm here to learn

[jmelson]

Q1 is supposed to ramp up voltage from 0 to 320v so the led doesn't burn before regulation and is supposed to do so when DUT is connected, other 2 transistors are supposed to limit current to ~ 10mA and ill power it from a 240 10VA isolation transformer.

OK, but it has a 100K resistor to the base of a bipolar transistor.  When it is pretty much ramped up, there will be only tiny base current available, so it will not turn on very well.

Jon

[spec]

well ty

I thought this to be very basic circuit but it turned out to have a lot of basics i didn't consider, but i'm here to learn

I thought it was a very basic circuit too, but I was surprised how tricky it was without the use of an opamp, and with the high voltages and the characteristics of the high voltage transistor, so you did a pretty good job.

[DarkZero]

Sorry to bring this old post up.

I actually built it and it works quite well. the problem occurs when i reverse bias the led strip or individual led it immediately kills them but in a funny way they illuminate but quite dimly. i measured that current limit is 6mA.

i think they are burning up because of high voltage reverse break down and that current that destroys the led is less than 6mA when reverse biased.
now i didn't try it but i was thinking i add some diode with high reverse voltage breakdown and low reverse bias current in series to stop them from going kaput but maybe im wrong so i tought i would ask here first

[PTR_1275]

What is the highest voltage strip you’ll be testing?

I did some reverse bias tests on a 3mm led (because why not) and it started to glow at around 30v, I got it up to a decent brightness before it lives no more.

Most led strips are 3 (or 6 for 24v) LEDs in series with a limiting resistor.

If you’re only testing 24v strips, reduce that input voltage and you shouldn’t have issues killing strips when reverse polarity.

[Brumby]

If you want to test the LED backlight of a TV, for example, you might want a much higher compliance voltage.

I checked out such a problem on a TV that had 5 strips of LEDs wired in series.  It required 160V to drive them normally and over 100V to get them to glow visibly.  You could check each strip individually, but it would require removing the complete front of the TV (LCD panel, filters and diffusers) to gain access to all the intermediate connection points, where 30V would be enough.

[PTR_1275]

Very good point.

I suppose it really comes down to the application and what the user will most likely be testing.

I have bench supplies that can do 120vdc and if need be I can always series a few up if I need higher voltages. Not everyone has that equipment though and it is something I take for granted

[DarkZero]

i made the circuit spec suggested in the post above and it works well, im powering it from 220/220 5VA transformer but like i said they die once i reverse bias them so im wondering if i can do something to stop that from happening like the tester in this video i know that 320v is well over reverse break down but i just cant think of a way to add some protection to it i lack some knowledge

https://youtu.be/IuHZBE2NP7c?t=442

[Zero999]

The only way to prevent reverse polarity from killing an LED strip would be to add a diode in series or reverse parallel with it. There's no way to make the LED tester which will not kill the LEDs, if they're connected up backwards.

[DarkZero]

im just trying to figure out how the tester in the video doesn't kill LED's when he reverse bias them and how does voltage doesn't rise above 18v in the video, its demonstrated that it can be done i just cant think of mechanism

[Zero999]

im just trying to figure out how the tester in the video doesn't kill LED's when he reverse bias them and how does voltage doesn't rise above 18v in the video, its demonstrated that it can be done i just cant think of mechanism

I doubt there's anything special about the LED tester being used there. Not all LEDs are killed by reverse voltage. AS long as the current is limited to a very low level, old red LEDs and lime green ones just act like a zener diode and non-destructively conduct, when subjected to overvoltage. Modern blue and white LEDs are much more sensitive and can be more easily destroyed, when reverse biased with too higher voltage, although some of the more expensive, higher powered ones have a built in reverse parallel diode to protect against reverse voltage.

[DarkZero]

ah ty Zero999 i didn't know all that. guess its impossible to implement it if its not built into the strip/diode

thx alot so much nice info