Kryton groove duty cycle @1A + SWITCH FIX inside!

Hi folks!
Today I got some tailclicky switches from DX in the mail.
Just what I needed to complete my Kryton groove 445nm build that
I got basically new and unassembled from someone who liquidated his hobby.
It had a nichia blue already pressed in its head, so I was kinda stuck
with that, not wanting to risk killing the diode by clumsily removing
it from it's press fit without proper tools.

The Kryton groove is a very slim and good looking host, but I needs some
fiddeling with to get it to work.
For one, mounting a spring for the positive battery contact needed some
macgyvering involving plastic, silicone rubber, springs and glue.
I guess I could have done it the quick way using a common flashlight-like
battery contact tab.

Also the tailspring needs some trimming and all-round support, lest it bends over
to the side, shorting against the barrel and thus turning the laser on all the time.

Now this was my first experience with a FlexDrive.
Pretty cool circuitry, might I add. Very compact, relatively easy to set
current, once I managed to solder blob those little SMD resistors.
So I set it at a relatively conservative 1A, hoping to get away without heatsinking the driver.

Now here's my "problem".
It lases like a winner, until after about one minute, when the laser starts
to slowly (1Hz) flash off and back on. This flashing becomes faster, until
I chicken out and turn off the laser altogether.
I read that this symptom is the FlexDrive thermal protection kicking in.
It's just as well because by that time the head where the diode sits
has gotten pretty warm, and you would have a bad feeling about
keeping that little diode going.

Hmmmmm.
I realize these hosts where designed at a time where the only >1W diodes where IR.
Still:

Is this 1 minute duty cycle considered "normal" for these hosts running a
445nm at 1A?
Am I just spoiled by my "porta-labby", which doesn't even know what a duty
cycle is in the first place, thanks to it's active cooling?
Mind you, this is my first HANDHELD 445nm.
Should I bother heatsinking the driver for a few extra seconds until the whole
thing has to cool down?
Should I better turn it down a notch to, say, 800ma?
:thinking::thinking::thinking:

EDIT: turns out the problem was two fold.
Heatsinking the driver with a coin, like in DTR pictures below, solved the blinking that started after a few seconds.

Then the switches kept failing, here's what I did about it. You can do it too.
To open the switch first pry each corner a bit via the "big" open slot on each side.
Next, slide a thin blade in each corner and cut through the 4 plastic poles.

You can see how the spring deformed when it got hot from the high (1.5~2A) current.

Give the electricity an extra path or two to flow. I guess the spring itself
now sees only 1/2 or 1/3 the current it had to handle originally.
I removed the here unnecessary wire insulation to not obstruct the switches action.
When properly done, you don't notice a thing in the clicky's feel.
It works as smooth as original.

Glue it back together. I'm using Pattex Contact glue, cyanoacrylate isn't strong enough.
Even if the glue job isn't perfect, once the switch is
installed in the tailcap, it will be held firmly in place and together.


I also soldered the big external spring to it's little tabs for reliabilty,
and trimmed the spring down a few milimeters to avoid bending and shorting
against the inside of the barrel.


If you want your high current Kryton to work consistently but don't feel like
performing this mod, you can also get such a reworked switch from me for 5$.

Re: Kryton groove duty cycle @1A

Have you thought about sinking it. Here is an excerpt from one of my ehgemus hosts build tutorials. I would think something similar would work for you and you could turn it to 1.5A if you wanted without worrying about duty cycle.

"I took a penny and ground it flat on both sides then bent the coin slightly on both sides.








I had to bend a little and check to see if it has contact with the host walls and repeat until I got a good fit.







Then I mounted a new flex set to max on the coin with some thermal adhesive.







Once the driver was set it was time to mount the coin in the host. Used an alligator clip to hold it against the host until the adhesive dried.







Then all I had to do is solder the diode to the new driver leads.

"

Re: Kryton groove duty cycle @1A

So how does the driver not contact the host body and short it with the penny?

Re: Kryton groove duty cycle @1A

Thermal adhesive is non-conductive.

Re: Kryton groove duty cycle @1A

Wolfman29 said:

So how does the driver not contact the host body and short it with the penny?

Click to expand...

The bottom of the chip you are sinking to the coin is not conductive plus you use thermal adhesive that also has a good deal of electrical resistance. Just make sure your wires are not protruding to far through the holes and you should be fine.

Here check out my coin mod thread. That is discussed in some of the replies.:beer:

http://laserpointerforums.com/f51/f...-driver-duty-cycle-1-5a-56404.html#post796360

Re: Kryton groove duty cycle @1A

Yeah thanks DTR, I knew about those coin mods, and I was thinking about doing something
similar. Too bad I don't have thermal adhesive on hand at the moment, only CPU thermal grease.

But: the driver heating is not the only issue, the Krytons head gets very warm too, from the diode.
Not much point in having a continuous duty cycle for the driver if I'm cooking the diode!
I was looking if someone who had a Kryton groove with a high powered diode could confirm this, and/or tell me what they did about that.

Is the fact that I'm getting the driver's negative input from the diode's case pin
contributing to the heat generated in the head? I think not, but wanted to ask anyhow.

Re: Kryton groove duty cycle @1A

I am a bit puzzled. I have a laser that is doing this as well using a microboost (unheatsinked). It would begin the slow blinking between 90 seconds and two minutes. I thought, "If I have to I'll just turn it down and sell it 'at cost' to someone."

I hooked it up to my bench PSU and ran a test to make sure what the output was before I started. Since I had it in front of the LPM, I decided to see how quickly the power was dropping with the heat.

The laser ran for 5 minutes without blinking, when hooked up to the bench PSU! The head was hot to the touch at the end of the 5 minutes. It was just the head and driver. It was not screwed into the body.

So it got hot enough to be uncomfortable to hold, and yet continued to lase without blinking. The bench PSU was set to 4.5V.

I am wondering if maybe we have some other phenomenon at work here. Could the problem be caused by something that is only present when the driver is being powered from a battery? Could it be that the battery cannot keep up? Could it be something overheating in the switch?

Peace,
dave

Re: Kryton groove duty cycle @1A

Dave could the driver be heating up quicker if it is boosting more. For example would it take longer to over heat boosting 3.5V in to 4.5V out as compared to 4.2V in to 4.5V out?


Also
Have you thought about just sinking it.

Re: Kryton groove duty cycle @1A

daguin said:

I am wondering if maybe we have some other phenomenon at work here. Could the problem be caused by something that is only present when the driver is being powered from a battery? Could it be that the battery cannot keep up? Could it be something overheating in the switch?

Click to expand...

OK I just replaced the tailcap real quick with a DMM, so I suppose I can rule out the switch overheating.

Yikes, I had no idea! Drawing 2.3A from the battery for ~1A at ~4.2V.

I'm pretty sure it's the driver overheating and not the batteries fault, because when
I let it lase until it starts to blink, and then quickly replace the battery with a "fresh" one,
it starts blinking again very shortly thereafter.

Yeeeeaaah I'mma heatsink that microflex somehow....:thinking:

Re: Kryton groove duty cycle @1A

anselm said:

OK I just replaced the tailcap real quick with a DMM, so I suppose I can rule out the switch overheating.

Yikes, I had no idea! Drawing 2.3A from the battery for ~1A at ~4.2V.

I'm pretty sure it's the driver overheating and not the batteries fault, because when
I let it lase until it starts to blink, and then quickly replace the battery with a "fresh" one,
it starts blinking again very shortly thereafter.

Yeeeeaaah I'mma heatsink that microflex somehow....:thinking:

Click to expand...

That current draw sounds really high. On my dual flex builds set to 1.8A it only draws about 2.4A.:thinking:

Re: Kryton groove duty cycle @1A

I don't know how these things work but maybe the fact that the microflex I used
has a big chunk of it's boost coil's core chipped away has anything to do with it
drawing lots of current?:thinking:
It's driver no. 1 in these pictures:
http://laserpointerforums.com/f67/help-identify-some-drivers-pics-61671.html

OK, now I'm really confused.
I used another DMM, and now I'm reading 1.3A slowly climbing to 1.7A.:wtf:
It is a cheap DMM, but the other one that I used before is much cheaper, really
the worst kind. I'm not trusting that one anymore, at least until it gets a fresh 9V battery.

Also I guess the choice of batteries is an important one too...

Re: Kryton groove duty cycle @1A

DTR said:

Dave could the driver be heating up quicker if it is boosting more. For example would it take longer to over heat boosting 3.5V in to 4.5V out as compared to 4.2V in to 4.5V out?


Also
Have you thought about just sinking it.

Click to expand...

I don't know about the electronics draw vs. heat question. Hopefully someone that knows the electronics will chime in.

I will heat sink at least one to see if that helps.

Peace,
dave

Re: Kryton groove duty cycle @1A

What voltage is your battery producing anselm?

Re: Kryton groove duty cycle @1A

@BionicBadger:
Well anything between 3.8 and 4.2 volt (no load) depending on which battery I use.
Under load, I'm sorry I couldn't tell you, I don't know how to get the batteries positive
side when it's inside the laser.
It feels like one would want to use a GOOD battery (low internal resistance) for a build like this.

Anyway...
That "£$%^ tailcap is a whole issue by itself. The first switch crapped out after two days
of just occasionally playing with it, the laser wasn't even assembled.
The replacement switch lasted exactly 2 days.
Sure I have 4 more of them, but this is totally unacceptable. I £$%^&ing HATE these
switches now officialy.
Does anyone know of a replacement switch with compatible dimensions and
that doesn't fall apart when yo uso much as look at it from an angle?

I did prepare a 5cent coin as a driver tray like in those pictures, but due to lack of
thermal adhesive, I packed it with foam on top and wrapped it in electric tape, so
the compressed foam will hold the circuit against the coin.

It now will hold out almost twice as long (1 minute now, opposed to 20-30s before),
but probably due to my shoddy construction it is nowhere near 100% duty cycle
for the driver.

After running it until it blinks about 3-4 times in a row, letting the driver cool off
inbetween (it cools off rather quickly), the HEAD however has become uncomfortably
hot anyways.

So my final verdict is this:
-heatsink microflex some more
-find quality tail switch
-live with short dutycycles for the diode.....

Re: Kryton groove duty cycle @1A

My Microboost with my 12x set to 450mA in my Cree c3 running off 1X 14500 Was working fine no problems.

Then One night it started to do a slow blink after a 30sec on time. It did this anywhere between 30sec and 1 minute. Hmm strange. Checked the voltage of the battery was still at ~4.18v after the 3rd time it did that.

So switched to a None protected 14500 cell and never had an issue again. However that UN protected 14500 cells was Farly New.

The First battery that i had in that started doing that slow blinking was about 1 month old and was a protected battery.

I do not know if it was due to the protection cuicuit or the battery being old.

Re: Kryton groove duty cycle @1A

OK so I did the coin thing like in DTRs pictures, and it's holdidng up much longer now, BUT:
The *^£$% DX clicky switch consistently craps out after a minute or two of use.

This is the offending POS switch:
http://www.dealextreme.com/p/clicky-switch-for-flashlights-16mm-5-pack-5632

Replace the switch, another three minutes of fun.
I opened one and found a tiny spring that carries the current AND is responsible for the
electrical contact, pressing a metal disc against some contacts....
My guess is that the ~2 amps that can flow, when the battery is getting low, heat up
the spring so much that it softens and deforms permanently, rendering it incapable of pressing that thing onto the contacts anymore.:banghead:

It shows this was designed at a time when 2 amp diodes were unheard of.
This switch is a serious flaw and renders the otherwise beautiful and nice laser worthless.


Does anyone know of a replacement switch of the same dimensions that is rated for ~2A??

Please?

Thanks.