Unpotting a new version ArcAAA (pics)

chimo

Flashlight Enthusiast
Joined
Sep 16, 2004
Messages
1,905
Location
Ottawa, Canada
I have seen lots of the old ArcAAA boards unpotted (I did my share) but had not seen what's inside a new one. I was curious about what differences, if any, were in the new ArcAAA-P (CS version) as compared to the much older ones.

I started with uncrimping, pushing out the pill and unpotting it. I used a heat gun and a sharpened dental pick.

I found that everything was mostly the same as earlier versions.

What's the same:
Same IC/schottky/inductor. The output capacitor has been relocated back to the legs of the LED from the main board like the early versions (perhaps for reliability reasons).

What's different:
There appears to be some added vias around the edge of the board to provide additional connections between the top and bottom grounding rings and the LED opening inside the head has an added bevel.

My next decision will be if I will put a GS or a DS LED back in and which driver board I will use.

Thanks for looking.

Pill pushed out


Pill out


LED O-Ring Exposed


LED O-Ring Removed


Tool Used


Output Cap exposed


Driver Un-potted (still functional)


Driver alongside older version and some potential replacements (Arcmania's 30/50mA)
 
Last edited:
Re: New version ArcAAA unpotting (pics)

Very interesting indeed chimo!
but i`m surprised at your choice of replacement led
i`d have thought you would have gone for a brighter alternative
btw the macro shots are superb :thumbsup:
 
Re: New version ArcAAA unpotting (pics)

Very interesting indeed chimo!
but i`m surprised at your choice of replacement led
i`d have thought you would have gone for a brighter alternative
btw the macro shots are superb :thumbsup:


Thank you. I want to keep this one driven off a primary AAA so I am going to stay with a low power LED. The driver will either stay stock or the 50mA arcmania driver. There seems to be a large overshoot with the 30mA one. I want to profile the 50mA driver before I install it in anything.
 
Looks good. About how long did it take you to carefully scrape away all that potting?

Curious to see how your driver tests turn out.
 
Looks good. About how long did it take you to carefully scrape away all that potting?

Curious to see how your driver tests turn out.

It probably took about 10-15minutes. It took much longer the first time I did one of these.

Paul
 
It probably took about 10-15minutes. It took much longer the first time I did one of these.

Paul
Is a heat gun better than boiling water as posted in another thread? My gun gets so hot, I'm concerned about ruining something. I don't want to destroy the board, I just want to replace a couple of LEDs.

Why do you want to replace the driver?
 
Hey chimo,

Good job. I guess this solves your question if it uses the 200mA regulator we discussed 3 yrs ago.

Do you think swapping the yellow 47uF tantalum output cap with a low ESR ceramic cap will affect the oscillation of the current regulator?

Some of my unpotted units don't have the output cap.
 
Last edited:
Is a heat gun better than boiling water as posted in another thread? My gun gets so hot, I'm concerned about ruining something. I don't want to destroy the board, I just want to replace a couple of LEDs.

Why do you want to replace the driver?

Good questions.

It's important to NOT keep the heat gun on the pill constantly. If you do, it will literally cook the board. Pick the epoxy gently while the heat is on - once it starts coming off easily move the heat off the target. As the epoxy starts to harden move the heat back on.

For the first few I did, I used the boil method. It takes much longer. The heatgun method is much faster if you are careful.

I'm not sure if I will replace the board. If the others perform better I may swap it out.

Cheers,

Paul
 
Last edited:
Hey chimo,

Good job. I guess this solves your question if it uses the 200mA regulator we discussed 3 yrs ago.

Do you think swapping the yellow 47uF tantalum output cap with a low ESR ceramic cap will affect the oscillation of the current regulator?

Some of my unpotted units don't have the output cap.

I don't think it will have a great effect. It's really a 5V boost regulator that has some inherent current limiting.

Some of the boards (bottom left of last pic in first post) have a small ceramic cap soldered on the board instead of the legs of the LED.

Cheers,

Paul
 
Hey chimo,

Good job. I guess this solves your question if it uses the 200mA regulator we discussed 3 yrs ago.

Do you think swapping the yellow 47uF tantalum output cap with a low ESR ceramic cap will affect the oscillation of the current regulator?

Some of my unpotted units don't have the output cap.

I must have been sleepy when I answered last. Yes, it does answer that question. :)

BTW, a while back I tried the higher powered chip and but I don't recall noticing much improvement, if any. I didn't do a great deal of measuring.

Did your unpotted boards have the LED installed? Were they the version with the small ceramic cap on the driver board? Cheers,

Paul
 
Hey Paul,

I have both the tiny ceramic cap and the tantalum cap versions. All of them are from pre-ArcLS Fraen era. Sorry I don't quite remember the time I got them. What interest you?

Did your unpotted boards have the LED installed? Were they the version with the small ceramic cap on the driver board? Cheers,

Paul
 
Hey Paul,

I have both the tiny ceramic cap and the tantalum cap versions. All of them are from pre-ArcLS Fraen era. Sorry I don't quite remember the time I got them. What interest you?


I was just commenting on your comment that some of your unpotted boards didn't have an output cap. Did they come off in the unpotting process or were they not there from the start. Cheers,

Paul
 
They are unpotted in the first place. Meaning no epoxy on them at all. So I believe there were not fitted with the ceramic caps. Some even weird they don't have any caps at all.
 
They are unpotted in the first place. Meaning no epoxy on them at all. So I believe there were not fitted with the ceramic caps. Some even weird they don't have any caps at all.


Interesting.... Was the LED installed?

The spec sheet recommends a 47uF to 68uF low-ESR Tantalum cap to reduce ripple voltage, however, that's not too big a concern for this app.
 
Hello again,

If I am not wrong, the leds are BS1 and BS2.
They are complete circuits but not installed in to the heads.
I remember I bought a bunch of them off a cpf member.
 
Thanks for detailing the swap Paul,

Last week I upgraded an older un-tapered head ARC AAA to GS spec and believe the original was BS spec. Just like Koala, there was no output cap soldered onto the legs (or anywhere else) and none of the other un-tapered HAIII heads I've modded have had this cap. They were boiled in ziplock baggies and I was sure to go slowly to not rip off any components during unpotting. Also fashioned a homemade tool to uncrimp the rim by taking an old kitchen paring knife and rounding off the tip. Very efficient at wedging + prying and made quick work of the task without deforming the shape.

On the topic of epoxy, can you recommend any commonly available substitutes for re-potting? Are RTV, JB Weld, 5-minute epoxy etc acceptable for the task?
 
Thanks for detailing the swap Paul,

Last week I upgraded an older un-tapered head ARC AAA to GS spec and believe the original was BS spec. Just like Koala, there was no output cap soldered onto the legs (or anywhere else) and none of the other un-tapered HAIII heads I've modded have had this cap. They were boiled in ziplock baggies and I was sure to go slowly to not rip off any components during unpotting. Also fashioned a homemade tool to uncrimp the rim by taking an old kitchen paring knife and rounding off the tip. Very efficient at wedging + prying and made quick work of the task without deforming the shape.

On the topic of epoxy, can you recommend any commonly available substitutes for re-potting? Are RTV, JB Weld, 5-minute epoxy etc acceptable for the task?

I haven't repotted any of mine - makes installing the next latest and greatest LED easier. :)

If you are going for waterproofing, I would use a silicone product like RTV.

JB Weld is not very "fun" to unpot and 5 min epoxy is a little thick.

The encapsulating epoxy used is injected into the little holes in the back. The o-ring on the LED keeps the runny epoxy from flowing out past the LED.

I would add an output cap when putting in the new LED.

Cheers,

Paul
 
Well, last night I made my choice. I used the 50mA(40mA) ArcMania driver.

That driver requires you to solder the LED to the surface of the board (i.e. no through-holes). The LED anode contact point is quite small so I decided to use a slihgtly different method:

I cut a couple of short pieces of 28 guage stranded wire and soldered them to the board contact points. I shortened the legs of the GS LED and curled them up. I then soldered the wires to the LED. This gives some shock relief to the assembly.

Before inserting the LED, I reamed out the LED opening to allow the GS LED to fit but not expend past the front of the bezel. I applied some silicone to ensure the LED stayed put.

I also soldered a small copper strip to the ground vias near the crimp to help make a more robust ground connection.

Why did I choose the ArcMania board? I decided to use LSD cells and this driver will ensure they are not over-discharged. I just did a run test and got 4hrs 15min to "blinking" with some older Rayovac 800mA Hybrids. (I charged these with a Duracell 15min charger and pulled the off when the fan stopped - there is a slower 20-30min charge that the charger continues to give after the fan stops that these cells did not get so they may not have been at quite at full charge to start the test). I will likely run another test with slow charged NiMHs as well as alkalines at some point.
 
Last edited:
Paul-
I have an unrelated (slightly related) question that I'd like your opinion on.

There has been quite a bit of controversy over new ARC AAA GS because of the slightly exposed tip of the LED that protrudes beyond the bezel of the head.

I'm of the belief that it is purely a cosmetic issue.

Accusations about potential damage to the LED because of the slight exposure I believe are completely bogus. There have been rantings about how flawed the design is, how short sited and uncaring the company is to deliver such a flawed product.

After removing a pill by hammering the LED back through the opening (with no damage to the LED) and seeing how it is potted, it appears impossible to damage it to me unless you were attempting to damage it, and that would take considerable effort.

On the other hand, personally, I prefer an ARC AAA newest generation LED, to have the same cosmetic appearance as other models, and I, too, would rather not have an even slightly exposed LED.

I'm guessing that when the LED became available, it was put into production on a limited basis for those of us who just had to have the latest and greatest ASAP. There's a big difference in choosing to like or not like or even keep a product because you don't like it's appearance from making wild, completely false accusations that have no foundation.

My question to you, is do you agree that there was no compromise in the integrity (functionality) of the GS model that was released and why or why not?
 
Top