Fault with HDS 17670(2x123) tubes with non 2x123 config..

When used with 17670 cells with the stock spring or 18650 using a modified spring a fault can be found. Have your light in a fully installed and ready to use state. Turn on your HDS then holding your light by the head section and the tail(battery) section facing up (between your index finger and thumb) swing the light. You will notice your HDS turn off and then flash back up to the last lit up state(indicating an error) you will loose button operation as well.

I have been unable to reproduce this using 2x123s but using any single cell it has been reproduced on all 3 of my tubes and all 4 of my HDS lights.

Henry if you see this it's something to consider in your design of the Ra Twisty's battery tube.

Seems like the forces acting on the battery is too much for the spring.

Is this with an unprotected 17670 or protected?

leukos said:

Is this with an unprotected 17670 or protected?

Click to expand...

Unprotected. But the difference between the cells can't be more than a mm. I shimmed up the cell thinking ok the difference between a 17670 and 2x123s can't be more that much more and it still happens.

Also I tested it with an 18650, actually I discovered the problem with the 18650 and using my modified spring. I thought it was my spring's design however when my 17670 came for more testing it was the same conclusion using both springs(stock and my design).

Ok further testing I shimmed up an 18650 to the same height as 2x123s.

At first I thought it fixed it however when I swung harder I was able to duplicate the same problem. Going back to 2x123s and swinging with all my might and in different ways the light remained 100% reliable.

EDIT: Ok I was able to find a fix for the tube while using a 17670 and stock spring. I had to use the Novatac tail cap while shimming the battery up to the same height as 2x123s. I used a screw from a computer to shim it up.

I had the same problem and did this same mod using this tube in conjunction with a Novatac head. I thought the issue was not using the intended head. I used a brass #6 screw and nut attached to the end of the tailcap (HDS tailcap) spring and filed it down smooth. There is a measurable difference between the HDS and Novatac tailcaps.

Tebore,

The Ra Twisty design is totally different.

First, there is no exterior spring to snag and lift off the contacts in the Ra Twisty. To the best of my knowledge, the battery tubes recently offered for sale were known rejects - which is why they had not been sold two years ago when they were made. We had problems getting the long exterior springs manufactured properly and eventually tabled the effort for lack of resources. A properly fitted spring will slide in the tube when gently pressed but will not fall out under its own weight. Further, if you push either end flush with the tube, the spring will always expand out the other end by roughly 1/8 - 1/4 inch. Finally, the battery will not bind on the spring and easily slide from end-to-end.

You may be able to get two of the short springs to sit one on top of the other - especially if they were soldered together very carefully. The ends are ground flat on the production 1x123 springs. Solder the flats together being very careful that the wires align side-to-side and that the loops are not soldered shut.

No spring contact is the reason for the loss of button operation.

Second, the Ra Twisty battery positive and negative contacts are very different and allow electrical contact to be maintained even if the battery shifts in the battery compartment. On the EDC Ultimate/Basic flashlights, the spring only generated about 3.5 pounds of pressure and the power-fail ride-through was long enough to cover most intermittent conditions but not long conditions. A "long" power outage lasting a small fraction of a second results in a reset when the flashlight is turned on.

Henry.

Button operation fails 50 percent of the time in the failures. I have a fix. 1 add buttons to non-button top cells. Then shim up the battery. I suggest using the NT tail cap in non-designed for cell configs.

Amazing how precise these lights are.

Looking at the ra the the two spring compartment was a great idea. A retrofit of the old lights might be possible to boost reliability.