This is what I was suggesting earlier, I guess I didn’t word it very well. lol
Instead of a round plate, a 1cm wide iron strip with a hole in the middle can be used to do the sandwich trick.
Despite the good performance of proto2, I ’m for going back to 7mm for the shelf (if it was indeed 7mm). There is plenty driver cavity already so I see no reason to have it reduced, it is not even a cost thing.
I hate to say it, but yeah, you didn’t. 
I even went back and read it now and it’s still not clear. For one, there are two sets of two threaded holes, and they are aligned. The “outside” holes are in the (extended inwardly) meat of the tube, and hold the driver on. The “inside” pair only go through the shelf.
That was my original design that we started with. Thing is that in 8 months time we have not been able to communicate this sandwich idea to Thorfire, and via several unclear designs from their side we now in this proto2 we have something that will work well enough.
If i can choose between two designs that will work, 1) our own design that will probably be best but has not been tested for practicality and that Thorfire does not understand, and 2) a design that has been proven to work well in proto2, that Thorfire has thought out themselves and that they are prepared to build, I know what to choose…
Found it! Post #3390, Shelf in proto #1 was 7.05 mm measured, this one is 4.92 mm, but I'll have to unsolder things and pull out the MCPCB and see what's goin on. This eve or weekend.
Ohh - on another good note, I was able to stretch out the lens o-ring and fit it easily around the reflector. Dunno what good it does outside the reflector, but that's how it fits. Have to check when assembled, if the lens is up againt the o-ring or reflector. I'm using the good, identical o-ring from proto #1 now. The cut o-ring will go in proto #1. Also trying a black thicker o-ring I have that's a close fit, just experimenting on proto #1 for now.
Dang, wanted post this earlier.
Proto #3, after tightening up the reflector lock down screw, looks like the LED's are better centered. When tightening it up, heard a pop sound, thought I broke something, but think that was an alignment piece popping into the hole., because they look great now.
On SONY VTC6's fully charged (~4.20V rested), lumens: 6,560 @start, 6,060 @30 secs, throw: 63 kcd @5m
Not bad...
I have been to busy and only have a few mins on the net per day. I know that my spelling and gammer have gone out the window recently along with the intent of what I am trying to say apparently lol.
I also agree, I was curious what happened to the shelf thickness, I like the thicker shelf.
Looking at the comparison pic of Tom
Hard to tell if they shortened the head a little (would explain thinner shelf.)
Well apparently TA is not the only one having troubles conveying a point but Djozz says it really nice.
You couldn’t communicate that via a drawing? Sounds like your guys need an engineer on the team to trade part drawings back and forth, the language barrier is reduced that way… 
You maybe haven't seen how tight things are packed in the buck driver. But there's still option 1.
Just asked for the CAD files used to make the V2 and V3 proto.
The head isn't shortened. Overall light length is 2 mm less - SS bezel is 1.25 mm shorter, remaining 0.75 mm? I'm guessing the battery tube screws into the head slightly bit more.
You can kind of see it comparing the aligned notches between the lights in the pic above.
The lack of a retaining ring could explain some of the shortness / deeper threading.
There was a clear drawing alright (not an official technical drawing but the message was very clear), but I think that there were reasons for them to dismiss the idea. And then the communication barrier was such that we never heard back why they thought it would not work. That back and forth idea sounds very attractive and constructive, but when back takes two months and forth never comes and then you get a surprise prototype, things go less constructive :weary:
10-4, gotcha. Language barrier can be frustrating, but unresponsive is worse. 
The amount of drawings we send is hard to count, it has been the way we made almost all clear.
Machinists, could the machine for threading call for a thinner shelf? (Though there were threads for ring in V1, could that finer threading been done with another machine/head/part?)
Not all goes wrong, I think we have a high current quad copper DTP board because of our clear drawings of how it could look like.
Yeah Djozz, and so much other things are there now, remember them keeping sending CADs with no shelf or the flat parts we wanted under the switch and tripod mount placed onder the row of small cut outs where the bottom small cutouts kept being placed under the switch and tripod mount … Sigh but in the end it got fixed
Has any of this changed? What has ThorFire said since about any f this?
Please add one more light (763)
Thanks