• You must be a Supporting Member to participate in the Candle Power Forums Marketplace.

    You can become a Supporting Member.

GDuP progress thread.

dat2zip

Flashlight Enthusiast
Joined
Jan 5, 2002
Messages
3,420
Location
Bay Area
Finally got some time on the board today. Put the micro on two boards and one came out crooked and I scrapped it. The other one I started loading more components when I double checked and noticed I had the pin 1 orientation incorrect so out with that one too! I reflowed on a third board and after adding one wire to power up the micro I was able to put it in the programming fixture and I was able to program the microprocessor. So, it's a green flag so far. Those are a couple big hurdles in putting this one proto together. I think the rest will go together without issues and it is now a matter of time to before it gets fired up, calibrated and the installed in another XR Can.

Maybe, I'll stick this on in one of the raw Aleph Mule heads Don just sent me. If this one fires up the next step will be to have the CM hand load a dozen or so and at the same time order production boards.

Revision 2 has the initial flash turn on correction which reduces the amount of initial flash and increases the control range. The one model will cover 1mA to 1A.

More details when the board is turned on.

The 10CC syringes showed up today along with a Digikey box and the patio has a nice roomy skylight.

Wayne
 

dat2zip

Flashlight Enthusiast
Joined
Jan 5, 2002
Messages
3,420
Location
Bay Area
I have completed assembly and functionality of the new board and the initial flash problem is still not better. The fixes or changes I have made have not addressed this to my satisfaction.

I feel this is a very strong issue and spent today looking into it and have determined the root cause of the problem After working on a solution that has minimal impact on the number of components needed for a solution and simulating the effect I mocked up the original prototype and modified the software for the changes. There is now no initial flash turn on at any level. This will be significant on level 1 or the lowest level as any flash tends to wipe out any night vision.

The question is how to fit on the required parts on the board. If I can manage the changes on the layout I will proceed with a new proto and or move this into production if the changes are minimal. At this point I doubt it will be simple.

I think I will attempt surgery on the rev2 board to add the changes and see if I can re-program the board to implement the changes on one rev2 board.

Wayne
 

PhotonFanatic

Flashlight Enthusiast
Joined
Dec 17, 2003
Messages
2,338
Location
western Massachusetts
Wayne,

Can you recap the physical specs that you intend to achieve for this two-board converter?

Are both boards going to be .55" in diameter? And is the total stack height, i.e., from the outside bottom board to the top of the highest component on the second board, going to be 5mm or less?

If both boards will be .55" in diameter, what would the height be from the bottom of the first board to the top of the second board?

Thanks.
 

dat2zip

Flashlight Enthusiast
Joined
Jan 5, 2002
Messages
3,420
Location
Bay Area
The GDuP will be 0.55" diameter.

I need to re-compute the thickness. It's not what I thought. Let me re-compute it and post when I get the correct info.

I just completed two of the revision 3 boards and programmed them. I seem to have tossed the revision 1 GDuP boards and the rev1 anode boards and have no anode boards to test these boards out. I will need to re-order the anode boards before I can make another LE and that should be mid next week when I can start testing the latest revision in a light.

Wayne
 
Last edited:

dat2zip

Flashlight Enthusiast
Joined
Jan 5, 2002
Messages
3,420
Location
Bay Area
I computed the GDuP to be 5.2mm for the current 1.6mm PCB I just measured is 5.25mm.

The total stackup of the anode board and the GDuP in the final version will be 5.2mm +/- a 1/10mm or so.

The board height could be reduced another 1mm by using a different inductor and assuming you can live with 0.5A max output.

I didn't post earlier, but, I was able to mock up the changes to fix the flash problem and made a revision 2 board that is functional and again since I don't have any anode boards it sits with the two revision 3 boards awaiting anode boards for final assembly into a LE.

Wayne
 
Last edited:

kenster

Flashlight Enthusiast
Joined
Nov 6, 2005
Messages
2,095
Location
San Antonio, Texas
The board height could be reduced another 1mm by using a different inductor and assuming you can live with 0.5A max output.

Wayne

I don`t know about Fred but that sounds great to me. These new LED`s are so dang bright 500ma for high is all you need in a small pocket type light. :thumbsup: Just my worthless .02 of course. :eek:

Ken
 

dat2zip

Flashlight Enthusiast
Joined
Jan 5, 2002
Messages
3,420
Location
Bay Area
Since my plan is the make the GDuP capable of driving 1A to the LED I will be using the larger inductor and since most of my LEs will be with the XRCAN the deeper depth of the XRCAN means the current 1.6mm PCBs on both the GDuP and the anode board fit in the XRCAN.

As for dropping the inductor height by the 1mm I doubt that will happen unless I can find a source for the inductor where I don't have to buy a reel of 3,000 pieces.

Wayne
 

kenster

Flashlight Enthusiast
Joined
Nov 6, 2005
Messages
2,095
Location
San Antonio, Texas
OK, I am deffinately in for some GDuP`s but I am not buying 3,000 of them so I like the inductor you have already picked out. :tinfoil: Really looking forward to this drivers completion! :thumbsup: Now we wait for Fred. :popcorn:

Ken
 

jch79

**Do Not Feed The Vegan**,
Joined
May 2, 2006
Messages
3,661
Location
On the asphalt.
OK, I am deffinately in for some GDuP`s but I am not buying 3,000 of them so I like the inductor you have already picked out. :tinfoil: Really looking forward to this drivers completion! :thumbsup: Now we wait for Fred. :popcorn:

Ken

:thinking: Let me think here... Ok, I'll buy 2, so you'll only have to buy 2,998! :nana: :poke: :wave:

Time for some
coffeetc7.gif
so I stop :sleepy: when I'm supposed to be
typinghx1.gif
.. :shrug:

Wayne, should I place an order with the Shoppe this week, or save on shipping and wait a little while longer? :whistle: :sssh:

Jeez, all these smileys :shakehead ... I think "The Eye"
cyclopsqt1.gif
is getting to me! :duh2: :devil: :duck:

:sleepy: john :broke:
 

TexLite

Enlightened
Joined
Jun 30, 2007
Messages
595
Location
Texas
Wayne,
A few questions on the GDuP,
Will it run off of two cells,Nimh or alkaline?
What about three cells,nimh or alkaline?
Will the levels be accessible on the fly or will the light have to be powered down to change levels?
Thanks
 
Last edited:

dat2zip

Flashlight Enthusiast
Joined
Jan 5, 2002
Messages
3,420
Location
Bay Area
Wayne,
A few questions on the GDuP,
Will it run off of two cells,Nimh or alkaline?
What about three cells,nimh or alkaline?
Will the levels be accessible on the fly or will the light have to be powered down to change levels?
Thanks

The GDuP is based on the GD driver and is a true buck/boost converter. It will run from 1.8V to 5V safely. So, yes to two cell Alkaline, NiMH or even three cell Alkaline/NiMH.

The current UI is an off state timing. Quickly blipping the light on/off is the signal to advance to the next level. Next time you turn it on it will be at the next power level. If you blip it again it will advance. So, blip, blip, blip will advance three times. If you click and leave it on more than 1 second it will stay at the current level and not advance.

I hope I explained this correctly, but, I believe there are other UI that behave similar to this.

Since it times from when you turn the light on it will cycle if you have the light on and then you blip it off/on/off and then on again. That will cycle it to the next level.

The only catch is if you keep blipping it and never leave the light on for more than 1 second it will accumulate the internal counter and when it reaches 15 it will go into calibration mode. If this happens and you don't want it to calibrate you can just abort it as it ramps up and down the LED 5 times. Turning off the light during this period where it is ramping up and down will abort the calibration.

Wayne
 

bluecrow76

Enlightened
Joined
Jul 10, 2006
Messages
638
Location
Tiger Town
Sounds great! I have an L1P with CR123 body and a Seoul P4 just waiting for this driver. It's going to be for my wife. Looks like this will be my first sandwich mod. Hopefully it will all fit in the can.
 

dat2zip

Flashlight Enthusiast
Joined
Jan 5, 2002
Messages
3,420
Location
Bay Area
Fridays update!

I had a major setback. On Wed I setup the bench to resume work on the GDuP. I had some code changes I needed to test and I tried programming the first board. It died. 2nd board. It died. Third board. It died. Oooops. No more boards to try! Something is wrong here.

I had taken the rest of the boards and kitted them up to be hand soldered by the local assembly house and I was short on the GD IC and had no spares. That bummed me out and I vegitated the rest of Wed.

On Thurs I removed an IC from a GD board and surgically repaired one of the GDuP boards. I checked all my calculations and reviewed the datasheet(s). I then scoped the board and found I had a stability problem which I quickly corrected before frying another IC.

With that I spent all Thurs fighting other setup and configurations, but, by the late afternoon I had a stable setup and could debug or re-program the GDuP board. I got all the bugs ironed out and now the board calibrates with the new high current range. So, now I have low around 10mA, medium around 180mA and high at 750mA.

I also fixed the delay on startup to eliminate the flash and moved it to not encroach on the UI timing. There is a delay to turn on which is around 60mS and some of you will notice this delay. If the light was turned on any sooner it would come on with some very bright value and then settle in to the final value some time later.

There is one remaining task to code in. It appears I have an offset in the LED current that is read by the uP. I need to zero this offset in SW to get the servoed LED current to the proper value. The low should be 3mA, but, it's calibrating to 13mA or so. No biggie for now. I still love it.

Ended up going to bed last night around 1am and I turned on the light on low and left it running. I will check it again when I go to bed again tonight, but, I expect it to still be running.

I UPS redline more ICs and they just arrived. I'll fix and re-program the other two dead boards today or this weekend.

Since the LED control range is now ~1mA to ~1A+ I'm sure there will be different needs and different drive level requirements. I'd like to keep the derivatives to a minimum, but, we can start the discussion here or in another thread.

Depending on what light this goes into some may want low lower. Some may want a mizer version while others will want the turbo version.

I will have some LEs officially available potentially near the end of next week and I'm looking for some Beta testers who would like to put the LE through their testing using primary and rechargeable. All feedback good or bad will be used to iron out any issues that may come up.

After the Beta period I should be getting stock in 3-4 weeks and will start making them available first in whole lights (Aleph Mule, A19) and kits (heads , clickie) for those that want to upgrade from a standard LE with two stage twistie to a GDuP.

There will be a lockout period during the phase. No GDuP board will be available to modders until after the lockout phase.

I have not thought about the price on the board or anything else so if you ask you'll need plenty of :popcorn:before you might get an answer because I'm still :banghead:.

Lastly, I re-ordered the anode boards and I just got shipment notification today and they should be here Monday. Once the anode boards arrive I'll be able to make more complete assemblies.

O, the LE I put together on Thursday I found one remaining anode board lying on the bench.

I won't let the cat out of the bag... In fact I'll leave it at that and you can speculate all you want.:nana: I'll just say it has nothing to do with Hammerhead.:devil:

Wayne
 

THE_dAY

Flashlight Enthusiast
Joined
Nov 28, 2003
Messages
1,820
Location
sfv, california
3mA, 80mA, and 400mA sound like nice levels, will you use these levels too?

will this board fit inside a dspeck firefly?
 

dat2zip

Flashlight Enthusiast
Joined
Jan 5, 2002
Messages
3,420
Location
Bay Area
3mA, 80mA, and 400mA sound like nice levels, will you use these levels too?

will this board fit inside a dspeck firefly?

Will this fit in a FF??? I know little about the FF3. The FF2 is sandwich compatible and I still have FF2 kits that are not selling. I'm thinking of making a run of FF2 with the GDuP if I can get around to it. Come to think of it I have some private stock of **CENSORED** that no one knows about that I might build out as well. The **CENSORED** is of course top secret. I will have to try making a Sandwich with the GDuP to verify it can actually be done. If it successful then of course anywhere a Sandwich fits the GDuP sandwich would work as well (with the correct battery combination).

If I can get rid of the offset and get closer to the expected cal values then the proposed table is for sure one that is a high candidate.

Wayne
 

mosport

Enlightened
Joined
Sep 21, 2005
Messages
395
Location
Toronto, Canada
Thank you for the development update and efforts Wayne,

3, 80, 400 mA drive settings sound like a great balance of brightness and runtime for small CR2 or 123 pocket lights, while providing good visual spacing. With the latest XRE and P4 flux bins my preferences lean towards runtime rather than all out brightness, I hope that both mizer and turbo versions will be offered if it is practical to do so.

Wondering if you anticipate the GDuP lockout period expiring before Christmas?

I've got a mod project that's been on hold since Spring, patiently waiting for the GDuP to be released...

:wave: :D


Derek
 
Last edited:
Top