Sold/Expired Feeler - Programmable Hotwire Regulated Driver drop-in for D-M@g

[Alan B]

First Article Hotwire Regulator (SN001) shown above, prototype below

​

​

For the past year I've been working on a Programmable Hotwire Regulator design. (A Hotwire is a high power incandescent). Many folks on the forums here have contributed including JimmyM, wquiles and others. The goal is to have several regulated hotwire driver choices available in the marketplace. These drivers provide protection for the bulb by soft starting and regulating the voltage (to prevent instaflashing - blowing the bulb, and overloading any battery protection) and to protect the batteries via low voltage and high temperature shutdown. They also allow the battery voltage to be greater than the bulb voltage and facilitate selecting the optimal voltage for the bulb. Bulbs last a lot longer and can be finely tuned for maximum intensity, and the light level will be constant until the battery voltage sags below the regulation level in which case the regulator becomes direct drive and still provides soft start protection for the bulb and battery. Additionally variable intensity is available in the specially developed User Interface to provide reduced output and lengthen battery as well as bulb life.

For those unfamiliar with M@g modified hotwires, you basically put in a higher voltage battery pack, change the bulb, socket, reflector and lens to build a high powered incandescent flashlight. Add this regulator and the battery voltage can be up to twice the bulb voltage for long runtimes and regulated output. A common example of a Hotwire is the so called Mag85, which employs a Welch Allyn 1185 30 watt medical bulb. It produces a very nice focusable white light, and can be overdriven to about 1200 hot tungsten bulb lumens.

This thread is to help me decide if I should build a few of my particular variant of the design for others. Originally I did not intend to build anything, but I have so much time invested and I want a few for myself, so I am considering it. wquiles and JimmyM are planning to build their variants. wquiles' is going to be in a battery carrier for the SF-M6, and JimmyM's is going to work with the KIU base in the D M@g. I will continue to support their efforts as well as developing my own.

My implementation is a D Maglite switch replacement (drop-in) that allows easily reprogramming the CPU without removing it from the M@g. New software may be uploaded or configurations changed easily with a cable to a computer. So each of the three projects are somewhat different and complementary.

At this point I have a working electrical and software prototype in a prototype aluminum "Sled" (shown above). The mechanical design is machined from a solid block of aluminum to fit the M@g and carry the electronics and the bipin bulb socket. The pushbutton switch plugs in through the M@g switch hole.

Standoffs on one end of the "MagSled(tm)" support the bipin bulb socket. The other end has the battery plus contact. The production regulator is designed to work with the existing stop-ring inside the M@g, sitting against it for support the same as the stock switch. A setscrew similar to the stock switch anchors the drop-in and makes the negative connection to the M@g body tube. The bulb socket standoffs can be changed for different bulb heights.

An alternate programming adapter plugs in replacing the switch board. The programming board sticks out of the flashlight and contains a programming socket extension and several pushbuttons for testing, etc.

​

Programming Adapter Installed
​

The circuit boards use no surface mount parts, so they can be readily repaired or modded if necessary. CORRECTION: The early boards were all through hole, but they have been redesigned to be a mix of surface mount parts and through-hole parts. The Regulator, Fuse, and FET are through-hole, as are the connectors. The Microprocessor is surface mount as are the resistors and capacitors. The parts are not super-small so service is still reasonable. The schematic is published. All electronic parts are readily available. There is a fuse onboard to protect the batteries in case of a short. This is especially important with unprotected high current cells.

The default user interface software has been optimized for Hotwires and includes variable intensity as well as voltage regulation, soft start, over temperature protection and low voltage cell protection, and a number of options. Changing these options requires reloading the software with a PC and programming adapter. Future versions of the software may have more features and flexibility.

The configuration of the basic regulator would be specified when the device is ordered. Several parameters can be specified - High and Low Bulb Voltage, Battery Cutoff Voltage and more. Standard configurations will be available. The regulator would be constructed, programmed and calibrated. It would be "drop-in", though some adjustments for bulb position within the reflector may be required (changing standoff height). The basic configuration would not include the programming/calibration board.

The optional programming kit would enable the user to reprogram the regulator with a computer. This requires some additional equipment and free software. One programming kit can handle any number of regulators. The programming kit can also be used to upload improved versions of software, change parameters and do software experiments and learn to program microprocessors. The existing software is written in Gnu C, so it is not necessary to work in assembly language, though it is supported as well. The software tools are freely available, though there is some complexity to develop and debug programs in a limited CPU environment such as this. Programming requires two parts - a programming dongle, available from many sources for about $35, and a programming/calibration board that reaches into the flashlight and extends the programming plug. Programming without the programming/calibration board can also be accomplished but it is more difficult to reach into the flashlight. If you wish to change the voltage settings a programming setup is required in the present software. Future software may allow changing some parameters without programming, but this is not yet planned. Customized versions of software by users or third parties are also possible.

Laptop Flashlight Programming in Action
​

The cost is yet undetermined. I expect to keep it under $100. Some part of this would go to benefit the CP Forums as it has been the vehicle that enabled this collaboration to occur. This detail has not been worked out yet.

This is a feeler thread, not binding in either direction. I may get distracted with other projects and have to put this on hold, and you may change your mind later. I will however use this feeler as the order in which to offer the limited quantities of hardware that do become available, so there is some motivation for posting to this thread and indicating your interest.

I need to know how many units folks are interested in, what battery voltage and type, bulb voltage, type and bulb current you would want, and other requirements you might have.


New Threads:

Discussion/FAQ/Documentation/Questions/Suggestions Thread:
https://www.candlepowerforums.com/threads/231656

Order/Sales/Interest Thread:
https://www.candlepowerforums.com/threads/231886


Links within this thread and other Historical threads:

Read this FAQ for further details:
https://www.candlepowerforums.com/posts/2805117&postcount=160

The Interest List is in post #5:
https://www.candlepowerforums.com/posts/2775605&postcount=5

The UI (User Interface) includes variable output, selectable levels, etc is in post #63:
https://www.candlepowerforums.com/posts/2785888&postcount=63

Voltage Variations listed in post #95:
https://www.candlepowerforums.com/posts/2788521&postcount=95

First Article Review:
https://www.candlepowerforums.com/threads/230519


Thanks for your support, comments and input,

-- Alan

Links:

Design Collaboration Part I:
https://www.candlepowerforums.com/threads/186291

Design Collaboration Part II:
https://www.candlepowerforums.com/threads/209098

Design Collaboration Part III:
https://www.candlepowerforums.com/threads/220475

wquiles SF-M6 variant:
https://www.candlepowerforums.com/threads/215806

JimmyM's D-M@g variant:
https://www.candlepowerforums.com/threads/216160


Summary of Plans

Design, prototype and test electronics and software (done)
Build first prototype mechanical to fit first prototype electronics (done)
First article PCB on order. Metal vendor selected. (done)
Build of first article electronics (done)
Final fit mechanical sled to electronics (done)
First Article out for Review (done)
SMT Redesign (done)
Preparing minor updates (done)
Order Production Metal (done 5/2009)
Production... (approx 7/2009)

Metalwork

Nanomiser is helping me with the mechanical final design and production. Here is a rendering from his CAD software:

​

5/2009 update:

Metalwork has been fitted and prototypes made under numerical control. I have a very small number of "Sleds". Nanomiser's adjustments to the design worked out very precisely. :thumbsup:

Production Machined Metal is ON ORDER.

 

[LED_astray]

Re: Feeler for D M@g Programmable Hotwire Driver

I've been following your project with fascination. Given more time & money, especially time, I'd have been in for one of the development boards already. I can't imagine I won't spring for at least one of your's or JimmyM's Mag D solutions, depending on price & features. (If I get laid-off tomorrow with a ridiculously large severance package I'm probably in for a development setup and a couple of each. I don't think that'll happen, but I can dream. So probably one, maybe more. Thanks!

(No slight intended wquiles, just no interest in the M6 right now.)

 

[Roland]

Re: Feeler for D M@g Programmable Hotwire Driver

I am looking for 1 or 2 regulated mag D switches.

From the description I am not sure how these voltage settings will work. Adjustable voltage or programming interface would be a must since bulb tolerances change and one might want to try it on different bulbs.

Philips 5761 needing 7,2 volt at 5,6 amps on 6 NIMH D-cells or 7 NIMH D-cells (yes 7D)

or a mag 623

 

[JimmyM]

Re: Feeler for D M@g Programmable Hotwire Driver

I'd love to see a couple of the pictures of the revised prototype.

 

[Alan B]

Re: Feeler for D M@g Programmable Hotwire Driver

I am superceding this old post to hold a list of the "Interest".

Please advise any changes required.

Thanks, Alan

Price: To be determined, goal is to stay under $100 for the pre-calibrated and configured ready to go Drop-In Regulator.

Bulb Interest List: (not complete)

wa 1111
wa 1164
wa 1185 11.0V

P5761 7.2

62138
64430
64458 21V
64610
64623 15V 16.5 m@g623
64633
64655 26.5


Regulator Interest List:

Group #1
LED Astray 1-2
Roland 1-2
donn_ 3
LuxLuthor 6-9

Group #2
cnjl3 1
Lips 2
BSBG 3-4
Starlight 3-4

Group #3
william lafferty 2
sylathnie 3
andygillis 1
ANDREAS FERRARI 2
Bullet Bob 1-2

Group #4
KiwiMark 1-2
rolltide 1
Drywolf 1-2
Iliger 1
Nanomiser 1-2
SR.GRINGO 3

Group #5
nailbender 2
SafetyBob 2
Cancerlad 1
Sober 1
Benson 1
AMD64Blondie 1
paulvw 1-2

Group #6
Anglepoise 1
QtrHorse 1-2
Brett80 1
ss2nv 2-3
Oli.Hall 1
Data 2

Group #7
DeLighted 1
full-beam 1
Joe_torch 1-2
syzygy 1
digitalpete 1 (email)
gswitter 1-2
DocD 1
petfulton 1-2

Group #8
mrQQ 2
LIGHTSMAD 1
KSC001 1
PhantomPhoton 1
moraino 1-2 (pm)
thefire 2 (pm)

Group #9
Raol Duke 1-2
niner 1-2

logged: 72-91 assembled regulators (as well as other interests) updated through post #444...


Discussion on kits started in post 402. I have some through-hole boards and parts that I may offer as kits. These would be advanced kits in that the instructions would be minimal. The interest is recorded here.

Kit Interest List Summary:

milkyspit: some prototypes for non M@g apps
niner: kit
Benson: kit, 1-2 non-mag prototypes
CancerLad: kit
^^Nova^^: prototype board


As mentioned in post #1, this interest list is non-binding in either direction. If/when product is produced this list will be used to offer it to potential purchasers. The released product may or may not meet all the purchaser's requirements, so make sure to check the specs when offered.


Original Post (edited) Follows:

The first post has been revised with additional information.

The new PCB design is 1.3 by 0.9 inches and uses the same parts as the existing prototype, so it looks quite similar, just compacted 15%. I need to check the revised layout carefully before sending out for more boards. I need to cut new aluminum for the sled. I will probably put the existing board in the new sled for more testing (and to actually have a working test flashlight), which will require the test mount to be longer than the stock switch, but there is plenty of room in the D M@g and the new design is easily adjustable for length by changing the rods and PCB mounting bar. If things go well perhaps I will have something built by this weekend.

 

[donn_]

Re: Feeler for D M@g Programmable Hotwire Driver

I expect I'll be in for at least a couple..

I'll want to run the usual lamps, on A123 power sources.

Any estimate on the resistance of the finished product?

Am I correct in my understanding..this can only be reconfigured with programming? IOW, there is no mechanical way to, say, change the bulb voltage to suit a different lamp.

 

[Alan B]

Re: Feeler for D M@g Programmable Hotwire Driver

I expect I'll be in for at least a couple..

I'll want to run the usual lamps, on A123 power sources.

Any estimate on the resistance of the finished product?

Am I correct in my understanding..this can only be reconfigured with programming? IOW, there is no mechanical way to, say, change the bulb voltage to suit a different lamp.

Thanks for your interest, donn_.

The FET in use is a high power TO220. It is around 5 milliohms. Not certain about all the remaining resistance, and the 15 amp fuse adds another 5. Traces on the board are heavy and both sides where possible and very short. The PCB is still in development, so effort to improve impedances is not complete. Perhaps wquiles could measure it in his breadboard to get an idea. I don't have a way to measure it accurately at the present time.

At this time the existing software has the voltages set prior to compile time. Different software could have features allowing other ways to set voltage values, but this creates a significant risk of blowing up bulbs - it is easier to set correctly and safely via programmed values on the computer than other ways. Measuring RMS from PWM is also problematic, so user adjustments are best done in the computer.

One way to select between a set of pre-chosen values is to use jumpers on the programming pins. This could be accomplished on the plug in pushbutton switch board, so several switch boards could be made that had solderable jumpers on them, and by plugging in the appropriate switch board a different set of values could be selected. If I recall correctly there are about ten jumper arrangements possible in this fashion. I will review that and consider it if there is enough interest. It would require a slight change to the switch pc board, but it is not too late for that now. How many selections do you need? (note - four configurations can be handled by swapping out the plug-in switch module.)

Lithium Iron Nano-Phosphate A123Systems cells are great. They do require lower voltage settings than other Lithium Ion. This is also easily accomplished in the program before loading into the chip.

The process of putting the software into the chip is basically editing a few values and pushing one button to compile the code, and another couple to load the chip. Depending on the programming hardware chosen the exact sequence varies.

To connect the programmer to the installed regulator, pry off the rubber switch cover, pull out the switch assembly and plug in the programming cable. After programming reverse the process.

Thanks,

-- Alan

 

[LuxLuthor]

Re: Feeler for D M@g Programmable Hotwire Driver

Alan, I see enormous value in each of your, Jimmy, and wquiles (M-6) projects.

I am 100% sure I would love at least 3-5 (likely more) from you and also from Jimmy. I have the programmer, and would use them in a variety of WA, Philips 5761, & Osram varieties. I'm comfortable doing the programming with dongle.

My deficiency without a foundation in electronics is knowing how the various resistors, diodes, capacitors, and other circuit components that you guys experiement with affect the overall circuit performance.

 

[Alan B]

Re: Feeler for D M@g Programmable Hotwire Driver

Thanks, LuxLuthor.

I just added the following to the first post to encourage folks to post in the thread:

I will however use this feeler as the order in which to offer the limited quantities of hardware that do become available, so there is some motivation for posting to this thread and indicating your interest.

-- Alan

 

[donn_]

Re: Feeler for D M@g Programmable Hotwire Driver

Alan,

I'd probably want 2 options on each switch, for fairly closely matched lamps.

For example, one could have both 64623 and 64633 options.

I figure I'd be using the same tube for each of those lamps, and just add/subtract cells/extensions as appropriate for the lamp used.

 

[Alan B]

Re: Feeler for D M@g Programmable Hotwire Driver

Alan,

I'd probably want 2 options on each switch, for fairly closely matched lamps.

For example, one could have both 64623 and 64633 options.

I figure I'd be using the same tube for each of those lamps, and just add/subtract cells/extensions as appropriate for the lamp used.

No reason to subtract tubes, run the max length all the time. The regulator will turn the extra power into runtime.

-- Alan

 

[cnjl3]

Re: Feeler for D M@g Programmable Hotwire Driver

The primary reason to switch tubes is that it is easier to carry a 2 cell host for a 1 or 2 cell light than lug around a 6 cell host that would power the same light. You dont 'always' need a sledge hammer to pound in a nail.

I'm interested in at least 'one' to start off with.

I like your idea of the use of 'jumpers' to select voltages but how does this option work with multiple battery sources? I would say that most CPF users have most sizes of MAG hosts (I know I do) 1D, 2D, 3D, 4D and possibly even higher MAG combinations with extensions plus some stampeding elephants and mammoths? So your 10 voltage selections sounds good to me with all the welch allen, osram and IRC bulbs around to try.

 

[Alan B]

Re: Feeler for D M@g Programmable Hotwire Driver

The primary reason to switch tubes is that it is easier to carry a 2 cell host for a 1 or 2 cell light than lug around a 6 cell host that would power the same light. You dont 'always' need a sledge hammer to pound in a nail.

I'm interested in at least 'one' to start off with.

I like your idea of the use of 'jumpers' to select voltages but how does this option work with multiple battery sources? I would say that most CPF users have most sizes of MAG hosts (I know I do) 1D, 2D, 3D, 4D and possibly even higher MAG combinations with extensions plus some stampeding elephants and mammoths? So your 10 voltage selections sounds good to me with all the welch allen, osram and IRC bulbs around to try.

Thanks for your input. I think I'll just build more flashlights to handle the different bulbs and batteries that I'm actually using. But the jumper setup can accomodate changes of sets of parameters.

The jumper (perhaps a switch if I can find one that will fit on the small pushbutton board) will select "parameter sets". These sets can include not just bulb voltage but other parameters as well, so if you wish to change battery cutoff voltage and bulb voltage they could both be changed in the different sets.

If we use a switch instead of jumpers the possible set would be reduced, so perhaps four sets of values would be practical. There are more possibilities with jumpers you can put an input pin to high, low or float, with switches that gets more dangerous since some combinations can damage the hardware, or it becomes more complicated to add protection circuitry to prevent it.

It sounds as though four sets of values may be enough, however, for most users. But perhaps not??:candle:

One thing to keep in mind here is that there are many options in the software, far more than we can ever handle with jumpers or switches. There are variations in the user interface, and there will be more options as new software is developed. To get the flexibility out of the CPU a good user interface is required, and pushbuttons or jumpers are not easy or efficient for this.

thanks,

-- Alan

 

[donn_]

Re: Feeler for D M@g Programmable Hotwire Driver

One thing to keep in mind here is that there are many options in the software, far more than we can ever handle with jumpers or switches. There are variations in the user interface, and there will be more options as new software is developed. To get the flexibility out of the CPU a good user interface is required, and pushbuttons or jumpers are not easy or efficient for this.

thanks,

-- Alan

ROFL! Typical tech weenie propaganda. I can't tell you how many times I've used a subset of that sermon trying to encourage users to study up on things like Visicalc, 123, Symphony, Quattro, Excel, Access, etc.. I have so much time in Excel I can make it emulate Word or Photoshop.

As I age, however, I find myself less interested in playing with software, and more interested in playing with bright lights and wooden boats.

Maybe some enterprising young hacker will offer a reliable and reasonable reconfiguration service for our switches. I'd rather pay to have it done than bother learning the software.

 

[Alan B]

Re: Feeler for D M@g Programmable Hotwire Driver

ROFL! Typical tech weenie propaganda. I can't tell you how many times I've used a subset of that sermon trying to encourage users to study up on things like Visicalc, 123, Symphony, Quattro, Excel, Access, etc.. I have so much time in Excel I can make it emulate Word or Photoshop.

As I age, however, I find myself less interested in playing with software, and more interested in playing with bright lights and wooden boats.

Maybe some enterprising young hacker will offer a reliable and reasonable reconfiguration service for our switches. I'd rather pay to have it done than bother learning the software.

Perhaps. However I see folks on these forums get very emotional about how they want or don't want levels, strobing, SOS, ultra low, variable output, multilevel output, low first, high first, etc. It even causes brand loyalty as some folks hate or love certain characteristics of the user interface of certain brands. Just the voltage settings on the battery protection alone with different cell counts and per cell minimums exceeds the practical flexibility of a reasonable number of jumpers. Plus each user has their own idea of how hard he (or she) wants to drive the filaments versus how much bulb and battery life, intensity and battery drain will result, and how fast they want it to "soft start". I tried to accomodate just some of these desires and the software already has many switches and parameter settings in it, and I doubt if users all want the same four settings (though that would be great if they did).

If it creates a market for hackers to customize the parameters or even the code, that's okay too. I see folks trying to hack the DX regulators, this should be a lot easier.

Once you get the settings the way you want them, they probably won't need to be changed. But each user has a different idea of what they want. I can set all this before delivering the regulator, not everyone will want or need to change these.

Last night I made a lot of measurements and spreadsheet calculations and tuned the new PCB layout for the physical layout. Perhaps tonite I can start cutting metal and try making the end discs that will fit in the M@g bore and support everything. I'll make the first prototype long enough to handle the existing prototype PCBoard, but there is space in the M@g tube for a little extra length anyway.

The present (default) variable interface:

click when off to goes to low (low voltage)
click when on low goes off

press when on to increase intensity,
stops when released or max reached

click when above low drops to low

press when off goes straight to full intensity (turbo boost)

a click is less than about 300 milliseconds
a press is more than about 300 milliseconds

Some folks probably have other interface modes they want. It would be interesting to hear about those as well (at least from the folks wanting regulators).

-- Alan

 

[Lips]

Re: Feeler for D M@g Programmable Hotwire Driver

I'd like one running a Wa1111 with 3 x 17670.


Will the switch be able to handle bi pin bulbs, PR bulbs, 100 watt bin pin bulbs?


Thanks

 

[Alan B]

Re: Feeler for D M@g Programmable Hotwire Driver

I'd like one running a Wa1111 with 3 x 17670.


Will the switch be able to handle bi pin bulbs, PR bulbs, 100 watt bin pin bulbs?


Thanks

Thanks for the input. Thus far I'm planning a bipin socket like the KIU adapter. A different socket for PR could be done, but I'm not sure how that would work mechanically with the standoffs supporting the socket.

I believe the switch will handle 25V 10A RMS, so a 100W bulb should be possible. The ratings are very conservative, but one thing to keep in mind is that when the battery pack voltage is higher than the bulb voltage the current pulses are also high. So if the pack voltage is double the bulb voltage the current will be double also, and the duty cycle will be 25% to get the RMS voltage and power correct. So a 10A bulb would have 20A pulses.

-- Alan

 

[BSBG]

Re: Feeler for D M@g Programmable Hotwire Driver

I would be in for one or two - an 85 for sure.

 

[Starlight]

Re: Feeler for D M@g Programmable Hotwire Driver

I would like one for the 62138/64610 combination and one for either a 64623 or a 64655. All will be run with up to 8 Li-on. If you can't make the high power version, then I would take 2 for the 62138.

 

[william lafferty]

Re: Feeler for D M@g Programmable Hotwire Driver

I would like one for use with one of the 20 v. Osram bulbs using 8 x IMR18650s as a power source. So I would need the driver to convert 32 volts to 20 or 21.