New driver from RUSH for the 1x123 "E-Series" lights.

[F89]

Re: New driver from RUSH for the 1x123 "E-Series" lights. The HIVE- TEMPORARY HOLD **

Hi All,

It is upsetting that there is an issue with the drivers that has slipped through, but seeing that there is a number of them that don't work as expected, Don has done the right thing to get me to fix this first before building more. I'm sorry for any one of you that might have to wait longer for their light now.
As Don has explained, the problem with some of the drivers is that there sometimes will be no light output on the lowest level of the standard configuration. This especially happens when the light has been off for some time and is then first turned on with this low level of output.
The cause lies in the code of the driver that uses some magic to allow the output to come on even on a rather low output level of a few mA. Well, this magic obviously doesn't use the ideal parameters to make it work on all of the boards. Thus i will have to change those parameters a little.

Those of you that have receive a HiveLD-S already and have not had a problem on the lowest output level (LL as Don called it) should be fine. Not all of the drivers have this problem.

Thanks for clearing that up.
I have a question regarding the LL power setting, I know that you mentioned the default settings but is there a specific mA that if below causes the problem? As in that the default LL is 12mA, is this the point where at or below the problem occurs?
I shortly after getting mine unlocked the programming and adjusted a few things and the LL was a main consideration as I didn't like the PWM. I now have settled on 3 modes at 14mA (no visible PWM), 313 and 800mA (level 21).
What I gather about the problem so far is that it only appears on the default settings on the LL?
I didn't use mine on the default settings for long enough to see if I have the problem or not and I'm assuming that if my LL (lowest level) is set to 12mA or less that I may have a problem be it on default or programmed levels?

 

[fyrstormer]

Re: New driver from RUSH for the 1x123 "E-Series" lights. The HIVE- TEMPORARY HOLD **

I'm guessing the current-regulator doesn't turn on correctly below a certain mA, but once turned-on it can continue running below that threshold, so the "magic" may be a split-second spike in power to turn on the current-regulator. I've seen similar things happen with LED lightbulbs connected to dimmer switches; they will run at the lowest dimmer setting, but if you flip the switch off, the lights won't turn on again until the dimmer setting is increased. Since the electronics that comprise the current-regulator are all made to a certain tolerance, the exact minimum mA at which the current-regulator will turn-on varies from one to the next. The only way to find the best compromise between reliable operation and minimizing visible "pre-flash" at power-on would be to test a large number of the drivers, and even then real-world use may indicate that changes need to be made.

TL;DR: "In theory, theory and practice are the same; in practice, theory and practice are different."

 

[archimedes]

Re: New driver from RUSH for the 1x123 "E-Series" lights. The HIVE- TEMPORARY HOLD **

Is that the reason for "pre-flash" (in other drivers, not necessarily this one) ?

 

[rush]

Re: New driver from RUSH for the 1x123 "E-Series" lights. The HIVE- TEMPORARY HOLD **

Fyrstormer has very well described the background of the issue, i couldn't have done it better. With this kind of converter, getting the output to come up properly at a low output current is difficult. Considering tolerances of components, the "magic" power boost has to be optimized based on a larger amount of drivers indeed. For these new drivers i have unfortunately not set the boost high enough as it now turned out.
I'm very sorry about any of you that have problems with your new light or LE because of this.

F89,
the brightness level 3 (LL, 12 mA) and below are indeed the most likely ones to show the problem. So the best way to make sure that your driver is okay is to switch back to the default brightness level 3 for a while.

 

[F89]

Re: New driver from RUSH for the 1x123 "E-Series" lights. The HIVE- TEMPORARY HOLD **

Fyrstormer has very well described the background of the issue, i couldn't have done it better. With this kind of converter, getting the output to come up properly at a low output current is difficult. Considering tolerances of components, the "magic" power boost has to be optimized based on a larger amount of drivers indeed. For these new drivers i have unfortunately not set the boost high enough as it now turned out.
I'm very sorry about any of you that have problems with your new light or LE because of this.

F89,
the brightness level 3 (LL, 12 mA) and below are indeed the most likely ones to show the problem. So the best way to make sure that your driver is okay is to switch back to the default brightness level 3 for a while.

Thanks rush,

That all sounds like I had imagined.
As long as the problem only occurs at around 12mA or less (which is where I found visible PWM) it shouldn't bother me as while I like low lows I don't like visible PWM so I'll stay at 14mA and above which is my current low (level 4 with a max of 800mA).
I may try level 2 or 3 (6 or 9mA) temporarily just to experiment and see if it fails.

 

[fyrstormer]

Re: New driver from RUSH for the 1x123 "E-Series" lights. The HIVE- TEMPORARY HOLD **

Is that the reason for "pre-flash" (in other drivers, not necessarily this one) ?

"Pre-flash". That's the term I was trying to remember, thanks.

It's certainly possible this is the cause of pre-flash in cheap drivers, though it's fairly uncommon to see cheap drivers that have "firefly mode" settings. It's also possible that for "buck" drivers the pre-flash is the result of the circuit running at full power for a split-second before the voltage-reducing circuit comes fully online and stabilizes at the correct voltage.

 

[KDM]

Re: New driver from RUSH for the 1x123 "E-Series" lights. The HIVE- TEMPORARY HOLD **

I've experienced the pre-flash in Foursevens and Cool Fall lights. HDS and Zebralight seem to have it figured out.

 

[gunga]

Re: New driver from RUSH for the 1x123 "E-Series" lights. The HIVE- TEMPORARY HOLD **

The nice (but costly) gdup circuit has preflash. Very annoying.

 

[PoliceScannerMan]

Re: New driver from RUSH for the 1x123 "E-Series" lights. The HIVE- TEMPORARY HOLD **

I was once so used to preflash I would cup my hand over the business end while clicking it on. I have not noticed pre-flash on my Tri-V2.

 

[rush]

Re: New driver from RUSH for the 1x123 "E-Series" lights. The HIVE- TEMPORARY HOLD **

A pre-flash can certainly be implemented deliberately but it can just as well be a side effect of the drivers' controller not starting to regulate the output quickly enough. Every driver design is different obviously. With the HiveLD a pre-flash or a quick boost of the output is necessary unfortunately.
A pre-flash is annoying for sure and thats why i try to make it an invisible as possible. But as it turned out i have reduced the boost too much to reliably get an output on the lowest brightness levels. I'll have to test with an even larger number of boards to make sure that all drivers work on the lowest level while keeping any pre-flash as unnoticeable as possible. Given the variances from one driver to the next, this is not easy.

The problem is really most likely to appear in the brightness levels 1 to 3, i. e. the modes that use PWM. So F89 you might not see the issue after you changed the output level to a non-PWM level.

 

[F89]

Re: New driver from RUSH for the 1x123 "E-Series" lights. The HIVE- TEMPORARY HOLD **

A pre-flash can certainly be implemented deliberately but it can just as well be a side effect of the drivers' controller not starting to regulate the output quickly enough. Every driver design is different obviously. With the HiveLD a pre-flash or a quick boost of the output is necessary unfortunately.
A pre-flash is annoying for sure and thats why i try to make it an invisible as possible. But as it turned out i have reduced the boost too much to reliably get an output on the lowest brightness levels. I'll have to test with an even larger number of boards to make sure that all drivers work on the lowest level while keeping any pre-flash as unnoticeable as possible. Given the variances from one driver to the next, this is not easy.

The problem is really most likely to appear in the brightness levels 1 to 3, i. e. the modes that use PWM. So F89 you might not see the issue after you changed the output level to a non-PWM level.

No worries.
I've had a bit of time to play with this driver (with the 119V LED) and my suggestion, like it or loathe it is:
1. Add a solder bridge on the menu access tab as standard. This will aid those who would like to play with the custom settings but don't have the tools to add the necessary blob of solder.
2. Raise the LL level to the first level without PWM (with a max of 1A this would be level 4 of 18mA).
3. Lower the max output, it gets pretty warm and chews through batteries.
I lowered my max output from 1A to 800mA but it still ran quite hot so I further reduced my high mode by going down to level 20 which is 628mA.

I now have 3 modes not too dissimilar from a stock 3S, funny that. A 14mA (PWM free) L, 254mA M and 628mA H which I'm happy with for now.
My settings suit my use and I'd say most, especially those using a 123 pak. I mainly use this driver on a 123 pak and sometimes a 2AA pak.
The only time I'd consider using the 1 or 1.2A capabilities of this driver would be using it with a 2x123 and liion 16650 or 2AA pak which would go really well with a Makai for example (which I don't have).
Either way I think points 1 and 2 are definitely worth considering.

Edit.
I just noticed, while tail standing the light as a candle in low mode, that even when you choose a low that doesn't show strong PWM as in choppy movement there is still a mild flickering. It appears that this driver isn't overly suited to low outputs? For someone that's picky like me anyway.

 

[fyrstormer]

Re: New driver from RUSH for the 1x123 "E-Series" lights. The HIVE- TEMPORARY HOLD **

The nice (but costly) gdup circuit has pre-flash. Very annoying.

I used the GDuP drivers in my Lunasol Centauri mods. I never saw them pre-flash. I just checked my LSC-27 and I still don't see any pre-flash.

 

[gunga]

Re: New driver from RUSH for the 1x123 "E-Series" lights. The HIVE- TEMPORARY HOLD **

I had and sold 4. I requested super low moonlight modes. They all had preflash. I talked to Wayne (the designer). He said they all exhibit preflash. It's part of the design.

 

[F89]

Re: New driver from RUSH for the 1x123 "E-Series" lights. The HIVE- TEMPORARY HOLD **

I had and sold 4. I requested super low moonlight modes. They all had preflash. I talked to Wayne (the designer). He said they all exhibit preflash. It's part of the design.

I've built two LE's with a GDuP and they both have preflash on low at first turn on, if at turn on you start on high or medium then cycle through to low then no preflash.
I really like this driver and its mode spacing (10,180 and 700mA) and have even come to not be bothered by the preflash or the occasional calibration that it does.

 

[fyrstormer]

Re: New driver from RUSH for the 1x123 "E-Series" lights. The HIVE- TEMPORARY HOLD **

I had and sold 4. I requested super low moonlight modes. They all had preflash. I talked to Wayne (the designer). He said they all exhibit preflash. It's part of the design.

Ah. When I ordered the drivers for the Lunasol Centauri mods, the lowest setting had to be bright enough to be seen over the flood emitters that were already powered-on, so I guess the circumstances that cause pre-flash with that driver didn't apply.

 

[rush]

Re: New driver from RUSH for the 1x123 "E-Series" lights. The HIVE- TEMPORARY HOLD **

Thank you for your thoughts and suggestions F89. I always appreciate constructive feedback.
If i understand 1) right you suggest to have a closed setup jumper as default? It was Don's decision not to have the setup mode active by default so nobody ends up in the setup mode by accident.
2) I had contemplated about this earlier on also but opted for the lower output vs. no PWM on the lowest output level. Not sure if the majority would share your preference...?
3) 1000 mA indeed produces a considerable amount of heat in a Haiku. But with the drivers' overtemperature output reduction and a stock 4 level configuration, there is the third output level that is often bright enough for many tasks.

You are correct that there can be some inconsistencies in the output for the very low brightness levels. There is still room for improvement of the driver hardware, but it will require lots of tests and optimization of the hardware. This would be on the agenda for the next redesign of the driver as i certainly also would love to loose the PWM for the low output levels.

As it is the pre-flash is necessary unfortunately, but i am trying to minimize the visible flash.

 

[F89]

Re: New driver from RUSH for the 1x123 "E-Series" lights. The HIVE- TEMPORARY HOLD **

Hi rush,

You're welcome regarding the feedback, just rambling through some thoughts.
I added a small blob of solder myself to unlock the programming, it didn't arrive that way. I thought it might be helpful to aid those without a soldering iron and could go unnoticed by those not interested as you have designed the programming not to be unintentionally activated.
I haven't noticed preflash as I do on the GDuP based LE's I've built but I have noticed an "unsteady" output on low drive currents. Many would not notice without looking for it. As I reported, I set my low to 14mA or above where no detectable PWM choppy movement could be seen but upon setting the light down and observing its beam on the ceiling I could see a flickering type of intensity of the light although quite mild and easily overlooked.
I'm definitely encouraging of your design and Don's for that matter and any thoughts and observations I make are positively motivated.
Keep up the good work mate.

 

[McGizmo]

Re: New driver from RUSH for the 1x123 "E-Series" lights. The HIVE- TEMPORARY HOLD **

...... I thought it might be helpful to aid those without a soldering iron and could go unnoticed by those not interested as you have designed the programming not to be unintentionally activated.
....

I went out of my way and RUSH's way for that matter in having him add a via for signal lead both on the converter as well as on a new contact board in order to have a jumper mod for programability. That might have been unnecessary but I want to be clear to any user that if they opt to get into the programming of their light that they are on their own to sort it out an not send the light back to me for program changes. To use a weak analogy, I didn't want to open this door but understand the utility of having such a door. If someone wants me to solder the jumper for them, I will do that when I build the light for them but it's then their baby to deal with any issues if they get into programming it. Another consideration for down the road is when one of these lights changes ownership and the new owner potentially has some issue with the light.

 

[F89]

Re: New driver from RUSH for the 1x123 "E-Series" lights. The HIVE- TEMPORARY HOLD **

I went out of my way and RUSH's way for that matter in having him add a via for signal lead both on the converter as well as on a new contact board in order to have a jumper mod for programability. That might have been unnecessary but I want to be clear to any user that if they opt to get into the programming of their light that they are on their own to sort it out an not send the light back to me for program changes. To use a weak analogy, I didn't want to open this door but understand the utility of having such a door. If someone wants me to solder the jumper for them, I will do that when I build the light for them but it's then their baby to deal with any issues if they get into programming it. Another consideration for down the road is when one of these lights changes ownership and the new owner potentially has some issue with the light.

I can understand where you're coming from there Don potentially opening such a can of worms.
It also makes more sense to solder on request as if it were solder by default then the jumper need not exist in the first place. Good for those interested to know that you'll put the blob of solder on for them to open up that can of worms for themselves but as noted if that can turns into a Pandora's Box, or jar or whatever, then they're on their own when it comes to blinking lights and Morse Code.

 

[RedLED]

Re: New driver from RUSH for the 1x123 "E-Series" lights. The HIVE- TEMPORARY HOLD **

I can understand where you're coming from there Don potentially opening such a can of worms.
It also makes more sense to solder on request as if it were solder by default then the jumper need not exist in the first place. Good for those interested to know that you'll put the blob of solder on for them to open up that can of worms for themselves but as noted if that can turns into a Pandora's Box, or jar or whatever, then they're on their own when it comes to blinking lights and Morse Code.

Maybe someone, one well respected, could take care of modding these things.

Selling a McGizmo is like exchanging gold bullion, and I would hate to see that change.

Henceforth, persnickety buyers could sink what was a smooth worry free deal, only to creat an absolute nightmare selling a McGizmo by turning into much more complex dealings, with lights being possibly being returned, with buyer not Liking the programmed light as the original person which purchased the light the way he instructed and asked for it to be built that way did.

All, you could when selling, and I hate to speak, of the second market here as this is Don's sales site! but, these lights are special, different even than other custom lights and I think Don will understand my point here. But, I do want to make a formal apology to Don here in his sales forum Point is, what if all the sold lights wanted to come back for any type of change. That could put a terrible strain on how Don conducts business. I love the fact he designed, made the light, packed it up and shipped it. Not many business owners with the statute Don carries who do it all themselves most amazing

There simply is nothing like having contact with the owner of the company.

so,unless a good modder would want a to take this on as a project Where that these lights may set exactly the way they want them, well I just don't know

We would all have to dis-claim LED drivers and the LED itself, selection itself in the sales thread.

frankly, I would love one of these already done from Don. I love LL's

LL (Known as moon glow, moon mode for the sake of translation, as not everyone speaks English).

This is just a thought:

I think we could live with: LL combined with couple of nice progressive medium levels, and a nice high level. That would be rather nice, with the levels coming from Don himself. I think it could be fantastic, if only temporally, and again these are just my thoughts.

But, I would have no problem keeping things as they are, or have Don builds us nice 4S until, such time an easily programmable light can be produced.

I don't blame Don, one bit for not wanting to not wanting to taking LED Level orders...orders , this is no different as someone calling in an order for pizza...what do you want on that? And the funny thing is, he gets the lights to you faster than any pizza joint!

Also, I always liked Don's business model on these are the LED's I offer, LED's, these are the only LED's we offer for installation in your light.

simple, easy for the new buyer or repeat buyer direct from The McGizmo LED Flashlight Factory research complex with campus located in Lahaina, Hawaii, and to the resale market unless modded, they are all the same. and that's it. It just seems nice and easy. Simple.

Simple!

But...

Now, however, It would be the same thing with the persnickety and possibly things made even worse, as you may have to include, for technical aspect of the different LED's /drivers the crowd, of LED color temperature academics and the uncompromising austere nature of the the color temperature tint ordained into the Order of This is the chosen one of LED's.

I mean with the programming of the levels, why not offer a variety of LED to choose from?

So, you can see why Don would not want to deal with the customizing of all these lights which the possibilities would be in the millions, an awful position for our beloved maker of our favorite flashlights to, be thrust in.

And, if you have not bought a light from the man himself, if you don't own a McGizmo, he should be addressed as Mr. McLeish.

I think it would be cool if he offered an extra level at purchase that he preset on all, with the new driver of has which has been posted in another thread but, nothing more, and if you do any modding, it is on you and not to be sent back, unless it is broken or a strange defect kind of thing, but that is different.

Any future buyers would have to take chances as to what they want in a light but not send it back to the maker for modifications, that would and should be left up, to others in fact many other industries work this exact same way, some for the same way Don had stated, some totally other reasons. So what he is, as a maker, is asking for is nothing unusual. And very common.


I guess that's it, somehow it will all get sorted out, and I am already planning my next Sun Drop.

There are other wonderful custom flashlight makers out there but, not any others that operate quite like Don McGizmo does. He is one of a kind no doubt. He has made, and is a huge part of modern LED flashlight history.

Best wishes,

RL

Don, just kidding about the Mr. Thing....although you do deserve it.