Grumpy Cat and RaceR86 tests LD-1 5A PWM-less driver (prototype)

Flashlight Modding and DIY Parts
1

Introduction:

Thanks for everyone who voted for me when I did my (cat) campaign. I wont let you down, so I invited Grumpy Cat for the review as well.

For some reason, my driver arrived late, a week ago. Most of the others seemed to have done most of their reviews or finished them. I was not too happy since I had good time to test it earlier but not when it arrived. As seen in the picture below, grumpy cat was very happy about the late arrival. Less stuff to cover about the driver since more was done by others. More time to be grumpy.


Useful links:

Led4power`s own thread about the driver (all basic info is covered here)

HKJ review

Tom E review

Djozz review

The pictures:

Click here if you need larger resolution picture

Connect the two half moons on "1" in order to disable moon. Connect half moons on "2" in order to activate turbo. These are very close. Easy to connect. Maybe not so easy to undo?

Mouseover with mirrored backside

E-switch UI1, UI2 and UI3 walkthrough.

One take video. Sorry about the bad language skills (rusty English, I barely speak it.), errors, strange hand gestures, some bad explanations, some issue that I did not understand, etc, etc, etc...

Hope the video is useful and understandable and that I did not do anything wrong. Oh and don't run the driver in open air like me.. 0:)

This is what Grumpy Cat thinks of this video:

RaceR86`s thoughts on the various user interfaces

UI - Mechanical switch only

Mode spacing is pretty much like a Qlite. I like Qlite and I naturally like this. Not sure if the moonlight lovers will like it. Not a massive difference between low and moonlight. I disabled moon. Seemed fairly similar to low for me, but I have not tried in it any real life use.

I measured:

5,02 (100%)

1,09 (21,8%)

0,07 (1,4%)

0,02 (0,004%)

That is pretty much the same as HKJ measured.

E-switch UI_1 - E-switch + mechanical switch

I like it. You can get it to start up in the mode you like with a switch on the tail, and you can easily change modes. Its quite similar to Star firmware for momentary switch. But with the added goodness this drivers hardware bring.

E-switch UI_1 - E-switch only

I prefer to use lockout in lights with E-switch only. With this UI the light will always light up when the driver gets power. I would prefer that it did not do that, but its not a big deal as long as I make it start in low. I approve.


E-switch UI_2

I really like good ramping UI`s. I don't like the ramping UI in this driver..

-Needs to be turned off and on again every time you want to change brightness. I prefer to be able to change brightness without having to turn if off.

-I often use lights on highest setting. This means I have to always ramp all the way from lowest to highest if the light have been off without power (no short cut to high)

-If the driver is in sleep and I have stored the highest setting. It will always do the bright blinks (telling its on highest setting) every time I start up. (Same if its started up at about 2,5A, but only one blink, which is less annoying). I do understand that some might like getting an idea of how much power the light consumes, and I agree that having some blinks to say when its at 2,5A and 5A can be useful. But I would prefer to not see those blinks every time I start up. I would prefer that the blinks were more subtle too. I would also prefer that the 2,5A blink were at lower amps (say 1A). Although, I personally don't need that extra blink.

-Linear ramping is not good. Going from 2,5A to 5A takes the same amount of time as going from 0,02A to 2,5A. Lots of visible difference the first halfway, not much difference the second half. Visually, the ramping should have a nice even increase in brightness all the way from lowest to highest. It can be improved

-Its jinxed. Several times when I go to this UI_2 I mess up the temperature sensor setting. Maybe its not the driver and just me, but its only happened when I switch to UI_2. As seen in the video, some strange behavior I did not understand happened too.

E-switch UI_3

Its simple like description says. Its nice to have optional UI`s, but Im not sure when I would use this over UI_1. I would prefer to see something more different compared to UI_1. Maybe a tactical UI or something just to make the driver more versatile. Its nice having a third UI option though.

LD-1 vs 5A Qlite out of regulation test

Made a 5A Qlite. Made sure it was out of regulation on highest setting due to "low" input voltage. It gave about 3,75A to the emitter. Changed driver to LD-1 with same setup. 3,73A to the emitter. Id say performance is similar in that test (0,02A difference could be power supply, soldering, crocodile clips) Just wanted to check out of regulation performance/resistance compared to something we all know. And its just as good.

Turbo + multiple emitters

3 XM-Ls in parallel. X voltage, no driver

3 XM-Ls in parallel. X voltage. LD-1 with turbo.

Almost 2A lower. The only thing added was the driver and a few centimeters of wiring. With a FET driver I would expect closer to 9A. Lower modes are naturally regulated at the usual amps. Temperature monitoring seems to work, although I would need to test it in a flashlight in order to better see how it works.

Mode spacing could be a bit strange with normal 1,1A medium and quite high turbo. (unless you use ramping mode, which I dont like on the prototype driver)

Testing overheat protection

I put the driver in a Convoy S-series light. Soldered to the pill in normal way. Ill go a bit by memory now, since its a long time since I did this testing.

If I used it with temperature monitoring disabled, it will still kick in when the drivers becomes too hot (85C?). That is good, grumpy cat disagree. With a freshly charged high discharge battery the driver will cut down on power and keep the light within reasonable (very high) temperature once its becomes too hot. But with a low voltage cell, the light will become too hot (beyond 60C). This means if the light is turned on accidentally in a bag or something, the light will become stupid hot at one point.

I tried a lot to set the tempereture threshold based on external tempereture, but it just doesn't work that way since the heat generated in the driver varies a lot depending on output and battery voltage. Basically, if I want the light to avoid overheating with a low voltage cell, overheat protection will kick in way too early when using fully charged cell.

Temperature monitoring works to some extend. It works to protect the driver, and not from an exterior heat point of view. In my light the overheating protection gave little extra safety.

LD-1 vs some other drivers

The "Bocian" driver, have a more sophisticated firmware for a mechanical switch only. Its also got tempereture monitoring in the version that uses Attiny25V. But based on my very limited testing, that did not work ideally either. Still, its IMO one of the better UI`s for a mechanical switch only. It can not go much past 4A for some reason.

DrJones have a lot of different drivers. In terms of UI his drivers are way more sophisticated. Especially when it comes to electronic switches.

Various other custom programmed 7135 based drivers. There are many options based on 7135/attiny13a. As an example, RMM sells some variations (if you don't program yourself) These are normally cheaper compared to the above. These have a decent UI`s which can also me customized to the user. Price is much lower.

FET based drivers, im talking about the ones that does pretty much 100% direct drive on highest setting. If you want max output, these are the way to go.

Qlite, if you are on a budget and just need the driver in a light that have a normal switch. A Qlite driver does pretty much the same (with extra 7135), just with lower efficiency.

Conclusion

Overheating protection does not really work from an external point of view. Its not really a driver fault, I just had some unrealistic expectations of what it could do.

Id like to see some improvement to the ramping UI.

"Moonlight" is not very low moonlight IMO.

Driver is best suited for single emitter/single cell lights.

I would like to see that it came with a lower output option directly from the seller (3A, 4A, 5A). For some lights/batteries 5A can be too much.

Efficiency, size, UI flexibility, off time memory, no pwm are some areas where this driver excels compared to many of its more common rivals.

Overall I think the LD-1 is a very good driver! Ive got several Convoy S-series lights. And the one with this driver is currently my favorite. Which have much to do with the driver. I also think this is a great driver for lights with both E-switch (momentary) and a mechanical switch due to the way the memory works.

Id recommend it!

2

This one should make GC smile, seems to be the perfect driver for a smallish EDC light. Look forward to hearing your angle on it…

3

Great detailed pics! I'm looking forward to hear your views on the UI. You got proto #3 looks like, think I got #2 Smile.

4

5

6

Subscribed, i have no interest in the drivers for now, but, i can’t miss the cat.

7

Large update to the review. Added lots of stuff!

8

Thanx for the cool review RaceR86.

Regarding the 3 x XML(2), it looks like about 1-1.5v is lost within the driver.

I assume you are running the power supply at 3.35-3.4v. What would happen if you set the power supply to 3.5-3.6v? Would the driver actually reach the 9A mark? (Hmmm, that means that somewhere in the driver 9-13W will have to be dissipated! Unless the extra voltage is because of something outside the driver)

Thanx

9

Thanks for the comment.

No, Im not able to get to 9A or above with the 3 XM-Ls (1st gen) Ive been testing up towards 4V and slightly higher. I can get to 8A+++, but it requires high input voltage, and the driver heats up instantly. Im not comfortable pushing it that hard for more than a couple of small seconds. Especially not when its just hanging in the air, and im using my finger at the back of it to "measure" heat.

I was thinking of how good this driver would be in something like a quad SRK type light. Having overheat protection would be a nice feature, but I get the impression that the driver would have to be really well attached to the light in order to work reliable. It does not seem to like the super high amp multi emitter turbo stuff based on the heat that is generated. So ill probably just avoid that. Its not what it was intended for either.

I have not changed to the correct capacitor, not sure if that could affect the results with the 3 emitters. My understanding is that single XM-L2 and single cell would work nicely with the hardware on my driver. And Im probably going for that combination. Or ill see if I can get my brother to change to the right capacitor. He have the correct capacitor and equipment for replacing it.

Right now, im thinking of just putting the driver in a Convoy S2+ as it is and in combination with a single XM-L2. Should be a nice light to play with the overheat protection. I just happen to have an S2+ in need of a driver and emitter. :)

10

Very good video,you explained all very well.

With ~40uApms of parasitic current,I don't think you need lockout,but that's just my opinion.I also don't understand why you don't like light up at power on,it's logical to me that if you connect power(unlock it),you'll use that flashlight immediately.

That's only true if you didn't press e-switch for more than ~1.5sec.So you can ramp up-down to set your desired level,but time between e-switch presses must be <1.5sec.

I'm listening to your suggestion how to turn off light in ramping UI.

I don't see why is this a problem,since ramping is e-switch UI,so you'll loose power only if you disconnect battery.

I can agree with annoying blink(s) on 50% and 100%,how would you solve this?

Believe or not,it's not linear:form 0-10% speed is 1x,from 10-50% speed is 2x,from 50-100% speed is 3x,but it seems it's still to slow on higher currents,so I'll increase speed in that area.

I saw in video,because you didn't cool driver driver reached 85C when you were at ~2.2Amps.One short blink was sign(I know I'll change this to something more annoying).That's why you weren't able to increase current.When you turned driver off,and on again,driver was still hot,and protection was active immediately. Driver burns most heat in 2-3Amps range(with typical set:xm-l2 and fresh battery).That's why you saw this only in ramping UI.

[quote=RaceR86]

E-switch UI_3

Its simple like description says. Its nice to have optional UI`s, but Im not sure when I would use this over UI_1. I would prefer to see something more different compared to UI_1. Maybe a tactical UI or something just to make the driver more versatile. Its nice having a third UI option though.

[/quote]

Suggest me one UI that you would like to have.

[quote=RaceR86]

3 XM-Ls in parallel. X voltage, no driver

3 XM-Ls in parallel. X voltage. LD-1 with turbo.

Almost 2A lower. The only thing added was the driver and a few centimeters of wiring. With a FET driver I would expect closer to 9A. Lower modes are naturally regulated at the usual amps. Temperature monitoring seems to work, although I would need to test it in a flashlight in order to better see how it works.

[/quote]

You run 3x xm-l2 at turbo with driver in air??

I said before,don't do that,turbo is predicted only for single led.

Turbo depends on driver voltage,theoretically it's around 10Amps when input voltage is 4.35V,and that current linearly drops to ~6.5Amp when battery is 2.8V.

But this isn't important any more,since I'll remove turbo as option(timer on high will probably be the new option).

11

Thanks led4power for detailed reply and good answers.

With 3 XM-Ls I only ran it long enough to take the picture. Otherwise I was holding my finger on the back of the driver and usually just doing 2-5 second runs depending on amps since it heated up so quickly. Someone had to do a quick turbo test with multiple emitters. 0:) At least I did advise others to not do what I did.

Hitting the overheat protection when ramping was clearly my fault. Sorry. Thanks for explaining. I thought I saw the 2,5A blink when ramping down, not overheating protection, so I did not understand what was going on. Bad timing and probably very easy to make it heat up fast there. (Yes, I should have had heatsinking on it. )

Ill get back to answering the other stuff when Im back at the computer tomorrow.

Timer on high instead of turbo??

12

Thanks RaceR for more testing on this driver, and thanks led4power for incorporating some suggested improvements into the final version . The picture of this driver is getting better and better. I will build my prototype into a flashlight and report about the experiences. But I already have a use for 1 or 2 more, especially if the issues that arose in the tests are adressed in the final version.

13

Thanks for the fantastic review RaceR86. The video meant more to me than a thousand words.

That GC of yours did not blink or move once the whole video. It looked like he (she) was in attack mode. Be very wary. Expect it to attack when you least expect it.

14

Good honest review of the kind that encourages improvement. I don’t think a fitseveryneed driver is possible but this one sure seems to fit the onecelloneled nitch very well. In a bigger light that could handle more heat a higher output driver would be nice but this seems like it should be a great fit with the X6 or anything else between a Micromag and C8.

@Led4power, will you be able to post a video showing the updated UI when it’s done? I like the suggestions on ramping. I know some prefer not to have a turbo timer but without it then some form of thermal protection will still be needed. Taskled and Sandwich shoppe market fitted sinks for their high power drivers and though it might be impractical to sell them if you could help us design some good shapes that could be fabricated easily that would be helpful.

15

I will remove turbo (higher current than 5A) and add timer on high(solder moon enable/disable).Timer should be useful as thermal protection on small lights.I explained why in Tom's thread.

With 4-layer pcb and increased overtemp threshold,in case of single battery-single led extra cooling shouldn't be required(except maybe in ramping mode),but it's always good idea to keep driver cooler,so you can use those tiny self adhesive copper heatsinks and potting stuff (silicone sheets,cubes,fujik...).

16

@led4power

I always use lockout if possible. Its just the way I do things. I know your driver is very good at safety, but still.

UI_1 is good as it is IMO. I prefer it like it is.

Or tecnhically, I would prefer it to not startup up when using with E-swtich only. But its just nitpicking based on my personal preference. (Also, most other E-swtich UI`s dont startup when you apply power, so its just what Im used to). When using it in combination with a mechanical switch I prefer that it does startup. Considering that your driver is so nicely suitable for both types of lights (E-swtich only, and E-swtich + mechanical) I prefer to have it as it is.

UI_3 is good. Having a tactical mode was just a suggestion to make the driver even more versatile. Personally im not into tactical modes, and I will not try to advise you on how to make the best strobe and setup.


UI_2, ramping

Here is an example of a fairly simple yet effective ramping UI with some shortcuts:

-When the driver gets power it will always start at lowest brightness (just like it does now)

-Short press turns on and off. Remembers last brightness setting. (If there is no last used setting stored due to having the power off, default could be medium/1,1A. )

-Holding button from sleep starts ramping from lowest brightness.

-Double click always takes you to highest setting. (Also works when driver is in sleep)

-Double click when you are in highest setting takes you to lowest setting.

-Long press always starts ramping (with the ability to go backwards and forwards like you have programmed it. Visually it should look like its always changing brightness evenly.)

-When ramping and hitting lowest or highest brightness setting, it will do a single "soft" blink to indicate that you have "hit the end". (If you start in lowest or highest, or use shortcuts to get there, it should not blink)

Again, these are just suggestions. But I would love to see a ramping UI like that! Always easy to change brightness (start ramping) even when it has been on for some time. Can always startup in lowest, highest and last used ramping setting. Easy to quickly go to highest or lowest output.

I hope to see that overheat protection is subtle and not anything annoying if you decide to change it. With the ramping UI above, It could just be as it is. That is if the "subtle" blinks when ramping to both ends is not similar to overheat protection.

Everyone reading, feel free to comment on my ramping UI suggestion.

17

@ Djozz. Yeah, I would not mind a spare driver for testing purposes. I asked if I could buy a spare one from led4power before he shipped, but he did not have spares. Gotta love the flexibility of the driver.

@MRsDNF. Im glad you found the video useful. I too sometimes find it a bit hard to fully understand various UI`s just by reading other peoples description. Its usually easier to see how it works. Your observations about GC are probably spot on. For saftey reasons I folded put him/she away.

@Rufusbduck. I agree. Fits lights of that size nicely. A video from led4power showcasing the final UI would be great.

@Led4power. You plan to remove user temperature threshold?? Why, and how exactly do you plan for a timer/overheat protection to work? Temperature monitoring and adjustability is for me the main interesting thing about this driver. Although, I have yet to properly test it.. But ill get to it..

Do you consider the temperature monitoring to be more of a driver protection/safety feature than a flashlight overheating protection since there is no external temperature sensor?

18

I was supposed to finish this review a long time ago, but lost "momentum" since there was no feedback/answers to previous posts.

I have now added the last sections to the review. :)

-Overheat protection

-vs other drivers

-Conclusion

Grumpy cat have gone missing. No conclusion from him. :_(