DB1000 and WA Solarc...

Heya,

I'm no electrical engineer, let's get that out of the way right off. But I learn fast, and generally understand what's being said to me.

That said, is there a compelling reason why I can't use, say, a DB1000 stepdown converter from the Shoppe and a stack of 3x18650s to drive a WA Solarc 10w ballast?

If I'm reading all this right...

An option for the Solarc ballast is 13v as an operating voltage. Another version will operate at 10.2v.

According to the spec sheet, current draw for the 13v unit is ~750mA. The 10.8v one draws roughly 1.0a.

I am uncertain of what voltage the DB1000 can put out - the Shoppe just says "Luxeon V see calculator," which is not hugely useful to me.

If the DB converters can drive voltages that a WA Solarc ballast wants...both the DB750 and DB1000 can drive the currents. it should be possible to regulate the thing, right?

[Alternately, would two stacks of parallel 3x18650s in series make a Shark a happier alternative?]

Or am I just totally off the deep end, here?

HIDs usually go through turn-on and then warm-up procedures during the first seconds after clicking on and during these stages the power needed may be very high compared to the steady state operation. DB1000 just dumps 1 amp to the ballast without "understanding" if the input voltage (and wattage) sags down to 8, 10 or 13 volts due to the power needed. In the worst case the DB + ballast combo starts to oscillate between on and off.

I suggest that you connect your ballast to the batteries through a dmm and check the input amps during all the stages of the ballast operation (high voltage turn-on strike, warm-up and steady state). If the amps stay at some constant level all the time (e.g. around 1 amp) and the input voltage does not sag down or kick up then your idea should work nicely in my opinion (if DB can only handle input and output voltages).

Pay also some attention to the input voltage vs input amperage characteristics of the ballast during each stage of its operation. If very small changes in the input current lead to very big changes in input voltage (and thus, in input power), constant current power source is unstable solution. And vice versa of course. This is exactly the problem with leds, although in reverse: very small changes in input voltage lead to very big changes in input amperage. That's why the preferred powering method for leds is constant current, not constant voltage.

You may want to have a look on my solution here.

Hope this helps.

-N

first of all, You will need a 4 cells in-a-row Li-Ion pack for a step-down converter to get to Your voltage.

Then I could tell You, that I have such a poject running for 2 years now and finally arrived at a driver using the National Lm2670 IC, because no other circuit worked with the WA-Ballast (and I tried SEVERAL!)
but i never thought about a DB, isnt this set on the ~6.7 volts for a Lux V ?

but the best way I can do NOW, is to tell You get Yourself 3 or 4 Cree/SSC P4 emitters and a SHARK and some reflectors instead.
WAY BETTER light for less price, the HID sucks.
The only thing really cool with the HID is the starting period. No other source gives this kind of show

I did this pic yesterday and it shows appearance in reality very good. The HID is a tiny! bit brighter, especially when used alone. But color is awful and You dont see red/brown/green very good.
Even the single Cree @ 800 mA emitter lights could compete with the W-A (which is the broader beam version, I admit, but still ...)

lights tested:


all about 15-20 meters away from wall:

Yellow a properly setup Welch Allyn HID will Blow away any single or tri-cree light. Your using a completely different version than used here on the forum.

The 10.2v ballast driven with 3x Li-Ion cells is pumping out probably well over 600 lumens.

Pick up the stand alone lamp and use a mag reflector and them make your judgement.

Mac

Yellow:

Pretty much what Mac said. I have one of his Mini-Mini HID units that runs on 3x R123 with no boost, buck, or regulator in between, and it screams. It's the standalone ballast and bulb, a real show-stopper.

Anyway! What is the max vOut of a Downboy? That's the biggest question right now...

The V out of a downboy is the Vin of the LED up to 6 volts, it varies with current, but is not high enough for a WA ballast, you might try using a shark converter board, then get the B10N008 Ballast, It will overdrive the bulb freshly charged, but I don't know of any board targeted to put out the right amount of power for a 12volt or similar ballast, I know using a downboy and then adding a shark in series would give you the right voltage but effieciency would be horrible, might be a place for a custom converter, I would recommend boosting the voltage up to 14.5 volts.

Actually, a Shark almost looks like a good thing, except I think my vIn is way too high in relation to vOut.

Freshly charged, 3x18650 should be a little more than 15v no-load, and be about 13v depleted. Even with the 13v ballast, that means Vin is higher than Vout. Can you set a Shark board to do a *tiny* boost - to, like, 16.5v? I wonder if that would fry the WA ballast...it would almost certainly shorten the ballast/bulb lifetime, right? Of course, more lumens...

Well, 3 Lithium's at 4.2Volts fresh, would be a fresh off charger V of 12.6 Volts, and at 11.V with medium load, and die around 9 volts. So you end up with a Vin range of 3.6 Volts, that's a pretty high range for a ballast, but you could check out some info about the shark by contacting member, dat2zip (wayne) and asking him about the J1 jumper which may get you enough voltage, he could probably help you out. I would probably recommend using some custom boost converter or a buck/boost that would put out a steady 12volts and use a B10N001 ballast.

really wonder but maybe its because of my "needs" are no thrower but more an small area light for biking :thinking:
I have a M10P002 13 deg. bulb and a B10N004 ballast.
http://www.walamp.com/lpd/files/datasheets/ISL305B2.pdf
The specs of the ballast are: nominal input: 10.2 Volts, max safe continuous Voltage 11 Volts.
Thats what my LM2670 pushes in: 9 V for "economy", 10.2 for "normal", 10.9 for "high" and there is a real difference in output with all these inputs.
Sure the light is bright, but color retention is bad, especially for an outdoor/woods applicaton like bike light.

Yesterday, when doing the pics, the 6P with Cree seemed way better, but the beam is more on the tight side, so maybe ...
but imho the tri or quad Cree/SSC emitter light, that is absolutely comparable in current consumption and price (and better in producing less heat), smokes this 10 W HID, even when I consider mine being a monday-model. Am planning to offset the three beams a bit with some central hotspot and a triangle like main beam, and am pretty sure to get the better light. Alone to be able to see if the wayfinding marks are red, green, brown or blue from a distance, gives this setup an advantage.

Unfortunately by now I was not able to compare this HID-mod against a "real" model (= a Lupine Edison) but what I remember from these, the optical appearance was relatively the same.


PS: it is four times 18650, not three times with a downboy (if this driver works here).
3 * 18650 (as 65535 typed) is 12.6-7.2 volts (very depleted), so buck/boost what does not work very good.

Sable, I used LM2670 in my maghid project. There is (at least) one disadvantage in LM2670: the maximum PWM duty cycle of the LM2670 is 91% (IIRC), not 100%. This means that it does not go to direct drive mode when the batteries get empty and the input voltage sags. In order to maintain full constant DC regulation with LM2670 you have to have decent difference between input and output voltage.

In my maghid the output voltage of LM2670 is set to 11V. When input voltage drops to 13,2 V the max 91% PWM duty cycle limit kicks in and the output voltage starts to drop following the input voltage drop. However, in my maghid that is not a problem because at 13,2 V input the average nimh AA cell voltage has dropped to 1,1 V and they are completely dead.

-N