With negotiations underway for an upgraded srk style light I thought it might be advantageous to discuss and exchange ideas for a driver to push the boundaries while minimizing physical modifications. The idea I found most appealing was a boost driver to allow drop-in conversion to 6v emitters to maximize the output from the Q8/Q6 project. It would also be useful as an upgrade for existing srk style lights. Consider this an appeal to the driver and firmware gurus on this forum - please help create a driver that will leave all who witness its performance awestruck and wanting one.
nice, subbed! Thanks.
With 4 batteries in parallel, current definitely won’t be an issue. It will be interesting to see what the driver guys here come up with. Is that one fellow with the 6v boost driver for a single cell still kicking around here?
If you’re going for a driver change out anyway, why not just change to a 4S buck driver. Richard at Mountain Electronics already designed a good, high-current buck driver which can be zener-modded to run 6V LEDs. All you need more is to mod the cell arrangement from 4P to 4S. Then again, if you instead mod it to be 2S2P, then you don’t need a buck or boost, and there are already several high performance linear drivers that work with 6V emitters! There are more people on BLF that can do PCB layout than there are that design new drivers. Perhaps someone could make a set of contact plates for 2S2P and for 4S as easy to install mod options for SRK’s
Bottom line: IMHO it’s goofy to use a boost driver with a multi-cell light because there are much better options.
Changing the cell arrangement is not as simple as changing out some contact plates. When converting an SRK from 4P to 2S2P or 4S you are not quite there with new contact plates, you run in trouble at the top side and it can only be done if you manage somehow to fix the top plate in one position aligned with the batteries while closing up the light. (I hope I made myself clear).
If i understand it correctly, I think Richard’s driver is maybe 6A total, so 1.5A per LED - not very powerful?
+1 with djozz - not so simple. Also, we would need an LDO driver, not zener-modded because of the e-switch.
Rich has only one boost driver, the Nanjg110 with~800mA from 1-2AA cells. The Taskled hyper boost is a 3200mA driver but requires 2 cell input minimum. the DQG boost driver drew 6A to get around 2.5A to the emitter. To get 8A to 4 emitters you would need 18-20A from the cells and have to redesign the light to handle all that current in all the wrong places. I found a 10A boost ic, the lm3478 that operates down to 2.97V but have no idea what the size or cost of the associated parts would be or even if it’s the right application. If it can be done in a host, the SRK is an ok place to try it if the driver pocket can fit the inductor. If you want this to happen, get out and push, offer to order some parts and pcb’s for testing. This kind of thing takes time, isn’t free, and often has to wait for someone’s time and interest to free up.
It just seems wrong from an engineering standpoint to have the greatest waste of energy in losses occur on the opposite side of the circuit from the point of use so you may be up against institutional resistance but this forum does appeal to the more radical, selfless, and open minded types so there’s hope.
The 10 micro Henry inductor in the lm3478 functional diagram might need to be up to 10mm high X 16mm square to handle the input current(don’t hold me to that as I’m only guessing based on Isat and Irms values and don’t know squat about either) but this is the main reason why it’s not going to happen in a smaller light. The ic lists at $2.12 and the inductor for a bit more. No sure about diode ratings but a 25A smd runs about $2 as well. This doesn’t include other parts or shipping and oshpark test boards are ~$13 for 3/46mm pieces. Please look at the pictures in the data sheets, over time some of the stuff sinks in. Not as well or as correctly as from a course in EE but at least helpful in identifying parts even when you don’t understand how they work.
I’m continuing to research this myself as time allows but unfortunately my now obsolete electronics training was over 30 years ago and little used since. Back then most components were through-hole mounted and surface mounted components were infrequent. For these reasons I thought a discussion of the possibility of creating such a driver might be beneficial. The 47mm driver size offers a lot of space compared to many other lights. If I could find my books and notes from the metrology course I could put myself through a refresher course on the circuit design portion.