StarMed StarLight medical headlamp

[ebastler (Moritz)]

Hi,

A few days ago my dad brought his headlamp from work home and complained about mediocre light quality and low brightness. I gave it a quick check, and figured you might want to see it, too.

The main controller body is completely sealed (guess some medtech things), and the battery looks like a generic 18650, but in a weird plastic case that is, ofc, locking you down to manufacturer-supplied modules. Which, no doubt, cost a small fortune.

The head is an incredibly lightweight, hollowed out CNCd aluminum structure, with a tiny PCB and a single domed 3535 LED, covered by a TIR lens. Oddly, the head is ventilated (so much for medtech requiring fully sealed electronics, eh?) and even dust can easily get in the head and between TIR and LED… Not great. There was 0 dust inside after years of use though… Guess hospitals are cleaner than my apartment

I quickly tested it with my Opple LM4, 5700K, 70 CRI, R9 of 0-20 bouncing around. I know, LM4 is pretty wack, but still. This is definitely no high CRI LED - a 5700K 90CRI XHP70.3 HI from Simon measures as 90 CRI and IIRC an R9 around 60.

How a medtech company can supply low CRI and low R9 LEDs to people who gotta spot fine differences in tissue is beyond me. Although maybe just his employer ordered the cheapest possible version.

Think I’ll let him take my 519A S2+ along for a test if it will improve color rendering, and then reflow a new LED in this thing. 5700K domed 519A probably. Neutral, high CRI, high R9, and pretty much as efficient as it gets. Hope it will play nice with the TIR.

Can’t do anything about the brightness without modifying the controller (which would void its ability to be used in medical environments), but if I can at least up the R9, that should help a lot.

EDIT: I obviously did not think about the fact that he can not use my S2+ for testing of a high CRI LED since he has a headlamp for a reason - he needs both hands to work
Guess I’ll just order a LED and will see.

[KnotSoMuch]

If your dad could convince the chief of surgery to let him perform a quick CRI evaluation test with your S2+, it would then have to survive a run through an autoclave to be certified for use in the sterile field. Pending that, his scrub nurse or the circulating nurse could be the one to hold the S2+ for the evaluation.

[ebastler (Moritz)]

@KnotSoMuch

autoclave

Not sure how it works exactly, but I am pretty sure that headlamp would not survive an autoclave either, and neither would stuff like magnification goggles or cameras (I know some people in the hospital used private cameras for surgery documentation, before the hospital got dedicated ones that never leave the rooms).

AFAIK only equipment that can/will come in contact with the patient or any of the equipment that will has to be autoclaved, the rest is probably sterilized some other way. The touch surfaces of the StarLight suggest some sort of chemical sterilization to me, but no way the lamp head would survive that - the LED is fully exposed, and even if stuff is not conductive, it would probably destroy either the TIR or the dome, or even the phosphor.

[zoulas]

I had to look up autoclave. Learn something every day.

[saltyboi6704]

Chemical autoclaves usually use Ethylene Oxide and I would be worried about it interacting with the dome and TIR, but if it were the correct material chances are it could survive a steam autoclave. The fact that it uses an FR4 PCB instead of aluminium or copper means that the LED likely isn’t being driven very hard.

I’d also use the Opple to test for flicker - if it’s bad it could be potentialy dangerous in an operation room since there’s a chance it reduces hand-eye coordination.

[gravelmonkey]

Cool pics, thanks. You sure that’s 3535? For some reason it looks like an XML to me.

Probably an XPG if it is indeed 3535.

That CNC body is awesome, shame about the waterproofing/sealing. I suspect that light costs an absolute fortune, I’m surprised some enterprising person isn’t doing small runs of decent medical lights- like the enthusiast caving headlamp builders do.

Med headlamp topics come up about once a year, iirc, sterilisation not required, bit of a wipe down and it’s good to go.

I know a guy who built his own using convoy S2+ shorty tube a plastic headband, possibly from AliExpress, and an external belt battery pack.

Edit to add: did you measure the lux in stock form? Lumens? If you measure the TIR, although you won’t be able to improve lumen output, one with a more narrow design would put more Lux in the spot, which might help.

I suspect you won’t have a problem if you reflow, but if you use another MCPCB, I wonder if you need to check to see whether the body is negative or positive?

[ebastler (Moritz)]

@gravelmonkey

I’m surprised some enterprising person isn’t doing small runs of decent medical lights- like the enthusiast caving headlamp builders do.

My guess is - regulations. Having your light approved for official medical use (and hospitals can only buy that) likely costs a fortune, and involves lots of legal chaos that only large specialized medtech companies can even afford…

@gravelmonkey

You sure that’s 3535? For some reason it looks like an XML to me.

Yeah, I managed fo squeeze calipers in the head, it’s a 3535. So I guess a 519A would fit.

@gravelmonkey

Med headlamp topics come up about once a year, iirc, sterilisation not required, bit of a wipe down and it’s good to go.

Yeah, talked to dad about it. Basically, everything that comes in contact with the patient needs to be autoclaved, but they can bring stuff like cameras, magnification goggles or headlamps as long as they don’t touch it themselves.

@gravelmonkey

Edit to add: did you measure the lux in stock form? Lumens? If you measure the TIR, although you won’t be able to improve lumen output, one with a more narrow design would put more Lux in the spot, which might help.

Not really since I have not yet figured a proper way to measure lux with the opple (gotta decide on a fixed working distance or it’ll be pointless). It’s not too bright, but very focused already. Pretty throwy optics. For now we’ll just try if a domed 519A will make their work easier by providing decent quality light, at least. If it’s not bright enough, I may end up making my own copper PCB and control/battery pack unit for him. Will also see if I can find TIRs with the right dimensions (have yet to measure).

@gravelmonkey

but if you use another MCPCB, I wonder if you need to check to see whether the body is negative or positive?

Good question, I gotta check. It seems like the MCPCB is only glued in with thermal glue, so probably insulated for now. It’s also tiny, I think 8mm diameter.

@saltyboi6704

I’d also use the Opple to test for flicker

Checked that, it has very little modulation and only in one power level, all others are pretty much flicker free.

[saltyboi6704]

@ebastler

Good question, I gotta check. It seems like the MCPCB is only glued in with thermal glue, so probably insulated for now. It’s also tiny, I think 8mm diameter.

I’m pretty sure it’s just a normal FR4 PCB with a shit ton of vias, you might have better luck stuffing a filed down MCPCB in there with a thermal pad.

[ebastler (Moritz)]

Unless I change anything about the drive current, that’ll be fine. If I end up making a new driver, I will probably also make a new LED PCB for the light.