Took the time today to test the driver from #7 of prototype run 3 a little bit. Soldered it into my test rig, consisting of four emitters on a chunky heat-sink mounted on a plywood base / cell holder. Running the latest NarsilM.
Old photo:
This setup, currently with AWG 14 wiring, pushes 17 A on a set of fresh 30Qs.
Scoped around the control signals, MCU supply voltage, LED current. No surprises. No real heat on any of the driver components in any of the modes or battery charge states either. The FET keeps up well at different charge states, no sign of excessive I2R losses. Rds does ramp up at low battery voltage, but the lower current draw (lower battery voltage) compensates. Did not see more than 34 mV across the FET, as can be expected for a 2 mohm FET.
The longer than usual wiring adds parasitic inductance, amplifying the inductive spiking seen at the drain of the FET. Looks scary on the scope, but the driver behaves well, regardless.
The 30Qs are almost drained now. Got the step down flash at 3.05 V (setting in firmware is 3.0 V). It steps down repetitively until the battery voltage recovers above 3.05 V, which currently is at 250 mA. Now just waiting for the LVP shutdown (set at 2.8 V in firmware), this will take a while.
> fixed it
Ruined the surprise, for anyone who didn’t already know what the link opens to.
Those who did know, didn’t hardly need to see it again full screen.
But bless your heart for trying to improve what I’d tried to make a subtle post.
————
Well Bless your little heart to.
All that talk from you the other day about you being old, I just figured you had a Senior Moment & screwed up your post.
But, not to worry… I can “un-fix it” as fast as I “fixed it” & your “surprise” will remain intact for most……
That… is a really nice test rig. Usually I just hold wires onto a battery with my fingers.
I’m going to have to set up something a little bit nicer soon though, for making alternate Q8 firmware. My usual approach is really awkward on a SRK-style driver.
I checked the first post but still got lost in all of the side conversations: Can someone catch me up and say where we are at in the development process and what still needs to be done?
Proper springs for longer term reliability
proper lamination of the ledboard, for the same reason
last Narsil version
manual updating based on latest Narsil
They don’t think that will be necessary. Plus it will add a lot more time which nobody wants to do (us or Thorfire). Everything is looking good except for those small issues.
Oh, please, add the time to check the changes. Count me as that one guy who wants the added time to be used.
Remind them: “There’s never time to do it right, but there’s always time to do it over” is not a good approach to QA/QC.
The danger is having just one employee who figures “oh, heck, that little glitch is not worth slowing down the deal, my boss would kill me if I stop the production run” and lets problems go on by.
Or one QA/QC person who figures the time needed to test the new change isn’t really available and just says hey, we fixed it, it has to be good.
We’ll never see the quality from handcrafted pride in production again, not since the world adopted Henry Ford’s assembly line.
But that’s why we need not to pick up the pace and rush through without careful attention to the last few changes.
It is never fun, but i have scrapped several days of work by 2 guys and a CNC lathe.
And i would not have people look any different on the work i do.
Sadly i managed to spend several days grinding and making a track on those axels before i found the error, so in total over a week work a CNC machine and a manual grinding machine and 3 people wasted.
But at least my work was fine, it was just made on top of something that was not fine :weary:
I always wanted to get into arospace as that’s my level of tolerances, but i never managed to get TERMA to hire me ( https://www.terma.com/ )
But that’s their loss not mine.
Hank, I hear you
With the issues , all the issues , of proto 3 the Q8 preforms very well
The Narsil bugs are fixed taking the most visible problems away.
So if the bigger springs are used and the ledboard manufacturer makes lamination right we can say it is good and done.
Thorfire understands reliability is important and see the Q8 as a showcase of what they can do.
None of the other Thorfire lights have lamination issues so I am sure it will get fixed right
Agreed, so long as TF does what they are supposed to do and get the proper springs as well as 100% solve the de-lamination of the MCPCB problem.
As far as adding time, I agree; nobody want’s that. :+1: BUT…… whatever time it takes to get it right is time well spent. :+1:
Whether that be a day, a week, a month, or a year.
BUT, with the MCPCB being fine on Proto’s 1 & 2 and then for whatever reason about half were a dud on Proto 3…… it could give the thinking person reason for pause.
hank makes some excellent & valid points here!
While I do not advocate another ProtoType, I totally advocated taking the time to make sure things are correct, whatever amount of time that ends up being. No, the BLF Q8 team is not responsible for this…… BUT hopefully at least a few MCPCB’s and chosen springs will be sent to key people on the Q8 team to evaluate. To me this only makes sense. ymmv
I hear ya’…… :+1:
Springs & MCPCB are the big ones to me. And yep, neither Proto 1 or 2 had MCPCB problem. Then all of a sudden Proto 3 did.
It makes me wonder why?
What if Proto 3 had been the production run :person_facepalming: Would disaster or possibly nightmare be appropriate words??
[EDIT: See Post #8851 for clarification ]
