Thank you very much.
Iām guessing they have a truck load of left over T-button caps. :partying_face: I did mention to Barry they need to use a smaller (tighter) o-ring. So far the USB cover is still attached to the prototype, but maybe they should include a spare with the light.
Running on 5 chips?? Canāt imagine how hot it gets running on 7 chips.
any ideas on what to try to get USB charging to work?
USB charging fixedā¦. not compatible with my usb C pixel chargersā¦( I have 3)
What is the purpose of the orange o-ring around the head?
probably ding protection if it tips over
two posibilities that I can think of.
One, just the switch LED isnāt working, a thought based on this:
For this just try plugging it in for an hour and see if the battery voltage changes using the clicks.
Two, time to pull the driver out and use an ohm meter/volt meter to see where the open or problem is. Unless you really feel comfortable doing this, I donāt really recommend this one, probably better to return to Sofirn for a replacement.
Probably not a lot of help, but troubleshooting from afar can be difficult.
thanks for the ideasā¦ I did leave it plugged in all day.
3.8 flashes no changeā¦
Looking at the brass ring with the head offā¦ There is a drop of solder where it says āchargeāā¦ ?
Does that have something to do with USB charging?
USB charging fixedā¦. not compatible with my usb C pixel chargersā¦( I have 3)
that sets the charging current. Connected is 1.5A max, open (no solder) is 0.75A max.
Do you have a meter to make electronics measurements?
drat
Thanks again
The LT1ās solder bridges allow the user to adjust the amount of power the lantern uses. They allow adjustment from 3x7135 (1.05 A) to 7x7135 (2.45 A), with the default set at 5x7135 (1.75 A). And they are connected in series, so they need to be connected in the correct order. Without pad 4 connected, pads 5, 6, and 7 do nothing.
The adjustments affect all brightness levels. Changing from 5 to 7 pads connected increases the brightness of every mode by about 40% (7 / 5 = 1.40).
The intended configuration, for both channels, is:
- 3/7: is just a label, does nothing
- 4/7: closed / connected
- 5/7: closed / connected
- 6/7: open / no connection
- 7/7: open / no connection
Since there were a couple reports of lanterns showing up with the wrong pads bridged, Lexel modified the driver design to put the pads in parallel instead of in series. That way, it shouldnāt matter what order theyāre in. This update has been sent to Sofirn, but itās not yet known if/when the update will go into production.
For now, itās recommended that people check the pads to make sure the right ones are connected. The lantern works regardless of that, but it may end up brighter or dimmer than expected, and the brightness might not be consistent between warm and cool modes if the two channels have different numbers of chips connected.
Is the unthreading for a mechanical lock out improved?
Thank You
It only takes a fraction of a turn to mechanically lock it out, about a 1/4 inch on mine.
It sounds like you have an extra chip connected on the āNWā channel. And since they go in order, thatād be the one labelled ā6ā.
About the change in current in the middle, thatās expected. It tries to keep power usage pretty constant while tint changes, but the curve isnāt completely flat. The amc7135 chips have a non-linear response curve, and the firmware tries to compensate for that, but it doesnāt have a way to actually measure the result so itās adjusting totally blind. Because of this, it only gets somewhat close to constant.
Hereās what I measured while doing that part of the code. The orange line shows power usage by tint without any correctionsā¦ and the blue line is how it looks after adding a correction factor.
Before correction, the curve looks like it has something similar to a triangle wave removed, with a magnitude of about 40% of the total power. So to correct it, I added a triangle wave of the same magnitude. The formula is, approximately: PWM = brightness + (brightness * 0.4 * triangle_wave(tint))
Basically, if it wants a total brightness of 100, and that power needs to be divided evenly between the two channels, it canāt just split the PWM signal 50/50. That doesnāt work. Instead, it needs to do roughly 70/70 because 70% PWM works out to about 50% actual power. Here are some other examples:
- Warm only: 100 / 0 (total 100)
- 25% tint: 90 / 30 (total 120)
- 50% tint: 70 / 70 (total 140)
- 75% tint: 30 / 90 (total 120)
- Cool only: 0 / 100 (total 100)
In more detail, the correction curves look something like this: (X axis is tint, Y axis is PWM level, both channels are graphed in the same image)
If itās getting too bright in the middle, that 0.4 correction factor may need some slight adjustment. The value itās using now is based on the performance of a prototype.
ā¦.Iām sure there are a lot of newbs and non - techies who are probably already overwhelmed by all this āsolder bridgingā stuff. And since it was reported that some lights did not come programmed as they were supposed to, hopefully something can be done for us without the technical know how. A close up drawing of what it should look like to make sure it came as it was supposed to would be nice. And then a procedure for increasing the power, if possible, for us layfolk. Perhaps this can go on the OP, or links to Youtube āhow to videosā in time?
many thanksā¦ā¦
Agreed. Itās disconcerting to hear of these quality control issues already.
I was on the list for 3, seems like everyone is getting theres - what am I doing wrong to not be asked to submit payment? Big dollahaz are waiting in my pocket.
Only the first 500 lanterns have been made so far; the rest of us have to wait until the other 2500 are made.
You only seem to be on the list for two lanterns, at slots 987 and 1366. See here:
Your right 2x lanterns (crap ah well)
Ok so they should get to us soon then got it!
Fingers crossed 
(My first lantern is at slot 876.)