The main difference is the driver voltage—it’s a similar design but with different components. They will both work at 50 and 60 Hz, but the US bulbs will only work at voltages found in North America (usually centered around 120V), while the European bulbs will only at European voltages (centered around 230 V).
The US bulbs immediately die at higher voltages. The European bulbs malfunction at lower voltages. Note that we do not recommend even trying this. It’s not safe.
Edit: I should add that they meet UL, CE, etc. so it shouldn’t kill you. But please don’t try it.
Sounds like you might want to check that the parts you’ve been buying are within specification.
Not to say that Chinese suppliers don’t do adequate QA and QC and ship poor products, but … chabuduo reins.
And do they come to you in ESD-protected bags, or just polyethyene?
Also check whether your supplier uses adequate electrostatic discharge control when building components. Zapping a semiconductor doesn’t generally kill it dead, it instead changes the level at which it’s a “semi” conductor — the line between being a conductor and an insulator, the threshold at which it operates, gets altered by electrostatic discharge. It’s been wounded.
One component a little off spec is probably tolerable. Several or a handful of components all out of tolerances adds up to malfunctions like flickering.
Yep, part substitution is something we’re looking into. From what we can tell so far, the components don’t seem to be ESD-damaged, but they may be slightly out of spec. Not sure yet, though.
If they’re consistently out of spec (like say plus or minus ten percent rather than five percent, when +/-five is what you thought you bought), then you may be seeing “”quality fade”:quality fade" at DuckDuckGo in action.
If they’re all over the map, that would sound more like “wounded” (ESD-damaged) components.
I believe that what is happening with the Bedtime bulb is that a large draw on the mains decreases the current to the circuit which is similar to what a dimmer does. The driver sees the decrease in current as a dimmer and decreases output. When the mains stabilizes after the in rush to the motor, the output returns to full. Lamps that don’t exhibit this behavior are likely not “dimmable”.
I don’t know if it’s possible for a dimmable lamp to not react in this way unless it could be programmed to not decrease output unless the current is decreased for >1 second. This would cause a delay when utilizing a dimmer which many users would not like.
We are finding your analysis to be mainly correct. We do believe it is possible to implement a dimmable driver that does not have this behavior, but it will likely require a different topology than what we currently use.
Yep, that someone was me. I discovered this thread by djozz called “SunLike 3500K 95+ CRI midpower led from Seoul Semiconductors tested”, and I said to myself “Hey, Remez builds with the same make and model of LED in 3000k, I should email Adam and see if he can source Remez in 3500k. But nope.
Well the (8) Remez I bought from you are quite impressive now that their tint has been fixed. My rooms take 5 bulbs each, so this is what pricing look like for me:
$35 per room for Remez 7 watt
$105 per room for Sunlike 7 watt
$232 per room for Sunlike 8 watt (with more efficiency)
That’s why I ordered the 3 “spares” to put in my bathroom. I can steal from there if I have some die in the office. Even with bad luck, they are still a better value.
5pcs R.9 would be $46, but you had to cancel that order because of flicker issues on 120v. So R.7 was my only option. And yea, there is a less efficient S.8 option for $111.