@NickAlexeev Wiki gives much higher values for white LEDs, so there must have been some progress.

Avago datasheets give luminous efficiency values. The first two companies I checked just gave luminous intensity (so including a factor for the beam divergence). If you click on, for example, HSMC-A100-R20J1's datasheet link, you'll get a datasheet covering multiple colors.

For yellow they give 590 lm/W.

For red, 240 lm/W is the best they have.

But this one does for a standard T-1-3/4 package. 145 for red, 500 for yellow and 595 lm/W for green. Not a bad improvement for 15 years work.

@NickAlexeev Very, very important to consider the receptor. Human eye or sensor or camera?

@ThePhoton Seems Avago have finally stolen/learned from Osram-OS

Moved away from Avago 5 or 6 years ago, because their datasheets were too much of a hassle to work with

hm, at home I recently slabbed together some smps like dummy load. basically it switches onto some resistors that dissipate the energy and then there is a filter in front of it that makes it look more resistive like. because of inrush etc. I thought about putting more or less the equivalent of a classical dc load in front of it for regulating that out... any other idea?

@PlasmaHH Depends, what frequency? I did once build a 2W full silicon load. Recently thought about making a 250W one, maybe with my SMU experience one that also does negative resistance :-D

@Asmyldof I dont care much about the frequency. The final goal is to be able to a) have the heat dissipating part far away from the dut and 3) be able to emulate a load of maybe 1-2kW .... and maybe one day VI) feed the power into an inverter and back into the grid

Hmmm I like your thinking

But maybe with batteries

<-- wants floating loads

maybe it can deliver 12/24/48V for batteries that are in your solar/wind power system (and then the inverter part is for free)

I once was present when they tested some 3kW power supply that was going into a huge dummy load... they opened the windows of the lab to get rid of the heat

hmmm

I did tell you guys about the 5kWh of LiFePO4 in my hallway, right?

would have been so much nicer to not waste it as heat but at least return it to the grid...

there was some talking about lifepos recently in here, didnt pay much attention ^^

You can have a feed back with insulation to allow floating operation (to within 1kV or so) but you'd need a buffer at about 20% of peak power

To allow good regulation

I'm building my own 6GHz SMA and SMB cables... fun!

on a totally different note I yesterday tried out the desoldering gun (and just unpacked the hot air station)... the base station is a bit bigger than I expected... and of course after using it for a while it clogged up completely

That shouldn't happen

yeah, well thats probably what you get for buying cheap stuff ;) the small pipe where the solder gets sucked in is only heated for maybe 3/4 of its length, and the back part where it comes out into he glass cylinder is much cooler, and when you suck up too much solder at once, it has a tendency to stick there.

That sounds like a fixable thing

you mean in design or in getting it out?

1. bit of resistive wire and/or a PTFE or PEEK replacement

2. 1.5 hour of time

I like how similar my cables look to the € 60,- ones

I solved it with a piece of solder and one of these cleaner sticks they provide within ~10 minutes ...

YEs... but...

It should not happen

frown

yep, it probably doesnt in the 400€ ones...

doesn't in dave's one

:-P

actually the one I have is the successor of the one dave has, the gun design is almost the same. I am sure it happens in daves ones too if he did what I did ^^

but anyways it is a really useful tool ^^

just the fan is noisy...

maybe tomorrow I will find time to test the hot air gun

wont expect much from 38 bucks...

@PlasmaHH Ahw balls... Only just made the order for shipping your present

@Asmyldof You might think that. I couldn't possibly comment.

@Asmyldof I have a dejavu... last week I said I can only receive it when I survive the camping with my girls... we then cancelled the camping due to severe thunderstorm warning (a wise decision). We try it again today. So, when I survive it, I will be able to receive it ^^

Half the datasheets for LEDs below 1W didn't even have the word lumen once. All mcd this, cd that

My office is moving and somebody binned an Inmac model 8340-1 "power supply". To me it looks more like an isolation transformer. Anybody know what it really is?

Pictures or didn't happen

@PlasmaHH Going to post it in a bit... might make it with the pickup of today...

@Asmyldof My thought was that if you're viewing the LED directly, you'd want an mcd spec. If you're using it for illumination, you'd want lm. But even parts recommended for backlighting seem to be spec'ed in mcd.

@Asmyldof I never timed how long it would take from NL to here...

@ThePhoton dont turn it on, take it apart?

@ThePhoton mcd is for signalling, maybe. although for signalling you can just as easily figure out how little current you want from lumens.

@ThePhoton quick google image search mkes me suspect its an isolation transformer too, or at least some kind of surge protection

But for any LED you'd ideally want to know lumens, and then if they also mention mcd, fine

@Asmyldof You mean indicators? If you're signalling to a photodiode I'd rather just have optical power (W).

@PlasmaHH Going to the shop now, the close one went out of business, so I might be just too late for pickup today

@Asmyldof Why? Who cares about lumens that are shining off to the side and not getting into the viewer's eye?

@ThePhoton Yeah, they are (or used to be) called signal diodes

@Asmyldof don't worry, I can wait, I still have plenty of dull boring black ones ;)

@PlasmaHH It's got a nice heavy transformer inside. That's about all I can tell without taking it apart. Which I won't do until I get it home.

Because no modern LED that needs a datasheet from an actual brand is so weak it's going to be used primarily for indication

@ThePhoton you could measure for continuitiy and inductance... but yeah I would guess its an isolation transformer... something that is missing from my home office "lab"

The back plate says "input: 120 VAC 1.1 A; Max total output 120 VAC 1.0 A"

@ThePhoton the one I found on the interwebs is 8340-5 and rated for 5A in 4.6A out, so sounds like isolation transformers

hm, totally makes me wonder what the fuck they neded that one for...

@Asmyldof High profile customers want all the LEDs they buy to have the same brightness so their displays don't look goofy. So they want a datasheet even for a low-power indicator. Notice all the brightness binning stuff in some of the Avago datasheets?

Yeah I know. Also something they didn't have back then

off to the wilderness, cya guys

My point just is, if you make 10000mcd PLCC2 LEDs, that's all cute, but if all you have in your datasheet is "10000mcd", "This radiation pattern", then I'm going to buy the brand that also tells me that it means it's 5lumen or whatever, because using that LED as an indicator would wash out anything else in the general vicinity of it

Bye, have "fun"

@Asmyldof Human eye will be the receptor.

@ThePhoton someone soldered in the wrong resistor network for one of the LED strings in something we shipped once....

We got the device back later and a report that it's "simply broken"

I think someone spent a day on diagnostics, testing every part of it before we realized... The blue LEDs were dimmer than the green ones.

That apparently constitutes "simply broken"

Hi all :)

@MickLH Japanese customer?

I have changed the above question of mine from opinion-based to fact-based. What do you think of it now?

@ThePhoton Oh damnit! I never realized it was skewed english for "broken in a simple way"

Just kidding no, they are in the USA

@Asmyldof Well the PLCC2 ones were the ones with lm/W specs. It's the 0402 and 0603 and T-1-3/4 ones that are mostly spec'ed in mcd.

@unforgettableid Is the apple watch waterproof?

@MickLH I dunno. Let me google.

@MickLH I've heard of product being rejected because the solder mask color changed from dark green to lighter green between lots. (On a VME card, so solder mask color was visible to cutomer without disassembling the product)

@ThePhoton That's.... I can't deal with client facing operations

@MickLH It's rated as IPX7 and marketed only as "water-resistant", but someone submerged it in a test chamber to 40 m (140 ft) and it worked fine even afterwards.

There you go.

@MickLH They (Japanese customer) just viewed it as us not having good control of our manufacturing process.

@ThePhoton Sounds like they don't have good control of their (human) logic.

Did they literally not consider that maybe... you have... more than one stencil printer?

@unforgettableid Now that is reason enough for me to do inductive charging. Good luck exposing power pins to water without one electroplating the other.

@MickLH Would the electroplating happen even if the power pins were exposed to water at a time when the phone was not being charged?

@MickLH Yes, but we should have specified the same soldermask material to both of them.

@ThePhoton I still think they are a little overboard, unless you guaranteed that kind of consistency

(Although I'd just replace it quietly, from a business point of view)

@MickLH We were willing to do it to keep a customer. This was a business with only 3 customers in the target market, so if you lose one of them, it's a bit painful.

@unforgettableid "Not really." (Technically yes but it can be made utterly trivial.)

But, you'll almost surely end up with plugs full of water.

And this is after you do a lot of mechanical engineering to design a plug that is both good and watertight in the first place.

@MickLH I don't understand.

@MickLH Which plugs would fill with water?

The ones that go underwater.

Also another issue I can see is heat

@MickLH So you mean the charger plug? The plastic device with metal pins which are designed to connect to the watch?

I mean the female receptor on the watch itself

The surface tension will lock a drop of water inside, if one makes it in somehow

And if the watch is wet, and you plug it in, you'll almost surely drag one in

But lets move on from that speculation, and go on to another speculation: I would keep the charging rate very low to avoid uncomfortable heat on the user's wrist

@MickLH Well, if people do silly things (such as charge a wet watch), they might get silly results (such as a damaged watch).

@unforgettableid People do silly things.

Like buying apple products for full price... ahem

:)

@MickLH Fair points, both about heat and about doing silly things.

@MickLH I, by the way, do not own any Apple watches, phones, tablets, or PCs.

Oh they aren't so bad, except on your wallet. They are that bad on your wallet lol.

@MickLH You do get very good after-sales service; plus maybe free OS upgrades, which is not something which Microsoft traditionally has offered.

Microsoft, on the other hand....

I love ambiguity, I never meant to finish that btw :P

:p

What they really need to do is make a spinning top with a magnet inside, that you can use to charge by spinning it on a little groove where it looks cool floating on your wrist

@MickLH :)

Efficiency of flywheel storage is not too bad actually, but the energy density might be too ridiculous

how many amps are coming out of a residental 120VAC 15A receptacle? I'll be embarrassed if its 15A

got it

@MickLH My neighbour told me about a van they used flywheel storage in, late 70

'late 70's

Guess what?

Cornering wasn't particularly easy

>.<

Doofi forgot one very important design spec

>> Flywheel direction and location are, like... important <<

@unforgettableid Also, I bet every mm is covered in that circuit. A connector would only take away real estate.

This fellow claims 20 years of real-world experience. At the same time, he asks this kind of question.

He must have had excellent staff, if he had manage to avoid power supplies for 20 years.

@NickAlexeev Can I upvote that observation?

I didn't want to comment..

Well, I actually did, but I didn't do it

Re the LED, on balance yellow is your best bet

On account of the curve of perceptibility

Naits