ebastlers electronics corner

SISS61DN, about 3~4mΩ.

Edit : you can calculate it from the table since the RPP/noRPP measurements were taken at the same time. (Although there is some trace resistance as well)

4 Thanks

That is exceptionally low, thanks! I did not know that FET. Perfect specs for RPP :smiley:

There is also PXP3R7-12QU, should be pretty close and a bit cheaper.

2 Thanks

I’m working on my wardrobe light right now. I bought 10 meters of 4mm wide COB stripe with 4000K and a CRI of 90 (at least according to specs, I will check with my Opple once it’s done).

Cut 14 55cm long pieces at first, then came the annoying part of getting it into the alu profiles. Those are a huge pain to work with, but with a bit of patience and the use of fine tweezers it turned out to be possible.

The profiles have some diffusing inserts, so I guess I wouldn’t even have had to buy COB stripes, but hey. They cost barely more than regular ones.

Specs say 5W/m, I am getting 2.6W (at 24V) for 55cm so sounds pretty accurate. Still 36W for the entire wardrobe. I will put one switch per door, however, so it’s divided in 4 separately hard-switched segments.

This light, together with some WW/CW + RGB stripes on top of the wardrobe, will be controlled by one of these:

A pretty simple 6ch LED PWM controller with an esp32-c6. Currently running Tasmota, I am hoping to switch to WLED soon for better smarthome integration. Although the wardrobe will be mostly a “set up once, never change” thing, but I like having options. It is supplied through a buck converter directly off the 24V LED input.

A second of these will control my main room lighting and light around the working desk, and that one is also equipped with a whole lot of sensors (humidity, temperature, pressure, VOC, CO2) - not really out of necessity, but if I already have a ESP32 running 24/7, why not.

More pics will follow when I actually make some progress - this is a pretty long running project because I’m way too lazy to tackle it :joy:

3 Thanks

Very cool. I’ve been toying around with an ESP32 here and there (arduino nano version). I’m interested to see what you do with it.

1 Thank

Oh my, a WLED ready 6ch driver! I might finally fix my desk lighting embarrassment of a setup!

Come to think of it, with your openness to feature creep in mind, a header for a wired ethernet extension would make this a full production ready tape driver for installations, would you consider that?

Not sure if it is, last time I checked wled did not support the esp32-c6 yet, but it’s been a while :smile:

Yeah looks like that’s still the case.

:slightly_frowning_face:

Finished soldering all the wires to the stripes (14 total not all shown), hope I’ll find the time to start putting them into the Ali profiles this evening.

1 Thank

Turnigy silicone leads, no expense spared! At first glance they looked like Lapp multinorm singlecores but it’s pricier stuff still.

1 Thank

8 strips done, 6 to go for tomorrow. Getting them into those super tight profiles is a major pain, but less so than I originally feared. After 2-3 I got the hang of it, and was pretty fast. I love how perfectly uniform the lighting of these profiles is, despite their small size.

CoB tapes help a lot in that. Are they high CRI? All I’ve seen in my line of work were ‘just’ 90+

R9070 according to my Opple LM4, and oddly enough 3500K (I already thought as much after turning them on for the first time) despite packaging, reel and shop listing claiming 4000K.

They can be a bit all over the place yea. I think it’s due to the phosphor/silicone application process being tricky to dial in and the manufacturers mostly not binning their tapes since they’d need to test the entire length, not just individual LED packages.

1 Thank

Currently working on a new MCPCB + AUX PCB for a FWAA, intended to be used with the 9V FWAA dual fuel boost by @thefreeman. 1.6mm Alu core PCB, 0.6mm FR4 AUX PCB, and then a TS10 optic on top.

The aux PCB serving also as a 0.6mm spacer should make sure the lens that was designed for the very thin CSP2323 won’t press onto the 3535 LEDs (DD 519A in my case). Since CSP2323 have the LES closer to the substrate than a regular 3535, it should not lead to a too big defocusing of the emitter. Will see just how well it ends up working.

According to a reddit post FWAA MCPCB + lens is 7.4mm thick, TS10 6.5mm. Add 0.6mm spacer, and use a 1.5mm oring instead of 1mm default, and it should work out height-wise.

0.4mm spacer and 1.5mm glass instead of the OG 1mm would be better (should be almost perfectly focused), but 0.4mm adds a lot of cost, and I would prefer to keep the OG glass lens, too.

Huge thanks to @Zwerglein02 for the TS10 MCPCB and lens measurements, and general advise and tips :slight_smile:

3 Thanks

You need resistors for each colors.

Oh, do people usually put them on the aux PCB? I thought they were usually on the driver :sweat_smile:

Good to know, will add.

Oh, does your driver expect common anode, or common cathode, or does both work with a software toggle?

EDIT: I think I will replace the RGB LEDs with 3 individual 0402s each. Downside, I need to have it assembled as soldering 0402 is a pain. Advantage, smaller and more flexible.

Anduril works with common cathode aux LEDs. Here’s one on LCSC : TJ-S1615CY6TGLCCSRGB-A5 TOGIALED | C601684 - LCSC Electronics

1 Thank

Ah, CC - makes sense, since most AVRs can only do internal pull-ups, but not pull-downs for some reason. Not sure if still a thing on avr32dd, but older AVRs often had that limitation. Thx also for the link, that one is smaller than 3 individual LEDs still. Nice LED, wonder why I did not find it. I hate jlc/LCSC part search (and the search hates me).

Back to fixing floodlights :smile:
My uncle’s employees are masters at breaking stuff, apparently. This one was sent to me because “it’s rattling”. Turned out the battery charging buck burned, and the inductor desoldered from the heat.

I thought the battery was dead, as it had 0V. However, when I removed the heat shrink I saw it had a BMS. Connected it to a power supply and it unlocked and went to 3.3V per cell, so seems to be fine. I can only charge the 2S2P pack with 1A at 3.3V idle voltage however, so internal resistance is likely sky-high already. It wasn’t sent to me for a battery replacement, so I’ll leave it as is. I’ll probably say that I can replace the battery eventually if he wants.


I resoldered the inductor and removed the burned chip for now, with my next LCSC order (when I will order parts for the FWAA) I will order a chip for this light and try to repair it. It’s not great, but my uncle’s employees like it, so why not.

1 Thank