Re-flashable hardware, current options

Newcomer to WLED here, looking to build a first project. I have a fair knowledge of electronics and microcontrollers, but no one knows it all.

On googling around, as one does, I found references to the “H801” hardware which is commonly available, based on ESP8266 or similar, and comes with level shifts, MOSFET circuits and a decent set of outputs. It looks good, although the price seems to have gone up a bit since 2017… anyway. As I understand it, it has high current output capacity on 5 channels that can be set for PWM (good for multi colour strips), and the W1 and W2 outputs can be used to drive WS8212/WS8211 addressable LEDS. It case housing, screw terminals etc. and the reflash capability is within my skill range.

I’m interested to know if this is still a good choice? It’s readily available via eBay vendors, for one thing, even if they don’t always mention the H801 branding. Are there any other hardware platforms with similar outputs, a prefabricated housing and so on, that are not too hard to reflash?

Thanks in advance for your thoughts.

Actually, google is your friend again, I did find this Device Type: Light | devices.esphome.io in ESPHome which at least gives me some device names. I would be interested in practical experiences with WLED, though.

Say something wrong on the Internet and the experts will bombard you with the correct answer.

I am not an expert, having the bruises from frequent bombardments.

I have no clue what “H801” refers to. My WLED projects are a mix of devices I made myself using the ESP8266-01 or the Wemos D1 Mini. If you want something “off the shelf”, I have a few Athom LS-4P controllers in use because they are so easy to set up and they already have WLED installed.

The Athom controller is also available cheaper at AliExpress.

On the outputs, you are mixing protocols. What do you want to control?

H801 is this: https://tinkerman.cat/post/closer-look-h801-led-wifi-controller . This, or clones of it, are extensively advertised in the usual places and it seems that they are relatively easy to flash WLED too.

Some project ideas need high-ish current PWM RGB. Other projects need addressable LEDs. I appreciate that these are different protocols, and for that reason it’s not necessarily the situation that one controller would fit all use cases. (However, it did appear that if the H801 has a total of 5 outputs it might be able to control both simultaneously on a small scale). Right now I am in a “discovery” phase of what’s possible.

As a side issue these are for small-scale visual expects on a small stage for amateur drama.

Seems some else has been down this rabbit hole as well:

H801 as a controller

Not the worst choice you could make given the PWM drivers builtin - if you need that.
The potential downsides I see:

  • May be limited to the number of LEDs you could drive as you’ll be using up RAM pretty fast because of the GPIOs used, you hit a limit around 300+ LEDs for a single controller.
  • No OTA updates of the firmware with 1M flash devices.
  • Probably don’t want to try and pass 5A of current through those board traces, though that may not be an issue if you’re doing small(ish) displays.

Much will depend on the actual displays you try and drive.
Personally, I’d add PWM boards externally where required and go for something like the SP511E’s or graduate to the ESP32 based boards out there which will be much more capable and open up the full power of WLED.

YMMV of course if the majority of your displays are analog this might be a reasonable choice.

Follow WLED down the rabbit-hole and a world of possibilities opens up :wink:

Yes, that’s a good link. I also note that you can take the W2 lines and opt not to use the MOSFET. That’s a level of soldering I could cope with. The lack of OTA isn’t a big deal for me, and a limit of 300 LEDs isn’t a problem for what I have in mind right now. Having said that, I still need to look at whether I get can addressable LEDs at suitable intensities… ah the rabbit hole.

I’m money-sensitive, but of course there’s no point in a false economy. So I will follow up on some of the ESP32 options.

One of the “quasi-solutions” for addressable intensity issues is to create your own density of LEDs.
Addressables are available in strips up to 144/m (unless some one has found higher?) and can be pricey.

An alternative is addressable strings which are closer to “Xmas light” style displays at much lower density, often 10/m or 20/m. They still give good effect and control and are available in RGBW which is nice to have for “clean” white displays. They’re also reasonably cheap, enough so that’s it’s worth considering overlaying multiple strings to increase the effective density back up to 30 or 60/m.

It really depends on what you’re envisioning for your stage props or background lighting or…

Keep exploring and try imagining some solutions you’d like to implement, the forum can be a great resource for what has worked and what could work.

Here’s what I have to work with - I am the director of an am-dram (community theatre) show in April. 200 seats, to give an idea of scale. The setting is a 1960’s English seaside town, which makes me think of rows of chaser lights. In 1960 of course these would have been rows of incandescent coloured lights controlled by relays or if you were super high-tech, germanium transistors (?).

Incandescent bulbs are hot, inefficient and coloured ones are harder to get hold of these days, so I’m thinking LEDs. Mains powered domestic lights would do, could be ganged up in say 4 sets and then allowed to simple chase patterns. Arduino + Relays maybe. Some lighting filter plastic would provide fixed colour. But the cost of bulbs and holders does mount up, as does the time to attach and wire them up. And I rather like using low voltage if I can.

If I used addressable RGBs, the wiring could be simpler. I could be clever and use different chase patterns and animations. And use less power. And easier wiring. The lighting isn’t just simple chase lighting but mood setting (wearing my director’s hat) so I can control the patterns on the board. Wearing my technical hat, of course, I can see that pre-built WS2818 and WS2811 strings probably won’t have a big enough impact. I don’t need to compete with all the stage lighting - it can be effective when the stage is dimmed for example, but I don’t want people having to squint and ask what the sparkly things are or it’s a wasted opportunity.

If I have to spend multiple hundreds of eurodollarpounds, it’s out of budget, and I guess I won’t be able to get enough brightness in addressable LEDs. After all, someone has to pay to hire the hall, and clothe the actors… It’s not a stadium prog rock concert.

I could forego the colour changes, and look for a solution based on COB white LEDs and coloured filters, I guess, but the wiring gets tiresome when you want to do chaser effects. Maybe the type used to illuminate stair runs or caravan interiors. I certainly don’t need colour changing for historical accuracy, but then again who says theatre needs to be accurate?

A Non-LED option is to use automotive lights like the 21W/5W which are still availble in bulk here n the UK because generic LED replacements are not yet legally authorised. But pretty hot, fragile glass, and so on.

Ideas appreciated, but anyone is going to post that my requirements and budget can’t be resolved, it won’t be a total shock.

Normally I’d push trying to build some individual elements (and that might still be the best direction)
But for your scenario it might at least be an idea to explore something like:

Tasmota bulbs

Especially now that they’ll come with WLED builtin.
The wiring simplifies in that all you need is AC as comms is WiFi
There are still a whack of questions, like:

What’s the total # will you need ?
Can you reasonably communicate via WiFi with that number
Is it reasonable to build the fixtures to house them
etc. etc etc.

It sounds like an interesting project overall.
Do you have any pictures or drawings of the space and/or your vision for it?

How about a simpler form of the same problem. Let’s say a standalone chaser light display about 6ft / 180cm, raised off the ground by 1ft/30cm. Discrete light sources about 10cm apart, so we’re talking about a “high teens” number of point sources. A typical WS2811 chain at 12V would be easy to do, although 10cm separation is a bit of a push without cutting and rejoining. A 50 led chain can be arranged so the pixels are in groups of 4, giving 12 sources. (4x12=48 and two left over). Maybe two chains of 50 WS8211.

As well as the LEDs, a controller that I could whip up from existing ESP8266 modules I have at the moment, and a PSU with suitable 12V oomph for the current. Off the top of my head I am not sure what that would be, let’s park that.

Cap each cluster of point sources with a simple diffuser. I looked at ping pong balls but they are … inflammable … are they not? But let’s ignore diffuser for now.

Would it be bright enough? I do have a 50 pixel chain on order so I’ll find out next week. I’m thinking it won’t be, but it’s an interesting baseline.

What would the electronics cost for that - obviously there’s mounting/woodworking as well.

I’d want at least two of these, and in my head I’d like four. Let’s start there as a foundation.

In terms of stagecraft I can’t have a lot of obtrusive stuff, it’s a small stage.

This would be much cheaper than that tasmota solution you mention, but i also think it wouldn’t be nearly bright enough.

I would think that a bunch of these in a line would pump out some decent brightness:

12V Led Pixel Module Light – Amusement Park Large Puck xConnect - Your Pixel Store they have them in 9 or 12 pixels per puck (address).

If you wanted a frosted look you maybe could scuff the surface with sand paper.

There may be similar options on Ali Express, as I don’t think you are in the USA.

I did not flash the SP511E, but I played with similar devices. It was - SP108E.
I flashed it according to the instructions from this article GitHub - psxde/sp108e-led-controller
I have two of these controllers. An ESP8285 is installed there with 2 MB of flash memory inside, but a separate controller (STM32F0) controls the CLK and Data lines.
The trick of the modification was that if the STM32F0 chip had its 7th leg pulled to the ground, it would be constantly in the “Reset” state. And now you can bypass the STM32F0, directly from the ESP8285, throw the wiring from CLK and Data to the TM74HC245 transceiver.
After this modification and WLED firmware, everything started up for me the first time. After some time, I began to notice that the Web interface was somehow jerky. Restarting helped for a while. Flashing with different versions of WLED also worked exactly for a while, then again everything slowed down and twitched. Until I accidentally touched the STM32F0 chip with my finger. I realized that you can smoke from him. )))
He was almost red hot! I checked all the soldering points, I made sure that I didn’t close anything anywhere. So I played with him for another day and he completely died.
I took a second such controller, again did everything according to the instructions, double-checked everything a hundred times. Reflashed the controller. After turning it on, it worked fine for about an hour. Later, the STM32F0 started to warm up and the brakes started. This time I did not wait for it to burn out completely, I unsoldered all this confusion and flashed the backup.
I still don’t understand why this happened, but the suspicion is that the STM32F0 cannot be kept in the reset state for a long time.
After that I bought LuaNode32, WeMos D1 mini, NodeMCU v3. Played with them as I wanted and had no problems.
So my opinion is that it is better to assemble an ESP32-based controller yourself than to gut someone else’s ready-made devices.

It is true that a device certified for use with stock firmware may not be reliable if hardware mods and alternate firmware are used. Reports from real experience are valuable, thanks.

This looked interesting and a good fit to my ideas

24V, 4W RGB, 3PIN JST connectors, Capochon housing (oooh sparkly)

This is based on a few google searches, so I’m not saying this is “the one”. Typical alibaba prices are sub $2 each, haven’t looked at the total cost including shipping yet. (to UK, which is where I am)

1 Like