I suppose you don’t have the original power source, if you did then measuring the output would be my starting point.

If it’s LED then it’ll probably be a single LED using PWM. In that case try figuring out where the switching circuit is.

If it’s not LED, you should measure resistance between the wires, that would tell you which one is the common. .

If you’re designing for production then talk with the assembly guys, they will have options and cost estimates.

If we’re talking a one of, then just try. It’s only 500mm so not that much can happen. Especially at the slow speeds you mentioned. 20kHz isn’t that fast compared to the +100MHz that a cat5e is designed for when using differential signalling.

Are plugs a must? If not I’d consider soldering the wires. At least for testing the cable’s suitability.

If you must use plugs, then don’t use rj45. Some yahoo, like yours truly, will come along and go “neat, an undocumented ethernet capability” and fry something expensive. Have a look around DigiKey or where you get your parts, and find a less common plug.

And if you end up needing more wires, then get an old parallel ATA HDD cable from the early 2000s. They have 80 wires and are free at the dump.

Don’t lick, that’s dangerous, but nobody said “don’t eat 60/40 solder”. So go for it. Just remember to put a few sheets of paper in the bowl before taking a dump, otherwise you might break it.

I whole heartedly agree with you on the virtue of simplicity in design, and this heater is a bit over the top for its basic function. But we wanted something that

1. Could do 400V
2. We could control remotely
3. Was available for sale in Denmark in case if an RMA

and at the time that narrowed it down to either the Clea or neo series from adax, and I believe one other product line from a second manufacturer. The remote control part was really hard to find in 400V.

A relay would do the job, but they also wear out faster and the sound is noticeable.

While writing that, I found myself thinking about how significant the life span difference is. So I did some back-of-the-envelope calculations.

The defective heater had been installed for roughly 2100 days. Sometime during April we shut the heaters off, and they’re not powered up again until October, so let’s call it 50% of the year. The schedule for the heaters is usually two power on cycles daily. Again let’s call it 50% of the time. The house is pretty OK insulated, but let’s say that the heating element is powering up 4 times each hour. In reality it’s probably closer to 25-30% and twice an hour, so I’ll use that as the lower bounds for this calculation.

1050 days is 25200 hours, meaning that the heaters will have done between 12k and 50k power cycles in the almost 6 years of operation. I’ve seen relays weld shut before 10k, but MTBF for suitably rated relays on RS starts at 50k power cycles.

So maybe a relay could actually do the job. But then there’d still be the issue of the noise, which, while negligible, could still be annoying for some, especially in a bedroom.

Wow, thanks, that is much appreciated. I was trying to find a SoMe channel. And knowing that Adax isn’t averse to support is even better.

I could rewire the heating circuit, but doing so without the backing of an authorized electrician would be illegal. I’m already on thin ice just installing the heater by self.

But the thought has crossed my mind, as I have noticed there being an order of magnitude more 230V heaters on the market than 400V. On top of having to rewire everything, I’d have to put in a new 3 phased breaker, otherwise I’d be pulling way too much current… Let’s just say that this 1200W heater isn’t the only one we have in operation in my house.

But you managed to pack my thoughts into 3 sentences. Unfortunately that means a new heater, because I don’t have a source for the board.

I was thinking about contacting the manufacturer, but they don’t really have any post sale customer support. That’s all on the stores, and besides the heater has been in operation for almost 6 years now. So I’d they’d tell me to go away at best.

I don’t know what stopped it. I wrote the occupant, but it was really just me. Neither RCD or fuses were blown. There’s a fuse (bimetal in a plastic box positioned above the heating element inside the heater) in series with the heating element, but that’s 400mm from the short, and its probably only there to prevent overheating and setting clothing coverie the heater on fire.

A new heater is 200€, because it has to be 400V as there’s no neutral wire available. That’s why I’d even entertain the idea of repairing the board.

Is the terminal emulator set up correctly in regards to baudrate and stopbits? I would try playing around with these settings.

I haven’t used Dell gear before, but you are sure that it’s not some proprietary protocol that’s running over serial? As in it requires some special software instead of a console?

I’m not sure what you mean by the signals being dirty. If it’s the slope after the initial fast rise you mean, I guess it could be remedied with an opamp, a push-pull pair, an optocoupler, or some other switching circuit with a suitable power capacity. But as long as it gets above the hysteresis threshold for long enough it should be fine. Whether that’s happens you should be able to tell from the scope (which I can’t see while writing in my client)

It’s not a matter of whether your meter can do 240V, can you do 240V? Because that can be the consequence of slipping with the probe.

Sorry, I don’t mean to come off all aggressive. I just want to make sure that you don’t hurt yourself because I wasn’t very clear in my communication.

Didn’t really understand, can you explain it a bit more? Sorry 😅 About your warning is it dangerous to measure a 230V current with my multimeter (with precautions, gloves mainly)

Depending on the meter and your approach it can be either safe or lethal. The more I think about it, the more I’d recommend that you just don’t do it at all. Working on 240V installations under voltage, that is something we teach electricians to do safely. Besides the training it involves some special and rather pricey equipment.

I cannot stress this enough, if you mess up, you will hurt yourself with enough severity that the can easily be fatal. If you don’t know what you’re doing, messing up will be exponentially more likely.

Desolder the transistor, does the rest of the board light up? If so replace the transistor. Otherwise, get trained assistance on person or replace the UPS. I’m all for learning electronics, but not on a live UPS.

1. Almost there is definitely a problem. It was working fine, but sometimes it would just not go to batteries and stuck in a fault with continuous beep. After that, I let it for 2 weeks, the batteries were 5V, but I tried to charge them and for now they went to 12.6V stable (as normal…), but even with that it does the problem described and does not works as before

That does sound like some part of the controller electronics has been dieing for a while, and has now finally keeled over.

3. I’ve tested the resistors, all proper to their written value, and relays, proper resistance too (85 ohms for one (just on the brown power plant cable), and 260 for the others)

You wrote originally, that you tested what you could with desoldering anything. Measuring resistance in circuit always renders a murky result.

The brown wire is likely what is known as phase or live. Blue will be neutral. But measuring resistance on the input only tells you how much current will flow in the present state of relays.

What could help you come closer to an answer is following the first law of troubleshooting “thou shalt check voltages”. With a device that operates two live rails this will not be both easy and safe at the same time, so don’t rush it. I suggest you figure out what voltages to measure, then solder wires to the relevant nodes. Terminate the wires in a terminal block, where you’re protected from touching the screw. Assemble the device as best you can with all the wires coming out, and then power it on, get your voltages and deduce from that.

Whatever you do, do not poke about a live UPS circuit with your multimeter probes directly.

My initial working theory would be that the transistor switching the fan is dead and I would be looking closely in that area. If that transistor has failed closed and is shorting the controller electronics VCC rail, that would explain a lot. Maybe I’d go so far as to test it without checking voltages (gasp!)

If you just desoldered the transistor and powered the UPS back on, that could give you a UPS with functioning display and an idea of what to replace.

Will try to find others batteries to try, but normally it charges batteries by lighting up the display properly + the fan does not start. And here the UPS is not connect to any AC power, only the batteries were plugged

If 12.6V is normal for your batteries, then there’s no reason to troubleshoot that route any further. 12.6 is little low for lead acid, but if it’s a different chemistry. Anyway keeping extra batteries is always a good idea.

1. Has it at anytime worked like you describe?
2. If it stopped working, what happened?
3. What have you tested, how, and with what results?

12.6V is low if it’s 12V lead acid. Could be that the UPS is just charging the batteries and attempts cooling them meanwhile.

I think tinkercad is a great place to start. It’s browserbased and doesn’t cost money. It may not have a lot of components, but it still have quite a few more than just resistors, caps and inductors. And it is aimed at newbies and hobbyists, which is reflected in range of the available components. Being able to drag an Arduino into your sketch and have it run your program is neat.

We used to use yenka, when I taught electronics. It was OK for teaching, but I don’t know if I’d recommend it for self-paced learning as a hobbyist. It costs money, requires software installation and is so much more than just electronics that navigating the program can be difficult. And default settings explodes components when you put too much current through them, that alway annoyed me.

So all we have is an unsourced entry in a list based article in a unnamed magazine?

I hope I’m not alone in having an issue with the statement then.

I’m sorry that I’m going to sound like a snob. If people claim to have health related information, I take it seriously. When I said I couldn’t find an article describing the correlation, what I meant was that I couldn’t find an article on pubmed. I went looking for abstracts for articles in peer reviewed medical journals, not pop culture entertainment in “octogenarians’ quarterly”.

Source? I can’t find an article linking PTFE with a metabolic change.

I can’t speak for Germany, but I guess that it won’t be much different than Denmark where my experience lies.

Almost all my recommendations has been mentioned before, but I’d like to point out:

• RS in my experience is reliable about 90% of the time. If you put in an order in can arrive in several shipments. I ordered 50 pieces of something got 46 from a UK warehouse and 4 from Germany. That’s another thing, some stuff ships from Frankfurt am Main, so no waiting for imports.
• Reichelt is good for cheap instruments. I’ve never used them for components, I prefer RS for that, but for instruments they carry some of the Chinese stuff that you can’t get other than on ebay or ali.
• Digikey stocks so much electronic components that sometimes you can’t avoid them. Where RS is really fast, their stock and prices are more suited for prototyping than production. Digikey is where you can get the specific series of cap and value you need for your repair project. Digikey will handle all the import and customs stuff as well.
• https://el-supply.com/ is new to the post. They’ve been in business since the 70s, but is primarily targeted towards Denmark. But they do have an international site. Some times they can be extremely cheap compared to RS and Digikey, especially with regards to stuff used for teaching electronics. I’ve been a customer since the late 90s privately, and I’ve placed several orders each year professionally since the early 10s as well. Never been let down.

Thanks for the update. In that case I would be looking at terminal block plugs. They can easily handle the current, they come in a wide variety including panel mount, and they’re cheap. What’s even better, they don’t require crimping and one I saw didn’t even use screw terminals https://www.digikey.com/en/products/filter/headers-plugs-and-sockets/370?s=N4IgjCBcpgHAzFUBjKAzAhgGwM4FMAaEAeygG0R5YB2BAFhCKtvgFZHKaBOMABg%2BY8ITbmABMA0QxHUuYhgF0iABwAuUEAGVVAJwCWAOwDmIAL7mgA

There’s also these that are rated for 300V 15A https://www.amazon.com/5-08mm-Terminal-Connector-Pluggable-Connector-Phoenix/dp/B0CM3Y1WS5/ref=mp_s_a_1_1?sr=8-1

I use something like this when repairing screw terminals that my students have applied too much torque to and broken the through hole pins on. It’s a great type of connector. But if we’re talking portability, then it’s maybe not the most ergonomic.