Double current transformer matryoshka. Amazing German engineering.

[wraper]

This is inside elemental analyzer which AFAIK costs >30k EUR (ELTRA CS580A). Was wondering why there were so many wires coming from that current transformer. I already knew that electronics in that device were designed by idiots but this is something on a whole next level.

[electr_peter]

Toroidal core seems to be used as a component storage space. And a component inside a toroid is a small rectangular transformer
Square peg into a round hole situation. Mechanical design lacks proper mounting solution.

[wraper]

Toroidal core seems to be used as a component storage space. And a component inside a toroid is a small rectangular transformer
Square peg into a round hole situation. Mechanical design lacks proper mounting solution.

Not the main problem though. They put a transformer into a current transformer. I don't know how large effect it has but it certainly distorts current measurement to some degree. Also it's an effing PCB mount transformer with wires soldered on and bent flush. It's an expensive laboratory device in serial production, not a one time job and I've opened 5 of them which all are made like this.

[uer166]

Toroidal core seems to be used as a component storage space. And a component inside a toroid is a small rectangular transformer
Square peg into a round hole situation. Mechanical design lacks proper mounting solution.

Not the main problem though. They put a transformer into a current transformer. I don't know how large effect it has but it certainly distorts current measurement to some degree. Also it's an effing PCB mount transformer with wires soldered on and bent flush.

Why would it, the wires are going in and out the same side so it cancels out like in a GFCI transformer.. The effect is most likely not measurable. Shit mechanical design of course but electrically I don't see an issue.

[wraper]

Toroidal core seems to be used as a component storage space. And a component inside a toroid is a small rectangular transformer
Square peg into a round hole situation. Mechanical design lacks proper mounting solution.

Not the main problem though. They put a transformer into a current transformer. I don't know how large effect it has but it certainly distorts current measurement to some degree. Also it's an effing PCB mount transformer with wires soldered on and bent flush.

Why would it, the wires are going in and out the same side so it cancels out like in a GFCI transformer.. The effect is most likely not measurable. Shit mechanical design of course but electrically I don't see an issue.

Transformer has magnetic field leakage. Placing a large piece of iron plates into a current transformer is not the best idea either.

[bdunham7]

Lazy bastards forgot the hot glue!

[TheMG]

Are those plastic comb bindings being used as cable management channels?

[electr_peter]

Toroidal core seems to be used as a component storage space. And a component inside a toroid is a small rectangular transformer
Square peg into a round hole situation. Mechanical design lacks proper mounting solution.

Not the main problem though. They put a transformer into a current transformer. I don't know how large effect it has but it certainly distorts current measurement to some degree. Also it's an effing PCB mount transformer with wires soldered on and bent flush. It's an expensive laboratory device in serial production, not a one time job and I've opened 5 of them which all are made like this.

I see three things that in my opinion are questionable, but technically are probably fine:

• device is placed into current measurement coil with no positive support, may rattle out. Technically may not be a problem.
• TH (or SMD?) part is mounted directly to wires with bent legs, but insulation and excution is good. TH/SMD inductors are interchangeable in practice with little bending.
• transformer and coil may interact in fields, although it may not be significant. Depends on how important is that coil.

My main objection is to 1) and to the lesser extent to 3). Technically it may be fine, at the moment I do not see hard rules why such thing would not be allowed. Though I would've expected better mounting than square peg into a round hole situation.

[wraper]

Are those plastic comb bindings being used as cable management channels?

Yes

[ace1903]

Does one know why only secondary has twisted wires going to the transformer? 

[Henrik_V]

If the transformer was fixed in any kind (to not leave the current transformer) ... maybe it's abused for some kind of compensation?
The current transformer seems a bit oversized for the current wire actually used otherwise

[wraper]

If the transformer was fixed in any kind (to not leave the current transformer) ... maybe it's abused for some kind of compensation?
The current transformer seems a bit oversized for the current wire actually used otherwise

Considering how the rest of it is designed, IMHO it may only have some idiocy compensation. I don't see any sort of good engineering anywhere in this device. You can see an oven temperature temperature PCB nearly, they use that golden RCA connector for a thermocouple  . Pump driver board failed at least once in almost all of them. Silicon carbide heaters are normally connected only by aluminum braid with a spring clamp placed around. Not here though. They decided it's a good idea to use insulated copper wires. And to avoid wire burning from overheating to pass 10A current through M4 steel screws, which are also heated by the heaters to 200-300^oC. What a nice decision, mechanical stress caused by stiff wires cause deformation of a fragile heater and eventual cracking but screws oxidize and burn out  .
And PSU is a crappy Chinese power brick with fake 330uF 450V electrolytic cap pulled from e-waste (dodgy sleeve, dents on the can, actually 220uF while having normal ESR). Also they glued it onto an aluminum plate instead to just using something like inexpensive enclosed Mean-well PSU without all of this BS.

[Kleinstein]

Using aluminium wire to connetec to a SiC heater would indeed be stupid. Chances are the breaded wire shown int he picture is more like nickle coated copper. Aluminium gets to soft to be usefull at higher temperature, gets poor contact from oxide and may react with the SiC at some temperature (like 500 C).

[wraper]

Using aluminium wire to connetec to a SiC heater would indeed be stupid. Chances are the breaded wire shown int he picture is more like nickle coated copper. Aluminium gets to soft to be usefull at higher temperature, gets poor contact from oxide and may react with the SiC at some temperature (like 500 C).

IIRC braid was soft, so should be aluminum. BTW aluminum is not stupid, it's how it should be done. Aluminum does not oxidize like copper at high temperature. Ends of the heaters are aluminum coated as well. However 2 thick and stiff copper wires connected to the 2 ends of the heaters ruin it all. Picture taken from a similar CS-580 without autosampler and a horizontal oven.

[mag_therm]

I had a couple of projects using silicon carbide at 1200 Celsius, years ago.
The ends are sprayed with aluminum, and must be kept below 250 C.
To do that, one way the manufacturer does it is by helix  machined in the hot length to increase the electrical resistance.
The braided connection tails are of aluminum for contact compatibility and  to allow movement of the ends.
The transition between hot zone and terminal is in the furnace wall.
As I recall, we bolted the aluminum tails to copper busbars. There are grease compounds for Al ~ Cu

[TheMG]

Are those plastic comb bindings being used as cable management channels?

Yes

I guess it's cheaper than Panduit but in a 30k EUR machine... Wow. That the sort of thing I'd expect to see in a DIY hobbyist build not a professionally manufactured piece of equipment!

[coppercone2]

I notice when you mix chemists and electronics you get some bizarre construction methods. I have a micro-pump that (used to) cost high dollars but inside it looks like it was soldered  by children, tangle of wires, etc... real piece of work. I am not complaining about service loops either, I just mean point to point trash, wires crushed under stuff, etc. It was soldered in a tangle too.. I did better work as a young kid with a wood engraver LOL. If I did not get it at 1/60th of the price I would have returned it

However I will say for some reason hot-plates are usually pretty sane, if you stick with the good brands (VWR)

[wraper]

I notice when you mix chemists and electronics you get some bizarre construction methods. I have a micro-pump that (used to) cost high dollars but inside it looks like it was soldered  by children, tangle of wires, etc... real piece of work. I am not complaining about service loops either, I just mean point to point trash, wires crushed under stuff, etc. It was soldered in a tangle too.. I did better work as a young kid with a wood engraver LOL. If I did not get it at 1/60th of the price I would have returned it

However I will say for some reason hot-plates are usually pretty sane, if you stick with the good brands (VWR)

I should mention another German manufacturer called IKA. They thought it's a great idea to put a bicycle brake cable into a bucket of water, connect a plug on its end, and a weak solenoid on another end. This was in C5000 calorimeter produced in 1998-2006. Then they replaced it with additional pump with a brushed motor which was even less reliable. This thing cost around 35k EUR.
Once ordered a new "brake cable" from them for a few hundred EUR. It rusted and got stuck in 2 weeks after replacement. Therefore repair usually was to take the wire out of the shell, clean the rust, lubricate and hope it is not left unused for more than a few days, otherwise it may stuck again  .