EEVblog 1572 - Haefely 16kV ESD Tester TEARDOWN + Repair Part 1

[EEVblog]

Teardown and Part 1 repair of a Haefely Trench PESD1600 16kV ESD tester.

Unboxing and testing: https://www.eevblog.com/2023/10/13/exclusive-haefely-esg-gun-unboxing/

[StAngus]

Hey Dave,

took me about 5 minutes to find your part:
https://www.ebay.de/itm/325838697537

As the text on the PCB was in german, I figured I should search for "Drehimpulsgeber "

Hope it helps you. If they don't sent to Australia, I could arrange that for you.

Best regards

[BrianHG]

Hey Dave,

took me about 5 minutes to find your part:
https://www.ebay.de/itm/325838697537

As the text on the PCB was in german, I figured I should search for "Drehimpulsgeber "

Hope it helps you. If they don't sent to Australia, I could arrange that for you.

Best regards

Website with datasheet:
https://www.buerklin.com/en/p/ebe/motion-sensors/48431-03103/75E480/?segment=b2c
Note that only the German datasheet has the pinouts & internal wiring.

[EEVblog]

took me about 5 minutes to find your part:
https://www.ebay.de/itm/325838697537
As the text on the PCB was in german, I figured I should search for "Drehimpulsgeber "
Hope it helps you. If they don't sent to Australia, I could arrange that for you.

 
Awesome, thanks!
You are the only one that found it out of Twitter, the forum, and now this video.

[f4eru]

Alternatively, you could very well take the corresponding Ak contact, put a logic inverter, and send the inverted recreated Rk signal to the circuit.
(perhaps add a litttle R/C to avoid recreating bounce)
the crazy encoder scheme was probably used because the software was to slow to reliably decode a standard A/B encoder.
With this scheme, if you miss steps, it just slows down, and does not move around randomly.


If you replace the encoder, please take apart the old one, it probably has some nice small german mechanical clutch magic to generate a separate signal per direction.


Company still exists, makes encoders, but does not seem to make this crazy modell any more : https://ebe.de/en/hmi.html

[Wollvieh]

Website with datasheet:
https://www.buerklin.com/en/p/ebe/motion-sensors/48431-03103/75E480/?segment=b2c
Note that only the German datasheet has the pinouts & internal wiring.

Wow, 94 Euros for a device that hardly works 30 years.

[BrianHG]

Website with datasheet:
https://www.buerklin.com/en/p/ebe/motion-sensors/48431-03103/75E480/?segment=b2c
Note that only the German datasheet has the pinouts & internal wiring.

Wow, 94 Euros for a device that hardly works 30 years.

Well, I know I'm not the engineer at Haefely who decided to use the mechanical rotary switch component as optical rotary knob encoders with steps option existed at the the time offering millions of guaranteed steps at less than half the price, new.  But I guess that ESD gun is a bottom of then end cheap disposable device where every penny of engineering was saved instead of lifetime sturdiness.

Maybe it was something else more stupid during the design phase.  (Looking at the part data sheet, I have my guess, but I am not saying...)

[Nifty]

Wow! That's the speed!

[newbadboy]

You should build one as you mentioned would be a great tutorial.

Ive just built one my self, it has some stuff that can be improved but it works'

https://sp-engineering.se/Projects.html

[f4eru]

Cool, thanks

How do you control the charge voltage ? no feedback ?

I would add a discharge path for safety
and a reversible polarity, like in this haefely example (you reverse the generator, or just the HV diode, then insert it upside down):

[Khaveer]

Idk if this is due to the lack of proper grounding, but the pulses in the unboxing video don't look like 16 kV. I did some ESD pre-compliance test back at my previous job and I remember the air discharges going through over larger distances that what you've shown in the video.

[thm_w]

Well, I know I'm not the engineer at Haefely who decided to use the mechanical rotary switch component as optical rotary knob encoders with steps option existed at the the time offering millions of guaranteed steps at less than half the price, new.  But I guess that ESD gun is a bottom of then end cheap disposable device where every penny of engineering was saved instead of lifetime sturdiness.

Maybe it was something else more stupid during the design phase.  (Looking at the part data sheet, I have my guess, but I am not saying...)

Just laziness or naivete no?
The datasheet claims "Rotary pulse generator for quick and easy setting of digital values. • High reliability thanks to gold contacts. • Inexpensive and space-saving solution for quick adjustment"

So I'm going to assume it was somewhat reasonably priced at the time it was designed in.

Idk if this is due to the lack of proper grounding, but the pulses in the unboxing video don't look like 16 kV. I did some ESD pre-compliance test back at my previous job and I remember the air discharges going through over larger distances that what you've shown in the video.

I don't think he's grounded the DUT to the gun at all, so it won't give a proper discharge. But 16kV is about 5mm gap in air, which seems right.

[thm_w]

You should build one as you mentioned would be a great tutorial.

Ive just built one my self, it has some stuff that can be improved but it works'

https://sp-engineering.se/Projects.html

Great work.
Now that its built though it seems the case could be half the size. Unless you need the airgaps for voltage tolerance..

[f4eru]

he air discharges going through over larger distances

Depends on the geometry of both sides.
With rounded surfaces, you get the high theoretical breakown voltage of the air,
With angular surfaces, the field concentrates highy at the angles, and breakdown occurs much ealier.
you can have 10x longer sparks using needles on both sides.

[tszaboo]

Website with datasheet:
https://www.buerklin.com/en/p/ebe/motion-sensors/48431-03103/75E480/?segment=b2c
Note that only the German datasheet has the pinouts & internal wiring.

Wow, 94 Euros for a device that hardly works 30 years.

Well, I know I'm not the engineer at Haefely who decided to use the mechanical rotary switch component as optical rotary knob encoders with steps option existed at the the time offering millions of guaranteed steps at less than half the price, new.  But I guess that ESD gun is a bottom of then end cheap disposable device where every penny of engineering was saved instead of lifetime sturdiness.

Maybe it was something else more stupid during the design phase.  (Looking at the part data sheet, I have my guess, but I am not saying...)

My guess is that 99% of the effort went into the ESD part, designing an injection mould and making it safe. And 1% of the effort into the assembly user interface and everything else. Could be that it was done by physicist, and not EEs, or EEs that specialize in something else.

[newbadboy]

Thx.

Yea that's true. I planned to do the HV part myself and not use a module. Therefore the casing was quite big. On the other hand as you mention big space and distance is god for isolation