Open HV Probe 40kV

[beanflying]

It is really simple SHOW ME EVIDENCE of the material or process not being suitable.

Engineers are used to it being the other way around.

Which is why it is up here and being tested against what available evidence there is and open for others to have their input. Science and Engineering are not based on who can hurl the loudest and most frequent abuse or post non evidence as fact.

PLA being a relative newcomer is worthy of additional scrutiny and as I have mentioned before there is always ABS which is a more known material.

[kony]

Defend your design, not your ego && faith, OP. The burden of prooving the due diligence being done for the design is on your side.

If you provide something that ressembles design calculations and documentation, I am fully willing to provide more detailed constructive feedback, but right now I don't even see any signs of understanding of the fundamentals.

[HighVoltage]

Nice design !

I would not worry too much about the insulation material. I have seem commercial HV probes with PVC insulation tubes and they are just working perfectly and are safe. Almost any good insulator material would work, as long as you keep the distance to your high voltage spot large enough.

[joeqsmith]

Some time ago, we had bought some sort of plastic to machine some parts from.   These parts were used in a high voltage application.   Once they were made and installed, we noticed that the system was having a problem.  The new plastic parts were presenting a decent load.   I pulled the remaining raw stock and sure enough, it was fairly conductive at DC.   The properties of the plastic chosen wouldn't have suggest it would behave this way.  In the end, we bought another slab of the same material and it was fine.   I never took the time to dig into why that one lot had a problem.  Something in the mix.   I can only think of one case like that in my entire career.     

Most of the commercial probes similar to what is being shown will state in their manuals that they are for CAT I only.  For my own personal use, pretty much everything it CAT I and very low energy.  Fairly low risk for the most part and certainly something I am willing to accept.

Back to the scheduled program...

[beanflying]

Defend your design, not your ego && faith, OP. The burden of prooving the due diligence being done for the design is on your side.

If you provide something that ressembles design calculations and documentation, I am fully willing to provide more detailed constructive feedback, but right now I don't even see any signs of understanding of the fundamentals.

Plastics vary electrically even engineered or injection molded ones and generally have upper and lower limits. As such no I have not done definitive calculations, in particular as there is limited data on PLA would fairly pointless. I have not seen upper and lower limits even for Engineered PLA just 'typical' or Calculated from Experiment. So Empirical evidence is what I am basing this design on using what limited data has already being linked in this thread. As I have said happy to look at EVIDENCE or Data if anyone has any.

Moving along again as I am not a huge fan of drumsticks and random lumps of wood from the corner of the garden shed with nails in the end  A first pass of a High Voltage Corona probe. Lanthanide Tungsten (DO NOT use Thoriated ones!)  Electrode, Ceramic insulator, HV rated cable. I have gone this way to increase physical separation to the hand to the HV and any resulting Arcs. Base of the Electrode is 300mm from the Hand.

And no you can't eliminate the second long barrel and shorten the design they are different Thread diameters.

WARNING Even if you are a Dumbass don't play with Microwave Oven Transformers with any stick or probe. Unregulated MOTS Kill !

[vk6zgo]

It is really simple SHOW ME EVIDENCE of the material or process not being suitable.

Engineers are used to it being the other way around.

Maybe in modern times, but Engineers were, in the past, very much in the forefront of finding what materials were most suitable for such use.
This was mainly using real world tests, so most of the evidence of "what works" comes from as much empirical testing, as from any knowledge supplied by the manufacturers of the materials.

[vk6zgo]

Some time ago, we had bought some sort of plastic to machine some parts from.   These parts were used in a high voltage application.   Once they were made and installed, we noticed that the system was having a problem.  The new plastic parts were presenting a decent load.   I pulled the remaining raw stock and sure enough, it was fairly conductive at DC.   The properties of the plastic chosen wouldn't have suggest it would behave this way.  In the end, we bought another slab of the same material and it was fine.   I never took the time to dig into why that one lot had a problem.  Something in the mix.   I can only think of one case like that in my entire career.     

Most of the commercial probes similar to what is being shown will state in their manuals that they are for CAT I only.  For my own personal use, pretty much everything it CAT I and very low energy.  Fairly low risk for the most part and certainly something I am willing to accept.

Back to the scheduled program...

There are many interesting properties of various materials, which sometimes seem counter intuitive.

An FM transmitter broke a porcelain standoff insulator in the driver stage.
The real part was forever away in Norway, so I dug around & found a piece of what looked exactly like  PTFE, & fashioned a new one.

Turning the Tx on, it came up OK, but after a few minutes dropped in power markedly.
On investigation, the insulator now looked like a "Cocoa Pop".
Whatever it was, it wasn't PTFE!

At home, later, I found a suitable porcelain insulator in a box of some National HRO receiver parts, so the beast was returned to service.
(A short interval for vintage Radio fans to burn me in effigy! ).

Cad plated steel screws usually don't fare well in RF environments,(both MF & VHF), & there are reports  of Nylon getting quite hot in the same environments.

[beanflying]

It is really simple SHOW ME EVIDENCE of the material or process not being suitable.

Engineers are used to it being the other way around.

Maybe in modern times, but Engineers were, in the past, very much in the forefront of finding what materials were most suitable for such use.
This was mainly using real world tests, so most of the evidence of "what works" comes from as much empirical testing, as from any knowledge supplied by the manufacturers of the materials.

Some of us also went through the Victorian Technical School system before it was Kennetted  so we learned a large bunch of practical hands on Engineering method as well as Maths and Science before getting to Uni.

[vk6zgo]

It is really simple SHOW ME EVIDENCE of the material or process not being suitable.

Engineers are used to it being the other way around.

Maybe in modern times, but Engineers were, in the past, very much in the forefront of finding what materials were most suitable for such use.
This was mainly using real world tests, so most of the evidence of "what works" comes from as much empirical testing, as from any knowledge supplied by the manufacturers of the materials.

Some of us also went through the Victorian Technical School system before it was Kennetted  so we learned a large bunch of practical hands on Engineering method as well as Maths and Science before getting to Uni.

Yeah, the WA Tech School was good value, too.
In fact, I think that was where I saw the EHT probe machined out of solid Perspex that I referred to earlier.

The main "Perth Tech" that I went to was an old building in St George's Terrace, Perth.

To get to the Electronics section, you walked through that building, down a lane between a bunch of Ex WW2 transportables & some weatherboard buildings, where all sorts of stuff went on, from training cooks, to running a massive steam engine for the guys doing "Heat Engines 1 & 2".(Steam would come out of the gutter)

In another building there was "hands on" use of a fair sized Radio Transmitter.(We did Semiconductors Lab in the same building, & a uAmmeter carefully set up to read base current would go full scale if the guys next door fired up the Tx)

At the other end was a red brick building, in the sub-basement of which Electrical Engineering had a large quantity of rotating plant, & ancillary test equipment, again, "hands on" for students doing the relevant courses.
From there it was  up two flights of stairs to where we had our lectures.

You really felt like you were learning stuff there, none of this "monkey see, monkey do" crap!

[joeqsmith]

WARNING Even if you are a Dumbass don't play with Microwave Oven Transformers with any stick or probe. Unregulated MOTS Kill !

Of course the cap can get you as well.   Just a quick search, many are attempting repairs, one some guy burning wood and one a kid trying to replicate a YT video.   

https://www.wkyc.com/article/news/local/lorain-county/teen-dies-from-homemade-science-experiment/145431142
https://www.northjersey.com/story/news/passaic/passaic-city/2017/08/24/passaic-man-killed-after-being-electrocuted-while-fixing-microwave/596748001/
https://www.union-bulletin.com/local/local-man-electrocuted-using-dangerous-wood-art-process/article_daf59a2a-3420-11e7-8508-c7eb2f3dc175.html
https://www.brownsvilleherald.com/news/local/man-fatally-electrocuted-while-repairing-microwave-oven/article_d0749396-3b1a-52f3-9201-65bce57008b3.html
http://www.walb.com/story/19823265/update-deadly-microwave-was-not-plugged-in/
https://dailyegyptian.com/21717/archives/student-dies-in-microwave-electrocution/
https://www.cdc.gov/niosh/nioshtic-2/20026141.html

[beanflying]

  So many needless ones and that is just the tip of the list. Lichtenberg wood burning is just asking to die unless you are very knowledgeable and well set up. For those that haven't seen it done High Salt content water and HV

OT but on the grounds it is my Topic. NST Jacobs ladder Saddle (will suit any 4" wide units) and slide adjustment for the electrodes. 50+mm minimum Vertical separation to any metal parts. Two small inverted heatsinks to make sure any heat is kept from the PLA mounted on two pairs of sliding mounts. Could easily be mounted on a board away from the NST too. About 1/2 way into the main saddle on the little Ender Pro.

The Corona Probe version is ready for a test too but I am tied up for the day but will get to it later maybe.

Edit Added photo. Starting to get the bits arranged for the inner of the voltage divider probe. Ohmite Resistors as discussed earlier and Tinned Copper sleeve.

[beanflying]

24 Hours of print time later the NST Jacobs Ladder Saddle is finished, tested and works as planned. I did try attaching on the front fin to Earth but didn't get any current reading which is to be expected as it has more isolation due to the ceramic insulator at the tip and increased distance. The slides on the saddle allow for shorted to 20mm+ of gap which won't jump without the probe.

Nothing new learned in particular just a few more safer/nicer bits than I used to have added to the kit. I will be stripping the probe shown down for the same epoxy coating treatment as the divider one. The NST Saddle will likely get the same treatment and time to repaint the NST outer shell.

** All Test Gear in the shack was unplugged during this playing.

https://youtu.be/Qx7eH9snu3U

[Helix70]

I work on MV switchgear products for the outdoor market. We rapid prototyped a housing using fdm printing using abs. It's dielectric withstand was terrible. It isn't the abs that's the issue, its the partial discharge due to the air trapped in the print.

I certainly would not hand hold a probe that was printed to the same dimensions as a proper vacuum molded silicon HV probe ( MV actually), I wouldn't trust it.

My two cents.

[beanflying]

I work on MV switchgear products for the outdoor market. We rapid prototyped a housing using fdm printing using abs. It's dielectric withstand was terrible. It isn't the abs that's the issue, its the partial discharge due to the air trapped in the print.

I certainly would not hand hold a probe that was printed to the same dimensions as a proper vacuum molded silicon HV probe ( MV actually), I wouldn't trust it.

My two cents.

Do you have any figures on that? Were you using a Hipot tester or what for your testing would be interesting? Thanks 

Sort of relevant too if you believe it was air causing the issue was your enclosure printed with walls thick enough to allow infill? As shown a page back the barrel tips have virtually zero infill until well back from the tip other than any under extraction (poor setup) that may occur.

[Helix70]

We have a 110kv capable transformer/capacitor in a chamber that we use for power frequency and partial discharge. We also tested at our lab in Germany for BIL of 170kv. Basically our magic number is 50kv for power frequency testing, and the partial discharge was really bad. Due to the size, our infill was not 100%, so yes that is not good, but even 100% infill will lead to bad partial discharge performance. That's why everyone in the industry uses silicon. HV stresses are just not good.

We ended up flashing over and blowing a hole in the side of the printed housing at about 75kv, but that's another story.

[beanflying]

Thanks for the extra information it's fairly hard to come by and first hand

Went and got my fairly ordinary Victor out of storage which I should have done before getting started on this project. The sections as tested have a chunk of 40% infill in them and are from the probe I was playing with in the video above. 5kV test on 200GOhm range.

Anyone with serious test gear and a printer and who is interested in punishing some bits PM me. Or even anyone in Oz PM me and I will post you a printed set to abuse.

Edit some dummy forgot the picture. 200+ Gohms at 5kV

[PartialDischarge]

and the partial discharge was really bad. Due to the size, our infill was not 100%, so yes that is not good, but even 100% infill will lead to bad partial discharge performance.

for MV or HV use things have to be pressure casted or vacuum filled, no other option

We ended up flashing over and blowing a hole in the side of the printed housing at about 75kv, but that's another story.

That's to expect. That's why I pointed in an earlier post that the kind of probe designed here would be ok only for low energy sources and to be used for small periods of time, because partial discharges are no problem for short term use, since it takes months or years for faulty materials to fail due to partial discharges.

That's why everyone in the industry uses silicon. HV stresses are just not good.

Silicone is mostly used in outdoor bushings for its superiority in non-stickiness to dirt, which is a problem in porcelain, and can not be casted as silicone rubber, when grading rings or other components are present inside the bushing.
By volume however I'd say that most MV 'stuff' (interior and exterior) uses epoxy.

[joeqsmith]

Have you made any further progress on your probe?   

[beanflying]

Life has been in the way.

Will assemble the Divider insert and take some pics of it as I have all the bits needed. The Probe front end itself is now epoxy coated and I have a few more Rolls of Black so I will reprint the handle end as I have tweaked some of it internally.

[ArthurDent]

Here are some photos of the old HV probe I mentioned in post #34 to show the construction details. In the photo that shows the metal shield on the handle you can see that the address has no ZIP code, which I recall happened around 1962, so that somewhat dates the probe. I believe the rating of this probe is 20KV.

The first two photos show the various parts and will give you an idea how they fit together. The shield of the probe cable securely contacts the brass tube that slides inside the entire handle. The small spring goes in the hole near the front of the handle to contact with the brass tube and you carefully hold it compressed while you slide the finned guard with the metal shield that the other end of the spring contacts with into position, then screw the front half of the probe into the handle. When the brass tube and the inner lucite insulating tube are seated all the way inside the handle there is about 0.75 inch of this inner lucite tube that slides inside the front lucite half of the probe so there is pretty good overlap and you won’t have a path for breakdown at the front of the grounded brass piece. So the way this probe is designed there is a complete metal shield between your hand and the HV. If there was an arc-over the grounded shield would protect your hand.

An interesting side note is that I went to a hamfest this past weekend and just before I left, at the last table I checked, was a Fluke 80K-40 HV probe brand new in the sealed plastic. The price was much higher than on eBay so I left without it.

[joeqsmith]

Life has been in the way.

Will assemble the Divider insert and take some pics of it as I have all the bits needed. The Probe front end itself is now epoxy coated and I have a few more Rolls of Black so I will reprint the handle end as I have tweaked some of it internally.

Abandoned your efforts?  I would sill like to see a few pictures of the coated parts.

[Kosmic]

I did some non-scientific electrical tests with a PLA filament. Connected 1.2 cm of filament to my Genrad 1644. Tested a 1Kv on the Tera Ohms range. The filament was
beforehand washed with IPA and gloves were used for manipulations.

- At room temperature I was reading 1000 TOhm (pretty much the limit of the instrument)
- Now heating the filament at 100C at almost no effect.
- Heating at 200C I saw a very little drop in resistance.
- Now at 300C for some time the filament was starting to melt and resistance dropped to something like 750 TOhm.

I was expecting a drastic change of electrical characteristics at high temperature but in the end, it was pretty moderate.

[Kosmic]

In comparison, with a similar piece of PTFE I can't detect any resistance (infinite) and heat do not really affect the results.

[trobbins]

I was a bit surprised that some simplified testing of the solid insulation resistance of significant sections of a probe didn't appear until post #115. But even that 5kVdc test, as shown by the photo, appears to be poorly configured.  Testing should I suggest be about wrapping foil around one surface region, and the other relevant surface region - whether that be for checking creepage as well as bulk resistance as shown, or by making an internal surface that coats the inner tubing, or about covering the entire handle region.   Foil wrapping was the standard technique for testing PD and breakdown of telephone handsets in days of yore - certainly for Telstra and EPR issues.

I would think the advantage of 3D printing would be more towards a lower kV probe, but with designed-in shielding to allow experimentation for flat compensation out to kHz plus range, as that is where the commercial probes get expensive for DIYers.  Designing in protective earth tubing, and even an additional redundant earthed handle would similarly seem to be worthwhile, plus some compartment allowance for compensation parts and additional protection parts such as MOVs to shunt any fault to redundant protective earths.

[tautech]

...............
Designing in protective earth tubing, and even an additional redundant earthed handle would similarly seem to be worthwhile, plus some compartment allowance for compensation parts and additional protection parts such as MOVs to shunt any fault to redundant protective earths.

Yet if someones scope is inadvertently floated due to a poor or NO mains earth there would be no protective gain.
Better to use the foil wrap test you suggested and so prove there is no HV leakage.