Open HV Probe 40kV

[vk6zgo]

Here's the electrostatic meter I referred to (3rd pix down), along with a lot of other oddities.

https://sites.google.com/site/squirrelsanonymous/scientific-laboratory-instruments

I used the HP 11042A along with a 410c to check the output power on TV transmitters of up to 2kW ( from memory).

[ArthurDent]

While not a 'probe' in the usual sense, this HV resistive oil-filled divider I picked up for $1 at a fleamarket works if used with great care and proper termination.

[beanflying]

Cool old tools.  I will resist adding them to my Evilbay searching 'for now' 

Finally off the printer after last nights random issue. I suspect it may have come loose from the print bed and made the Z axis skip up so this one was printed with a Raft and Brim (3D nerdishness) to help it stay stuck to the bed without any other supports. Empty like this it is just over 200g in weight and balances just in front of the Black hand guard. Bit more Cleanup and Epoxy coating next.

[joeqsmith]

Looks nice.  Testing it will be fun..   I think I would just try your other parts with a car coil and see how it behaves.  You should be able to get 40KV pretty easy this way.   

This is another antique HV meter.  I think it can read up to 100KV.

[beanflying]

I think the PS on my small Laser Cutter is 15-20kV and I have a Neon transformer that will be over 12kV and I can run it up to over 1kV on my Calibrators first so soft start then get something more to get medieval JoeQSmith on it 

[joeqsmith]

A 50/60Hz neon sign transformer would be a good one to try.   I wouldn't even assemble with the resistors.  Maybe just put the wires on on the two ends.  You could attempt to measure the current through it.    Obviously a bit more risk.   Anytime I play with these, I do it remotely..   

[beanflying]

What is your thoughts on a Uni-T 210E measuring current on 10kV Plus wiring?  or  or 

[joeqsmith]

I think if you are measuring mA+ leakage current through your plastic parts, something major is wrong!!

With your 12KV neon sign transformer, that would only be 12M and your plastic parts would dissipate 12 Watts!    Then again, my math is not great so double check.

[BravoV]

I think the PS on my small Laser Cutter is 15-20kV and I have a Neon transformer that will be over 12kV and I can run it up to over 1kV on my Calibrators first so soft start then get something more to get medieval JoeQSmith on it 

Really tempting, as I just acquired recently a NOS 15kV neon transformer too, thanks for sharing these.

[ArthurDent]

Warning! If I recall correctly, a neon sign transformer that has an output above about 7.5KV has a secondary that is centertapped and the centertap is grounded to the metal base. If you plan to use a neon sign transformer, then check it out to see if this is the case. I don't have a neon sign transformer but check it out before you power up the transformer. Oil burner igniter transformers may be centertap grounded as well.  If you have a c.t. transformer and connect your meter common lead to one of the transformer terminals you could have a big problem.

Google it for more info. Also trying to disconnect the c.t. and ground one of the output terminals can cause transformer insulation failure

[vk6zgo]

I'd like to have one of these:-
https://www.ebay.com.au/itm/L-M-Ericsson-Pty-Ltd-Trimax-Division-Model-52C-Ionisation-Tester-Type-G1B-/202602664145

Just about everywhere I worked had one, so there were plenty around.
The eBay one is a bit too costly for me, though!
Grays Online auctions have one, too, but they always insanely overprice everything.

[beanflying]

Warning! If I recall correctly, a neon sign transformer that has an output above about 7.5KV has a secondary that is centertapped and the centertap is grounded to the metal base. If you plan to use a neon sign transformer, then check it out to see if this is the case. I don't have a neon sign transformer but check it out before you power up the transformer. Oil burner igniter transformers may be centertap grounded as well.  If you have a c.t. transformer and connect your meter common lead to one of the transformer terminals you could have a big problem.

Google it for more info. Also trying to disconnect the c.t. and ground one of the output terminals can cause transformer insulation failure

You are correct at least in the case of mine about the centretap. While that information is available fairly easily it might be a good idea to add some additional appropriate safety information/links on HV probe use and safety but there are a lot of testing cases to cover. There is also a lot of bad information out there to try and counter 

[beanflying]

Just a few design details on the handle end. Currently takes 80x20mm PCB's with plenty of room above the board location and is setup for a small cable gland on the bottom. I might change the sizing and part for it as some idiot used a 13x1.25mm thread pitch on this one   Creating non standard threads in Fusion is a PITA!

[joeqsmith]

I was thinking just start by applying the signal across the two ends of an empty probe body and see if you can measure any current through it.   

If you have an electrometer, maybe try and measure it that way.   

Yes that shield on my probe is there for noise and prevent coupling to any stray objects.  It has nothing to do with safety.   Even the GDTs are not there for safety.  My only concern is to protect the scope's front end.   

[beanflying]

Really Quick and dirty test at 1kV AC and DC uncoated and uncleaned after being handled trimmed/sanded etc by me without gloves. Poor leads and poor technique

DC 0.00mA @ 1kV.
AC @50Hz 1kV 0.00mA
AC @1kHz 1kV 0.24mA

Given the setup that stray current may be coming from elsewhere into the 121GW other than via the Valhalla. That said the Valhalla needs a Cap replacement and a proper going over and I haven't hooked up my Micsig HV probe to look at the wave at higher voltages. My 344501A and 34461A are not back from Keysight Cal yet (should have them next week). May is Recalibration and testing month for the fleet.

[joeqsmith]

DC 0.00mA @ 1kV.
AC @50Hz 1kV 0.00mA
AC @1kHz 1kV 0.24mA


You lost me.   The meter's switch is set to  A/mA and the leads are connected to the Volts input. 

If I assume that everything was setup correctly when you took the actual measurements,  with the 1KHz@1KV signal,  if you just removed the red wire from the probe body as shown and let it rest on the wooden stool, does the current drop?

I made some posts about testing some capacitors for leakage.  One problem I ran into was how much my test jig would leak (the colored boots).  Ended up floating them from the case with some Teflon.  With you working in the 40KV level, I assume you are going to have some head room.  Maybe double that?   You may find that the materials cause an error in your measurements. 
https://www.eevblog.com/forum/testgear/using-low-voltage-levels-to-test-hv-capacitors-leakage-current/25/

[beanflying]

  oops brain fart due to being in a rush grabbed the meter off the bench and forgot to swap. Tomorrow job I will have some free time and do it properly 

[joeqsmith]

This makes more sense.  Look forward to the actual numbers.  At 1KV, I am expecting the 121GW not to detect any thing but you never know.

[kony]

Someone notify the Darwin awards commission, they might get a special serial entry brought you by the open source community (TM).

No, NO, and FUCK NO to the material and manufacturing selection for the insulator of the probe. Having dealt with reliability of manufacturing processes of 20kV PSU assemblies in the past, I am very painfully aware of how deep the rabbit hole goes to even start thinking about passing the design for production after several iterations of debug and validation of BOTH the design and manufacturing process. And you are attempting to get this out in a way such everyone can wing it with the leftovers just lying around? Gotta be shittin me! I am sorry, but I am not sorry for my languague, it is fully apropraite for the magnitude of the possible consequences this can have.

OP, do you have a single fucking clue about the E field distribution on the probe insulation, what air pocket inclusion or surface defect or grit contamination will do to it? Sensitivtity to moisture, long term stability of the plastic (selection of which you, BTW, have absolutely zero control over)? Corona resistance of the material should partial discharge occur - and what certanity is there it won't avelanche in the degradation? How about air gaps in the insulation to the center isntalled divider or elsewhere - do you at least understand what relative permitivity discontuinuity implies for local electric field magnitude?

Can you imagine the consequences of some poor bastard using such thing handheld on hard power supply?

[beanflying]

Moving right along to some actual practical testing and data. Two Coffees this morning after a 4.30am Possum Roof visit  , wiped the seat and Neon Transformer over with IPA, probe as per yesterday uncleaned etc. 121GW plugged into the mA jack and had the correct range selected 

** I do have an issue with the Valhalla it seems on the 1kHz range it is clearly spraying noise as I was getting a current reading with the 121GW unconnected as per the pics below.

Feeling a little brave for the inner smoke on the 121GW I broke out the Neon Transformer to raise the stakes a little. 6+kV to Ground and 0.0000 mA so I am feeling happier about where I am heading. Forgot to take any snaps of in on the Fluke but Zero @ 1kV DC.

Photos remind me I must clean up and repaint the outside of the Neon.

[joeqsmith]

Looks like a nice old transformer.   

You could try using the uA range as well.   Can you go across the output of the transformer instead?  You could also try a shorter distance.  Maybe try going through the sidewall of the case.

[beanflying]

Transformer is an Allanson 277V, 60Hz, 15kV, for anyone playing along at home (Running nominal 240-250V in my case) 

I have started prepping the Probe for Epoxy coating so no more testing for a bit. Floating a decent DMM that far above ground was worrying me so I didn't. I do have a couple of old Victor 86B's with mA and uA ranges that I care quite a bit less about so maybe I will look at that when the coating is done.

Having a look at what ArthurDent raised I think I will extend the inner tube further toward the handle by a bit at a reduced diameter again and add some Tinned Copper Sleeve and bond it to the Earth and LV connection to the meter to it. Belts and braces a bit but it certainly cant hurt.

I have ordered some more PLA which should be in tomorrow and I have planned to print a second one as a Chicken stick for the Neon Transformer with HV cable going in right at the tip part of a Lanthanum TIG electrode (longer barrel again). While I am going I can do a second barrel with multiple points to connect to along the Barrel.

Easy to add these as small raised knobs or holes/divots to help connect to? Let me know what may work best?

For interest I have attached some cutaways of the actual Sliced Gcode for the upper Barrel. It was 6 inner and 6 outer layers and it is well down into the tip before there is any infill/voids. They could be eliminated easily if it looks like being an issue with a simple setting change to the infill from 40% to 100%.

[tggzzz]

Someone notify the Darwin awards commission, they might get a special serial entry brought you by the open source community (TM).

No, NO, and FUCK NO to the material and manufacturing selection for the insulator of the probe. Having dealt with reliability of manufacturing processes of 20kV PSU assemblies in the past, I am very painfully aware of how deep the rabbit hole goes to even start thinking about passing the design for production after several iterations of debug and validation of BOTH the design and manufacturing process. And you are attempting to get this out in a way such everyone can wing it with the leftovers just lying around? Gotta be shittin me! I am sorry, but I am not sorry for my languague, it is fully apropraite for the magnitude of the possible consequences this can have.

OP, do you have a single fucking clue about the E field distribution on the probe insulation, what air pocket inclusion or surface defect or grit contamination will do to it? Sensitivtity to moisture, long term stability of the plastic (selection of which you, BTW, have absolutely zero control over)? Corona resistance of the material should partial discharge occur - and what certanity is there it won't avelanche in the degradation? How about air gaps in the insulation to the center isntalled divider or elsewhere - do you at least understand what relative permitivity discontuinuity implies for local electric field magnitude?

Can you imagine the consequences of some poor bastard using such thing handheld on hard power supply?

Some of us have tried asking pointed questions, and been rebuffed, and bowed out of the discussion.

[beanflying]

I chose to roll past Mr Fuck this that or the other before as he offered no factual information to back up spurious claims that would be true of HV probes commercial or otherwise in the hands of a numpty. If we use this thinking anyone who spends money on evilbay or aliexpress and buys Probe X can injure of kill themselves and that's ok?

It is really simple SHOW ME EVIDENCE of the material or process not being suitable. When offered video of the material being tested under HV for breakdown you didn't watch and then went on again without looking further for yourself. You chose to dimiss the people as 'Talking Heads'. When offered further information about the material you went on again with no evidence to back up your claims about it.

PLA and many thermoplastics are hygroscopic to some extent. Unless proven otherwise I would assume that would change the electric properties over time, e.g. it becomes more conductive.

I have not seen (because I haven't looked for) any measurements indicating whether or not that would be sigificant in this application.

I have done some additional research into moisture and all plastics absorb water to a greater or lesser degeree with Teflon being a bit of an exception. The ABS for example used in a lot of these probes commercially  will hold water in the order of 0.2-0.3% based on my reading and for PLA it seems to vary but at around the 0.4-0.6% range is what I have read and all plastics absorb more moisture as the Temperature rises. What have you done and what did you find?

You have sought at every possible point to add zero constructively to the project and only tell the world it cant and shouldn't be done because you say so. 

As I thought may be obvious by now I have a set of files ready for release that work and quite deliberately haven't and won't be generally releasing them until I am happy that tests carried out be me and hopefully a few others independently give more confidence in the design.

[The Soulman]

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.