EHT power supply design and construction

@Aleph(0) and @shortbus

Although my 'recovery' from exasperation corollary to certain 'online annoyances' (Spec: non-registering developers and especially @Jazz2C's recent 'faux pas' elsewhere on these fora:rolleyes:) is nearly complete -- I apologize that my participation here will be rather limited prior to mid/late May (i.e. following 'Dayton') --- In the meantime @Aleph(0) and (one can but hope) @theodoravain will be available to respond to questions/comments -- thanks for understanding!:)

But you are incorrect in saying stainless is only available in fine thread
Such represents the 'routinely stocked' items available from my local 'bricks and mortar' 'True Value' home center (sans 'special order') -- I own the misunderstanding inasmuch as @Aleph(0) was merely relating my 'report' on the 'materials availability situation' lo these two years...

If soldering is involved why not use brass hardware?
Succinctly: Said alloy is 'softer' with correspondingly compromised structural properties (specifically rigidity and 'thread life') -- Note also that --quite contrary to the majority of electrical applications-- effects corollary to introduction of small parasitic resistances are negligible Re: the EHT indicator project (owing to the resistance ratios involved) -- Hence the acceptability and, indeed, desirability of steel fasteners...:cool:

While places like Home Depot or Lowes are much larger than an Ace Hardware, they former don't have the selection of fasteners that an Ace or maybe a True Value hardware has
Agreed!:) As you are no doubt aware, 'classic' hardware retail outlets often feature 'specialty hardware' (in small chests of drawers along the fastener aisle) -- While some may balk at the 'price structure' of said items - I would argue that such is more than compensated by the products' ready availability!

Ace or maybe a True Value hardware has, though I haven't been in a True Value for years.
We yet have them 'round here (N.E. MN and W. WI):cool: - And 'Hardware Hank' too!:D Surprisingly, 'Mills Fleet Farm' (being, as they are, rather 'big boxish') likewise carries specialty hardware...

Next question(s). Why use Teflon for the 'fault path ribs'? If PVC is good enough for the rest of the tower, why isn't it good enough for the ribs?
Ribs fashioned of PVC, Acrylic, Polycarbonate and, curiously, polyimides and PVDF (tested to 1 cm thickness) rapidly present with marked porous degradation (up to and including full thickness perforation and crumbling) owing to a process perhaps best --but in any event adequately:oops:-- described as 'pyrolysis incited electrolysis/chemolysis' secondary to corona... --- Of the substances tested only (cured) silicone adhesive, PTFE, glasses, fused silica and mica showed no detectable degradation following Ca. 100 hours (automated) exposure to 80kV DC (resistor/wax 'friendly' duty cycle observed!:cool:) --- Please don't ask which microcontroller platform I employed Re: automation of the tests -- it's just... too... Um... -- Oh! Best not to dwell upon it:oops::oops:o_O --- Laziness is a lamentable vice -- hypocrisy is much worse!:oops::(

As to why the PVC tubing (denoted 'Tower Body' in the annotated image) is not similarly effected, one must consider the role of discontinuities as regards their influence upon EMF gradients/field 'focusing' and, hence, corona -- Please be mindful of the fact that the resistor is 'potted' central to the PVC tube in hydrocarbon 'wax' - said resistor being effectively 'grounded' at its lower end --- Hence, for all practical purposes, any corona incident upon insulator surfaces owes soely to EMF gradients secondary to electrical leakage along said surfaces (principally via inevitable electrolyte condensates/contamination) ... Hence corona (conditions permitting) will manifest at discontinuities (i.e. points/areas of significant field intensity)... --- Please note that the upper end cap's rim --although representative of a rather 'stark' discontinuity at a (seemingly) unfavorable location - is, as a practical matter, of little consequence for its already very close proximity to the electrically elevated terminal --- That said, slight 'erosion' of same is notable following a single 'session' of (the above described) protracted test excitation...

All of which brings us to...
Also don't understand the use of silicone glue to hold things together.
Although the PTFE ribs do not, themselves, degrade -- they nonetheless form a discontinuity with the highly 'corona vulnerable' PVC tube -- Hence measures must be taken is degradation of said structure to be inhibited-- Silicone adhesive represents a viable solution for two reasons:

1) Electronic grade silicone is one of the very few long-term 'corona resistant' materials tested (as noted above).

2) Via proper application one may effectively 'dress out' (as a practical matter greatly minimize) the discontinuity introduced via the rib-to-tower interface all the while shielding the 'endangered' PVC via a film of silicone at the margins...

As far as "rational notation" over "conventional notation"
'Fraid we'll have to 'agree to disagree' on that particular 'note';)--- The way I learned 'rithmetic:
AB=A*B (as opposed to A+B) ---- Similarly:
1¾ = 1*¾ = ¾ (as opposed to 1+¾)o_O --- Please note that I do not dispute your assertion that 'convention' dictates (Re: rational notation) that 'N p/q' = 'N + p/q' --- My stance is, rather, a reflection of my extreme distaste for said notation which being far too strong to permit my use (and, hence, tacit support) of same!:rolleyes: ---- Then too 1+¾ is utterly without ambiguity!:cool:

using medical and scientific terms for mechanical things,
But then these tutorials/projects are intended as 'scientific' exercises:) -- as opposed to 'recipes' and/or the 'fringe science' and 'hocus-pocus' having veritably lain siege to much of the (amateur) online EHT (and related topic) community...

if you or any one were to take any of this somewhere to a machine shop that wasn't tied to a university they wouldn't have a clue to what you wanted. I know we've had this conversation before but it's just how things are.
Agreed! But then you forget (or perhaps we neglected to declare?:oops:) that (inasmuch as such would be all of unduly burdensome to the reader and contrary to the goals of this effort) none of our projects require out-sourcing of fabrication (beyond mere purchase of inexpensive, commonly available, ready made, 'raw materials')!:) -- Moreover, our projects are so conceived that only 'standard' tools are required! Whether for direct application to a project or for fabrication of specialized instruments required for operation/construction thereof (e.g. construction of the {perhaps $200 all told} EHT indicator -- as opposed to requiring the reader's purchase of a $60k 'Keithley' or even a $5k radiography 'voltage divider')...

Nor do we require or expect that the reader is equipped with a 'home workshop' -- Granting that possession of, or ready access to, a quality automated additive manufacturing system (AKA '3D-Printer'), lathe, milling machine, band-saw, etc... may well ease certain tasks -- such is neither required nor expected! -- Indeed! In addition to common manual tools: a small drill press (≥ 6" 'depth'), a variable-speed hand-held power drill (½" chuck) and a reciprocating saw is all that are required in the way of 'shop tools')...

using medical and scientific terms for mechanical things,
@shortbus I apologize that, having reread our recent posts, I'm uncertain as to the 'specifics' of your reference? -- Is your comment with regard, perhaps, to our our use of 'costa' vs. rib?:confused: and/or 'fasteners' vs. 'nuts and bolts'?:confused: --- am I missing something? -- If not - methinks you underestimate hardware sales associates!?;):)

Pursuant to the above -- please be advised that @Aleph(0)'s reference to thermodynamics (apparent near the close of post #1999) was by way of what amounts to an 'inside joke' (i.e. that we might somehow 'right' the neigh-on ubiquitous 'upside-down' correlation of 'perceived temperature' to photon energy (and, hence, 'colour temperature') of visible spectra -- Please know that said comment was wholly off-topic and bears in no way upon the project!:)

HP since I TOTALLY know people will be wondering and just for my curiosity how do you calculate optimal intercostal distance? So I get that if ribs are too close they'll basally act as only one. So arc would just bypass intercostal space jumping from edge of one rib to edge of other! Also being too far apart allows steep enough gradient to set up corona possibly initiating arc! So o/c in perfect world it's basically just normalizing from interval of convergence but in REAL world with unrestricted, non-ideal gas (by which I mean air) as _part_ of insulator it's way more complected than that! So I'm basically asking is there just like _rule of thumb_?
As per your own observation, the matter is complected by too many contingencies to admit of a straight-forward mathematical solution --- While empirical observation (i.e. experimentation) with various insulating materials employed as fault path extension barriers applied to DC systems up to 800kV tends to suggest that optimal intercostal length (radius of the largest rib)/2 (where said rib is positioned nearest the 'electrically elevated' terminal) I have, as of yet, neither given said observation careful consideration nor hypothesized a theoretic basis therefor --- But to directly respond to your question:

1) Based upon the above described observation, an intercostal distance of 25mm was chosen as the 'starting point'...

2) During empirical examination (two rib 'tower' -- 10 sweeps each 'test distance') 20mm was found to give the highest flash-over EMF...


------From my notes------
Re: present tower geometry/dimensions/materials (sans 'Mini-Fed' well):

---Test parameters---

Method:
Tower mounted vertically to ground plane.

Sweep rate = 5kv per second, each sweep followed by 30 seconds at ground potential (0v across tower).

Electrically elevated terminal exposed to atmosphere (i.e. 'Mini-Fed' well absent).

Ambient atmospheric:

Temperature ≈ 290°K
Barometric pressure ≈ 100kPa
Relative humidity ≈ 60%

--------------------------------------------------------------------------------------------------
Profile runs -- 5mm increments:

Intercostal distance = 25mm (mean flashover EMF (10 sweeps) = 146kV
Intercostal distance = 30mm (mean flashover EMF (10 sweeps) = 137kV
Intercostal distance = 20mm (Mean flashover EMF (10 sweeps) = 183kV
Intercostal distance = 15mm (Mean flashover EMF (10 sweeps) = 160kV
Intercostal distance = 10MM (mean flashover EMF (10 sweeps) =133kV

'Bookend runs' -- 2mm increments:

Intercostal distance = 22mm (Mean flashover EMF (10 sweeps) =165kV
Intercostal distance = 18mm (Mean flashover EMF (10 sweeps) = 176kV

As may be seen it's quite (IMO surprisingly) abrupt/critical near optimal separation...

Very best regards
HP:)
 
Last edited:
Maybe you can get some construction tips from this:

"These [single phase] transformers feed into a transmission system at a voltage level of 11,00kV Direct Current"
---Emphasis Added---
:confused::confused::confused:

What's wrong with this pictureo_O? -- Postterm April fools joke? 'Impaired' narrator? --- I dunno...

Perhaps time variant DC and a special non-saturable (insatiable?;)) core are implied? -- Possible? Of course! Still.. I doubt such is the case - Even as I doubt the costs of R&Ding the pictured apparatus could be held to anywhere remotely approaching the vicinity of a mere $2M (as claimed)o_Oo_Oo_O

Very best regards
HP:)
 
Last edited:

Ya’akov

Joined Jan 27, 2019
9,069
---Emphasis Added---
:confused::confused::confused:

What's wrong with this pictureo_O? -- Postterm April fools joke? 'Impaired' narrator? --- I dunno...

Perhaps time variant DC and a special non-saturable (insatiable?;)) core are implied? -- Possible? Of course! Still.. I doubt such is the case - Even as I doubt the costs of R&Ding the pictured apparatus could be held to anywhere remotely approaching the vicinity of a mere $2M (as claimed)o_Oo_Oo_O

Very best regards
HP:)
As unlikely as it seems, HVDC links are used because in some ways they offer higher efficiencies. In the US, there is a major HVDC link along California. There are several schemes for this, which are laid out here: https://circuitglobe.com/different-types-hvdc-links.html

The strange English that made it seem as though the cost was ~2 million dollars really intended to say that it was “double digit millions” to develop. The native language of the narrator is German, and almost certainly the writer as well if not the same person.

I must say, it is quite an engineering feat all around, and I wouldn’t want to test if it was live with the age old method of using the tongue, à la 9V batteries.
 
HVDC links are used because in some ways they offer higher efficiencies.
Yes, Indeed! I'm familiar with DC transmission lines!:) --- My dubiety owed to the notion of (ostensibly magnetic) 'DC transformers' and to the concept of 'phase' Re: DC phenomena -- Of course 'DC' needn't mean 'non time variant' - Still...:confused:

The strange English that made it seem as though the cost was ~2 million dollars really intended to say that it was “double digit millions” to develop. The native language of the narrator is German, and almost certainly the writer as well if not the same person.
---Emphasis Added---

Ah! That explains many things!:cool: -- Perhaps discussion/mention of (post-transformation) rectification was likewise prey of said language barrier?

I must say, it is quite an engineering feat all around, and I wouldn’t want to test if it was live with the age old method of using the tongue, à la 9V batteries.
Or indeed a 'classic' 40kV 'high voltage' probe!:eek::D

Sincerely, always a pleasure @Yaakov!

Best regards
HP
 

Ya’akov

Joined Jan 27, 2019
9,069
Yes, Indeed! I'm familiar with DC transmission lines!:) --- My dubiety owed to the notion of (ostensibly magnetic) 'DC transformers' and to the concept of 'phase' Re: DC phenomena -- Of course 'DC' needn't mean 'non time variant' - Still...:confused:
Ah, I see. I just assumed they were referring to the input to the rectifier from the generating plant so I didn't catch it as a question. I suppose that's because it wasn't a technical presentation but a marketing one, and I have long since stopped expecting technical accuracy from marketing material, particularly aimed at "higher" decision makers—though I am pleasantly surprised when I find it.
 

shortbus

Joined Sep 30, 2009
10,045
'Fraid we'll have to 'agree to disagree' on that particular 'note';)--- The way I learned 'rithmetic:
AB=A*B (as opposed to A+B) ---- Similarly:
1¾ = 1*¾ = ¾ (as opposed to 1+¾)o_O --- Please note that I do not dispute your assertion that 'convention' dictates (Re: rational notation) that 'N p/q' = 'N + p/q' --- My stance is, rather, a reflection of my extreme distaste for said notation which being far too strong to permit my use (and, hence, tacit support) of same!:rolleyes: ---- Then too 1+¾ is utterly without ambiguity!:cool:
That is the reason most blue prints don't use fractions as such. Outside of wood working. Instead of 1+3/4 they would say 1.75. And I can't see anyone doing something scientific using common fractions. The 1/4- 20 is a thread description not meant as a fraction.

So then to my real question, you don't want any ideas on how to position the Teflon used on the fault path ribs?
 

Thread Starter

Aleph(0)

Joined Mar 14, 2015
597
That is the reason most blue prints don't use fractions as such. Outside of wood working. Instead of 1+3/4 they would say 1.75. And I can't see anyone doing something scientific using common fractions.
Shortbus I basically agree with what you are saying! So _common fractions_ (by which I mean _rational fractions_ which is important to disclaim because being nitpicky technical about it expansions are _fractions_ too) are totally essential in math and ABSTRACT physical science and definitely anywhere fractional expression contains variables or functions! Also since it's totally more intuitive, it's common practice in math to leave fractions with integer component in _improper_ format! So 1¾ is written as just 7/4:cool:

But I totally agree with you that for practical purposes with just numerals as numerator and denominator decimal expansions are a LOT clearer to most people:)!

So then to my real question, you don't want any ideas on how to position the Teflon used on the fault path ribs?
Shortbus it's vry important for you to understand that we TOTALLY value your input! And so totally hope you keep it coming:)! So if it seems like we don't need as much advice on EHT meter project it's just that it's basically completed like 2 years ago so most of the _bugs_ are already squished:cool:! And @shortbus YOU definitely squished a HUGE _bug_ showing us how to mount meter inside box:)! So where we're at now with EHT meter project is basically showing them how to put it together plus providing info (like formulae for people who can't find same resistor values we're using and like that). But if you see anything that looks like possible opportunity for improvement or even just making it easier for builders including even how we're describing things (like with bolt pitch) please don't be afraid to comment! Cuz now with CDT basically _fired_ (except for @theodoravain who isn't very active on here and @Jazz2C who's basically on _involuntary leave of absence_ from tutorials:rolleyes:) it's basically just HP and me plus all the help we can get from very kind REGISTERED AAC members! Like you and @cmartinez @The Electrician @Yaakov (and including some other AAC members too) who even though we don't know any of you personally you've been way more help than CDT team ever was!

@shortbus now just so you know, next _tihidi_ project after EHT meter is unregulated LVPSU which even though electrical and electronic design is basically done and tested definitely needs a LOT of structural design work! So when we get to that you'll probably get totally tired of us asking for advice and help:eek::)!

Although my 'recovery' from exasperation corollary to certain 'online annoyances' (Spec: non-registering developers and especially @Jazz2C's recent 'faux pas' elsewhere on these fora:rolleyes:) is nearly complete -- I apologize that my participation here will be rather limited prior to mid/late May (i.e. following 'Dayton') --- In the meantime @Aleph(0) and (one can but hope) @theodoravain will be available to respond to questions/comments -- thanks for understanding!:)
HP I say HUGE THANK YOU:)! Now being honest I thought you prolly couldn't stay away after like 4 years _investment_ of your time here but it's definitely a huge relief seeing you say u plan to return:cool:!
 

Thread Starter

Aleph(0)

Joined Mar 14, 2015
597
Ribs fashioned of PVC, Acrylic, Polycarbonate and, curiously, polyimides and PVDF (tested to 1 cm thickness) rapidly present with marked porous degradation (up to and including full thickness perforation and crumbling) owing to a process perhaps best --but in any event adequately:oops:-- described as 'pyrolysis incited electrolysis/chemolysis' secondary to corona...
So here's for people who don't understand how corona (which feels just warm) can heat something enough to make it start breaking down into other compounds:

*****Disclaimer******

Now just so everybody knows! For now we're keeping the math and science _light_ (npi) to keep series interesting and accessible to people at different stages in their education process! So I'm going to totally avoid explaining in terms of black body radiation and analytic math! So this isn't a dumb-down at all! It's just getting everyone on board before leaving port! So we'll definitely get to the math and science of all these topics in appropriate time and place in series! So for anyone having problem with _initially qualitative_ approach @Hypatia's Protege says she has totally open mind on subject! So please just let her know how you feel:)!
**********************

Anyhow qualitative explanation is that plasma ions can be extremely hot! To see how hot just look on color temp of spectra emitted by electrical corona in air! So as example just azure is like 10,000°K (so about 9,727°C) and corona has emissions way into UVA spectrum which are much _hotter_! Now the reason the corona plasma doesn't seem hot is cuz high energy ions are basically few and far between so average temp is basically like _unimpressive_ but the high energy ions damage insulator at molecular level breaking down compounds some of which reform into electrolytic _conductors_ so then the intense electric field basically _takes it from there_ electrolysing pyrolysis products producing reactants and solvents plus more electrolytes causing total _runaway_ disintegration:eek:

So as example of plasma having ions with way higher energies (temps) than average of volume, you can think on fact that even though just yellow flame (like on match or candle) has average temp of like 500°F (so 260°C) it can ignite combustibles with much higher kindling temps! So that's why it's basically impossible to ignite treated lumber with dull red steel bar (which has temp of abt 820°C) but it's very easy with just lazy yellow flame with average temp of just 260°C:)! Now for people saying _wait a minute! The sun is just big ball of plasma and its average temp is totally terrifying!_:D All I can say is two things!

◊Just like you'd expect, ion population density increases with plasma pressure! That's main reason why low intensity discharge lamps (like linear florescent) run way cooler than HID lamps:cool:! Now even though average density of sun is low that doesn't change fact of much higher central density! So you can think on composition of sun as high density fusion reaction surrounded by larger low density spherical _cloud_ of ejecta:)!

◊A lot of the sun's spectra is very high energy (in xray and gamma ray _region_)! So even at sun's super high pressure core, temperature (which is about 15M°C) doesn't totally _live up_ to energy of spectra:)!

Also for people wondering why PTFE doesn't degrade like even other fluoropolymer Kynar: We're not 100% sure but we're guessing maybe it just doesn't form conductive pyrolysis products so only degradation is from direct action of corona ions which by itself is so gradual that it doesn't even matter:cool:!
 
Last edited:

shortbus

Joined Sep 30, 2009
10,045
Aleph, I get mixed messages from you and HP on my input. One seems to welcome it and the other doesn't seem to want it. That said I will draw up what I came up with for the towers and if doesn't suit EITHER of you, say so and I will delete with no hard feelings. I do understand that the both of you may have built yours but doubt that any others have.
 

Thread Starter

Aleph(0)

Joined Mar 14, 2015
597
Aleph, I get mixed messages from you and HP on my input. One seems to welcome it and the other doesn't seem to want it.
Shortbus I'm sorry it seems like that to you because that's not case at all:confused:! So only major disagreements HP and I have are about where and when to use quantitative vs. qualitative approach cuz qualitative explanations can seem like _dumb-down_ but being strictly quantitative early in series or in _basic_ discussion can scare people off who just need a little time to get like _acclimated_ to the math:)! So anyhow you can see how _touchy_ subject is by just looking on how HP and I have like _changed sides_ on that more than once since start of tutorialso_O!

Shortbus anyhow that has nothing to do with you at all or with us wanting your input! We BOTH totally value your feedback so I don't know how you're coming to conclusion that one of us doesn't:confused::(? Unless it's just cuz of our different writing styles causing confusion?

So you asked:
you don't want any ideas on how to position the Teflon used on the fault path ribs?
So being 100% clear: Short answer from both of us is: _We don't need advice on design of towers because we totally worked that out like two years ago and present design has stood _test of time:)!_
So reason for _verbose_ explanations is so you (and other readers) will know we definitely value advice and comments by totally explaining basis for our decisions! So it's just that all those issues with towers have been explored to total extent practical! So it would be totally rude of us to encourage you to do work in vain:(! Shortbus please carefully look on HP's post 2002 and my post 2009 again so you can see it's not really a structural or mechanical issue! Cuz even though structure definitely comes in to it, it's really just electrical, thermochemical, chemical and electrochemical issue:)!

Shortbus I totally understand how off-putting looking on a _mile long_ post can be:oops:! But one huge difference between us and a lot of other people soliciting advice is that we're very happy to spontaneously explain our rationale for adoption or rejection of ideas:cool:!

So we're definitely not like JOKE that calls itself _ SCOTUS_ who basically makes habit of just refusing to hear cases w/o explanation:mad:! Cuz totally unlike them we respect our constituency (by which I mean our readers and advisers including you!) Also, like I said, _core development team_ is gone! We gave them ultimatum of registering on AAC or _kissing off_ I say you can guess how that turned out:eek::D! So anyhow now it's just @Hypatia's Protege, @theodoravain, (eventually) @Jazz2C and me! Now _Alpha test team_ still exists but they're just _proof builders_ so they aren't involved with design decisions at all! Also you can totally believe that NOBODY on AAC would want them as members:eek:! So whole point is that we're definitely NOT giving deferential treatment to non AAC registrants or non EHT thread participants! So I totally promise everybody with direct influence on any aspect of tutorials is AAC member who posts their ideas on this thread or tags us on other PUBLIC AAC threads:)!

Shortbus now I'm going to tell you something that we're very grateful for:cool:!

Your criticism of _team_ (by which I mean CDT) kibitzing from _shadows_ was huge part of our arriving at _epiphany_ that their _hearts weren't in it_! So I say it's totally fair to say you saved series:)!

So anyhow just a for general info here's some of what we tried and considered designing towers:

So first this is very important! Corona we're talking abt is very subtle! It can only be seen as faint blue glow in totally dark room after dark adapting eyes for like 20 minutes! Spraying or hissing corona or any corona bright in dark room is _symptom_ of major issue that you need to solve first! Also weak corona is not problem UNLESS it's between two areas with big potential difference (like anywhere between top and bottom of tower) because it can start runaway electrochemical reaction (which is brought up by HP in post 2002 and elaborated on by me in post 2009) and it also totally compromises ability of air to hold off flash-over (cuz plasma is NOT insulator)!

So about towers, just being clear: Only practical way of improving towers is to replace PVC tube with PTFE or fused silica tube! IMO glass is iffy cuz it's not not crystalline (it's amorphous solid) which means basically very high viscosity liquid all the way down to 0°K! So even though it performs ok as electrical insulator up to temps of like 300°C, so called _solid_ glass is still just _thick electrolyte solution_! So just personally I don't like that!

So first thing we tried (which was before finding out PVC and most other polymers rapidly degrade in strong electrical field) was to just cut shallow grove (2.54mm wide X 3mm deep) around PVC tube and size ID of rib so that it basically _snaped_ into grove for very tight fit with just a film of silicone in grove to make it _ionized-air tight_. So just aesthetically it looked great until tube got so porous and carbonized it basically self-destructed:oops::(

So next thing we tried was _dressing-out_ discontinuity at rib/tube junction with epoxy which helped but since it's not geometrically possible to totally _shift_ discontinuity to rib, PVC degradation was still way too fast! So by then we knew silicone could solve problem by dressing out discontinuity AND masking PVC margins but we still wanted more elegant solution!

So we thought on source for glass or fused silica tower tubing:

So glass tubing with usable diameter can be salvaged from FxxT12 linear florescent lamps. (available in 17", 2', 4' and 8' length).

Fused Silica tubing can be salvaged from from FGxxT12 linear UVC lamps (available in 17" and 4' length).

But here's problems with light tubes:
Glass (florescent) tubes can be very inexpensive but walls are so thin and fragile they're basically impossible to work with:mad:! Also phosphor powder can be toxic if accidentally inhaled and even very slight heavy metal (Hg) contamination totally spooks _chemophobic_ ppl:rolleyes: plus overcomplicating legal disposal of _rinsate_ from pressure washing inside of tubes.

Fused Silica (UVC) tubes have advantage of stronger wall and no phosphor to clean off. Also at _room temp_ all non-gaseous Hg is in solid amalgam beads so much easier to collect for proper disposal. So _down side_ is even though walls are stronger than glass tube they're still very fragile and cutting them w/o shattering is very difficult (it can be done with like _Dremel cutoff wheel_ but it's definitely NOT easy!

Now we don't suggest either of these cuz even when it's all done and filled with wax it's so fragile even tension on test lead can break it!

Now for people wondering about PTFE or silicone tubing, here's problem:

PTFE tubing would be awesome IF it was easy to source in usable size (which it isn't) also making it from sheet stock is vry hard cuz only good way of joining it to itself is fusing but since its mp and pyrolysis temps are vry close that requires special equipment:(!

High purity silicone tube would definitely solve _porosity_ issue but it totally lacks needed rigidity!
Also, in case anybody's wondering, even though it can totally protect PVC at margins of _dressed-out_ discontinuities, coating whole PVC tube with silicone won't work at all cuz of creating interface discontinuities and so high voltage antinodes causing _walking_ gradient patterns all along tower:eek:! So for people saying _that doesn't make since for DC circuit_ I say please remember corona is type of _disruptive discharge_ and so time varying in nature which gives it properties of AC even if it doesn't change polarity! That's also why _DC_ arc like lightning (which is also example of disruptive discharge) can cause AC phenomena (like standing waves) too:)!

So anyhow after a LOT of testing we just decided on PVC towers and PTFE ribs w/ silicone functioning as _rib lock_/_discontinuity dress-out_ beads AND PVC protectant on margins of discontinuity:)!

So anyhow @shortbus it's not that we don't want your input it's just that towers are done and debugged two years ago:)! So that doesn't mean we're not overlooking something (like when we were calling bolts by wrong name:oops:) but EHT Meter project's _sciencey stuff_ is basically done deal now:cool:!
 
Last edited:

shortbus

Joined Sep 30, 2009
10,045
I'll just make one more point. After the explanation of why the "ribs" needed to be Teflon, I get that. But my Idea is a way to make the positioning and anchoring of the 'ribs' stronger and more easily done by using the same materials, but with a small change. The same would be where the "fed" connector joins the tower.

On another similar note, did you try painting the PVC parts with the old fashion 'corona dope' that was used back when CRT TV repair was still done? I think it is or was still available.
 
So next thing we tried was _dressing-out_ discontinuity at rib/tube junction with epoxy which helped but since it's not geometrically possible to totally _shift_ discontinuity to rib...
LOL! I'm bound to remark that reflection upon said 'process' places me in mind of FB Long's 'Hounds of Tindalos' - and who will deny that failure of same is with the lamentable fate of Halpin Chalmers but upon the towers?!:eek::p

But my Idea is a way to make the positioning and anchoring of the 'ribs' stronger and more easily done by using the same materials
FWIW Our original plan (once having quite decided to implement the tower bodies via PVC tubing) was to cut PVC rings/spacers having inside diameters marginally larger than the outside diameter of the tower tubing such that the ribs might be rigidly 'clamped' between same -- such would have produced a 'rock solid' assembly but at the cost of further discontinuities -- Happily, electronic grade silicone bonds quite well to PVC -- Hence our development of the following technique (each rib):

1) De-glaze PVC tubing in areas ribs are to be mounted (to increase surface area and hence bond strength)
2) Form a single silicone bead at the appropriate 'latitude' about the PVC tower.
3) Allow said bead to cure undisturbed for at least 48 hours (ambient temp≈18°C).
4) Slide rib into apposition with bead while maintaining sufficient pressure to keep the latter in elastic deformation.
5) Form a second silicone bead about the and PVC tubing on the plane --and in immediate physical contact with-- the (central) rib surface opposite the first bead.
6) Allow said second bead to cure à la step #3

Note that the rib is now elastically bound by the silicone beads

7) As much as possible 'dress out' the discontinuity with silicone taking to care 'flank' the margins...

The same would be where the "fed" connector joins the tower.
Inasmuch as the 'Mini-Fed' wells are 'outside' the poles - corona (although present) in not an issue -- Hence, instead of silicone, they (the CPVC Wells) are secured to the PVC end-caps via 'high temp' acrylic adhesive --- Although I've not attempted it, I suspect the PVC and/or CPVC would break prior to separation of the bond! -- For all its 'agro' acrylic is surprisingly 'tough stuff':cool:

On another similar note, did you try painting the PVC parts with the old fashion 'corona dope' that was used back when CRT TV repair was still done? I think it is or was still available.
I am familiar with the product to which you refer -- Indeed is is yet available as GC 10-4702:) - A useful compound indeed! --- Unfortunately it is inapplicable to the presently discussed application for reasons similar to those noted by @Aleph(0) (Re: silicone) quoted immediately below:

coating whole PVC tube with silicone won't work at all cuz of creating interface discontinuities and so high voltage antinodes causing _walking_ gradient patterns all along tower:eek:! So for people saying _that doesn't make since for DC circuit_ I say please remember corona is type of _disruptive discharge_ and so time varying in nature which gives it properties of AC even if it doesn't change polarity! That's also why _DC_ arc like lightning (which is also example of disruptive discharge) can cause AC phenomena (like standing waves) too:)!
Stated otherwise --- 'layering' of 'atmospherically-immersed' solid insulation is --to the extent possible-- best avoided above Ca. 45kV --- Nor is it that the "rules change' -- merely that what is negligible at 35kV is often highly pronounced at 60kV -- said 'criticality' being prominent among the reasons 'high voltage' equipment is generally 'potted' in conductive containers...

Well hey @shortbus, I logged in today at @Aleph(0)'s request, as it seemed you felt one of us undervalues your suggestions? --- On my word! Such is not the case!:) Heck! Even at this 'late date' you saved us driving hardware retailers to distraction via reader's nonstandard requests!:oops::cool: --- Shortbus please rest assured! -- If we say it you may confidently believe that we believe it!:) Your assistance and willingness to assist us is greatly appreciated!:)

Very best regards!
HP:)
 
Last edited:

shortbus

Joined Sep 30, 2009
10,045
FWIW Our original plan (once having quite decided to implement the tower bodies via PVC tubing) was to cut PVC rings/spacers having inside diameters marginally larger than the outside diameter of the tower tubing such that the ribs might be rigidly 'clamped' between same -- such would have produced a 'rock solid' assembly but at the cost of further discontinuities -- Happily, electronic grade silicone bonds quite well to PVC -- Hence our development of the following technique
What you tried is very similar to what my idea is. But instead of trying to find a tube that will work for the spacers, I was going to suggest using PVC "unions" some times called "couplings". They are made to fit over the original pipe , they have a slight internal rib to center two pipes when using in their normal job, but that is easy to take out with an Exacto knife or a Dremel tool.

Inasmuch as the 'Mini-Fed' wells are 'outside' the poles - corona (although present) in not an issue -- Hence, instead of silicone, they (the CPVC Wells) are secured to the PVC end-caps via 'high temp' acrylic adhesive --- Although I've not attempted it, I suspect the PVC and/or CPVC would break prior to separation of the bond! -- For all its 'agro' acrylic is surprisingly 'tough stuff
My suggestion for the Mini Fed connection is to use one of the 'unions' mentioned above on the end of the tower. Then using a PVC reducer fitting inside the open end of that 'union'. This will make the fed connection much less prone to getting broken off or giving another radial joint to worry about. The reducers are available to go from the 1" diameter of the tower down to as small as 1/2"(0.5") PVC pipe size.

It may have been mentioned before but if not, you are aware that your CPVC mini fed connector tube doesn't fit into PVC fittings. The two different pipe materials aren't compatible so they made the sizes so they couldn't be interchanged in common use.
 

cmartinez

Joined Jan 17, 2007
8,218
It may have been mentioned before but if not, you are aware that your CPVC mini fed connector tube doesn't fit into PVC fittings. The two different pipe materials aren't compatible so they made the sizes so they couldn't be interchanged in common use.
yeap ... CPVC's dimensions are made to same standards as old copper tubing, while PVC's are of the "schedule" type
 
What you tried is very similar to what my idea is. But instead of trying to find a tube that will work for the spacers, I was going to suggest using PVC "unions" some times called "couplings". They are made to fit over the original pipe , they have a slight internal rib to center two pipes when using in their normal job, but that is easy to take out with an Exacto knife or a Dremel tool.
FWIW My (failed) approach entailed cutting several snug-fitting PVC 'washers' from 3/16" flat stock, thence placing one such 'washer' above the upper rib, another below the lower rib (having first 'stacked':oops:o_O several of same between said ribs) -- with predictably lamentable results:oops:

Well hey, @shortbus! I like your idea!:cool: -- For its requirement of merely three PVC spacer/clamps per tower (i.e. First sleeve against top of upper rib --- Second sleeve intervening upper and lower ribs --- third sleeve against bottom of lower rib).

That said -- Instead of a union, it seems 'snugger' fitting sleeves might be fashioned of PVC end-caps via removal of their end-closures --- Thoughts?:)

you are aware that your CPVC mini fed connector tube doesn't fit into PVC fittings. The two different pipe materials aren't compatible so they made the sizes so they couldn't be interchanged in common use.
yeap ... CPVC's dimensions are made to same standards as old copper tubing, while PVC's are of the "schedule" type

Correct! The CPVC well is merely 'glued' the PVC end-cap:cool: -- Said end-cap being centrally fitted with a 'banana plug' socket -- such details to be fully reveled in the forthcoming construction article:cool:

As an aside - Re: the CPVC plug and well -- Inasmuch as a plasma-tight seal is required (and unavailable via most polymers) we snugly _reach the sides_ of the well via formation of 'alignment ribs' upon the 'mini-fed' plugs (via turns of PTFE 'thread tape') as illustrated below... -- In addition to establishment of a far superior 'seal', said arrangement is readily maintained via addition and/or replacement of PTFE 'tape' as required:)

Detail of mini-fed plug (note the 'alignment ribs' formed of PTFE 'thread tape): ----- Post continued below image


Please be advised that I will be absent from these fora prior Tuesday/early Wednesday (Off-line 'life' again:rolleyes:) -- @Aleph(0) expects to resume participation here no later than early Monday...

Very best regards and many thanks!
HP:)
 

shortbus

Joined Sep 30, 2009
10,045
That said -- Instead of a union, it seems 'snugger' fitting sleeves might be fashioned of PVC end-caps via removal of their end-closures --- Thoughts?:)
A union has the same exact inside diameter as the end cap, within manufacturing tolerances, since they both have to fit the same pipe. The union is also two end caps long and can be cut to any length needed, or one could be used instead of two for a longer separation of the ribs. to my mind easier than sawing the end off of a cap. There is usually a small radius in the bottom of a cap also.

The reducers also have another benefit, the center hole is a little(don't know the actual amount) than the inside diameter of the reduced size pipe so the banana plug may fit better.
 

shortbus

Joined Sep 30, 2009
10,045
As an aside - Re: the CPVC plug and well -- Inasmuch as a plasma-tight seal is required (and unavailable via most polymers) we snugly _reach the sides_ of the well via formation of 'alignment ribs' upon the 'mini-fed' plugs (via turns of PTFE 'thread tape') as illustrated below... -- In addition to establishment of a far superior 'seal', said arrangement is readily maintained via addition and/or replacement of PTFE 'tape' as required
While I've never had much luck getting teflon tape to stay in place other that on a thread, and even that only after the first wrap, there are such things as teflon "O" rings. They would last much longer than the tape, and a groove could be made in the outside of the mini fed connector by putting it in a drill chuck and holding a file against it while it was spinning.

Are you saying that the plasma will travel up through the gap in an unglued pipe to fitting joint? That is hard to believe. There is only about 0.001" difference between the two diameters of that joint, ~0.0005 per side. And over a 1" length, with a direct connection to the internal circuit of the meter.
 
Last edited:

cmartinez

Joined Jan 17, 2007
8,218
While I've never had much luck getting teflon tape to stay in place other that on a thread, and even that only after the first wrap, there are such things as teflon "O" rings. They would last much longer than the tape, and a groove could be mad in the outside of the mini fed connector by putting it in a drill chuck and holding a file against it while it was spinning.

Are you saying that the plasma will travel up through the gap in an unglued pipe to fitting joint? That is hard to believe. There is only about 0.001" difference between the two diameters of that joint, ~0.0005 per side. And over a 1" length, with a direct connection to the internal circuit of the meter.
The problem with solid teflon o-rings, is that it's a material with very poor elastic properties. So here's another alternative: teflon-coated buna-n (an excellent elastomer) o-rings.
 
Please be advised that I will be absent from these fora prior Tuesday/early Wednesday (Off-line 'life' again:rolleyes:) -- @Aleph(0) expects to resume participation here no later than early Monday...
Embarrassingly, it seems 'off-line life' has -once again- made 'liars' of us:rolleyes: -- That said, I can imagine no scenario preventing either of us returning to regular participation next week (i.e. that of Mon May 13, 2019).

In the meantime I'm pleased to have found made time for the following post addressing many of your recent questions/comments/suggestions! -- Hope you find it useful:)

Please note -- By way of eliminating duplication, I here advise that this post continues after ALL images!:D

So... With no further ado:


While I've never had much luck getting teflon tape to stay in place other that on a thread, and even that only after the first wrap
The 'Trick', as it were, is in winding the first several turns with a degree of tautness sufficient to achieve firm constrictive tenure to the CPVC all the while avoiding undue distortion of the 'tape'...

there are such things as teflon "O" rings. They would last much longer than the tape
The problem with solid teflon o-rings, is that it's a material with very poor elastic properties. So here's another alternative: teflon-coated buna-n (an excellent elastomer) o-rings.
While O-rings and their ilk would indeed prove more durable -- PTFE film (i.e. 'pipe thread tape') is a better choice - given that:

1) Preclusion of plasma migration demands intervention by a high-strength dielectric fluid or fluid-like 'seal' between 'plug OD' and 'well ID' surfaces... -- Properly wound, 13mm X 77μm PTFE film freely 'flows' under moderate pressure filling microscopic abrasions, 'pores', etc to less than 400pm (pico meters) as well as macroscopic irregularities. --- Please note that the function of the 'tape' is preclusion of air/ion exchange with the outside atmosphere --as opposed to-- plug-to-well isolation...

Note:
The outer turns are rather loosely wound such that said 'plastic fluidity' is greatly enhanced via 'rugae action'...

2) 'Tat' is a non-issue (for the ready, swiftly effected, remediability of same) -- As a point of information, each PTFE film application is 'good for' at least fifty plug insertion/withdraw cycles - Said transience representing only negligible inconvenience in that common/good practice will generally leave the test leads attached to the instrument while idle...

As a point of interest - our tests have found that --for the purposes under discussion-- 'rugae-wound' PTFE film is markedly superior to 'dielectric grease' and only marginally inferior to ($300 per ounce) 'vapour proofing' compound - both of which being 'messy' gels...

Are you saying that the plasma will travel up through the gap in an unglued pipe to fitting joint? That is hard to believe. There is only about 0.001" difference between the two diameters of that joint
Re: the conditions under discussion -- Ion exchange becomes detectable (sans instrumentation) at channel major-axes > 500pm (picometers) and unacceptable at major axes > 650pm (Note that 0.001" ≈ 25.4 million picometers or ≈ 39 thousand times the maximum acceptable pore-channel major axis [650pm]). ---- To understand this - you need merely entertain the perspective that 'plasma' is as much a process as a 'substance':cool:

How truly: What's optimal from a structural/mechanical standpoint is, not uncommonly, less so from an electrical point of view! -- Especially at EMFs ≥ 40kV...

@shortbus - Re: the PTFE pipe 'unions': -- I believe this is what you have in mind?:) --- Let the reader be warned that Menards (sic) calls them 'couplers' whereas 'their' PVC 'unions' are akin to 'flare nut' assemblies:confused: -- Should your locale be bereft of Menards home centers - You may count such as a blessing sans the merest risk of taking liberties!:rolleyes::D



Note to readers: Inasmuch much as the 'stop-ridge edge' -to- tube rim distance (Ca. 25mm) is more than accommodative of PTFE (fault path extension) rib clamping and spacing purposes - you need merely cut the 'coupler' immediately each side of said 'stop ridge' -- Note, however, that a 'square' cut is essential!

Such may readily be achieved through one of (at least) two methods:

1) Best choice: via a 'clamp and wheel style' pipe cutter having a 'wheel' radius sufficient to fully penetrate the tubing wall sans wheel-hub/tubing-OD apposition...

2) Acceptable method: Via employment of a saw jig (commonly marketed as a 'mitre box').

Note that 'Shear style' tubing cutters are not recommended -- for their distortion of the tubing and, not uncommon, production of 'out of square' cuts!

Inasmuch as there seems to be some confusion as regards attachment of the 'banana plug socket' to the upper end caps, I offer the following by way of clarification...

The procedure:
1)
Locate center of the end-cap.
2) Drill a central bore of appropriate diameter.
3) Install the banana plug socket.
4) Secure the installation via fixing the bur with 'hot melt adhesive'

Re: 'Step 1' above -- Inasmuch as I have been unable to locate a commercial source for 'circle-center locator jigs' I was compelled to conceive/assemble my own (as follows):

Required tools: A 'square' and a 'right-triangle' (note that, quite despite nomenclature, the 'square' is likewise a 'right triangle' sans a (physical) hypotenuse...




Tools assembled such that the 'hypotenuse' of the triangle-tool forms angles of π/4 (aka 45°) with the 'adjacent' and 'opposite' sides of the square. Of course said hypotenuse will define a diametric path over the plane of a circle whose circumferential surface being in apposition with both sides of the square --- Hence, said position being established, (non-coextensive) lines corresponding to the path of said hypotenuse drawn upon the circle (CIP the top of a PVC end-cap) will intersect at center of said circle...



Illustration of an end-cap -- center located as per the above described method. Note also the uninstalled 'banana plug' socket and associated hardware (included for clarity)




Banana plug sockets installed (Top/outside view)


Banana plug sockets installed (Bottom/inside view) -- Note the 'hot melt' adhesive covering the burs --Please rest assured that the function of the small compression springs will be discussed in the course of the construction presentation!:)


Very best regards
HP:)
 
Last edited:
Top