Thanks Dan.

One side of anode has been milled. One to go.

Some cleaning and deburring to do but its as close to accurate as I can get manually. Progress anyway

In focus this time

Something like this. My cathode is not seamless, so I have ordered some seamless tube to replace this. That should be a simple job to replace.

Spectral analysis tests, both anode and cathode seem to be resonant at the same frequency (within 100hz) with slots milled in anode. More testing to be done. Test completed by exciting through striking and measuring with spectral analysis. Very interesting. It is possible that i now need to spend some time adjusting anode slots to tune more precisely. The difference (circa 100Hz) may be in part due to some deviation in my part vs the target dimensions. I will however replace the cathode with seamless tube before doing any more tuning.

Further work with demo cell 3D design. Making the body a smaller diameter (40mm) this will make it much cheaper to make when using delrin/acetal and faster to 3D print, too. Images to show more clear cutaway detail and connection point/holes added for electrical connections. For printing you only need two files: Top-part and Body.

cool nice post like the layout  how is the gap between the inner and outter in mm ?

Hey Dan. Thanks. Do you mean between anode and cathode or the gap between the 3D printed parts?

If you mean cell distance, the anode is 12mm OD and the cathode is 15mm ID so 1.5mm gap. For 3D printed parts, I have modelled a gap of 0.1mm between parts that should fit together. I am not sure if thisnis enough for 3D printing yet, as it is not so accurate but I may have to adjust that by hand

I intend to machine this on the cnc lathe too so want the model to be accurate as possible so do not mind adjusting the prints by hand a little.

between anode and cathode

It's 1.5mm/1.6mm Dan. Depending on the accuracy of the stock.

I made an error on the previous models so am correcting it here. I may have to make some additional adjustments for the inaccuracy of the 3D printer, but I don't like to do that in the model itself. New files attached

cool

Whilst I await tooling and material. I have completed some more theoretical work. Based on some of my research from others (Petkov) I have started to plan towards a VIC. I began also with the VIC measurements from the estate in Russ's posts on this site. I have created an image to help with construction and a calculator that can estimate turns to a desired mH value. I have used this many times in EQ design. It's close to within 10-20% normally.

In order to get accurate values, there are notes towards measurement steps to define Al value for the core in question. Variations can be +/- 25% and that does not account for the 'broken' core. Measurement is the only way. Working files attached.

Values in the xls currently are based on 3400nH core Al value taken from the data sheet (which is certainly not the case with a broken core).

I think it is important to specify things in this manner, as cores become unavailable having a target mH value (and Q) allows the process to be translated to other cores. At least, that is my preference.

Interestingly, the winds suggested by Petkov are equal for C1 and C2 whereas the original unit measurements show deviation here. This could be due to inconsistency in winding but should not be discounted.

C1: 76.28mH @ 6.27Q (1kHz open)                  C2: 64.26mH @ 5.79Q (1kHz open)
C1: 1218.8mH @ 69.5Q (1kHz on core open)   C2: 1093mH @ 68.9Q (1kHz on core open)

A difference of around 10-15%

please consider to always make cell pairs  series

as it fixes several items like balances the resistance between inner and out tubes across 2 cells
please consider gap placment on ferrite havles to be same as meyer
use u cores
insure you add the pickup coil into total resistance   
the winding can be various turns based on the build equasions of cells
remember transistor or faster
 the way you wind is also important if you are carful to line or stak wire on bobbin
or use square wire you will gain 15% + performance from coil capcitance
so way you wind is very important not just on a lathe random
the wire from vc core= booins to cell should be carefully selectred with correct
gold joiner like thin HV anntenna coaxial  for VHF HV   
 these are just notes there are threads on this forum explain each of those items
of coarse it can work with simple build but better build gets better results
dd

Thanks Dan, very good points.

I plan on making a pair of cells to start for testing as you suggest. Will like start with simpler inductors and higher input voltages and work towards this, even though I have the cores etc.

While I have built some coils not sure I did it right as coils are not my thing.  I am in process of trying to take measurements of coils I have just not sure I am doing it right. Just purchased a signal generator to help with taking measurements so I start with a clean know signal. There is a lot going on with coils. 

Best thread I found on this is the one by Ronnie Walter who works through my it works (he did detailed math in another thread) and also talks about what you need to do to balance everything.   I have read this thread multiple times (including again this week) and understand more each time I read it.   As the thread is extremely long, I have copied key parts into a word document and that is still over 100 pages long.  This document covers most of issues Dan mentioned above.  While I have already attached this document in another thread, I attached it here as I know how hard it is to find things in here.

On of key things is you must get a DC voltage bias of 1.23v across cells correct before things with work.

I also attached a second collection from another thread that where Ronnie walks you through the power and impedance calculations.  He says using this information you can scale your system to different number of tubes like the 6-tube system he has working.

These 2 threads I found to be the most helpful, and threads by NAV have a lot of information about testing he has done on coils.

Earl

 Nice work

DD

note coil i have bifilar professionally for e cores
wound and working is ID 15 mm

Quote from Evengravy on January 6th, 2023, 11:08 AM

Hey Dan. Thanks. Do you mean between anode and cathode or the gap between the 3D printed parts?

If you mean cell distance, the anode is 12mm OD and the cathode is 15mm ID so 1.5mm gap. For 3D printed parts, I have modelled a gap of 0.1mm between parts that should fit together. I am not sure if thisnis enough for 3D printing yet, as it is not so accurate but I may have to adjust that by hand

I intend to machine this on the cnc lathe too so want the model to be accurate as possible so do not mind adjusting the prints by hand a little.

NOTES
Length input
4 inches=2020 volts
5 inches=1900 volts
6 inches=1780 volts
7 inches=1660 volts
8 inches=1440 volts
9 inches=1320 volts
10inches=1200 volts
11inches=1080 volts
12 inches= 960 volts
13 inches= 840 volts
14 inches= 720 volts
15 inches= 660 volts
16 inches= 540 volts
17 inches= 420 volts
18 inches= 300 volts

Thank for that Earl, I'll keep these docs safe somewhere and will digest asap.

Thanks also, Dan. I had a good understanding that smaller cells would require increased voltage for the same effect but these figures are very useful. I am starting with the smaller cell to do some analysis of it directly and learning a lot by replicating the original cells as closely as I can which will be of help in other realms. I actually intend to build the larger tube set as a test cell for first practical tests (again for learning).

As an update, I think I have found a suitable source for E52 core via mouser int (listed as EC52):

https://www.mouser.co.uk/ProductDetail/EPCOS-TDK/B66341G0000X127?qs=%2FsLciWRBLmC%2Fyp7MKxtOxQ%3D%3D&countrycode=US&currencycode=USD

Perm is 2000 and it is ungapped, but can play with gapping distance manually. Dan, could you confirm the size of this will match your cree core?

No worries if not, I can wind my own.

Notes from estate. 8xa

Comments from Don Gabel: ..."On Stans original coil, he used thin cardboard between each row of wire. The count of 14 turns was an estimated guess. I didn't unwind it to verify. I just measured the wire size and the width of the core and calculated that 14 turns would fit."
Don Gabel

Seems to be a tight fit for 14 Turns, 13 more likely (although the extra space might be due to gapping the core halves). Should be easy enough to test via inductance measurement. (even)

On the coil measurement front, Earl. I use these meters: https://eleshop.eu/de-ree-de-5000-lcr-meter.html

They are not exactly cheap but are accurate and very cost effective compared to lab grade tools. I use mine a lot but modified them with Kelvin connectors which makes them even better. It's simple to do: https://www.youtube.com/watch?v=3mHemRWq_eg&t=0s

I am going to need a differential probe for HV (any hints welcome). The most affordable one I can see is this: https://www.sainsmart.com/products/micsig-high-voltage-differential-probe-dp20003

That's the one I have. The one issue I have with it is my scope does not have a matching probe setting.  Probe is either 200 or 2000 and my scope only has 100 or 1000 settings.  I can still use it on my scope reading are just off by a factor of 2.  They have another model that matches my scope but cannot read higher voltages.  So, issue is not with probe but with my scope as it is an older model and I cannot update probe settings on it.

Thanks for the information on meters. I did check on-line and Amazon has version with 2 different types of connectors.

I have found differential probe was very helpful when looking at connection to cell as there is no ground reference.  I did take readings using just scope but had to tie ground side of scope to system ground which effected the readings.  I found I was using 2 probes to try to create a differential reading using math function and with only a 2-channel scope I then lost ability to see input signal reference to the coils.

Thanks Earl. I can live with the factor of two difference, I think. I expect my scope would be similar but I think it would make things much easier for comparison purposes and I'm not that interested in accurate measurements (more comparisons I expect) so ill try put some funds aside for one soon. I could design a dif probe but last thing I need at the minute is another project.

 Gap IS important as it assists several things and creats capacitence and step chargingand creatses a form of cold voltage

spec not sure youhave to try eveyr thing works but  to what degree better or worse is unknown and too meany variable
weall try exact replication first

DD