We received the pins, plans for burn-in machine. F project | EP#6
Вставка
- Опубліковано 23 січ 2024
- The pins for glass stems arrived, we are planning another machine - another piece of the F tube puzzle!
Thank you for watching and for your continuous support!
NEWSLETTER
► www.daliborfarny.com/newsletter
WEBSITE
► www.daliborfarny.com/
INSTAGRAM
► / daliborfarny
FACEBOOK
► / daliborfarnycom - Наука та технологія
This entire process you have been through up to this point has been quite incredible. The amount of manufacturing knowledge that has essentially been lost and that you and the team are re-engineering, as well as the novel aspects of your designs that have NEVER existed or been attempted before, is amazing, and I'm very thankful you are so public with this journey. It is clear you all work towards the reviving and cost reduction of these tube displays simply because you love them (as do I), and for that your work is much appreciated.
Thank you for the comment! There is a large amount of very specific knowledge that is needed for manufacturing of the nixie tubes - this surprised me when I started.. For instance learning how the glow behaves (plasma physics) was hard, even plasma physicists (sputtering business) were not able to answer the questions.. So, thanks for following the journey!
That robot arm handled that singularity transition like a champ! Looking forward to seeing future developments!
2:32 yup. I have a couple of 1914 glass bottle molds and they're also made from cast iron. A very, very graphite rich cast iron that's you only need to touch to get that dusty-oily coating on your fingers. I believe they were also used in the early plastics industry, specifically for the Bakelite molds. Same reason as far as i know.
👍
Glad this supplier worked out, those pins look beautiful!
I’m pretty stoked that American parts are gonna be in the new tubes.
Mee too honestly.. we try to source materials from where our customers are (USA, EU), but the price difference is usually like 1000% compared to Asia. in this case the price is very good and we will stay with them..
@@daliborfarny
I'm just so happy to see that we are able to maintain at least some manufacturing specialties relevant to the world in this country.
Love your work, you guys are at the cutting edge of vintage nixie-tube re-engineering, so impressive.
Wow, thank you!
velmi zajímavé
Awesome! 1:34 they look really good, but then again they're already shoe'd in the industry so they know what they're doing. If the price's right, this is a big score for the project.
8:36 looks really good! Smart move that you didn't go with a single table and kept it modular.
Always great to see your progress....cheers !
Wow! Great progress. Excitibg to watch.
So the Nixie tubes are 0-9 only, or also some 0-F?
Very nice!
Hi Dalibor,
As usual, it's incredibly interesting to follow the steps of the development of your tubes.
I simply love the newly arrived pins - it's obvious that the company delivering them has pretty deep experience. And I doubt this could be easily reproduced in ways that were considered, so I bet that you would prefer to stick to new pins at the moment.
When it comes to fusing them within glass I have some thoughts, as usual ignore them if I overcomplicate, as I tend to do it :D
You mentioned that there's the etching process involved later, due to the oxidation issue. I may have some solution here, that would solve both fusing and etching, yet take it with a grain of salt, as I don't even know the exact alloys used in the pins, exact type of glass used in tubes and their respective termal expansion coefficients, wetting angles and other parameters. I can only guess that they are within some thresholds.
So, it's quite a common practice to use copper in glass seals, due to its tensile weakness. There are multiple processes leveraging this property of copper. In this particular case it could make sense (or could not) to bulk electrolyticly or chemically plate pins with copper and then fuse them in glass. This would provide an additional benefit of shielding pins from oxidation at this state (copper would oxidize on the surface instead). Then, etching would only attack the copper plating and etching formulas could be assembled to only attack this layer and not the alloy beneath. This would strip the plating only and stop at this point. Meanwhile, glass enveloped layer of copper would remain intact and allow for a tight seal.
Bulk cladding should not be an issue. The chemical only process (without electrolysis) is used in plating holes in circuit boards so both the reagents and the process implementation should not be an issue. On the other hand electrolysis can be used as well as pins are metallic.
As usual, my ideas may not be suitable in your process and you are doing a great job to design the best scenario. Hence, ignore the above if there are better ways.
Hi! Thank you!
- pins: yes we will stick to the professional manufacturer, they know their job very well - we need to spend our resources elsewhere!
- seals: the alloy is "Alloy 52", which is a 50:50 Ni-Fe alloy. We might try a hydrogen reduction mini-chamber right after the sealing while the wires are still hot. If this doesn't work, there are plenty of recipes for etching/electropolishing of the pins in a book by "Fred Rosebury". The copper is a good idea, we would practically obtain "dumet" wire which is also a NiFe core with thin coating of copper. We use dumet in the colon tubes in our clocks and it is super reliable. But still, you need to do a descaling of the oxidized copper after making the seal.
Let's try to keep it stupid and simple first, and if it doesn't work we can always add complexity :-)
@@daliborfarny
After descaling copper if the copper scenario is considered, it can be left without etching and plated with gold electrolitically. It would make it more reliable in terms of contacts' resistivity and corrosion resistance, plus would make it look cool.
@@daliborfarny Somehow my longer entry wasn't posted, or got removed.
You can Descale copper of oxides with acetic acid in 35 deg. C. Also, I hope it can be done without the Dumet way.
Damn thats very cool!
I wish you the best of luck with your projects!
Looking forward to the new tubes! One question, will your Umashimenkana Artwork be visible somewhere now that the Hiroshima Alfredo Jaar exhibit closed?
I dont know, I guess next event will be again in Japan..
Dzięki!
Thank you
I am curious to see how you will solve the problem of separating the pins and inserting them in the mold in a quick way, since doing this in production would be really slow by hand.
we will start doing it by hand and later add automated feeder - like a vibratory feeder bowl or something in this manner.
@@daliborfarny One way to make inserting the pins faster would be to create a semi-automatic dispenser, like a handheld device that holds a bunch of pins and when you press a button it dispenses one, this could be made fully mechanical. Loading a device like that with a bunch of pins and then dispensing them one by one would already reduce of time spent loading the pins. For the mechanism, look at nail gun mechanisms for inspiration
Good idea, this might be a nice first step before something automatic. There is a video from crt production on YT where the operators use a jig like you describe.
Super jako vždy Dalibore, už se nemohu dočkat až mi eFka budou svítit u mě doma 🙂
Díky, intenzivně na tom pracujeme :-)
Are the rejected tubes completely unusable or would it be possible to sell them cheaper without warranty?
They work normally, but have higher risk of early failure. We keep them for a bigger installation I have in mind ;-)
Привет Далибор. Подскажите вдруг знаеье... Какой схемой получить стабильные 150 вольт вне зависимости от входного напряжения. Спасибо.
Why having the whole rpi on the table taking space in the arm's path ? Can't the rpi be mounted underneath, while a compact USB hd webcam the only thing that sticks out ?
Good idea! I have a suspicion that rpi camera might not be enough and something more complex will be needed - to eliminate different lighting scenarios (day and night) etc.. Let's see!
@@daliborfarny Take a look at openPNP they are using the cameras for checking the SMD component locations, it uses an open source protocol from memory
If tubes are prepared (inserted in socket) for that testing stand manually, how will you prevent that operator put pcb with tube in incorrect position?
I mean they can accidentally rotate it for 90 degrees, for example.
You would need some additional step to check this (up to machine vision). Or you can add a key to the pcb, so it can be placed in the frame of the box in one position only.
The operator will be trained to place it correctly, but the machine will need to be ready for that problem.. The routine will start with placing the tube to the visual inspection socket - the arm will rotate the tube by 90degrees until it recognizes a serial number or rotates by 270 without finding it.. The keying is good idea!
@@daliborfarnyThere could also be an additional contact on the boards, if the adapter is in the correct orientation it would short 2 pogo-pins, informing the machine that it's oriented correctly and the adapter board is also correctly seated. This could also be used to add a resistor on the adapter instead of a short and allow the system to check if it's the correct adapter or the machine could have 4 sets of pogo-pins, depending on the orientation of the adapter board one of them would get shorted as the adapter is put in it's place and tell the machine if it's oriented correctly or if it needs to rotate by 90 180 or 270 degrees, reducing the time needed to find the correct orientation
Hi Dalibor,
I know this is still early in the development of the manufacturing process but have you considered using Computer Vision to do Quality Assurance on components of the F tubes? Once the factory is all setup would you consider having automated checks on the glass and glass metal joints and placements of the digits etc? Modern automation could help reduce and indentify systemic errors? What do you think?
Hi! Yes, we already started creating tube rejection criteria - based on shapes of the tube. To be able to tell what is straight and what is already crooked (esthetically, to human eye). Each tube would go through the visual inspection and get its rating.. e.g. if the angle between the grid and the envelope is more than 0.3 degrees, add -50 to rating. If the angle is above 0.6, add -150 to the rating. If the rating is below -200, reject the tube. While 0.7 degrees is already enough to reject te tube from our experience. Do you have any experience with such a systems?
@@daliborfarny maybe the Keyence Vision-sensors could help you with the Task. Also take a look to the Basler pylon tool.
Hi Dalibro! Compliments for your Copanyy! . I have been following your efforts since almost from the start. I am a (relatively) young experimental physicist dabbling in surface science, cathalisys and UHV, and I have used (legacy) nixie based counters in my lab experiences. Mostly HP hardware and such, i guess from the 60s and 70s.
Can you give us a benchmark on how your wonderful nixies perfmorm in terms of static power consumption, switching power requirements and durability (average light hours per figures or average numbers of figure switche?)
Hi, welcome! Nice to hear the good old nixie instruments are still in use! The power consumption of the R nixie tube is under 1W per tube, including the resistors around. The lifespan of our tubes seems to be at least comparable to old nixie tubes - the oldest tubes that were running 24/7 for around 7 years now have a very light deposit on the glass - but observable only when side by side with a new tube. Old tubes were rated for 200k hours (22 years), I can imagine ours will live that long but deteoriation will be apparent after that time.
yes i rememer it was an HP 5245 Family of freq counters. old but good! eheheh. Didn´t know which nikies they are using tho, but they were of the "axial" form factor (number showing from above the tube.
Thank you for the info and keep up the good work! @@daliborfarny
@@daliborfarny I had no idea they last that long. That is quite impressive contrasted to the modern planned obsolescense of consumer products.
very impressive. That means you already managed to control and reduce the main bottlenecks on the tube quaity and performance, i guess metal quality and etching procedures, and vacuum/getter controls. what in your experience is the most important factor to control and reduce? @@daliborfarny
Do you guys plan to try to make a cheaper version of your clocks at any point in the future? I really want one, but it's just so expensive.. :(
Thats the point of this project, cost effective production, lower price!
@@daliborfarnyOH! So this is for the new clocks! Ok, didn't realize that, my bad. I love it guys, I am so exited.
It’s not “way how to.” It’s just “way to.”
I thought of you, really! So, at first I said it wrong, but at the end of the video I used it correctly. I am proud I make progress :-)