| Electronics > Projects, Designs, and Technical Stuff |
| Cutting open tiny filament bulbs |
| << < (3/4) > >> |
| jpanhalt:
Water has been mentioned as an aid to spreading a scratch to become a crack. A little spit works better. Another trick is to make your fine scratch. The take a glass rod (2 mm or less) and heat it to make a little red blob on the end. Then touch it to the crack. Again, spit helps. That is a safe was to open evacuated ampules. When you make a scratch, don't saw back and forth. That decreases its effectiveness. Back to your basic problem, glass works entirely differently under water than dry. I can even demonstrate cutting it with scissors under water. Grinding is great for cutting glass bulbs under water. |
| richnormand:
I never got reproducible results with scoring and cleave or the Dremel approach (has too thick a blade and not that good at low speed). Never mind hand holding the motor either. If you are going to do several specimens I would be looking at an old low speed diamond wafering blade saw from Buehler Isomet. It has a micrometer adjustable arm and variable pressure that would be ideal for what you are trying to do. Precise adjustment of the depth, distance and angle of your cut. The blade is water cooled. They are sometime available on fleabay for a fairly low cost, even including the diamond blade. Another suggestion is a diamond wire saw. I made my own saw from a small low speed 12V motor driving a gear to give a see-saw movement to the diamond wire attached to a pulley with a weight for tension and the sample in between. Again the diamond wire is cooled and lubricated with drips of water. Also you can buy small length (2m or so) on ebay and do it by hand or motorize the process. In both cases the quality of the sawed surface depends on the speed. In both setups it is easy to wait an hour but you get excellent results. I can post photos should you wish to explore further. Good luck with it. |
| james_s:
The belt sander method is really quick and easy, I think it took about 5 seconds to sand the end off leaving a roughly 1/4" hole. At the time I was experimenting with a vacuum pump I had repaired and wanted to see if it would pull a good enough vacuum for an incandescent lamp. Incidentally it worked, the lamp ran for several hours on the pump and I suspect the main reason it did fail was a combination of being a tired old bulb and the vibration from the pump. |
| Globe Collector:
Hmmm, TCD for GC....interesting project. I'd take the chemical approach. Cover the lamp in wax, scrape away the wax where you want the bulb removed and dunk it into HF and etch the glass away. I'd monitor the filament resistance and detect the change once the HF gets through and use this end point to sound an alarm or activate a solenoid to add a lot of sodium hydroxide or lift the lamp to halt the process to prevent the HF etching through the fine filament. I reckon you would only loose a few lamps in trials to be able to refine the process to be reproducible and reliable. All the mechanical methods mentioned above are quite valid...but amazingly difficult to implement on such a small lamp where just holding onto it to take all the reactive forces of dremel blades or WC scratching tools will be an issue in itself not to mention the sheer size of the cutting edges compared to the bulb diameter itself. When I saw this post it bought back memories of Chemisrty at Uni. I remember learning about FID's (Flame Ionization Detectors) in an undergraduate lecture and at the end if the lecture I told everybody that I could probably build one out of bits from the junk box...they all "poo-hooed" it as "impossible", which was, of course like a "red rag to a bull". So I built it, i used a nozzle from a disposable cigarette lighter and cut the ends off...and pushed the "jellyroll" out of an axial electrolytic capacitor to get the cylindrical sensing electrode around the flame, I used a high voltage corona wire unit from a photocopier and I recall that the amplifier used a 12AU7 valve as the gain stage because of its high input impedance and immunity to high voltages. I took it to Uni a few days later and convinced them to let me connect it to the output port of the column in one of the GC's there...it worked just like the "real" thing and they all had to "eat their words". |
| james_s:
I just went and tried this right now since I realized I still had a burned out miniature lamp in my pocket. This took about 10 seconds on my belt sander, it was trivially easy just holding the lamp in my fingers and buzzed the end off. I wouldn't even consider messing around with HF when there's such a simple and easy way of doing it without any nasty chemicals. |
| Navigation |
| Message Index |
| Next page |
| Previous page |