AoE x-Chapters, section 9x.23, power on a light beam.
formatting link
For my electrospinning project (recall 15kV+4kV=19kV, now 30kV), we need to power an LED voltmeter module, to measure the nano-thread production current on the high side of a 6kV power supply. Measure to 0.1nA with a 10k resistor. DATEL (now Murata) DMS-40PC is a 4.5-digit meter; it needs 5V at 35mA = 180mW.
I don't trust the long-term reliability of isolated dc-dc modules at 6kV, and we might need to do 25kV.
In section 9x.23, we used a 100-watt COB LED with a 65x65mm solar panel, spaced at 25mm, and got 750mW. We got 300mW from a 25x35mm panel, with LED at 35W.
On Oct 6, 2019, Winfield Hill wrote (in article ):
and silicon photocells. If the LED is 20% efficient and the photocell is 10%, one would expect about 2% transfer efficiency, better than double the above
As for the solar cell, the reflection from the face may be larger than expected. A standard dodge is two cells arranged in a Vee, face to face, with IR beam entering the wide opening and bouncing from face to face until fully absorbed.
Yes, that's what we get with an inefficient optical design, using eBay parts. When I needed an isolated 150-volt supply on top of 25kV nax, RIS-623, we wound a transformer, using a machined bobbin with 0.25-inch thick plastic walls, on a TV flyback ferrite core. It was rather large, but worked well.
Now I'm thinking for 6kV, I can try RM10 with split- bobbin core having a 0.6mm divider. It's large enough I can add extra layers of Kapton tape. I could soak it at 8 or 10kV for a few days to see what happens.
On Oct 7, 2019, Winfield Hill wrote (in article ):
I was going to suggest something like this, but a big pot core can work.
I assume that the bobbin is nylon. If so, 0.6mm seems a bit thin, but
that it will be arranged so there is no safety issue if it sparks over.) Can you vacuum pot this in silicone rubber, like an ignition coil? Or in a wax-filled can? Those old flyback transformer were often vacuum wax potted.
If the pot core is large enough, a three-section bobbin with the center section empty could also work. Joe Gwinn
Maybe I should pot the finished transformer. A few years ago I re-supplied with Dow Corning Sylgard 184, etc., recommended for HV potting, better than the stuff you get at Micheals, awwkk!, but other labs use it for PDMS, "borrowed" my stuff, and now I'm out again.
Why a LED display? LCD is much lower power. Or send the data wireless to a 'remote' display. Bluetooth is very low power. Since you are spinning something there - cannot you tap some mechanical energy? I remember a Van de Graaff generator, the top end of the belt drove a dynamo inside the ball for all the measurement power supplies. Data was sent down digitally over fiber links. Mega-volts? No problem.
The LCD version with backlight actually takes more power. I'm making a small attractive instrument, with two LED display meters. Thanks, but my new plan to make a 6kV transformer takes away any advantage for 2.5mA vs 35mA. I'm making the RM10 bobbin PCB footprint right now.
I did 5 watt 100 kHz 6kV rated transformers on RM10 cores using wire-wrap wire, and one extra layer of teflon tape (flows better than kapton) around the secondary wiring. Did not use the bobbin pins but used the long wires.
I tested each of them 6kV AC 1 minute before actually soldering them into the circuit. It was for medical grade CF isolation (direct to the hearth) and I had to validate and sign each report personally...
Why do you want to run the display at 19 kV potential ?
Why not run a minimum circuit at elevated potentials which performs an ADC and sends back the result as serial data, to be displayed on a 7 segment display close to ground potential ?
On EHT power lines, the current transformer sits on the elevated 400 kVac potential. The measurement is often taken down by shining a laser into one fiber and using a solar cell at the EHT potential to power the ADC and an other fiber is used to transport down he measuring result in serial format.
To avoid the inefficiency of running the downlink transmitter LED, you could consider extracting a part of the uplink light power to a mirror and LCD shutter to control the amount of light forwarded back to the downlink fiber, controlled by the serial data.
Ah, you're a good person to have in the conversation! A part rated with 6kV for a standard one minute test, probably shouldn't be used over 3 or 4kV continuous. I'll increase tape layers, and pull inner bobbin pins, but leave two outer ones for mounting. They're only 5.5mm from the core, so I'll use long wires like you. I add teflon sleeves to the wires inside the winding, so the four exiting wires will have teflon sleeves. Kapton tape over windings, to insulate from the core. Then cross fingers, and do long excess-voltage tests. But I'm also preparing a HV potting scheme, using Dow Corning Sylgard 184. That should handle it!
Win just got close to something I did about 10 years ago for my former supe rvisor, Professor Reneker. We had very high control of our E-Spin station' s environment so we just stuffed 8 AA or D cells into a project box and fl oated the whole current meter. I can see the desire to have a finalized in strument with a laser pumped power supply, and we briefly looked at it.
One of the problems when you have twelve or so grad students working wi th up to 60 Kv is they will often get in a hurry, forget a procedure or a g round clip, and fry the serial port or USB port on a PC, So I very quickly adopted plastic fibers for isolation in the lab. That and I kept a pile of cheap serial port bus cards in a shielded, grounded, drawer.
So I figured I can finally disclose the simple technique as the lawyers rejected the concept of patenting it when Electro-Spinning was a very hot topic a decade and change ago. There were similar patents on using a NE2 in series with a load to sense on/off but none working over six to nine ord ers of magnitude. Despite the boss's best attempts, the lawyers declined.
I'm going to try to publish the new version shortly so I can explain the o ld version today. Better to get it out there and in use.
You optically couple a neon lamp to a plastic fiber. Hand Selected NE
2s will often have a hot region on the upper part of the electrode which ha s enhanced emission. If your lucky this is on the side of the electrode fac ing the bulb wall. It becomes a simple matter of gluing the large plastic f iber to the side of the selected lamp. You may need to strip a clear insul ating coating off the lamp with a little abrasion. NE series Lamps were of ten dipped in a hydrophobic coating in the old days.
Works fine down to sub-nanoamp and the light output is VERY proportional to the current down in the starved region of the lamp's operation.
The GE Glow Lamp manual has a wonderful plot of the lamp voltage all the wa y down to 10.0 to the minus 15 Amperes. So I'm not the first to discover this . I found the plot after I tried the idea.
Keep the photodiode, PMT, etc at a constant temperature and it makes for a remarkable instrument. Bandwidth is more then sufficient for the motional current components in a classic E-Spin jet.
LEDs did not fair well in this application, protection clamping diodes then available did not react fast enough. A typical ESpin frame or apparatus ha s a lot of stray capacitance, so it does not take much to pop an LED before it starts to conduct. The mechanism is puncturing of the die. Large die LE Ds for LED lighting were just coming onto the market at the time, and were notoriously static and reverse voltage sensitive.
Steve
What I'm working on now will not use a stock NE2. I can make my own glass t o metal seals so I have something newer that has a few added techniques.
Could you use the measured current as the power source? Maybe charge a cap and make a relaxation oscillator. Pulse a diode laser into a fiber when the cap fills up.
I had a similar problem a while back, trying to blink an LED from a very low current. That needs a relaxation oscillator with basically zero standby current.
One could cheat and send sampling/breakdown pulses to the relax gadget, optically or magnetically.
I used to work on ships that had a strain gage on the big (like 3' diameter, 32KSHP) prop shaft, to measure torque and horsepower. Power was transmitted to the rotating shaft magnetically, and the signal was returned capacitively.
After the noise and heat and chaos of the engine room, the shaft alley was a cool, serene retreat, with that big shaft quietly turning.
How about using lithium batteries that have to be replaced every 10 years, and fiber?
On Thursday, October 10, 2019 at 12:18:32 PM UTC-4, snipped-for-privacy@gmail.com wro te:
pervisor, Professor Reneker. We had very high control of our E-Spin statio n's environment so we just stuffed 8 AA or D cells into a project box and floated the whole current meter. I can see the desire to have a finalized instrument with a laser pumped power supply, and we briefly looked at it.
with up to 60 Kv is they will often get in a hurry, forget a procedure or a ground clip, and fry the serial port or USB port on a PC, So I very quick ly adopted plastic fibers for isolation in the lab. That and I kept a pil e of cheap serial port bus cards in a shielded, grounded, drawer.
rs rejected the concept of patenting it when Electro-Spinning was a very ho t topic a decade and change ago. There were similar patents on using a NE
2 in series with a load to sense on/off but none working over six to nine o rders of magnitude. Despite the boss's best attempts, the lawyers declined .
old version today. Better to get it out there and in use.
NE2s will often have a hot region on the upper part of the electrode which has enhanced emission. If your lucky this is on the side of the electrode f acing the bulb wall. It becomes a simple matter of gluing the large plastic fiber to the side of the selected lamp. You may need to strip a clear ins ulating coating off the lamp with a little abrasion. NE series Lamps were often dipped in a hydrophobic coating in the old days.
to the current down in the starved region of the lamp's operation.
way
is.
a remarkable instrument. Bandwidth is more then sufficient for the motion al current components in a classic E-Spin jet.
That's very interesting! Thank you Steve. Is E-spinning done at AC or DC? (or both?)
en available did not react fast enough. A typical ESpin frame or apparatus has a lot of stray capacitance, so it does not take much to pop an LED befo re it starts to conduct. The mechanism is puncturing of the die. Large die LEDs for LED lighting were just coming onto the market at the time, and wer e notoriously static and reverse voltage sensitive.
The LED's blew from reverse voltage? (I measured some leds that could take >100V before conducting in reverse.) But at low currents the led light is not at all proportional to current.. I measured something like the 2/3rds power. George H.
(I found GE glow lamp manual here..)
formatting link
%20Edition.pdf
to metal seals so I have something newer that has a few added techniques.
supervisor, Professor Reneker. We had very high control of our E-Spin stat ion's environment so we just stuffed 8 AA or D cells into a project box an d floated the whole current meter. I can see the desire to have a finalize d instrument with a laser pumped power supply, and we briefly looked at it.
g with up to 60 Kv is they will often get in a hurry, forget a procedure or a ground clip, and fry the serial port or USB port on a PC, So I very qui ckly adopted plastic fibers for isolation in the lab. That and I kept a p ile of cheap serial port bus cards in a shielded, grounded, drawer.
yers rejected the concept of patenting it when Electro-Spinning was a very hot topic a decade and change ago. There were similar patents on using a NE2 in series with a load to sense on/off but none working over six to nine orders of magnitude. Despite the boss's best attempts, the lawyers declin ed.
he old version today. Better to get it out there and in use.
d NE2s will often have a hot region on the upper part of the electrode whic h has enhanced emission. If your lucky this is on the side of the electrode facing the bulb wall. It becomes a simple matter of gluing the large plast ic fiber to the side of the selected lamp. You may need to strip a clear i nsulating coating off the lamp with a little abrasion. NE series Lamps wer e often dipped in a hydrophobic coating in the old days.
l to the current down in the starved region of the lamp's operation.
e way
this.
or a remarkable instrument. Bandwidth is more then sufficient for the moti onal current components in a classic E-Spin jet.
The standard cable on many US made HV psus is a PTFE diectric coax with a s olid core that is almost a half inch in diameter covered with 100% braid. Something like 60-90 pF per meter. A few meters of that or HV30 stranded w ire at 18 pF per foot stores a lot of charge and discharges very quickly.
Couple that to a very well built six stage multiplier in the PSU and large pulses form. As our psus were built for a minimum of 1 mA at full voltage, output, even forward biased leds suffered.
The collector plate or spinning. drum for collecting the fibers probably ad ds 30-60 pf, especially when grounded via a 10k or 100k current sensing res istor.
Hindsight now tells me a string of large die leds with equalizing caps or resistors across them might have had a chance... but hindsight is 20-20.
Our isolated little 9$ lcd current meters had a 0.1 uf 3Kv bypass cap ac ross the sensing resistor, and series resistors in the leads. Usually i'd get nine months to a year out of a 300 mV FS lcd. Either 10k or 100k would be used as a shunt, and I prefered metal film for this application.
Velleman water resistant plastic enclosures with clear lids and the sealin g gasket are expensive , but work well for floating current meters.
The psus had their own current sensing shunts in the ground end of a float ing multiplier assembly, so that was probably 10k more of Burden.
My preference is for DC, but AC will work with certain polymer solutions. Ac is sometimes used in certain bulk production situations, but for that I prefer to use DC and mechanical motion with wet rough surfaces instead of nozzles to create more Taylor cones or Taylor cones with multiple active je ts, which we refer to as "sisters".
The humble neon served me well.
There is as much art in espin techniques as there is science.
7% by weight, 300,000 average molecular unit, polyethylene oxide in water, is a good starter solution, and will make fibers with a starting voltage of 6-7 kv and a running voltage of around 2200-2500v at 200-300 nanoamps for a single jet. That is within range of the very inexpensive E100 Psu brick w hich has an output proportional to its DC input.
Add a tiny trace of Rhodamine 6G or Kiton Red dye and you get about a three month storage life for the solution, which otherwise only lasts a week or so from UV / visible light breakdown or bacteria eating it.
The current is so low you can collect the fibers on your hand, and they wil l be 200-300 nm in diameter. You can stick your hand in a well designed res earch setup and collect fibers, which leads to a lot of wound treatment res earch using the fibers as scaffolding for cell growth.
Not a recipe for making a structural material, but for research and demos i ts a great starter. Takes 2-3 days for the powder to dissolve, but wonderf ully non toxic.
The E-Spin lab was on the campus school tour, so I had the pleasure of givi ng 20-40 minute demos to several thousand school kids and their parents.
If you have the right lighting a clump of fibers floating in open space during the process is quite visible to the ey e and a camera, and with a bit of care 200-300 nm fibers are very visible t o your eye streched across your fingers. The individual strands are quite t ough, breaking strengths on the order of a gigapascal have been measured.
ISTR we used off the shelf module isolation current sources for powering filaments on mass spectrometers at 8kV (and 5A precision variable). They were pretty reliable in service but I don't recall any more details.
If you have the space at the receiving end then a simple flux concentrator will get you a worthwhile improvement in energy transfer for little effort. The simplest is putting the PV at the base of a half hexagon with mirrors at 60 degrees either side (source at infinity).
\_/
The one which allows the greatest gain is a truncated dismembered parabola with the foci at the opposite edge of the PV cell. I have had a usable 10x optical gain using this slightly bulky configuration
formatting link
You might want to go back to the original papers for non focussing flux concentrators as some parts of the Wiki article look a bit dodgy...
A near point source high power LED emitter and a lens will also allow you to deliver a lot more energy directly to the PV.
We're using PMMA (acrylic), for solute. Our collector currents are only 3 to 5nA at 15kV. It's very hard to see the jet. We moved the collecting surface (aluminum foil for now) far enough away to get a 2-cm deposit patch. With the very low current, I worry that the fibres are smaller than we need. We can't see them with our lab microscope, need to use one with higher magnification.
ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here.
All logos and trade names are the property of their respective owners.