Those are merely the default colors that OrCad used back then.
You are being silly. They are not "my colors", and my response to john was in defense of your poorly focused 'image'.
Wrong. In fact green on black was deemed decades ago to be the most readable, and least wasteful, and that remains true. Amber on black was a close second.
But anyway, the colors don't mean shit. I could have printed it out, and taken a snapshot of it, like you did with your hand crap, and it would have been far better, because I also know how to use a camera, or at least keep shooting till the right shot gets taken. sheesh.
You're a complete dweeb. and an idiot. IBM used black screen displays for their AS400s and they were only 16 color, and you can bet they had high visibility BECAUSE of the rendering being on a black background.
Something you need to re-study.
I'll bet you had to think about that first.
Colormetry is not a word. AND you are wrong, as it relates to the display technologies that were around back then. Everything was CRTs and white backgrounds caused short display life and they were not cheap on CAD workstations.
Of course we all use white or cream or beige or even neutral gray for backgrounds now, but this is now, and that was then, and if you had any true observational capacity, you would have noted that I was using legacy apps from the old days under DOSBox.
My current EDA design software does not sport such color schemas. Maybe you are just a newbie to not recognize legacy OrCAD when you see it. Or just too dumb (or lazy) to remember the details, which raises questions for someone in this industry.
I'll bet that most of the others in here remember legacy OrCAD and it color schemas, and the problems associated with changing them.
On a sunny day (Sun, 05 Dec 2010 13:40:21 -0800) it happened life imitates life wrote in :
Take some of your own advice. As you know shit about colormetry. Who cares what who did before you were toilet-trained. Try drawing on a piece of A4 with pencil. Van Gogh did not use Orcad either. Had he donnit nobody would have paid what they pay now for his work.
There was a WWII sniperscope (IR rifle sight) that ran for months on one D cell. The tube needed a kilovolt or two. It was a flyback switching regulator.
On a sunny day (Mon, 06 Dec 2010 12:00:25 -0800) it happened John Larkin wrote in :
series
Sure, but this also does a MAX 232, a LM317 with bleeder resistors, a HV divider resistor chain, some voltage divider stuff, and I did not even bother to optimise, simple used a TIP140 (darlington with build in reverse diode) as switch. It's Vce sat is already a volt or so, driving it via 100 Ohm from a 5 V PIC output, all that small stuff eats milli amps, so I was sort of surprised at the low power consumption. Never intended it to run on battery, but it looks like it could. This is actually also a (TV) flyback. The hystereretic voltage control implemented in the PIC with it's hardware comparator and hardware PWM unit works beautifully. No special chips needed...
Hm.. what was the switching device? Electromechanical?
Best regards, Spehro Pefhany
--
"it's the network..." "The Journey is the reward"
speff@interlog.com Info for manufacturers: http://www.trexon.com
Embedded software/hardware/analog Info for designers: http://www.speff.com
On a sunny day (Mon, 06 Dec 2010 12:00:25 -0800) it happened John Larkin wrote in :
series
PS, the scintillator is working! Some radium test shows huge signal! Was planned for the Christmas, ahead of schedule. I went to the supermarket and bought a roll of magnetron 'food' foil. it is pure polyethylene, seems to work, just put layer after later over the PMT front. Still needs more testing, with and without the foil, all late night (12 o'clock) play .. But I am happy with the results so far, had a little dark current problem, turned out to be a wiring error on the PMT PCB, now very good, into the nA range I think.
Yes. It was a watch spring that closed a contact briefly, about once a second. That drove a step-up transformer in flyback mode, and then a cold-cathode rectifier and a filter cap.
There were, maybe are still, some Russian night-vision viewers where the switching supply is literally a battery and a pushbutton driving the flyback transformer. You pump it until you get the brightness you want.
"However, the radioactivity is a major concern for those people involved with the manufacture of mantles and with contamination of soil around some former factory sites [4]. All of these issues have meant that alternatives, usually yttrium or sometimes zirconium, are used in some countries although they are either more expensive or less efficient."
formatting link
"Coleman, the best-known U.S. manufacturer, phased out thorium gas mantles in the early 90s. (Today its mantles are made with nonradioactive yttrium.)"
On a sunny day (Mon, 06 Dec 2010 20:49:28 -0800) it happened John Larkin wrote in :
range I think.
I just had a chance to test it without the polyethylene and I do not see much difference. I now think the PMT sees light flashes from the radium watch hands I am using, not from the poly-thing, although I do not see anything with the normal eye even in the dark. What confirms this is the fact that I get little from the Uranium sample. Setup: ftp://panteltje.com/pub/sc_pic/PMT_with_radium_in_bag_img_2482.jpg
I has this idea this morning that it would be cool to make those pulses visible. that would require making them a bit longer, and then feeding them into a normal TV video input (of the PC). ftp://panteltje.com/pub/sc_pic/radium_to_TV_2_img_2478.jpg The width is about 60 us, so I made the pulses really long with a large RC time with the little haystack circuit at the bottom, a 2 transistor pulse shaper and interface to 75 Ohm.
30 to 40 pulses per 20 mS frame makes say about 2000 pulses per second. Varies strongly, nice random generator too.
The idea is to sort the pulses by amplitude, and make something like this on the PC:
If you really dark adapt, half an hour or so, you can see the individual alpha flashes from a radium watch hand. A magnifier helps.
The alphas probably don't make it out of the phosphor mix in the watch hands, but the light does. Try an alpha source without the phosphor, smoke detector americium maybe.
Incidentally, americium is fissionable, so you could make your own nuke with the stuff you could recover from some billions of old smoke detectors. Or trillions maybe.
On a sunny day (Tue, 07 Dec 2010 11:14:31 -0800) it happened John Larkin wrote in :
difference.
even in the dark.
Yes, I was thinking about what sort of a mistake I made, and it was very simple. I left the Uranium glass pieces in the specimen bottle, and the bottle stopped the alphas. So, after some thinking, I wrapped some (just 7 small pieces) of the uranium glass in some polyethylene foil, and put it in the 'body bag', in front of the other polyethylene that I already had covering the PMT.
I am blown aways, there are not many pulses per second, but fewer, but extremely strong pulses, of different amplitude. The strongest one I have seen was more the 50 V from the PMT! Many smaller ones. This is absolutely beautiful as now I can make the bins with different amplitudes and counts to make the spectrograph. you can almost do it by hand this way! Scope triggers if you want it or not, such strong signals. Absolutely beautiful :-) Why would anyone spend 100$ on a scintillation crystal when an Euro 1.50 roll of magnetron food foil is 40 % as good? hehe Thanks for the help, very inspiring, made me think what I was doing.
Are you planning to digitize individual pulse heights? That's the best way to get energy spectra. You can use window comparators, but it's very slow.
The best foil would have a thin opaque aluminum layer on top, to keep out stray light and reflect scintillation light back down into the PMT. Matching index of refraction is probably too much to ask for.
On a sunny day (Tue, 07 Dec 2010 15:39:00 -0800) it happened John Larkin wrote in :
amplitudes and counts to make the spectrograph.
of magnetron food foil is 40 % as good?
Yes, I need to digitise each pulse, and sort by amplitude and number of counts. The 'slow' can be an issue, but consider this, now we get a bit technical, but OK.
First the requirement, it looks like if one makes spectra from say uranium ore, there is a maximum to the number of impulses per unit of time, say per second. There is indeed a probability that 2 impulses will occur very close together, or even overlap, but that is a rare occurrence in a measurement period, and missing one of those in say a one minute or longer measurement period would cause a only very very small error. Also the measurement is purely relative, the pulse amplitudes relative to each other, as all is set by distance to the object under test, the quality of the scintillator (the poly-thing here), the PMT HV voltage, and PMT sensitivity, etc.. So a lot of things are NOT interesting. As I mentioned in some other posting some time ago, when playing with vidicons in the sixties when designing a camera, I found that one can treat things like that, and PMTs, as current sources. In fact it seems a waste to feed the PMT electrons, as that is what they are, into a 50 Ohm resistor, and then waste them there, and then using expensive low noise amps to amplify the tiny voltage drop over that resistor to get the nice short pulse (low RC time in fact), to then filter that to make it longer, the works, IF those pulses happen only 2000 x per second (as I observe here) most of the time. It seems much more logical to use them precious electrons, the product of all the electron sex in the PMT, in a more efficient way. After thinking about that for a few moments I came up with this:
coax with capacitance C waveform here PMT anode
------=======================------------------------------------------- PIC comparator input | | | | | -- interrupt R | | 4.9 V zener /--
+4.9V PIC comparator reference | 680k | | | | transistor switch sample input | | | | | | ----------------------------------------------------- +5 V do conversion | result to FIFO
The way it works is that if a light pulse is detected, then the electron multiplication process starts, and a CURRENT, so electrons from the PMT, charge the coax capacitance + other parasitic capacitances, up to a high negative voltage in a very short time (nS). The voltage will STAY high for a very long time if nothing is done, and C will slowly discharge via the 680 k resistor. The hardware comparator in the PIC will detect the pulse if it is more then 100 mV (as example) negative, and trigger an interrupt. The interrupt routine samples the voltage, starts the conversion and stores the result in a FIFO buffer. The main routine sends the FIFO as 8 bits values via fast RS232 to the PC. After the conversion, the interrupt routine discharges the capacitance by means of a transistor switch, and the system is ready for the next pulse, Now we make optimal use of the electrons we got, we do not waste any heat in 50 Ohm resistors, and can work at the maximum speed of the hardware, for example a PIC, or a fast FLASH ADC with FPGA, can work at. The software on the PC side can set the PMT voltage remotely via the same RS232, and will set it in an initial calibration period so the max pulse amplitude measured is always just below the max steps (in fact
0 steps, as we measure upside down) of the ADC.
Pulses that occasionally happen so close to each other in time that the serial link (or UDP ethernet link) is too slow, are simply accumulated in the FIFO buffer, and are all send in quieter periods. Probability sets the size of the FIFO, and in fact the error rate,
Measurement error is not what you may think, as the input capacitance of the sampler circuit (a few pF) with the cable capacitance forms a simple capacitive divider, a FIXED divider, so that should be a constant for all pulses, and cancels out. In this application the absolute timing of the pulses, as well as the delay from light pulse to output to the computer is irrelevant, I think I make good use of the signal with simple means. This solution is here by published into the public domain, and hereby made prior art for anyone who wants to go and patent it.
Keywords: prior art, photo multiplier, spectrograph, PMT, sucks, patent, Balmer, Jobs, IBM, HP, boss, money, NSA, CIA, FBI, president, Iran, wikileaks, N Korea, S Korea, China, underwater, area 51, area 52, areas 53, add as required.
I hope that answered your question. Patenting it costs 5000 Euro, so I will use that to buy chocolate,
The best foil is probably tin foil, I have a roll somewhere, maybe good for a nice hat :-)
On a sunny day (Tue, 07 Dec 2010 15:39:00 -0800) it happened John Larkin wrote in :
amplitudes and counts to make the spectrograph.
of magnetron food foil is 40 % as good?
Yes, I need to digitise each pulse, and sort by amplitude and number of counts. The 'slow' can be an issue, but consider this, now we get a bit technical, but OK.
First the requirement, it looks like if one makes spectra from say uranium ore, there is a maximum to the number of impulses per unit of time, say per second. There is indeed a probability that 2 impulses will occur very close together, or even overlap, but that is a rare occurrence in a measurement period, and missing one of those in say a one minute or longer measurement period would cause a only very very small error. Also the measurement is purely relative, the pulse amplitudes relative to each other, as all is set by distance to the object under test, the quality of the scintillator (the poly-thing here), the PMT HV voltage, and PMT sensitivity, etc.. So a lot of things are NOT interesting. As I mentioned in some other posting some time ago, when playing with vidicons in the sixties when designing a camera, I found that one can treat things like that, and PMTs, as current sources. In fact it seems a waste to feed the PMT electrons, as that is what they are, into a 50 Ohm resistor, and then waste them there, and then using expensive low noise amps to amplify the tiny voltage drop over that resistor to get the nice short pulse (low RC time in fact), to then filter that to make it longer, the works, IF those pulses happen only 2000 x per second (as I observe here) most of the time. It seems much more logical to use them precious electrons, the product of all the electron sex in the PMT, in a more efficient way. After thinking about that for a few moments I came up with this:
coax with capacitance C waveform here PMT anode
------=======================------------------------------------------- PIC comparator input | | | | | -- interrupt R | | 4.9 V zener /--
+4.9V PIC comparator reference | 680k | | | | transistor switch sample input | | | | | | ----------------------------------------------------- +5 V do conversion | result to FIFO
The way it works is that if a light pulse is detected, then the electron multiplication process starts, and a CURRENT, so electrons from the PMT, charge the coax capacitance + other parasitic capacitances, up to a high negative voltage in a very short time (nS). The voltage will STAY high for a very long time if nothing is done, and C will slowly discharge via the 680 k resistor. The hardware comparator in the PIC will detect the pulse if it is more then 100 mV (as example) negative, and trigger an interrupt. The interrupt routine samples the voltage, starts the conversion and stores the result in a FIFO buffer. The main routine sends the FIFO as 8 bits values via fast RS232 to the PC. After the conversion, the interrupt routine discharges the capacitance by means of a transistor switch, and the system is ready for the next pulse, Now we make optimal use of the electrons we got, we do not waste any heat in 50 Ohm resistors, and can work at the maximum speed of the hardware, for example a PIC, or a fast FLASH ADC with FPGA, can work at. The software on the PC side can set the PMT voltage remotely via the same RS232, and will set it in an initial calibration period so the max pulse amplitude measured is always just below the max steps (in fact
0 steps, as we measure upside down) of the ADC.
Pulses that occasionally happen so close to each other in time that the serial link (or UDP ethernet link) is too slow, are simply accumulated in the FIFO buffer, and are all send in quieter periods. Probability sets the size of the FIFO, and in fact the error rate,
Measurement error is not what you may think, as the input capacitance of the sampler circuit (a few pF) with the cable capacitance forms a simple capacitive divider, a FIXED divider, so that should be a constant for all pulses, and cancels out. In this application the absolute timing of the pulses, as well as the delay from light pulse to output to the computer is irrelevant, I think I make good use of the signal with simple means. This solution is here by published into the public domain, and hereby made prior art for anyone who wants to go and patent it.
Keywords: prior art, photo multiplier, spectrograph, PMT, sucks, patent, Balmer, Jobs, IBM, HP, boss, money, NSA, CIA, FBI, president, Iran, wikileaks, N Korea, S Korea, China, underwater, area 51, area 52, areas 53, add as required.
I hope that answered your question. Patenting it costs 5000 Euro, so I will use that to buy chocolate,
The best foil is probably tin foil, I have a roll somewhere, maybe good for a nice hat :-)
parasitic capacitances, up to a high negative voltage in a very short time (nS).
slowly discharge via the 680 k resistor.
mV (as example) negative, and trigger an interrupt.
result in a FIFO buffer.
of a transistor switch, and the system is ready for the next pulse,
Ohm resistors, and can work at the maximum speed of the hardware,
RS232, and will set it in an initial calibration period
0 steps, as we measure upside down) of the ADC.
link (or UDP ethernet link) is too slow, are simply accumulated in the FIFO buffer,
sampler circuit (a few pF) with the cable capacitance forms a simple capacitive divider,
from light pulse to output to the computer is irrelevant,
prior art for anyone who wants to go and patent it.
IBM, HP, boss, money, NSA, CIA, FBI, president, Iran, wikileaks, N Korea, S Korea,
nice hat :-)
The other way to digitize the pulses would be dump the PMT pulse into a cap. Pull the cap up to, say, +3 volts with a resistor or a current source, and drive a comparator. The comparator output will be pulses whose width is a function of the PMT pulse amplitide. A uP timer channel can process that. That amounts to sort of a Wilkinson ADC.
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.