e
uctor solution,
test
tors can be really small.
n 10 cm.
lower RFI,
We buy these little PMT supplies from Hamamatsu (C8991) that have the Cockroft-Walton mulitplier built into the socket. There is a bit of ripple on the anode, but not too bad. The whole thing runs from +15V input and claims 8mA (max) of current, (with no light), but I've not measured it.
George H.
On a sunny day (Thu, 13 Oct 2011 05:58:37 -0700 (PDT)) it happened George Herold wrote in :
I think I found it with google,
monitor (no lens needed).
and in sync,
at.
rather long.
eavesdropping.
This is fun:
Get an old analog oscilloscope, make a medium speed horizontal sweep. Connect a PMT to the vertical input of the scope and aim it at the screen. If you get things adjusted right, you can put objects, like your fingers, in front of the screen, between the screen and the PMT, and the scope will trace their outlines along the upper edge of the shapes.
If you cut a waveform out of a piece of black paper, it will trace that too. Arbitrary waveform generator!
John
in
.som=3D
ind=3D
lso=3D
aci=3D
thi=3D
ing=3D
Yup that's it. Looks like we pay ~$180 for quantities of 50-100.
Yeah, all the anodes have the same voltage difference between them, you can change the total voltage though, up to 1100V or something like that.
George H.
Probably not. Links used to work fine, and mostly still do, but Firefox up-dated itself a day or two ago (to 7.0.1 according to the "help" screen) and my guess is that was the result of one of the "improvements" in the up-date.
-- Bill Sloman, Nijmegen
:
than 1.1 MOhm...
battery - for a reasonable amount of time, say a 12 hour day
kind
always
the cleanest.
power.
I think you meant - voltages instead of + voltages?
?-)
On a sunny day (Thu, 13 Oct 2011 07:44:22 -0700 (PDT)) it happened George Herold wrote in :
That is a lot of money, and does not do all my voltages. So I played a bit today, to see if I could sort of duplicate their circuit, and get extra voltages at that. I came up with this circuit:
Hey, I suggested something like that!
John
of caps.
from about 3 V,
air.
experience with so few diodes.
On a sunny day (Fri, 14 Oct 2011 08:09:19 -0700) it happened John Larkin wrote in :
Yes, and I thanked you for the pdfs in an other reply. Your help is very greatly appreciated. But the way this is driven is slightly different from the normal HV multiplier. But I think for the higher stages the drive capacitors can be in series, in the conventional way, and low voltage types. Add an other 250 turns and save 2 diodes and 2 caps, but need a HV cap... Will play some more in the weekend so I know what to order. The nasty thing is that without any load these HV caps stay charged for a very long time. They bite unexpectedly when you then solder on them.
A battery powered PMT HV supply for PMTs with not equal voltages between the anodes:
So, I calculated a new transformer, using the old secret magic formula of 1 turn per volt (2 Vpp). At 5 V supply that gives:
5 turns primary for 5 V max supply (remember we adjust the DC supply to set the HV). 2 turns feedback to the base for the BC547B sine wave (ahum) oscillator. 79 turns to get 79 / 5 * 10 = 158 V for the smallest anode voltage stages. 40 additional turns to get 238 V for the odd anode stages.E core has an Al of 1200, no airgap.
The last anode, or first if you want to count from the PMT cathode, so A1, gets 2 x 158 V from 2 stages of the multiplier ladder.
This is all for nominal 1750 volt overall, although I only tested the first few stages. I also over-wound the secondary 10% to compensate for diode drops and 'other things (tm)', and to allow control of the output with 5V or lower. All very easy to wind by hand. The circuit uses about 10 mA at 3.3 V, at that point it already reaches the target output voltage.
Reducing the turns dramatically relative to the previous experiment upped the frequency to above 100kHz, nice low ripple!
Of course with all the multiplier stages added, the current consumption will likely be more. But this sort of power consumption is acceptable for a battery operated device. At least I have some idea now what caps and diodes to order. The transformers are OK, got the right ones, transistors too, diodes FR1Y too.
Usenet patent 0Z010#010$0101 hupsadecimal Copyright Jan Panteltje 2011-always. You have seen it here, prior art. Donated into the public domain.
PS the feedback winding to the transistor base is 2 turns, I would have used 1 turn, used to do that sort of oscillator with potcores, was not sure about these Ecores (potcores are rare these days). Could perhaps improve efficiency or/and or waveform to use 1 turn, will test it later.
s,
RM cores replaced pot cores a few decades ago.
-- Bill Sloman, Nijmegen
Really? How come people keep selling and using them? I guess word hasn't gotten around.
John
anodes:
turn per volt (2 Vpp).
Funny how often 1 t/v works.
HV).
stages.
things (tm)',
target output voltage.
I like this one:
ftp://jjlarkin.lmi.net/LC_osc.jpg
P-P swing at the collector is almost exactly twice V+.
John
-et=3D
e),
t,
n out of caps.
ts up from about 3 V,
in the air.
Grin, It's always funny when some one else hurts themselves.
The ripple on the hamamatsu is ~140kHz, (IIRC) if that helps.
Hey, if you can sell us one with the base for less money, I'm all ears. But we only have a small volume.
George H.
(pp).
g experience with so few diodes.
On a sunny day (Sat, 15 Oct 2011 13:36:44 -0700 (PDT)) it happened George Herold wrote in :
Yes, I wound a new transformer with fewer turns, and am now at > 100kHz.
That base is it a standard item you can buy by itself? This is for small battery powered gamma spectrum analyser, Will see what happens. See also my other posting for the latest version of the circuit. I hope to order all those caps I need to make a final prototype next week, usually takes a week to arrive too.
On a sunny day (Sat, 15 Oct 2011 13:18:59 -0700) it happened John Larkin wrote in :
anodes:
turn per volt (2 Vpp).
One of the things in my school days I did was to wind my own TV output transformer. I spend many hours doing calculations, until finally one thing became very clear, 'look at it from voltage rations only'. The old BW TVs had a DY87
the HV).
few stages.
things (tm)',
target output voltage.
voltage.
Yes, I have seen that, the voltage from the feedback loop in the emitter however subtracts from the total in your circuit, so it can never be 0 to 2 x supply swing? Or do I see that wrong?
On Oct 16, 3:10=A0am, Jan Panteltje wrote: ...
however
ply swing?
...
Note the polarity - it adds, not subtracts.
When the collector is coming down to ground, the feedback winding is driving the emitter below ground.
kevin
anodes:
turn per volt (2 Vpp).
transformer.
clear,
the HV).
few stages.
things (tm)',
target output voltage.
voltage.
swing?
At peak negative collector swing, the emitter is in phase, so it swings below ground.
It starts with too much base bias current, so oscillations build rapidly. The base is about +0.65 or something like that. As the oscillation amplitude hits 2x V+, the negative collector swing starts dipping down to ground, and the c-b junction forward conducts on the negative swings, which steals base current, which pulls Vb down and reduces transistor gain. It's a temperature compensated AGC circuit.
At the negative swing peak instant, when the c-b junction is forward conducting, the emitter will be below ground, which also steals some of the base bias voltage. It works well when the voltage across the feedback winding is small, a tenth of a volt or so.
I first used this for generation the excitation for an LVDT-like device, actually an inclinometer, for the C5A. The sinewave amplitude is very stable and has essentially zero TC.
Hey, they still make the Talyvel transducers!
It was fun to play with. I had a steel barrel full of sand, with a massive steel plate on top, with a big lever driven by a micrometer, to generate inclinations in arc-seconds. Any time anybody walked around in the room (on a concrete slab!) I'd see the building flex.
John
On a sunny day (Sun, 16 Oct 2011 08:54:50 -0700 (PDT)) it happened kevin93 wrote in :
mm looks like you are right :-)
On a sunny day (Sun, 16 Oct 2011 11:06:42 -0700) it happened John Larkin wrote in :
however
swing?
Yes, I see my error now. Very clever circuit. I will definitely have a try with that too.
Making zero TC sine wave oscillators is difficult, the core material of the coils change permeability with the slightest of bias. For radio frequencies I always use this: + | | |---- --------->| | | |--- L === | | |---------| | === R | | | /// /// ///
10 arc seconds, not much ...It makes me think o that Foucault pendulum, NASA had some test done many years ago to see if it swung differently during a solar eclipse. Sometimes it did... I have thought at times that was because the ground moved a bit because of people running about to see the eclipse... An inclinometer that sensitive would help to eliminate such a possible systematic error.
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.