Hello

I cannot afford the big iron. Well, I can, but then I would end up limited to one project ever and have nothing to hobby with again but doublers I can afford and they do well enough in small projects for both radio and oscilloscope. However, before I post a WTB I thought I should run this by you folk first. I saw this item on eBay not long ago and thought "I could do with that" but not for the 20 some dollars it would cost to buy from eBay. I set up a simulation of it and worked some values (as nothing was on the original) that are values I have typically seen in doublers (and this one being made for a crt I went with that for load, the 1nF for the crt capacitance I chose was nominal in a dozen spec sheets I looked at), and have decided on a parts list IF they are right. But, posting here first to see if anyone who's had hands on experience with doublers can give any suggestions on this before I commit to purchasing the parts and trying.

Specifically of interest to me is the veracity of the 68nF cap at over 800v which is the only 'hard part' to find that I have discovered so far (well, easy if I want to spend 20 dollars or more on one capacitor but then if I wanted to spend 20 dollars in one go I would have just bought this from the seller in the first place ...) (I am planning on asking for a non-ceramic 68nF@2500v or better actually as I need about 1000v out from this for one project and about something else I just realized I didn't make note of that and oh I go digressing again >.<).



(PLEASE don't shoot me I know there are pencil and paper maths ways of doing this, and I have done them and found the sim to give nearly the same results in general) This is the sim of it (save as a txt file and save, open the txt file in the stand alone of this program as so many browsers have such a cow over java) http://www.falstad.com/circuit/

$ 1 5.0E-6 382.76258214399064 50 5.0 50
c 128 48 192 48 0 4.7000000000000005E-7 -76.55974658681487
c 192 48 256 48 0 4.7000000000000005E-7 -41.56974594232236
c 256 48 320 48 0 4.7000000000000005E-7 -23.442871740806922
c 160 96 224 96 0 4.7000000000000005E-7 -41.44704255752039
c 160 96 160 160 0 6.8E-8 22.205035247359152
c 224 96 288 96 0 4.7000000000000005E-7 -23.534035529677837
c 384 208 448 208 0 6.8E-8 -130.93742096185858
c 384 240 384 304 0 4.7000000000000005E-7 197.80227183481105
c 224 304 288 304 0 4.7000000000000005E-7 197.85092821780663
c 288 96 352 96 0 4.7000000000000005E-7 -26.37390996894706
d 128 48 160 96 1 0.805904783
d 256 48 288 96 1 0.805904783
d 192 48 224 96 1 0.805904783
d 320 48 352 96 1 0.805904783
d 160 96 192 48 1 0.805904783
d 224 96 256 48 1 0.805904783
d 288 96 320 48 1 0.805904783
r 352 96 448 96 0 10000.0
w 160 208 384 208 0
w 448 96 448 160 0
r 288 304 384 304 0 10000.0
d 288 304 256 256 1 0.805904783
d 256 256 224 304 1 0.805904783
w 448 240 448 336 0
w 448 336 352 336 0
w 352 336 352 384 0
w 384 304 416 304 0
w 416 304 416 384 0
d 160 336 224 336 1 0.805904783
w 224 336 224 384 0
w 160 336 160 384 0
w 160 240 160 304 0
w 160 304 224 304 0
w 80 256 80 336 0
c 80 256 144 256 0 4.7000000000000005E-7 119.41909421726496
w 80 48 128 48 0
w 144 256 256 256 0
w 80 336 160 336 0
R 160 384 160 416 0 1 60.0 120.0 0.0 0.0 0.5
g 48 384 48 416 0
O 224 384 224 416 0
O 352 384 352 416 0
O 416 384 416 416 0
c 448 336 448 384 0 1.0E-9 130.93742096185858
g 448 384 448 400 0
g 496 352 496 368 0
c 496 304 496 352 0 1.0E-9 -197.80227183481105
w 416 304 496 304 0
w 160 208 48 208 0
r 224 384 304 384 0 500000.0
g 304 384 304 416 0
w 48 208 48 384 0
w 160 160 160 208 0
w 160 208 160 240 0
w 448 160 448 208 0
w 448 208 448 240 0
w 80 256 80 48 0
w 384 208 384 240 0
x 398 251 431 255 0 14 250V
x 118 228 151 232 0 14 120V
x 273 254 306 258 0 14 250V
x 313 33 346 37 0 14 250V
x 398 173 431 177 0 14 830V
x 244 408 292 412 0 14 No Idea
x 87 383 135 387 0 14 120V in
x 506 346 615 350 0 14 CRT capacitance
x 337 446 375 450 0 14 -230V
x 406 446 439 450 0 14 830V
x 136 489 666 495 0 24 All the cap voltages have to be doubled for safety
x 170 154 203 158 0 14 120V
o 40 64 0 34 160.0 9.765625E-5 0 -1
o 41 64 0 34 1197.2621413014756 9.765625E-55 1 -1
o 42 64 0 34 320.0 9.765625E-5 2 -1

This type of voltage multiplier works well as a power supply for CRT. Voltage will be high but available current low and poorly regulated. We used something like this on the Friden first electronic calculator, around 1962. Worked great powering a 5" CRT.

That is good to know Norm.

Most radios only go as far as one doubling at least i dont recall anyone using a trippler. Regulation does fall so does current. If my maths have been right its somewhere on the order of four per stage. 1 Amp in doubled is 0.25 amps out etc. What i need for the PE65 is around +750/-350 thats from memory. For the radios though max 350V B+



If you think that 68nF@2500V is good to wtb i will

How much current do you need to get out of it? Those capacitors look mighty small for 60 Hz unless the current is REALLY low.

The photo you provided looks like a voltage multiplier for a TV or computer monitor. It probably is made to run in the 15 to 35 KHz region.

One problem with voltage multipliers is that they don't provide any isolation from the AC input line. If your project is completely self-contained, that isn't a problem. But if it needs to connect to something external, there will be a shock or fire hazard.

Jim Mueller

Hello Jim.

First my apologies if this is a long or rambly reply. I over did my Sunday this week and am paying for it. Got a bunch of muscle relaxers in me feeling better now the cramps have again subsided but half a nights sleep leaves me tired on top of the tired of the muscle relaxers leave me but I deleted SO MUCH TEXT I hope that what I left isn't too bad :C (oh gee an edit >.< yes this part was listed and described as 'taken from a working oscilloscope' on eBay)

I've a few ideas on the go. The big voltage ones would use the doubler as presented first, but the smaller ones would be fine with just a single stage of doubling which is as I've seen fairly commonly done by folk with 120:120:12 volt transformers (or two 120:12 back to back) to get around 240 or so from them.

You are exactly right about AC line separation too. I want to make a modified version of this using two compactrons (two 6M11's or a 6M11 and a 6J11/6BN11) and sand for the detector and B+ http://www.electronixandmore.com/projec ... index.html The max for the 6M11 is 330V, wont need that much I don't think, Bob Weaver has some JPG files of the Unicorn that use just one 6M11 and that is powered at 125vac. http://s199.photobucket.com/user/BobWea ... t=3&page=1

And yes, a few are home brew Oscilloscope(like) projects. One CRT that I don't have a data sheet for (3FP7A and if anyone has a data sheet on that I'd like a copy) I want to put in this http://g4oep.atspace.com/toycro/toycro.htm isn't too far different than the CRT presented in that article (I will also give the HV in that one a go as well I've the E/I cores in a box for that). (and an edit here, I saw this years ago, and I think this will display nice with the long persistence phosphors http://frank.harvard.edu/~paulh/misc/lorenz.htm )

One project calls for either 3BPl. 3EPl. or 3GP But, I am getting a Russian one from the 8lo series most probably 8LO29i On one CRT the only web page stating current at 300uA

http://translate.google.com/translate?h ... 2F7lo1m%2F

The data sheet otherwise I don't see containing a current rating for it http://frank.pocnet.net/sheets/084/8/8LO29I.pdf

For my small crt DB3-91 the current should be around 0.5uA if I read the data sheet right http://www.shinjo.info/frank/sheets/030/d/DB3-91.pdf (this crt I got to do what I found of an Italian text version of http://www.albinarrate.com/content/sche ... DH3-91.jpg (big image) the mini English scan is on this page http://www.radiomuseum.org/tubes/tube_1cp1.html the original article isn't online as a pdf but eventually I may purchase a paper copy as they are around for 15 to 30 dollars .... yea Im going to chuckle about that too ... )

I also found this guys articles kinda inspirational, google translated too:

http://translate.google.com/translate?d ... z/osc.html

http://translate.google.com/translate?d ... elmon.html

and he added sound

http://translate.google.com/translate?d ... lmon2.html

Of course Jon also has http://www.electronixandmore.com/projec ... index.html and I have the transistors and op-amps for that.

I've got most of the capacitors now, need a few switches, got more than a few 120v:12v transformers hmmm mostly covered finally, taken me a while to purchase things and then my health tanks and I take a break. Going for a couple 1watt and 3watt resistor assortments that will cover all the resistors across all these that will be either this months pay check or next months pay check (I get 1 chance to buy stuff once a month and I owe Alan 15 dollars and a seller on eBay around 20 dollars and that I think may be this months available funds spent got a bit of wiggle room available though, we will see OH and I turn 45 in a couple weeks and SO want something for my birthday so that has to go in the budget!!!). Ahh well, I will persevere onwards as the song goes, crush my heart into embers I will reignite https://www.youtube.com/watch?v=re32xnyYP3A

Consider getting a copy of the ARRL Handbook from the mid 50's to 60's. These cover in detail the design of voltage doublers, in fact a number of power supplies.

GL

Chas

Chas!!! I will. I had one from that era but time hid it from me. Any particular year that is best or any from that time span?

Unfortunately, it doesn't scale that way.

Output current depends on the size of the capacitors used. Most voltage multipliers provide only microamps of output current. You will need capacitors in the mF range to get even a few mA output. Also, with small caps, the ripple and regulation will be very bad.

I built a doubler to get 24V out from a 12V transformer. Had to use several hundred mFd to get decent ripple with 30 mA output.

Rich
http://www.vishay.com/docs/88842/anusingr.pdf

Agreed, there needs to be sufficient I_in and C to support any type of load - as well as to accommodate the unavoidable and inevitable leakage. In addition to safety, leakage is also one more reason why many of these are potted or covered. Dust, dirt, and film will add up over time.
BTW: The output @ the "no idea" 500k resistor will simply be pulsed DC.
- Or, will it be 1/2 of an AC sine wave?

The "no idea" was just a load value that I put in that isn't part of the original part, sometimes I label things as a "note to self" more than as meaningful info. I should have cleaned it up yes, is just unfiltered half wave out at that point (unless it wasn't a sin wave in yup). I think what I had in mind was "I've no idea why this would have been used here in this part rather than as a separate part". (though to keep it dust free would be a good reason. I would think that the voltage going in would have been at a higher frequency then if that were the case that electrostatic dust build up was an issue... Hmmm Im going to keep thinking about this some)


Rich most doublers I have seen range 220uF to 470uF for radios or switch mode for CFL so I am guessing by your comment they are woefully under valued? My needs being less than 1ma with that end of the HV I didn't think I really needed that much capacitance (but I am mulling over frequency now). When I was 10 or 11 the first TV doubler I tore apart doing the "what is this and why?" explore use 0.01uF ceramic caps. However yes if I want to drive 30mA I gather much bigger caps would be needed?

My simulation shows that you can change all of the caps in the top string, and the two 68nF caps to ground at each end of this string, to 47uF and then you can load both DC outputs to 1mA and still have right decent regulation. The outputs will be -223 and 764 V and the current from the 120Vac source will start at ~ 1A p-p spikes (!!!) that settle in to about .2A p-p within 4 seconds time.
The lower string should be able to provide 1mA as is.

Again, this depends on how much current you need to draw and how much ripple you can tolerate. I doubt that CFLs use doublers. The filter on the incoming DC is usually about 200 uF. I don't think CFLs care much about ripple.

Rich

I used a voltage doubler power supply for my 2x 6M11 stereo audio amplifier project, because I only had a 120V isolation xfmr available for the power supply. It puts out 250 volts at around 50 mA, with no hum problems. The power supply schematic is shown near the bottom of this page:
http://electronbunker.ca/TubeAudio_01b.html
No ripple problems with this power supply. Hum in the output is inaudible.
It could actually put out a bit more voltage than 250V, but I deliberately cut it back to keep the tubes operating in their safe plate dissipation range.

Another thing to consider when designing a power supply for a CRT, the current through the voltage divider (brightness and focus controls) needs to be significantly larger than the current through the tube to provide decent voltage regulation. This is where most of the current actually goes.

Jim Mueller

Here is one crt use that I would need to consider because of no way getting that transformer

This is the other, these are two that I am working on building now (the original full size image) http://www.albinarrate.com/content/sche ... DH3-91.jpg (and this one really because the CRT is cute as beans and I have one that I lucked out on getting a few years ago and simply want to use)

And for no reason, the Italian typed out. Though there is an English original to this around.

De todos los instrumentos de medición el osciloscopio es el que mejor se presta a la miniaturizatión ya que sus dimentions están en functión de los del tubo. Los tubos de rayos catódicos de menor tamaño (1CP1, DH3-91) tienen un diámetro de pantalla de 3 cm suficient para permitir el examen de una forma de onda, y no necesitan grandes tensiones en sus electrodos. Por esto ha sido atractivo construir un osciloscopio cuyas dimensiones son de sólo 18x12x6 cm.

El control de ganancia vertical está dado por R1 destacando que una señal de 1.5 V eficaces a la entrada es suficiente para barrer por completo la pantalla; sobre la respuesta del amplificador vertical, se conserva lineal hasta 55 kc/seg. y, aceptando una caída en la curva de un 30 %, la banda pasante llega a los 300 kc/seg.

La base de tiempos cubre en cinco pasos las frecuencias comprendidas entre 20 ciclos/seg. y 35 kc/seg., con el potenciómetro R19 como ajuste fino de la frecuencia de barrido. La sincronización se obtiene por medio de R12 que actúa sobre la señal previamente amplificada por V1.

El transformador de entrada está previsto para proporcionar 2 x 250 v a 25 mA y 6.3 V a 1 A

The 1CP1 is a nice loctal CRT with operates on fairly low voltage. If you can't find one look for 1DP1, different base.

Norm!! I have. I have one wooooo me I realllly got lucky, they typically are going 30 dollars each and up and I won at 20 dollars!! I've seen sellers with them into three digits each. Really insane considering. It really is a neat as beans CRT!! But, that is a 500 volt transformer. That may have been an off-the-shelf item from Radio Shack in the 50's today is a very different story. But, that would should be replaceable by a doubler even with noise, it isn't a really solid scope more a curiosity machine I think.

What are you going to use that CRT for ?

If you plan to attach it to some external input signal, then it should be isolated, or else you're going to get a big surprise ! If so, do the voltage multiplication after an isolation transformer.

Most scopes of that time used a high negative supply. That's what one of your drawings shows. Doing so, simplified the circuit, because the deflection plates need to have voltages close to that of Anode 2. The amps to generate the deflection signals, swing between zero to a typical B+, of 250, or 300V. But a CRT needs a higher voltage of 750V to 3KV, between anode 2 and the cathode. To get that, the cathode is parked down to some negative voltage to make up that difference. Here is a little charge pump circuit that you can try:

You may need to increase those capacitor values a little. The pump circuit, as shown, runs at ~450Hz. This frequency is still low enough, so common 1N4007 rectifiers can be used.

Here's a CRT test jig, I made:

I generates just under 2KV, but its a switcher running at ~20KHz. It needs ultrafast MUR160 rectifiers.

Isolation always goes without saying!

Charge pumps I have seen but not read on. I shall do so

As to signal in I have no idea on the miniature CRT one. Probably nothing more stunning than the audio out of my old Android. The bigger PE65 one that is supposed to be good to 25mhz which should be good enough to work on the AA5 type radios. I am not after 'high end bang on' just good enough at the gate will do. Once I get one working, I will work on the next then the next. The Toy Cro one I will feed it with a Lorenz Strange Attractor though, that has been planned for a long time. Have a look here http://frank.harvard.edu/~paulh/misc/lorenz.htm

https://www.youtube.com/watch?v=xnF-jKqW6kI