ERNIE - Electronic Random Number Indicating Equipment
Found something, ERNIE, the first lottery computer, maybe you have already come across these pages, at least the part of the schematics you added can be found in the PDF link below, skimming quickly through I couldn't see any capacitor values but maybe some other data might give a hint of what type of capacitor values it may concern.
ERNIE, the first lottery computer
http://www.radiomuseum.org/forumdata/upload/ernie.pdf
Found something, ERNIE, the first lottery computer, maybe you have already come across these pages, at least the part of the schematics you added can be found in the PDF link below, skimming quickly through I couldn't see any capacitor values but maybe some other data might give a hint of what type of capacitor values it may concern.
ERNIE, the first lottery computer
http://www.radiomuseum.org/forumdata/upload/ernie.pdf
Excellent question that I can't answer
It's not an audio noise generator, but the front end of a digital random number generator. I have absolutely no idea what frequency it operated at. There is scant information available for it. It could have been anything, but I'll only consume an audio level signal below 20kHz if possible.I'm going to proceed on the basis that the amplification stages should deal with <20kHz signals. I'm no valve expert, but this circuit (and component sizes) looks very similar to the preamps I see on this forum.
For what it is worth, my electronics experience started in 1954, yes I am old as dirt. While the units have existed all my life, I have never seen millifarads until maybe the last decade. I can't imagine anyone in his right mind would write 1uf as 0.001mf. Certainly in the era of tube amps, no one had millifarad caps. The idea of a 1 farad cap was pretty much just theoretical at the time. Now people have multifarad caps in their car stereos. Times change.
So 0.01 and 0.001 are most surely microfarads. And I can't imagine anyone was doing much at gigahertz freqs back then either. Especially not just for a random number generator.
Two pentodes in a row ought to have a TON of gain. Not sure how audio will do coursing through that. Apply some and find out.
C2 is a decoupling cap, the B+ ought to already be plenty smooth by the time it gets to this stage. 2uf ought to be enough. A 2000uf cap in the 1950s? Not in a commercial circuit. 100uf was a big cap then.
If you have the circuit, what the heck, power it up and FIND OUT what frequency range it spits out noise.
Schematics like this tended to, but did not always, have a note section in a corner somewhere, and a common note was: all resistances in Ohms and all capacitors in microfarads unless otherwise noted. I would have had to have a REAL strong reason for not thinking the caps are in uf and the resistors in ohms. Note they even spelled out that the k meant 1000. Further clue that resistances were in ohms.
I now pretty much consistently say picofarads, but it took me a long time too to transition from uuf to pf. I think mf was fading out when I was learning, in favor of uf. Both meant microfarad. Same with mmf and uuf.
It was common for techs and ham operators to refer to mmf as "mickey mouse" in casual conversation. A 27 mickey mouse cap. Going to uuf killed that small fun.
But the one that took me the longest was to start saying Hertz instead of "cycles per second", or cps. I still often find myself saying, "put a 400 cycle note through it". Cycles per second was a descriptive term, and lent itself to math. Hertz is just a name which has no correlate to a value.
So 0.01 and 0.001 are most surely microfarads. And I can't imagine anyone was doing much at gigahertz freqs back then either. Especially not just for a random number generator.
Two pentodes in a row ought to have a TON of gain. Not sure how audio will do coursing through that. Apply some and find out.
C2 is a decoupling cap, the B+ ought to already be plenty smooth by the time it gets to this stage. 2uf ought to be enough. A 2000uf cap in the 1950s? Not in a commercial circuit. 100uf was a big cap then.
If you have the circuit, what the heck, power it up and FIND OUT what frequency range it spits out noise.
Schematics like this tended to, but did not always, have a note section in a corner somewhere, and a common note was: all resistances in Ohms and all capacitors in microfarads unless otherwise noted. I would have had to have a REAL strong reason for not thinking the caps are in uf and the resistors in ohms. Note they even spelled out that the k meant 1000. Further clue that resistances were in ohms.
I now pretty much consistently say picofarads, but it took me a long time too to transition from uuf to pf. I think mf was fading out when I was learning, in favor of uf. Both meant microfarad. Same with mmf and uuf.
It was common for techs and ham operators to refer to mmf as "mickey mouse" in casual conversation. A 27 mickey mouse cap. Going to uuf killed that small fun.
But the one that took me the longest was to start saying Hertz instead of "cycles per second", or cps. I still often find myself saying, "put a 400 cycle note through it". Cycles per second was a descriptive term, and lent itself to math. Hertz is just a name which has no correlate to a value.
Last edited:
The values in schematics of that era were almost exclusively in micro farads if not marked otherwise. It was common to omit μF or mfd since the use of milli, pico, or nano was not standardized at the time.
The use of milli farads is still frowned upon because of the use of mF in the past to signify micro farads.
A 2000μF 200V+ capacitor back then would have been massive and absurdly expensive.
Assuming the schematic you are working from is this: http://www.radiomuseum.org/forumdata/upload/ernie%2Epdf
That would also make C3 100000μF at 200V+. Even today such capacitor would more than twice the size of a pop can and cost hundreds of dollars new. This is the closest thing I can find to such a capacitor (at only a 160V rating): http://www.mouser.com/ProductDetail/Nichicon/LNR2C104MSE/
100μF is a much more sane choice than 100milli-F or 100nF.
Copper was cheap back then and a 20H filter choke would not be uncommon. Chokes were cheaper than a large value capacitor. Here is an example of a 20H filter choke still in production today: http://www.mouser.com/ProductDetail/Hammond-Manufacturing/193C/
The use of milli farads is still frowned upon because of the use of mF in the past to signify micro farads.
A 2000μF 200V+ capacitor back then would have been massive and absurdly expensive.
Assuming the schematic you are working from is this: http://www.radiomuseum.org/forumdata/upload/ernie%2Epdf
That would also make C3 100000μF at 200V+. Even today such capacitor would more than twice the size of a pop can and cost hundreds of dollars new. This is the closest thing I can find to such a capacitor (at only a 160V rating): http://www.mouser.com/ProductDetail/Nichicon/LNR2C104MSE/
100μF is a much more sane choice than 100milli-F or 100nF.
Copper was cheap back then and a 20H filter choke would not be uncommon. Chokes were cheaper than a large value capacitor. Here is an example of a 20H filter choke still in production today: http://www.mouser.com/ProductDetail/Hammond-Manufacturing/193C/
Last edited:
I remember the term "mickey mike" being used too.
If the output of this circuit was being fed into some type of gated counter, accentuating the high frequencies would help keep the rising edges sharp. The rest of the signal wasn't important, just the edges for triggering the timing / counting circuits of the day.
Depending on what your actual use will be, your values may need some adjustment.
BinaryMike's 13.7 seems to tally with my 16kHz minimum frequency response based on RC time constants for the cathode shunt capacitors. If we can trust Wikipedia to say the time constant should be 10x the minimum frequency response, but we can can't we?
QS92/10 valves are in the air from Spain as we type. Should know more soon when I knock it up. I have a hope that the QS92/10 will spit out so much noise that I won't actually need the pentodes to hear it. I saw a data sheet that mentioned a max. noise of 100mV (rms or peak?). I have a 20V Zener diode that you can physically hear through a speaker with only a voltage follower in between.
QS92/10 valves are in the air from Spain as we type. Should know more soon when I knock it up. I have a hope that the QS92/10 will spit out so much noise that I won't actually need the pentodes to hear it. I saw a data sheet that mentioned a max. noise of 100mV (rms or peak?). I have a 20V Zener diode that you can physically hear through a speaker with only a voltage follower in between.
Well I've found out thanks to this forum. I had been fooled into thinking that a QS92/10 is a neon bulb as that's what the schematic symbol suggests. It's actually a cold cathode voltage regulator tube, and hence subject to microphonics and general vibrations from stomping about in the ERNIE office.
QUESTION: Does anyone here put their amplifier valves into anti-vibration mounts? The military minded types perhaps? Seems that someone could market and sell this feature (useful or not) as a product differentiator in an expensive amp.
Check out this informational web site, among various dampers they have also MIL stuff dampers.
Vacuum Tube Dampers / Vacuum Tube Coolers For Sale
For anyone following along and still confused:
ERNIE - A Random Number Generator
this article has a nice background on how the random number generator worked.
ERNIE - A Random Number Generator
this article has a nice background on how the random number generator worked.
I think the same symbol is often used for any cold cathode valve. Any discharge tube is likely to suffer in the same way from vibration. I guess ERNIE wanted to be sure that it was no only incorruptible but seen to be incorruptible.Paul Uszak said:
Anti-vibration mounts are rarely needed for ordinary valves in ordinary circuits.
Hertz = cycles per second when you describe what it is
It is more in analogy to Volts, Amperes, Watt, Siemens, Tesla, Henry etc. etc. all named after their inventors. It makes international standardization easier when compared to descriptive terms. What "revs per second" is in a country might very well be "revs per minute" in another country. Not the best example, I know but I lack the creativity now to find a better one
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.