21 Aug
2008
21 Aug
'08
8:13 p.m.
> A multi-stage direct coupled vacuum tube
circuit.? Fun.
Yes, be prepared for some fun, in both senses of the
word: it is fun to play
around with, but the implemention of tube logic can be problematic or
unreliable, at least in the way the ABC tried to implement both NAND and NOR
gates with resistive input circuitry.
Atanasoff makes it sounds easy in his paper, but if one reads it closely
it's not quite so, at least as measured by modern standards where one comes
up with a gate design and then simply repeats it ad infinitum.
After reviewing the ASM circuit on your web page, it became readily apparent to me that it
would likely be a major challenge to fill in the blanks for those missing resistor values
and end up with something that actually worked reliably.
The ABC reconstruction and the original required (at
least some) hand-picked
resistors in the gate circuits.
I bought the excellent book ?The First Computers ? History and Architectures? mainly
because of its content on Konrad Zuse?s relay-based machines. However, in its included
paper on the reproduction of the ABC, the authors state:
?In building add-subtract modules, we found the circuits very demanding of precise
resistor values. We have evidence that Berry hand-selected resistors from bins until he
found ones that worked, and we attempted the same tactic. In measuring the
characteristics of 10% resistors, we discovered the distribution about the nominal value
shown in Fig. 3. (It?s a graph of a bell curve between -10% and 10% with a big notch in
the center of the curve). Apparently, the manufacturer had already segregated the
resistors close to nominal value. Hence, we found it necessary to use 1% tolerance
resistors.?
So, here?s a 14 gate, multistage, direct coupled vacuum tube circuit that apparently
requires various hand-selected values of 1% resistors to work properly. Whew!!! I think
I?m going to start with something a great deal simpler, like a single flip flop or astable
multivibrator using an inexpensive 9-pin miniature dual triode at low voltages to see if I
can even get that to work.
Some of the issues:
1. Tubes can be insufficiently non-linear. Driving between saturation and cutoff
? can take a bit of swing and saturation is soft (curved).
2. Plate circuit impedance is high. If the difference between the plate circuit
impedance (lower better) and the grid circuit impedance (higher better) is
? ? insufficient, then different fan-outs in the logic circuit (loading), and
? ? varying tube/component characteristics, can pull the logic levels away from
? ? the design targets and upset everything.
Going to the low plate voltage is a nice idea from a practical view but I
wonder if it may compound issues of point 2 above. It will reduce the voltage
shift between the plate and grid circuits that needs to be accomplished, but it
may also reduce the voltage swing between the 0/1 logic levels. How it works
out in the balance will be interesting.
Thanks for that info and it will be interesting indeed to see if this will work. I was
surprised to find nothing on low voltage tube logic when there?s so much on the web about
low-voltage tube audio circuits of all kinds. True, tube logic is a fairly esoteric
subject with no usefulness beyond that of a novelty, but with all of the audio low voltage
stuff, I?d expected some tube enthusiast somewhere to have already done something on the
logic side just for the heck of it.
Straying completely away from technical issues, and
I'm somewhat loath to
mention this, but the animosity engendered by the early-70s court battle
continues decades later. The somewhat nasty inter-personal battle
found it's way into the Amazon book reviews as recently as 2004.
Another bizarre twist in the ENIAC patent saga on the legal/social side.
Yes, I saw that when I went to look at the Amazon entry for Burks' "The First
Electronic Computer." Wow, it?s a long-time, knock-down feud for certain.
Bill
21 Aug
21 Aug
8:26 p.m.
William Blair wrote:
The other factor ... this was a proof of concept here. OK , it worked
We got our Grant, Fame,PHD what ever and it shelved as it was for
achademic use only.
Like Zuse, and the other computer-calculators of the time had
paractical problems to solve.
The other minor factor ... solid state diodes did not happen untill
about the middle of WWII.
Ben,
> A
multi-stage direct coupled vacuum tube circuit. Fun.
>
Yes, be prepared for some fun, in both senses of
the word: it is fun to play
around with, but the implemention of tube logic can be problematic or
unreliable, at least in the way the ABC tried to implement both NAND and NOR
gates with resistive input circuitry.
Atanasoff makes it sounds easy in his paper, but if one reads it closely
it's not quite so, at least as measured by modern standards where one comes
up with a gate design and then simply repeats it ad infinitum.
After reviewing the ASM circuit on your web page, it became readily apparent to me that
it would likely be a major challenge to fill in the blanks for those missing resistor
values and end up with something that actually worked reliably.
The ABC reconstruction and the original required
(at least some) hand-picked
resistors in the gate circuits.
I bought the excellent book ?The First Computers ? History and Architectures? mainly
because of its content on Konrad Zuse?s relay-based machines. However, in its included
paper on the reproduction of the ABC, the authors state:
?In building add-subtract modules, we found the circuits very demanding of precise
resistor values. We have evidence that Berry hand-selected resistors from bins until he
found ones that worked, and we attempted the same tactic. In measuring the
characteristics of 10% resistors, we discovered the distribution about the nominal value
shown in Fig. 3. (It?s a graph of a bell curve between -10% and 10% with a big notch in
the center of the curve). Apparently, the manufacturer had already segregated the
resistors close to nominal value. Hence, we found it necessary to use 1% tolerance
resistors.?
So, here?s a 14 gate, multistage, direct coupled vacuum tube circuit that apparently
requires various hand-selected values of 1% resistors to work properly. Whew!!! I think
I?m going to start with something a great deal simpler, like a single flip flop or astable
multivibrator using an inexpensive 9-pin miniature dual triode at low voltages to see if I
can even get that to work.
8:48 p.m.
The other minor factor ... solid state diodes did not
happen untill about
the middle of WWII.
Yes, and they sucked anyway.
Still, if there was enough money behind a computer project, I bet RCA
or Sylvania could have been convinced to make a run of diode gating
tubes using 6H6 elements.
--
Will
10:15 p.m.
On 21 Aug 2008 at 23:48, William Donzelli wrote:
Yes, and they sucked anyway.
1N21s could be pretty touchy mechanically and not at all uniform. As
a "microwave diode", I don't think they would have even been
considered for digital logic.
They weren't the first packaged solid-state diode. Copper-oxide
rectifiers were around since the 1920's, but have lousy reverse-
voltage ratings (are they still used in the Simpson 260 VOM?).
Selenium rectifiers were around a short time before the war, but not
popular till afterward (I can still recall the smell of a failing
one).
Cheers,
Chuck
22 Aug
22 Aug
7:15 a.m.
>>>> "Chuck" == Chuck Guzis
<cclist at sydex.com> writes:
Chuck> 1N21s could be pretty touchy mechanically and not at all
Chuck> uniform. As a "microwave diode", I don't think they would
Chuck> have even been considered for digital logic.
Chuck> They weren't the first packaged solid-state diode.
Chuck> Copper-oxide rectifiers were around since the 1920's, but have
Chuck> lousy reverse- voltage ratings (are they still used in the
Chuck> Simpson 260 VOM?). Selenium rectifiers were around a short
Chuck> time before the war, but not popular till afterward (I can
Chuck> still recall the smell of a failing one).
Did Univac 1 use selenium diodes? I certainly remember them as
rectifiers.
My father had an article, or data sheet, describing a copper-oxide
FET. The article was from well before WW2. I haven't seen it in
decades, unfortunately. But that fits with a transistor timeline that
showed up in this week's edition of one of the EE trade rags, which
says that the FET was patented in 1925 "but never built". I wonder
about the accuracy of that "never built". Or maybe vaporware was
invented way back then, too? :-)
paul
21 Aug
21 Aug
10:45 p.m.
William Donzelli wrote:
Just a datapoint I'm aware of: within a few years (late 40's) Whirlwind used
'crystal diodes' in the 'control matrix'.
The other minor factor ... solid state diodes
did not happen untill about
the middle of WWII.
Yes, and they sucked anyway. Still, if there was enough money behind a computer
project, I bet RCA
or Sylvania could have been convinced to make a run of diode gating
tubes using 6H6 elements.
(There was a triple diode released in the 50s or 60s)
9:55 p.m.
Just a datapoint I'm aware of: within a few years
(late 40's) Whirlwind used
'crystal diodes' in the 'control matrix'.
By the late 1940s, crystal diodes were much better animals.
(There was a triple diode released in the 50s or 60s)
6JU8? The Bridge-in-a-bottle?
Moot by then.
--
Will
11:39 p.m.
William Donzelli wrote:
Flipping through the 1960 RCA tube manual.. things like the 6BC7 or 6BJ7
independant triple-diodes (intended for FM disc. and color TVs .. signal
diodes), noval base. I'm noticing there are some interesting tubes such as the
6BN8 (fully independant twin-diode-and-a-triode) that might be turned into
2-input diode-logic gates.
6JU8 isn't here, must be later.
Just a datapoint I'm aware of: within a few
years (late 40's) Whirlwind used
'crystal diodes' in the 'control matrix'.
By the late 1940s, crystal diodes were much better animals.
(There was a triple diode released in the 50s or
60s)
6JU8? The Bridge-in-a-bottle? Moot by then.
..yes
22 Aug
22 Aug
5:57 a.m.
Flipping through the 1960 RCA tube manual.. things
like the 6BC7 or 6BJ7
independant triple-diodes (intended for FM disc. and color TVs .. signal
diodes), noval base.
I forgot about those - dumb of me, as I have about 100 or so of the things.
I'm noticing there are some interesting tubes such
as the
6BN8 (fully independant twin-diode-and-a-triode) that might be turned into
2-input diode-logic gates.
It appears that none of these designs ever had computer rated variants.
--
Will
6:43 a.m.
Computer grade tubes did have to be tougher by far than the signal
amplifiers. ENIAC had a problem with burning out tubes rather quickly
because of that issue. Transisters were probably adapted much more
quickly in switching networks than in signal processing for that
reason and because no extra work had to be done to handle the stray
capacitances that tubes suffer from. Transisters in signal processing
had not yet achieved a state of the art that lent them to be used for
much beyond those rather tinny sounding transister radios by 1960.
I remember a book in the Old Dominion University library with a
printing date of 1954 that showed practical schematics for vacuum
tube logic gates with the necessary ancillary passive components. Like
RTL but with more capacitors.
bs
On Fri, 22 Aug 2008, William Donzelli wrote:
Flipping
through the 1960 RCA tube manual.. things like the 6BC7 or 6BJ7
independant triple-diodes (intended for FM disc. and color TVs .. signal
diodes), noval base.
I forgot about those - dumb of me, as I have about 100 or so of the things.
I'm noticing there are some interesting tubes
such as the
6BN8 (fully independant twin-diode-and-a-triode) that might be turned into
2-input diode-logic gates.
It appears that none of these designs ever had computer rated variants.
--
Will
7:42 a.m.
Computer grade tubes did have to be tougher by far
than the signal
amplifiers.
Not really. The ruggedized nature of computer tubes was because most
were design originally for military equipment. Extra mica, beefier
filament, and so forth - things not related to the cathode purity
issues.
ENIAC had a problem with burning out tubes rather
quickly
because of that issue.
ENIAC had problems because the tubes were not computer grade. I also
have to wonder if the team was given tubes of lesser quality - the
"M-R" types from the war years. These were tubes that were made to
feed the demand outside war production, and were sometimes seconds and
tubes that the War department inspector would not pass. We may never
know what tubes they were given originally. When I looked at ENIAC
pieces some time ago, I did notice that the thing did not use JAN
qualified tubes - but they could have been swapped out years ago.
Transisters were probably adapted much more
quickly in switching networks than in signal processing for that
reason and because no extra work had to be done to handle the stray
capacitances that tubes suffer from.
Ultimately, yes, but in the 1950s, the machines were slow enough that
the tiny capacitances in the structure of tube elements did not matter
much.
--
Will
29 Aug
29 Aug
11:55 a.m.
(Picking up where we left off.. a few messages sent to the list last week seem
to have been dropped during the outage.)
William Donzelli wrote:
I think, rather, that that's a good portion of the explanation of why they
were so slow.
Ballpark example, take a 12AU7: the sum of the grid-to-plate and
grid-to-cathode capacitance is around 3 pF. Suppose the network resistance
feeding the grid circuit is 250 KOhm, that's an RC time constant of 0.75uS,
a little better than just 1 MHz. (R can be reduced of course but power
consumption is then on the climb.)
(Not to say there weren't other reasons they were slow..)
Transisters were probably adapted much more
quickly in switching networks than in signal processing for that
reason and because no extra work had to be done to handle the stray
capacitances that tubes suffer from.
Ultimately, yes, but in the 1950s, the machines were slow enough that
the tiny capacitances in the structure of tube elements did not matter
much. 
11:04 a.m.
New subject: Vacuum-tube speed [was Re: Schematics of Atanasoff-Berry Computer logic circuits?]
Ballpark example, take a 12AU7: [...3pF
capacitance...RC time
constant...]
(Not to say there weren't other reasons they were
slow..)
That brings up something I've occasionally wondered about: how fast do
electrons move in a vacuum tube? (I'm talking about the free-flight
path between cathode and plate.) In particular, what's the time delay
between emission from the cathode and reception by the plate? I
imagine it depends on the plate voltage; does that make enough of a
difference to care about? I don't remember enough of the constants to
figure out what sort of acceleration a gradient of, say, 200 V/cm
imparts to a free electron....
/~\ The ASCII der Mouse
\ / Ribbon Campaign
X Against HTML mouse at rodents-montreal.org
/ \ Email! 7D C8 61 52 5D E7 2D 39 4E F1 31 3E E8 B3 27 4B
11:27 p.m.
New subject: Vacuum-tube speed [was Re: Schematics of Atanasoff-Berry Computer logic circuits?]
der Mouse wrote:
I think you need study a broken tube to find the distance
Ballpark
example, take a 12AU7: [...3pF capacitance...RC time
constant...]
(Not to say there weren't other reasons they
were slow..)
That brings up something I've occasionally wondered about: how fast do
electrons move in a vacuum tube? (I'm talking about the free-flight
path between cathode and plate.) In particular, what's the time delay
between emission from the cathode and reception by the plate? I
imagine it depends on the plate voltage; does that make enough of a
difference to care about? I don't remember enough of the constants to
figure out what sort of acceleration a gradient of, say, 200 V/cm
imparts to a free electron....
from the cathode first.
/~\ The ASCII der Mouse
\ / Ribbon Campaign
X Against HTML mouse at rodents-montreal.org
/ \ Email! 7D C8 61 52 5D E7 2D 39 4E F1 31 3E E8 B3 27 4B
30 Aug
30 Aug
12:44 a.m.
New subject: Vacuum-tube speed [was Re: Schematics of Atanasoff-Berry Computer logic circuits?]
I don't
remember enough of the constants to figure out what sort of
acceleration a gradient of, say, 200 V/cm imparts to a free
electron....
I think you need study a broken tube to find the distance from the
cathode first. 
1 Sep
1 Sep
4:08 a.m.
New subject: Vacuum-tube speed
bfranchuk at jetnet.ab.ca wrote:
der Mouse wrote:
The final speed of an accelerated electron only depends on voltage, not
the distance. A quick calculation gives the final speed close to 2% of
the speed of light at 200 V acceleration. That means it travels 10 cm in
aproximately 30 ns. (not counting relativistic effects and putting mean
speed to 1% of c)
In all practical cases the real distance is much smaller than that,
tubes usually isn't so big, at least not in a digital circuit. The x-ray
tube I have is 50 cm, but that is a totally different story. :-)
/G?ran
That brings up something I've occasionally
wondered about: how fast do
electrons move in a vacuum tube? (I'm talking about the free-flight
path between cathode and plate.) In particular, what's the time delay
between emission from the cathode and reception by the plate? I
imagine it depends on the plate voltage; does that make enough of a
difference to care about? I don't remember enough of the constants to
figure out what sort of acceleration a gradient of, say, 200 V/cm
imparts to a free electron....
I think you need study a broken tube to find the distance
from the cathode first.
30 Aug
30 Aug
2:45 a.m.
New subject: Vacuum-tube speed [was Re: Schematics of Atanasoff-Berry Computerlogic circuits?]
der Mouse wrote:
Standard Disclaimer: It's been 30 years since I've done such calculations, so,
if I still have the concepts right:
Given a B+ (cathode-plate differential) of 200V, an electron accelerated from
velocity of 0 at the cathode, to the plate, will have acquired 200 electron-Volts
(eV) of energy when it hits the plate.
Circumventing force/acceleration calculations and simply going backwards from the
energy/mass/velocity relationship:
Given the mass of an electron and now knowing it's kinetic energy, we can calculate
the final velocity of the electron when it hit the plate (with a conversion factor
inserted to convert eV to Joules (m^2Kg/s^2):
E = m v^2 / 2
v = sqrt( 2E / m )
= sqrt( 2 (200 eV) (1.6e-19 (m^2Kg/s^2)/eV) / (9.1e-31 Kg) )
= ~ 8,400,000 m/S
Supposing the plate-cathode distance is 3 mm, and given the average velocity over
that distance is half the final velocity since the electron started at 0:
t = d / (v/2)
= 0.003 m / ( (8.4e6 m/S) / 2 )
= ~ 0.7 nS
.. seems valid perhaps .. Of course, it's not really the cathode-to-plate time
that matters, but the grid-to-plate time (for amplifying tubes), but anyways..
Ballpark example, take a 12AU7: [...3pF
capacitance...RC time
constant...]
(Not to say there weren't other reasons they
were slow..)
That brings up something I've occasionally wondered about: how fast do
electrons move in a vacuum tube? (I'm talking about the free-flight
path between cathode and plate.) In particular, what's the time delay
between emission from the cathode and reception by the plate? I
imagine it depends on the plate voltage; does that make enough of a
difference to care about? I don't remember enough of the constants to
figure out what sort of acceleration a gradient of, say, 200 V/cm
imparts to a free electron....
noon
New subject: Vacuum-tube speed [was Re: Schematics of Atanasoff-Berry Computerlogic circuits?]
Standard Disclaimer: It's been 30 years since
I've done such
calculations, so,
if I still have the concepts right:
Given a B+ (cathode-plate differential) of 200V, an electron
accelerated from
velocity of 0 at the cathode, to the plate, will have acquired 200
electron-Volts
(eV) of energy when it hits the plate.
Circumventing force/acceleration calculations and simply going
backwards from the
energy/mass/velocity relationship:
Given the mass of an electron and now knowing it's kinetic energy, we
can calculate
the final velocity of the electron when it hit the plate (with a
conversion factor
inserted to convert eV to Joules (m^2Kg/s^2):
E = m v^2 / 2
v = sqrt( 2E / m )
= sqrt( 2 (200 eV) (1.6e-19 (m^2Kg/s^2)/eV) / (9.1e-31 Kg) )
= ~ 8,400,000 m/S
Supposing the plate-cathode distance is 3 mm, and given the average
velocity over
that distance is half the final velocity since the electron started at
0:
t = d / (v/2)
= 0.003 m / ( (8.4e6 m/S) / 2 )
= ~ 0.7 nS
.. seems valid perhaps .. Of course, it's not really the cathode-to-
plate time
that matters, but the grid-to-plate time (for amplifying tubes), but
anyways..
Seems plausible. If you take a "lighthouse" tube, the cathode to plate
distance is smaller and the voltages higher, so they are good up to 3
GHz or so.
Another consideration: velocity modulation. The changing grid voltage
changes the electron speed, and if the transit time is a significant
fraction of the signal period, you end up with the electrons clumping
together. For "regular" tubes that sort of behavior is undesirable and
will make for strange results. (For klystrons, it's how they work,
which is why those are good for microwave amplifiers.)
paul
29 Aug
29 Aug
11:13 a.m.
I think, rather, that that's a good portion of the
explanation of why they
were so slow.
Ballpark example, take a 12AU7: the sum of the grid-to-plate and
grid-to-cathode capacitance is around 3 pF. Suppose the network resistance
feeding the grid circuit is 250 KOhm, that's an RC time constant of 0.75uS,
a little better than just 1 MHz. (R can be reduced of course but power
consumption is then on the climb.)
Well, OK, i could buy that, but from what I have seen from the
circuits and construction of tube logic circuits is that the 3 pF is
not much compared to all the stray capacitance kicking around due to
construction techniques.
--
Will
11:59 a.m.
On 29 Aug 2008 at 14:13, William Donzelli wrote:
Well, OK, i could buy that, but from what I have seen
from the
circuits and construction of tube logic circuits is that the 3 pF is
not much compared to all the stray capacitance kicking around due to
construction techniques.
In particular, the tube socket. Not as bad as the phenolic based
tubes, but still a factor. In early HF work, you'd see tubes with
the base removed or with hacksaw slots in the base between pins--and
connections soldered directly to the pins.
Your comment about space-charge tubes brings up the question "Did any
AM/FM auto radios used them?" An FM application might indicate some
speed potential.
Cheers,
Chuck
12:12 p.m.
>>>> "Chuck" == Chuck Guzis
<cclist at sydex.com> writes:
Chuck> Your comment about space-charge tubes brings up the question
Chuck> "Did any AM/FM auto radios used them?" An FM application
Chuck> might indicate some speed potential.
I've used 12 volt ("car radio") tubes in a shortware broadcast
receiver. Worked fine as far as I remember.
paul
1:03 p.m.
On 29 Aug 2008 at 15:12, Paul Koning wrote:
I've used 12 volt ("car radio") tubes in
a shortware broadcast
receiver. Worked fine as far as I remember.
Were these the special "space charge" tubes? e.g. 12J8, 12U7, 12DK7,
12DL8, 12CX6, 12DE8, 12CN5, 12BL6, 12DS7, 12AC6, 12AD6, 12AE6, 12AE7,
12AF6, 12AJ6 or 12AL8 (I've probably left out a few)?
Cheers,
Chuck
1:10 p.m.
>>>> "Chuck" == Chuck Guzis
<cclist at sydex.com> writes:
Chuck> On 29 Aug 2008 at 15:12, Paul Koning wrote:
> I've used 12 volt ("car radio")
tubes in a shortware broadcast
> receiver. Worked fine as far as I remember.
Chuck> Were these the special "space charge" tubes? e.g. 12J8, 12U7,
Chuck> 12DK7, 12DL8, 12CX6, 12DE8, 12CN5, 12BL6, 12DS7, 12AC6, 12AD6,
Chuck> 12AE6, 12AE7, 12AF6, 12AJ6 or 12AL8 (I've probably left out a
Chuck> few)?
I don't recognize the term "space charge". They were tubes explicitly
specified in the data sheet for operation with 12 volt anode voltage.
Almost certainly they were Philips -- EFnn and ECHnn for suitable
integer nn... (I did that project in Holland.)
paul
2 p.m.
Chuck> Were these the special "space
charge" tubes? e.g. 12J8, 12U7,
Chuck> 12DK7, 12DL8, 12CX6, 12DE8, 12CN5, 12BL6, 12DS7, 12AC6, 12AD6,
Chuck> 12AE6, 12AE7, 12AF6, 12AJ6 or 12AL8 (I've probably left out a
Chuck> few)?
I don't recognize the term "space charge". They were tubes explicitly
specified in the data sheet for operation with 12 volt anode voltage.
Almost certainly they were Philips -- EFnn and ECHnn for suitable
integer nn... (I did that project in Holland.)
I think nn=83 :-). I am really digging into the depths of my memory, but
I seem to remember the ECH83 (triode hexode frequency changer) and EBF83
(double diode and pentode). Maybe at least one other type too.
-tony
2:14 p.m.
>>>> "Tony" == Tony Duell <ard
at p850ug1.demon.co.uk> writes:
Chuck> Were these the special "space charge" tubes? e.g. 12J8, 12U7,
Chuck> 12DK7, 12DL8, 12CX6, 12DE8, 12CN5, 12BL6, 12DS7, 12AC6, 12AD6,
Chuck> 12AE6, 12AE7, 12AF6, 12AJ6 or 12AL8 (I've probably left out a
Chuck> few)?
> I don't recognize the term "space
charge". They were tubes
> explicitly specified in the data sheet for operation with 12 volt
> anode voltage. Almost certainly they were Philips -- EFnn and
> ECHnn for suitable integer nn... (I did that project in Holland.)
Tony> I think nn=83 :-). I am really digging into the depths of my
Tony> memory, but I seem to remember the ECH83 (triode hexode
Tony> frequency changer) and EBF83 (double diode and pentode). Maybe
Tony> at least one other type too.
Sounds right. I don't remember EBF83, I thought I had a semiconductor
diode for the AM detector stage. So that would have made an EFnn IF
amp.
paul
31 Aug
31 Aug
8:59 a.m.
As per the 1982 Amateur Radio Handbook, ECC83 (and ECC82) are double
triodes. Type 83 is described as a half wave rectifier in the 1951
edition. None of the 12* are listed in the '51 book though similar
numbers indicate double diodes and triodes as do base descriptions
in the 1982 book but I cannot come up with more precise info.
bs
On Fri, 29 Aug 2008, Tony Duell wrote:
Chuck>
Were these the special "space charge" tubes? e.g. 12J8, 12U7,
Chuck> 12DK7, 12DL8, 12CX6, 12DE8, 12CN5, 12BL6, 12DS7, 12AC6, 12AD6,
Chuck> 12AE6, 12AE7, 12AF6, 12AJ6 or 12AL8 (I've probably left out a
Chuck> few)?
I don't recognize the term "space charge". They were tubes explicitly
specified in the data sheet for operation with 12 volt anode voltage.
Almost certainly they were Philips -- EFnn and ECHnn for suitable
integer nn... (I did that project in Holland.)
I think nn=83 :-). I am really digging into the depths of my memory, but
I seem to remember the ECH83 (triode hexode frequency changer) and EBF83
(double diode and pentode). Maybe at least one other type too.
-tony
11:39 a.m.
As per the 1982 Amateur Radio Handbook, ECC83 (and ECC82) are double
triodes. Type 83 is described as a half wave rectifier in the 1951
Correct, but I think you've misinterpetted what I was saying.
Philips/Mullard valve numbers are actually quite informative, and can be
decoded as follows :
First letter gives the heater/filament rating :
A = 4V
B = 180mA???
C = 200mA
D = 1.5V
E = 6.3V
G = 5V (early-ish ones, later used for miscellaneous ratings)
H = 12.6V?
K = 2V
O = Semiconductor (no heater)
P = 300mA
U = 100mA
Y = 450mA
(Where a current is given, the valve was intended for series-string
operation)
Subsequent letters give the electrode structure. For multi-section vales,
they're given in alphabetical order
A = diode
B = double diode
C = triode
D = power/output triode
E = signal tetrode
F = signal pentode
H = hextode/heptode
K = heptode (phantom-cathode tpye of fequency changer)/octode
L = power/output tetrode/pentode
M = tuning indicator ('magic eye')
N = thyratron (gas-filled triode/tetrode)
Q = nonode
X = gass-filed full-wave rectifier
Y = half-wave rectifier
Z = vaccume full-wave rectifier
The first digit gives the base type
None/1 = miscllaneous (often only a single-digit number means P side
contact base)
2 = B8B 'Loctal' For 3-digit numbers, 2 = B10B
3 = International Octal
4 = B8A Rimlock
5 = B9G. For 3 digit numbers, 5 = B9D
6,7 Subminiatores
8 = B9A Noval
9 = B7G (7 pin miniatore)
Other digits distinguish between valves where the rest of the code is
the same.
Now the oroginal poster was talking about some 12V-anode-voltage valves.
He'd rememebrs are frequency changer -- a triode hexode -- with an ECHnn
bumber. I beleive the full number of that valve is the ECH83. And that
the EBF83, and possibly the EF83 are als 12V anode types.
This has nothing to do with the the American type 83 rectifier valve
-tony
1 Sep
1 Sep
11:53 a.m.
Tony is correct. I sell Tesla type ECC-83s to the audio tweaks. It is the
functional equivalent ( except better ) of a 12AX7 dual triode. The filament
on the ECC is on pins 4 and 5 with a center-tap connected to pin 9 such that
you can run the filament at either 6.3 or 12.7 Volts.
Best regards, Steven
As per the
1982 Amateur Radio Handbook, ECC83 (and ECC82) are double
triodes. Type 83 is described as a half wave rectifier in the 1951
Correct, but I think you've misinterpetted what I was saying.
Philips/Mullard valve numbers are actually quite informative, and can be
decoded as follows :
First letter gives the heater/filament rating :
A = 4V
B = 180mA???
C = 200mA
D = 1.5V
E = 6.3V
G = 5V (early-ish ones, later used for miscellaneous ratings)
H = 12.6V?
K = 2V
O = Semiconductor (no heater)
P = 300mA
U = 100mA
Y = 450mA
(Where a current is given, the valve was intended for series-string
operation)
Subsequent letters give the electrode structure. For multi-section vales,
they're given in alphabetical order
A = diode
B = double diode
C = triode
D = power/output triode
E = signal tetrode
F = signal pentode
H = hextode/heptode
K = heptode (phantom-cathode tpye of fequency changer)/octode
L = power/output tetrode/pentode
M = tuning indicator ('magic eye')
N = thyratron (gas-filled triode/tetrode)
Q = nonode
X = gass-filed full-wave rectifier
Y = half-wave rectifier
Z = vaccume full-wave rectifier
The first digit gives the base type
None/1 = miscllaneous (often only a single-digit number means P side
contact base)
2 = B8B 'Loctal' For 3-digit numbers, 2 = B10B
3 = International Octal
4 = B8A Rimlock
5 = B9G. For 3 digit numbers, 5 = B9D
6,7 Subminiatores
8 = B9A Noval
9 = B7G (7 pin miniatore)
Other digits distinguish between valves where the rest of the code is
the same.
Now the oroginal poster was talking about some 12V-anode-voltage valves.
He'd rememebrs are frequency changer -- a triode hexode -- with an ECHnn
bumber. I beleive the full number of that valve is the ECH83. And that
the EBF83, and possibly the EF83 are als 12V anode types.
This has nothing to do with the the American type 83 rectifier valve
-tony
12:18 p.m.
Notice that in the bad old days part numbers could be decifered to tell
you what the hell they actually did? We don't have those conventions
anymore it seems.
bs
On Mon, 1 Sep 2008, Scanning wrote:
Tony is correct. I sell Tesla type ECC-83s to the
audio tweaks. It is the
functional equivalent ( except better ) of a 12AX7 dual triode. The filament
on the ECC is on pins 4 and 5 with a center-tap connected to pin 9 such that
you can run the filament at either 6.3 or 12.7 Volts.
Best regards, Steven
As per
the 1982 Amateur Radio Handbook, ECC83 (and ECC82) are double
triodes. Type 83 is described as a half wave rectifier in the 1951
Correct, but I think you've misinterpetted what I was saying.
Philips/Mullard valve numbers are actually quite informative, and can be
decoded as follows :
First letter gives the heater/filament rating :
A = 4V
B = 180mA???
C = 200mA
D = 1.5V
E = 6.3V
G = 5V (early-ish ones, later used for miscellaneous ratings)
H = 12.6V?
K = 2V
O = Semiconductor (no heater)
P = 300mA
U = 100mA
Y = 450mA
(Where a current is given, the valve was intended for series-string
operation)
Subsequent letters give the electrode structure. For multi-section vales,
they're given in alphabetical order
A = diode
B = double diode
C = triode
D = power/output triode
E = signal tetrode
F = signal pentode
H = hextode/heptode
K = heptode (phantom-cathode tpye of fequency changer)/octode
L = power/output tetrode/pentode
M = tuning indicator ('magic eye')
N = thyratron (gas-filled triode/tetrode)
Q = nonode
X = gass-filed full-wave rectifier
Y = half-wave rectifier
Z = vaccume full-wave rectifier
The first digit gives the base type
None/1 = miscllaneous (often only a single-digit number means P side
contact base)
2 = B8B 'Loctal' For 3-digit numbers, 2 = B10B
3 = International Octal
4 = B8A Rimlock
5 = B9G. For 3 digit numbers, 5 = B9D
6,7 Subminiatores
8 = B9A Noval
9 = B7G (7 pin miniatore)
Other digits distinguish between valves where the rest of the code is
the same.
Now the oroginal poster was talking about some 12V-anode-voltage valves.
He'd rememebrs are frequency changer -- a triode hexode -- with an ECHnn
bumber. I beleive the full number of that valve is the ECH83. And that
the EBF83, and possibly the EF83 are als 12V anode types.
This has nothing to do with the the American type 83 rectifier valve
-tony
2 Sep
2 Sep
9:49 a.m.
Notice that in the bad old days part numbers could be
decifered to tell
you what the hell they actually did?
You do not use tubes much, do you?
Most of the old numbering systems had so many special cases, mistakes
in registration, and other weirdness that the systems basically could
not be trusted.
Everyone was to blame - the American RMA, the various British systems
(the worst offenders), the Continental system, the German systems, the
Japanese systems...
The best system turned out to be the 5500 series non-system in use
from roughly 1950 to 1970.
--
Will
1:37 a.m.
On Mon, 2008-09-01 at 11:53 -0700, Scanning wrote:
Tony is correct. I sell Tesla type ECC-83s to the
audio tweaks. It is the
functional equivalent ( except better ) of a 12AX7 dual triode.
It's not really an "audio tweak" valve, at least not in the UK. It's a
bog standard entry-level dual triode used in the preamp stage of just
about any guitar amp you care to look at ;-)
The fact that audiophiles love them seems pretty funny. Us amp builders
love them for their second-harmonicy crunch...
Gordon
4:44 a.m.
Gordon,
If I had meant audiophiles I would have stated such. The "tweaks" are the
ones that can hear the nuances in an amplifier that cannot be quantified
because they don't exist. They describe an amplifier as if it were a bottle
of wine... Somehow they are never willing to submit to a double-blind test
to check their palette.
Best regards, Steven
----- Original Message -----
From: "Gordon J. C. Pearce" <gordonjcp at gjcp.net>
To: "General Discussion: On-Topic and Off-Topic Posts"
<cctalk at classiccmp.org>
Sent: Tuesday, September 02, 2008 1:37 AM
Subject: Re: Schematics of Atanasoff-Berry Computer logic circuits?
On Mon, 2008-09-01 at 11:53 -0700, Scanning wrote:
> Tony is correct. I sell Tesla type ECC-83s to the audio tweaks. It is
the
functional
equivalent ( except better ) of a 12AX7 dual triode.
It's not really an "audio tweak" valve, at least not in the UK. It's
a
bog standard entry-level dual triode used in the preamp stage of just
about any guitar amp you care to look at ;-)
The fact that audiophiles love them seems pretty funny. Us amp builders
love them for their second-harmonicy crunch...
Gordon
8:47 a.m.
Gordon J. C. Pearce wrote:
any guitar amp you care to look at ;-)
The fact that audiophiles love them seems pretty funny.
That is not me ... I hate
them for see above err below.
Us amp builders
love them for their second-harmonicy crunch...
Still with out the guitar amp people, I don't think any kind of valves
would be around for people like myself to play music through.
Gordon
PS: I want quality parts ... not 1 cent Jap parts or $100 snake oil
'audiophile'
components.
8:56 a.m.
On Sep 2, 2008, at 11:47 AM, bfranchuk at jetnet.ab.ca wrote:
Did all the tubes disappear or something? LOTS AND LOTS of tubes
were made during their heyday. I know a few people who have
hundreds, and in one case thousands of tubes, and I've got a few
dozen myself. Don't be paranoid, Ben. ;)
-Dave
--
Dave McGuire
Port Charlotte, FL
Us amp
builders
love them for their second-harmonicy crunch...
Still with out the guitar amp people, I don't think any kind of valves
would be around for people like myself to play music through.
9:12 a.m.
>>>> "Dave" == Dave McGuire
<mcguire at neurotica.com> writes:
Dave> On Sep 2, 2008, at 11:47 AM, bfranchuk at jetnet.ab.ca wrote:
>> Us amp builders love them for their
second-harmonicy crunch...
>>
> Still with out the guitar amp people, I don't think any kind of
> valves would be around for people like myself to play music
> through.
Dave> Did all the tubes disappear or something? LOTS AND LOTS of
Dave> tubes were made during their heyday. I know a few people who
Dave> have hundreds, and in one case thousands of tubes, and I've got
Dave> a few dozen myself. Don't be paranoid, Ben. ;)
Last time I was at a hamfest (admittedly some years ago) tubes were
still being sold by lots of people. Ads show up in ham radio
magazines, too.
So avoid audio-nut circles and look to hams, then tubes shouldn't eba
problem.
Then again, if all else fails, make your own:
http://blog.makezine.com/archive/2008/01/make_your_own_vaccum_tube.html
paul
9:27 a.m.
Dave McGuire wrote:
Did all the tubes disappear or something? LOTS AND LOTS of tubes
were made during their heyday. I know a few people who have hundreds,
and in one case thousands of tubes, and I've got a few dozen myself.
Don't be paranoid, Ben. ;)
Good ... send me a DOZEN AD1's to play with :)
-Dave
I know there are a lots of TV & Radio tubes around. Just find
a good IF
transformer
now days, let alone a Non-talk AM radio channel.
9:33 a.m.
On Sep 2, 2008, at 12:27 PM, bfranchuk at jetnet.ab.ca wrote:
Well yeah, if you want to work with the super-rare ones.. ;)
Did all the
tubes disappear or something? LOTS AND LOTS of
tubes were made during their heyday. I know a few people who have
hundreds, and in one case thousands of tubes, and I've got a few
dozen myself. Don't be paranoid, Ben. ;)
Good ... send me a DOZEN AD1's
to play with :) I know there are a lots of TV & Radio tubes
around. Just find a
good IF transformer
now days, let alone a Non-talk AM radio channel.
IF transformers are easy to wind!
-Dave
--
Dave McGuire
Port Charlotte, FL
9:53 a.m.
Did all the tubes disappear or something? LOTS AND
LOTS of tubes were
made during their heyday. I know a few people who have hundreds, and in one
case thousands of tubes, and I've got a few dozen myself. Don't be
paranoid, Ben. ;)
I am down to about 10,000. At peak, I had roughly 40,000.
And I am small potatoes...
Some tubes did disappear. I pity the man that wants to get a power
supply for an IBM 709 up and running. C16Js, in used UNTESTED
condition, go for a few hundred bucks. If you can find them.
Same with anyone that gets a computer using 6386 dual triodes.
--
Will
10:17 a.m.
William Donzelli wrote:
I am down to about 10,000. At peak, I had roughly
40,000.
How did you get so-many?
And I am small potatoes...
From Idaho too? :)
Some tubes did disappear. I pity the man that wants to
get a power
supply for an IBM 709 up and running. C16Js, in used UNTESTED
condition, go for a few hundred bucks. If you can find them.
Same with anyone that gets a computer using 6386 dual triodes.
That would be nice to see ... are there any valve computers
up and running still?
--
Will
12:53 p.m.
William Donzelli wrote:
Does the CHM have a complete 'large'-category tube system that could
conceivably be made to run again?
When I spoke to Paul Pierce a few years ago he hoped to get his 709 running
again, but I think he was viewing it as a retirement project.
I would hope some of those smaller drum machines (Bendix, LGP-xx, etc) which
are still around in some number would run again. (Yes, there's the one in
Germany). IIRC, Tom Jennings had an LGP-<something> he was working on.
That would be nice to see ... are there any
valve computers
up and running still?
In Britain and Germany, at least. I know of none in North America.
12:04 p.m.
Does the CHM have a complete 'large'-category
tube system that could
conceivably be made to run again?
Al?
When I spoke to Paul Pierce a few years ago he hoped
to get his 709 running
again, but I think he was viewing it as a retirement project.
Certainly a 709 is possible. Depending on the condition of the tubes,
or how well one has been squirreling them away, it could be expensive
as well.
I would hope some of those smaller drum machines
(Bendix, LGP-xx, etc) which
are still around in some number would run again. (Yes, there's the one in
Germany). IIRC, Tom Jennings had an LGP-<something> he was working on.
Yes, the smaller drum machines should not be to bad. 400 small tubes
in a system is not a big deal at all. Things like the condition of the
drum, however, can be a big deal.
I think there is a working IBM 650 in Germany as well.
--
Will
12:11 p.m.
Dave McGuire wrote:
On Sep 2, 2008, at 11:47 AM, bfranchuk at jetnet.ab.ca
wrote:
Did all the tubes disappear or something? LOTS AND LOTS of tubes were
They're still being made, mostly in the former USSR. After Mullard
stopped producing valves in the 1980s, a company in the USSR bought the
whole lot up and shipped it across. If you buy a Sovtek ECC83 or EL34,
it's exactly like an old Mullard one.
Gordon
Us amp
builders
love them for their second-harmonicy crunch...
Still with out the guitar amp people, I don't think any kind of valves
would be around for people like myself to play music through.
12:56 p.m.
Gordon JC Pearce wrote:
They're still being made, mostly in the former USSR. After Mullard
stopped producing valves in the 1980s, a company in the USSR bought
the whole lot up and shipped it across. If you buy a Sovtek ECC83 or
EL34, it's exactly like an old Mullard one.
I buy 6B4G's,5889's,6BQ7A's and throw in a 12DW7 and a 0B2 for good
measure. :)
Gordon
And plenty of Big Iron ... Oh wait you want puter big iron.
12:33 p.m.
New subject: vac tubes / was Re: Schematics of Atanasoff-Berry Computer logic circuits?
Dave McGuire wrote:
Did all the tubes disappear or something? LOTS AND
LOTS of tubes
were made during their heyday. I know a few people who have
hundreds, and in one case thousands of tubes, and I've got a few
dozen myself. Don't be paranoid, Ben. ;)
Just to put a number on it, in 1955 alone, 480,000,000 receiving tubes were
produced (Electronics, Sep. 1956, Figures of the Month). I presume/guess that
was U.S. production.
At the radio museum we get boxes and boxes of what we call "TV-junk" tubes, a
variety of types that were targetted for, or commonly used in, 50/60 era TVs,
and which didn't get used much in home radios. The result, because these days
there are far more people restoring radios and doing tube audio compared to
restoring TVs, is that there are lots and lots of these tubes and a very low
demand for them.
It has occurred to me that if one were setting out to build a (small?) tube
computer today for fun, to look into selecting the primary tube(s) from that
class; the question being if one would do any better than just using
duo-triodes, which are in higher demand.
29 Aug
29 Aug
12:26 p.m.
In particular, the tube socket. Not as bad as the
phenolic based
tubes, but still a factor.
Yes, but even the wiring. Looking at a Burroughs module, I see they
liked to run the signal wires right up against the metal chassis of
the module.
In early HF work, you'd see tubes with
the base removed or with hacksaw slots in the base between pins--and
connections soldered directly to the pins.
Cheap ass hams (that is, 90 percent of them) would run a 10 WAY beyond
what they were made to do, so they cut the sockets (and tube bases) so
they would not burn up due to high voltages. Sometimes you can find a
10 (or other triode) that has not been cut, yet the plastic of the
base is all bubbly from the leakage.
You could find this with some 1960 audio tubes as well, where the amp
maker pushed too hard.
--
Will
2:42 p.m.
On 29 Aug 2008 at 15:26, William Donzelli wrote:
Cheap ass hams (that is, 90 percent of them) would run
a 10 WAY beyond
what they were made to do, so they cut the sockets (and tube bases) so
they would not burn up due to high voltages. Sometimes you can find a
10 (or other triode) that has not been cut, yet the plastic of the
base is all bubbly from the leakage.
Heh. It was the spendthrifts who splurged on a 6V pilot lamp as a
plate current indicator. The rest of us just tuned for the least
amount of orange glow from the plate. :) Ah, for the days of
building CW transmitters from a junker TV chassis.
There were various power tubes (6L6G's included) that could be had
with ceramic bases.
Cheers,
Chuck
2:47 p.m.
William Donzelli wrote:
I'd say point-to-point wiring becomes problematic when you start getting going
into the higher shortwave / low VHF, and granted it becomes an issue in a
physically large system, but it may not be bad as you might think. The 10MHz
stage of the HP 524 counter is done with 'rats-nest' point to point wiring with
components on tag boards separated by wires from the tube sockets, even for the
grid circuits. The key was they went to high-frequency pentodes instead of
duo-triodes, low-R circuits, diode clamps, 1% resistors: hi-speed was
'possible' but very expensive to accomplish.
I think, rather, that that's a good portion
of the explanation of why they
were so slow.
Ballpark example, take a 12AU7: the sum of the grid-to-plate and
grid-to-cathode capacitance is around 3 pF. Suppose the network resistance
feeding the grid circuit is 250 KOhm, that's an RC time constant of 0.75uS,
a little better than just 1 MHz. (R can be reduced of course but power
consumption is then on the climb.)
Well, OK, i could buy that, but from what I have seen from the
circuits and construction of tube logic circuits is that the 3 pF is
not much compared to all the stray capacitance kicking around due to
construction techniques. 
2:04 p.m.
On 29/08/2008 22:47, Brent Hilpert wrote:
I'd say point-to-point wiring becomes problematic
when you start getting going
into the higher shortwave / low VHF, and granted it becomes an issue in a
physically large system, but it may not be bad as you might think.
Sometimes even an asset. I recall a DIY project in Practical Television
in the UK, for a Band III (about 190-200MHz) VHF TV preamp, built in the
proverbial tobacco tin with a VHF pentode (IIRC), and using a length of
coax for its tuned circuit.
--
Pete Peter Turnbull
Network Manager
University of York
8:59 p.m.
On 29 Aug 2008 at 22:04, Pete Turnbull wrote:
Sometimes even an asset. I recall a DIY project in
Practical Television
in the UK, for a Band III (about 190-200MHz) VHF TV preamp, built in the
proverbial tobacco tin with a VHF pentode (IIRC), and using a length of
coax for its tuned circuit.
Before that, you'd find acorn tubes using copper pipe for a tuned
circuit at 112 MHz. Parallel tuned lines made of brass or copper rod
weren't unusual back then for the 2.5 meter band and down.
Cheers,
Chuck
21 Aug
21 Aug
10:37 p.m.
"bfranchuk at jetnet.ab.ca" wrote:
The other factor ... this was a proof of concept here.
OK , it worked
We got our Grant, Fame,PHD what ever and it shelved as it was for
achademic use only.
That is a gross misrepresentation of Atanasoff's objectives. He was attempting
to produce a functioning computing machine to support his 'real' research
objectives (climate modelling, IIRC). It may have been ambitious in that he
wasn't an EE by background, and it may not have been entirely successful, but
it was not a vanity project. He had nowhere near the resources or audience of
other computing projects of the period such as the Stibitz/Bell relay
calculator, Harvard Mark 1, ENIAC or SSEC. (TMK) He already had a PhD. Fame
didn't enter the picture until decades later and unintentionally at that.
8:52 p.m.
Thanks for that info and it will be interesting indeed
to see if this will work. I was surprised to find nothing on low voltage tube logic when
there's so much on the web about low-voltage tube audio circuits of all kinds. True,
tube logic is a fairly esoteric subject with no usefulness beyond that of a novelty, but
with all of the audio low voltage stuff, I'd expected some tube enthusiast somewhere
to have already done something on the logic side just for the heck of it.
Yes, but remember to always question the findings of a tube audio
enthusiast. Yes, there are good ones, but there are a lot of flakes
out there. A LOT.
I invite interested parties to join the TCA (Tube Collectors
Association). We have an email list, and would LOVE talk like this.
The list is full of REAL tube men - the guys that were there.
--
Will
10:26 p.m.
On 21 Aug 2008 at 20:13, William Blair wrote:
Going to the low plate voltage is a nice idea from a
practical view but I
wonder if it may compound issues of point 2 above. It will reduce the voltage
shift between the plate and grid circuits that needs to be accomplished, but it
may also reduce the voltage swing between the 0/1 logic levels. How it works
out in the balance will be interesting.
Just before transistors began to be used in auto radios, "space
charge" tubes made a brief appearance, running their plates and
heaters from the 12v battery. No vibrator supply. Some of the
earliest transistor designs used a transistor only in the audio
output stage, among them the Delco "Wonder Bar" models.
Jeff Duntemann has a nice page on this stuff, along with datasheets:
http://www.duntemann.com/12vtubes/12vtubesindex.htm
Cheers,
Chuck
22 Aug
22 Aug
12:29 a.m.
Chuck Guzis wrote:
Just before transistors began to be used in auto radios, "space
charge" tubes made a brief appearance, running their plates and
heaters from the 12v battery. No vibrator supply. Some of the
earliest transistor designs used a transistor only in the audio
output stage, among them the Delco "Wonder Bar" models.
I remember working on, or dismantling, some of those as a little kid. Memorable
for the absence of hi-V supply, hybrid design, the big heat sink for the
transistor and the oddball case style of some of the transistors (some were
TO-3 but there was another "big-button" style that I don't know the
designation of).
5731
days inactive
5742
days old
55 comments
14 participants
participants (14)
-
ard＠p850ug1.demon.co.uk -
axelsson＠acc.umu.se -
bfranchuk＠jetnet.ab.ca -
cclist＠sydex.com -
gordonjcp＠gjcp.net -
hilpert＠cs.ubc.ca -
mcguire＠neurotica.com -
mouse＠Rodents-Montreal.ORG -
Paul_Koning＠Dell.com -
pete＠dunnington.plus.com -
schwepes＠moog.netaxs.com -
steven.alan.canning＠verizon.net -
wbblair3＠yahoo.com -
wdonzelli＠gmail.com