Topics

7854 vs 7D20


Cliff Carrie
 

I already have a working 7854 and a good assortment of 7000 series plug-ins. What would a 7D20 add that the 7854 cannot do or can only do poorly or with difficulty? Is the 7D20 compatible with the 7854?


Cliff Carrie


Chuck Harris
 

The 7D20, and the 7854 handle digitization in very different
ways. The 7D20 uses a CCD to store the analog waveform, and
then reads it out slowly into a slow ADC to digitize the waveform.
At sufficiently slow speeds, it directly digitizes the waveform
using the ADC... below 50KHz, I think.

The 7854 goes directly to the ADC for slow speeds, below 500KHz,
I think, but for higher frequencies, it behaves as a sampling
scope, taking samples in random positions after the trigger,
until the entire waveform has been sampled.

I use the 7D20 a lot more often for digital duty than I do the
7854.

-Chuck Harris

Cliff Carrie cliffcarrie@... [TekScopes] wrote:

I already have a working 7854 and a good assortment of 7000 series plug-ins. What would a 7D20 add that the 7854 cannot do or can only do poorly or with difficulty? Is the 7D20 compatible with the 7854?


Cliff Carrie


Harvey White
 

On Wed, 13 Dec 2017 02:54:49 +0000, you wrote:

I already have a working 7854 and a good assortment of 7000 series plug-ins. What would a 7D20 add that the 7854 cannot do or can only do poorly or with difficulty? Is the 7D20 compatible with the 7854?
To the best of my knowlege, the 7D20 was used to add digital storage
rather specifically to the 7600 scopes, giving you a large screen
digital scope. I think that the digital bandwidth is about 20 Mhz or
so.

With the additional processing in the 7854, the higher bandwidth, I
can't think of anything in particular the 7D20 would give you.

if you're looking for relative oddities, then a 7A42 logic plugin (4
trace, dual width, somewhat of a logic analyzer, might be an idea.

Harvey



Cliff Carrie



 

Hi Harvey and Cliff
,
The 7D20 is spec'd at 70MHz and 40MSamples/Sec. It is also GPIB compatible
so it is programmable.

The 7854 is also GPIB compatible but in addition it has its own built in
vector processing programming language. That was another thing that made the
7854 revolutionary when it came out. It was the first scope with its own
vector processing language. Unfortunately electronics engineers were
programming illiterate in those days and that feature actually discouraged
sales so Tek quietly stopped promoting it in their ads and in their
literature.

To this day I have not found more than a handful of 7854 programs in spite
of doing an exhaustive search of all of the EE magazines in the 1980s and
1990s. However, all that changed a few months ago when Holger Lubbin (a
member of TekScopes and a dedicated fan of the 7854) wrote what must be the
largest 7854 program ever written. It calculates the FFT of the signal
stored in the 7854 memory. Unfortunately the 7854 did not have floating
point math, a fast microprocessor, or lots of storage (3 things you need to
do an FFT) so it is not practical except as a proof of what the 7854 was
capable of.

Dennis Tillman W7PF

-----Original Message-----
Sent: Tuesday, December 12, 2017 8:25 PM
Subject: Re: [TekScopes] 7854 vs 7D20

On Wed, 13 Dec 2017 02:54:49 +0000, you wrote:

I already have a working 7854 and a good assortment of 7000 series
plug-ins. What would a 7D20 add that the 7854 cannot do or can only do
poorly or with difficulty? Is the 7D20 compatible with the 7854?

To the best of my knowlege, the 7D20 was used to add digital storage rather
specifically to the 7600 scopes, giving you a large screen digital scope. I
think that the digital bandwidth is about 20 Mhz or so.

With the additional processing in the 7854, the higher bandwidth, I can't
think of anything in particular the 7D20 would give you.

if you're looking for relative oddities, then a 7A42 logic plugin (4 trace,
dual width, somewhat of a logic analyzer, might be an idea.

Harvey

Cliff Carrie
------------------------------------
Posted by: Harvey White <madyn@...>
------------------------------------


Craig Sawyers <c.sawyers@...>
 

Unfortunately the 7854 did not have floating point math, a fast microprocessor, or lots of storage
(3
things you need to do an FFT) so it is not practical except as a proof of what the 7854 was
capable of.

Dennis Tillman W7PF
The 7854 uses a TMS9900 processor, with a maximum clock speed of 3MHz. Takes me back to 1981 in my
first paid job, designing video switching networks driven by that processor. Massive ceramic package
that needed a great amount of force to push into a turned pin socket.

Because it was so sluggish, that processor design worked right off. I put all the bodge logic into a
PAL. The one gotcha is that Texas had different division for processors and memories. And each of
them labelled their buses the opposite way. Bastards. One has A0/D0 as LSB and the other had them as
MSB - so there was a massive bit scramble on the buses. While I was moaning about my not noticing it
and beating my chest, and just about to run at the circuit board with a knife and wirewrip, our
brain-on-a-stick software guy said "Don't do anything - I'll just do a bit scramble in software.
It'll be a security feature!". I could have kissed him; well not really, but you get the idea that I
was a relieved guy.

That software fella was of course a bit weird. You could never catch him doing any work. He was
either drinking coffee, was at the pub, or eating smelly food at his desk. But magically 100 lines
of bug-free and beautifully commented code appeared at the end of each day. Blow the compiled
finished code into a PROM and it would work right off. Damndest thing. I only once accused him that
his code didn't work. He asked what the symptom was, and immediately without noticeable pause and
looking over his pint of beer said "bit 4 of the third BCD switch from the left - the wire has
fallen off". He was of course absolutely right.

Craig


 

Unless a 7B87 timebase is used, the 7854 always operates in random
sampling mode. The only difference between it and a sampling
oscilloscope is the linear elements like amplifiers and attenuators
before the samplers. I say samplers because sampling oscilloscopes
and the 7854 sample both the vertical input and the horizontal sweep
unlike digital storage oscilloscopes like the 7D20 which sample only
the vertical signal and rely on a digital timebase. One later
oscilloscope, the 2252, works the same way as the 7854.

When a 7B87 timebase is used, then the 7854 operates like a digital
storage oscilloscope but its maximum sampling rate is pretty low at
like 500 ksamples/second.

On Tue, 12 Dec 2017 23:03:24 -0500, you wrote:

The 7D20, and the 7854 handle digitization in very different
ways. The 7D20 uses a CCD to store the analog waveform, and
then reads it out slowly into a slow ADC to digitize the waveform.
At sufficiently slow speeds, it directly digitizes the waveform
using the ADC... below 50KHz, I think.

The 7854 goes directly to the ADC for slow speeds, below 500KHz,
I think, but for higher frequencies, it behaves as a sampling
scope, taking samples in random positions after the trigger,
until the entire waveform has been sampled.

I use the 7D20 a lot more often for digital duty than I do the
7854.

-Chuck Harris


 

On Wed, 13 Dec 2017 07:37:27 -0000, you wrote:

The 7854 uses a TMS9900 processor, with a maximum clock speed of 3MHz. Takes me back to 1981 in my
first paid job, designing video switching networks driven by that processor. Massive ceramic package
that needed a great amount of force to push into a turned pin socket.

...

Craig
Many people may have run across the TMS9900 in the TI-99/4 and
TI-99/4A home computers from Texas Instruments. I was surprised to
find it in the 7854 but immediately recognized it.


Daniel Koller
 

Dennis,
  Is that FFT program available to us?
  Dan

On Wednesday, December 13, 2017, 1:41:34 AM EST, 'Dennis Tillman' @Dennis_Tillman_W7pF [TekScopes] <TekScopes@...> wrote:

 
Hi Harvey and Cliff
,
The 7D20 is spec'd at 70MHz and 40MSamples/Sec. It is also GPIB compatible
so it is programmable.

The 7854 is also GPIB compatible but in addition it has its own built in
vector processing programming language. That was another thing that made the
7854 revolutionary when it came out. It was the first scope with its own
vector processing language. Unfortunately electronics engineers were
programming illiterate in those days and that feature actually discouraged
sales so Tek quietly stopped promoting it in their ads and in their
literature.

To this day I have not found more than a handful of 7854 programs in spite
of doing an exhaustive search of all of the EE magazines in the 1980s and
1990s. However, all that changed a few months ago when Holger Lubbin (a
member of TekScopes and a dedicated fan of the 7854) wrote what must be the
largest 7854 program ever written. It calculates the FFT of the signal
stored in the 7854 memory. Unfortunately the 7854 did not have floating
point math, a fast microprocessor, or lots of storage (3 things you need to
do an FFT) so it is not practical except as a proof of what the 7854 was
capable of.

Dennis Tillman W7PF

-----Original Message-----
Sent: Tuesday, December 12, 2017 8:25 PM
Subject: Re: [TekScopes] 7854 vs 7D20

On Wed, 13 Dec 2017 02:54:49 +0000, you wrote:

I already have a working 7854 and a good assortment of 7000 series
plug-ins. What would a 7D20 add that the 7854 cannot do or can only do
poorly or with difficulty? Is the 7D20 compatible with the 7854?

To the best of my knowlege, the 7D20 was used to add digital storage rather
specifically to the 7600 scopes, giving you a large screen digital scope. I
think that the digital bandwidth is about 20 Mhz or so.

With the additional processing in the 7854, the higher bandwidth, I can't
think of anything in particular the 7D20 would give you.

if you're looking for relative oddities, then a 7A42 logic plugin (4 trace,
dual width, somewhat of a logic analyzer, might be an idea.

Harvey

Cliff Carrie
------------------------------------
Posted by: Harvey White <madyn@...>
------------------------------------


#yiv9191084253 #yiv9191084253 -- #yiv9191084253ygrp-mkp {border:1px solid #d8d8d8;font-family:Arial;margin:10px 0;padding:0 10px;}#yiv9191084253 #yiv9191084253ygrp-mkp hr {border:1px solid #d8d8d8;}#yiv9191084253 #yiv9191084253ygrp-mkp #yiv9191084253hd {color:#628c2a;font-size:85%;font-weight:700;line-height:122%;margin:10px 0;}#yiv9191084253 #yiv9191084253ygrp-mkp #yiv9191084253ads {margin-bottom:10px;}#yiv9191084253 #yiv9191084253ygrp-mkp .yiv9191084253ad {padding:0 0;}#yiv9191084253 #yiv9191084253ygrp-mkp .yiv9191084253ad p {margin:0;}#yiv9191084253 #yiv9191084253ygrp-mkp .yiv9191084253ad a {color:#0000ff;text-decoration:none;}#yiv9191084253 #yiv9191084253ygrp-sponsor #yiv9191084253ygrp-lc {font-family:Arial;}#yiv9191084253 #yiv9191084253ygrp-sponsor #yiv9191084253ygrp-lc #yiv9191084253hd {margin:10px 0px;font-weight:700;font-size:78%;line-height:122%;}#yiv9191084253 #yiv9191084253ygrp-sponsor #yiv9191084253ygrp-lc .yiv9191084253ad {margin-bottom:10px;padding:0 0;}#yiv9191084253 #yiv9191084253actions {font-family:Verdana;font-size:11px;padding:10px 0;}#yiv9191084253 #yiv9191084253activity {background-color:#e0ecee;float:left;font-family:Verdana;font-size:10px;padding:10px;}#yiv9191084253 #yiv9191084253activity span {font-weight:700;}#yiv9191084253 #yiv9191084253activity span:first-child {text-transform:uppercase;}#yiv9191084253 #yiv9191084253activity span a {color:#5085b6;text-decoration:none;}#yiv9191084253 #yiv9191084253activity span span {color:#ff7900;}#yiv9191084253 #yiv9191084253activity span .yiv9191084253underline {text-decoration:underline;}#yiv9191084253 .yiv9191084253attach {clear:both;display:table;font-family:Arial;font-size:12px;padding:10px 0;width:400px;}#yiv9191084253 .yiv9191084253attach div a {text-decoration:none;}#yiv9191084253 .yiv9191084253attach img {border:none;padding-right:5px;}#yiv9191084253 .yiv9191084253attach label {display:block;margin-bottom:5px;}#yiv9191084253 .yiv9191084253attach label a {text-decoration:none;}#yiv9191084253 blockquote {margin:0 0 0 4px;}#yiv9191084253 .yiv9191084253bold {font-family:Arial;font-size:13px;font-weight:700;}#yiv9191084253 .yiv9191084253bold a {text-decoration:none;}#yiv9191084253 dd.yiv9191084253last p a {font-family:Verdana;font-weight:700;}#yiv9191084253 dd.yiv9191084253last p span {margin-right:10px;font-family:Verdana;font-weight:700;}#yiv9191084253 dd.yiv9191084253last p span.yiv9191084253yshortcuts {margin-right:0;}#yiv9191084253 div.yiv9191084253attach-table div div a {text-decoration:none;}#yiv9191084253 div.yiv9191084253attach-table {width:400px;}#yiv9191084253 div.yiv9191084253file-title a, #yiv9191084253 div.yiv9191084253file-title a:active, #yiv9191084253 div.yiv9191084253file-title a:hover, #yiv9191084253 div.yiv9191084253file-title a:visited {text-decoration:none;}#yiv9191084253 div.yiv9191084253photo-title a, #yiv9191084253 div.yiv9191084253photo-title a:active, #yiv9191084253 div.yiv9191084253photo-title a:hover, #yiv9191084253 div.yiv9191084253photo-title a:visited {text-decoration:none;}#yiv9191084253 div#yiv9191084253ygrp-mlmsg #yiv9191084253ygrp-msg p a span.yiv9191084253yshortcuts {font-family:Verdana;font-size:10px;font-weight:normal;}#yiv9191084253 .yiv9191084253green {color:#628c2a;}#yiv9191084253 .yiv9191084253MsoNormal {margin:0 0 0 0;}#yiv9191084253 o {font-size:0;}#yiv9191084253 #yiv9191084253photos div {float:left;width:72px;}#yiv9191084253 #yiv9191084253photos div div {border:1px solid #666666;min-height:62px;overflow:hidden;width:62px;}#yiv9191084253 #yiv9191084253photos div label {color:#666666;font-size:10px;overflow:hidden;text-align:center;white-space:nowrap;width:64px;}#yiv9191084253 #yiv9191084253reco-category {font-size:77%;}#yiv9191084253 #yiv9191084253reco-desc {font-size:77%;}#yiv9191084253 .yiv9191084253replbq {margin:4px;}#yiv9191084253 #yiv9191084253ygrp-actbar div a:first-child {margin-right:2px;padding-right:5px;}#yiv9191084253 #yiv9191084253ygrp-mlmsg {font-size:13px;font-family:Arial, helvetica, clean, sans-serif;}#yiv9191084253 #yiv9191084253ygrp-mlmsg table {font-size:inherit;font:100%;}#yiv9191084253 #yiv9191084253ygrp-mlmsg select, #yiv9191084253 input, #yiv9191084253 textarea {font:99% Arial, Helvetica, clean, sans-serif;}#yiv9191084253 #yiv9191084253ygrp-mlmsg pre, #yiv9191084253 code {font:115% monospace;}#yiv9191084253 #yiv9191084253ygrp-mlmsg * {line-height:1.22em;}#yiv9191084253 #yiv9191084253ygrp-mlmsg #yiv9191084253logo {padding-bottom:10px;}#yiv9191084253 #yiv9191084253ygrp-msg p a {font-family:Verdana;}#yiv9191084253 #yiv9191084253ygrp-msg p#yiv9191084253attach-count span {color:#1E66AE;font-weight:700;}#yiv9191084253 #yiv9191084253ygrp-reco #yiv9191084253reco-head {color:#ff7900;font-weight:700;}#yiv9191084253 #yiv9191084253ygrp-reco {margin-bottom:20px;padding:0px;}#yiv9191084253 #yiv9191084253ygrp-sponsor #yiv9191084253ov li a {font-size:130%;text-decoration:none;}#yiv9191084253 #yiv9191084253ygrp-sponsor #yiv9191084253ov li {font-size:77%;list-style-type:square;padding:6px 0;}#yiv9191084253 #yiv9191084253ygrp-sponsor #yiv9191084253ov ul {margin:0;padding:0 0 0 8px;}#yiv9191084253 #yiv9191084253ygrp-text {font-family:Georgia;}#yiv9191084253 #yiv9191084253ygrp-text p {margin:0 0 1em 0;}#yiv9191084253 #yiv9191084253ygrp-text tt {font-size:120%;}#yiv9191084253 #yiv9191084253ygrp-vital ul li:last-child {border-right:none !important;}#yiv9191084253

[Non-text portions of this message have been removed]


Harvey White
 

On Wed, 13 Dec 2017 07:37:27 -0000, you wrote:

Unfortunately the 7854 did not have floating point math, a fast microprocessor, or lots of storage
(3
things you need to do an FFT) so it is not practical except as a proof of what the 7854 was
capable of.

Dennis Tillman W7PF
The 7854 uses a TMS9900 processor, with a maximum clock speed of 3MHz. Takes me back to 1981 in my
first paid job, designing video switching networks driven by that processor. Massive ceramic package
that needed a great amount of force to push into a turned pin socket.

Because it was so sluggish, that processor design worked right off. I put all the bodge logic into a
PAL. The one gotcha is that Texas had different division for processors and memories. And each of
them labelled their buses the opposite way. Bastards. One has A0/D0 as LSB and the other had them as
MSB - so there was a massive bit scramble on the buses. While I was moaning about my not noticing it
and beating my chest, and just about to run at the circuit board with a knife and wirewrip, our
brain-on-a-stick software guy said "Don't do anything - I'll just do a bit scramble in software.
It'll be a security feature!". I could have kissed him; well not really, but you get the idea that I
was a relieved guy.
I ran into the same thing for the TMS9914 in a very early 488 bus
control design. The hardware engineer "guessed" wrong. I was the one
doing the software, so it was easy enough to redefine the commands and
flip the data in software.




That software fella was of course a bit weird. You could never catch him doing any work. He was
either drinking coffee, was at the pub, or eating smelly food at his desk. But magically 100 lines
of bug-free and beautifully commented code appeared at the end of each day. Blow the compiled
finished code into a PROM and it would work right off. Damndest thing. I only once accused him that
his code didn't work. He asked what the symptom was, and immediately without noticeable pause and
looking over his pint of beer said "bit 4 of the third BCD switch from the left - the wire has
fallen off". He was of course absolutely right.
There are people like that. Not sure how he managed it.

Harvey



Craig


 

The primary operating difference is that the 7D20 operates as a real
time digital storage oscilloscope where the displayed waveform is
continuously updated while the 7854 literally stores the waveform for
later viewing and measurement. The 2252 works this way also.

Usually this means operating the 7854 in analog mode and then pressing
one of the waveform aquisition buttons to capture and store a waveform
for further display and processing.

On Wed, 13 Dec 2017 02:54:49 +0000, you wrote:

I already have a working 7854 and a good assortment of 7000 series plug-ins. What would a 7D20 add that the 7854 cannot do or can only do poorly or with difficulty? Is the 7D20 compatible with the 7854?

Cliff Carrie


Holger Lübben
 

Hi Dan!

Yes, the FFT program is available. You'll find a link to a zip archive at the end of this message.
But the program needs some explanation.

The project was only a proof of concept and the program is a very limited prototype.
Some of the limits are:

- There is no error handling
- It's only tested on a 7854 with 8k of ram.
- Some sort of waveform scaling is missing, at the moment it works only with 1V/DIV
- It's very slow. Calculationg the FFT of a waveform with 128 points takes about 12 minutes.

The last point is the reason why I stopped the project. The 7854 is a fantastic scope - but the programming language has too many limits for this kind of calculation.


The zipfile contains 4 files:

1) fftprog.txt
the fft program for the 7854

2) inputwaveform_128_points.txt
My input waveform for testing

3) input.jpg
A screenshot of the input waveform

4) result.jpg
A screenshot of the correct result

And here is the link:

http://www.methodyn.de/fftprog.zip

Holger Lübben


redarlington
 

Hah, love it. You got me thinking that because of the speed, I could just
do a READX over gpib and pull the data in, process, and SENDX it back out
for display. Or, you know, display on the PC or use the specan 3 feet
away... Either way, this is very cool and much appreciated work. It also
might be neat to add a Raspberry Pi or some similar single board computer
externally to the 7854 to handle doing this without much additional bulk.
Hit a button and get the FFT.

-Bob

On Thu, Dec 14, 2017 at 1:12 AM, tekscopesinput@...
[TekScopes] <TekScopes@...> wrote:



Hi Dan!

Yes, the FFT program is available. You'll find a link to a zip archive at
the end of this message.
But the program needs some explanation.

The project was only a proof of concept and the program is a very limited
prototype.
<snip>