scope jitter

No big deal, got the same results using a bessel with Nyquist bw on the trigger channel- ages ago.

You can get a perfect reconstruction of the signal within the bandpass which may not be the same as the signal on your board. Believing the scope signal is the board signal is always a stretch no matter what type of scope you are using. In that sense, a digital scope can lull you into a false sense of security or accuracy. The analog scopes typically don't have means to blow up the trace so much.

Do your volume knobs go to eleven too?

Actually, I expect the trigger circuit is not an ADC. Why does the trigger need to be a sampled signal just because the other inputs are sampled? Actually it is sampled (which it must be at some point to be used in the digital domain), likely *after* the comparator where it is a single binary bit.

--

Rick

Were those scan-conversion scopes any good? I saw a photo of one recently and the trace seemed to be disappointingly thick.

-- Adam

I think I used an old Iwatsu in Japan a couple years ago. It used a scan converter tube, and then digitally processed the traces so they came out in color, had digital cursors, etc. I thought it worked QUITE well.

I tinkered with a Tek scan converter rack mount scope that used a video monitor for display. While it worked, and by tinkering with the video adjustments, you could get a fantastic writing rate, it wasn't what I really needed.

Then, there was the Tek scope that played a fan beam over a bunch of electrodes, so the target of the tube was actually the ADC! That supposedly worked well, but the fast memory was rather pricey.

Jon

The 7 GHz Tek scan conversion scope was actually made in France. It's still available:

Bernard must have a shed full of old scan conversion tubes.

Greenfield Technology - name, logo, most of the product line - is literally a copy of Highland Technology. Bernard is a nice guy so we don't mind.

Digitizer-based scopes are now available up to 100 GHz, so scan conversion is obsolete.

--

John Larkin         Highland Technology, Inc 
picosecond timing   precision measurement  

jlarkin att highlandtechnology dott com 
http://www.highlandtechnology.com

Here's a DSO5054A for realtime

and random sampling.

The measured jitter seems much worse than your 6054, which is interesting.

Using the external trigger didn't seem to introduce any discernible jitter (judged by looking at the trace smear in a persistence mode).

I bought a bunch of them in the late '70s and early '80s at IBM. We had two or four of them connected to PDP-11s for the processing (it was a trick buying them ;-), along with a 7704 with a sampler inserted between the plugins and display. I had 10 or 20 GPIB instruments connected to the PDP11, as well. Neat systems (Tek called them "Signal Processing Systems", out of their test systems group).

The early Bell Telephone PCM systems digitized voice signals by deflecting a fan beam onto a set of gray-code anodes, shaped to do the code compression.

--

John Larkin         Highland Technology, Inc 
picosecond timing   precision measurement  

jlarkin att highlandtechnology dott com 
http://www.highlandtechnology.com

A friend of mine from grad school went to work at IBM Almaden in about

1987. He was already a crusty old VAX bigot, and couldn't adjust to VM/CMS. (He said that it was just too much for him to work with a system that had a virtual card punch.)

At the time, Tektronix had an ATE product line that was built around a MicroVAX. John persuaded them to generate a Tek part number for just the MicroVAX, which enabled him to sneak it past the IBM purchasing department.

He didn't last long at Almaden, but went on to great things at SLAC--he invented a beam cooling scheme that every accelerator in the world now uses.

Cheers

Phil Hobbs

--
Dr Philip C D Hobbs 
Principal Consultant 
ElectroOptical Innovations LLC 
Optics, Electro-optics, Photonics, Analog Electronics 

160 North State Road #203 
Briarcliff Manor NY 10510 

hobbs at electrooptical dot net 
http://electrooptical.net

I designed the MIK11, which was an LSI-11 board hidden inside a CAMAC crate controller. People could buy it and end-around organizational rules that didn't allow them to buy computers.

It mostly ran RT-11, a DEC os that sort of originated the command-line set of CP/M and MS-DOS. RT-11 supported only contiguous disk files, which was interesting.

--

John Larkin         Highland Technology, Inc 

lunatic fringe electronics

It was too small to run VMS, I gather? (I quite liked VMS back in the day.)

Cheers

Phil Hobbs

--
Dr Philip C D Hobbs 
Principal Consultant 
ElectroOptical Innovations LLC 
Optics, Electro-optics, Photonics, Analog Electronics 

160 North State Road #203 
Briarcliff Manor NY 10510 

hobbs at electrooptical dot net 
http://electrooptical.net

VMS was a VAX os. PDP-11 had DOS-11, RT-11 (kinda like CP/M), several RSX-11 "realtime" variants, the wonderful RSTS/11 timeshare system, and of course the first UNIX. The c language sure looks like it was designed as a high-level PDP-11 assembler, which is in some ways unfortunate.

--

John Larkin         Highland Technology, Inc 

lunatic fringe electronics

Kinetic Systems? I have one of those on the shelf behind me. Can't bear to throw it out.

Jon

PDP-11's were 16-bit computers. VMS ran on the VAX, a 32-bit virtual memory system, successor to the PDP-11. The LSI-11 was a very stripped-down PDP-11, implemented on 4 or 5 LSI chips made by Western Digital (not a subsidiary of DEC, despite the name). The LSI-11 came out in 1975.

Jon

No, Kinetics was our arch rival. I worked for Standard Engineering in Fremont, which later became DSP Technology and moved to Detroit to do auto test gear.

Ours was prettier than theirs.

Fremont was hot and boring so I quit after one year.

--

John Larkin         Highland Technology, Inc 
picosecond timing   precision measurement  

jlarkin att highlandtechnology dott com 
http://www.highlandtechnology.com

The rules forbidding the purchase of "Data Processing Equipment" were staggering but buying DEC just wasn't done. ...so I bought it from Tek and made sure the contract was worded such that it had no DEC labeling on the outside and was done up in Tek blue instead of Dec pink.

The systems we had ran Tek SPS BASIC. Nice signal processing commands built on top of BASIC. Later I pulled the same trick, buying VAX-11/780s to support other test systems. They ran VMS, of course.

The HP9845 was one of the most capable computers of its class, being a large desktop computer with full keyboard, graphics monitor and good i/o capability.

HP officially defined it as a calculator.

I asked the reason, and it was explicitly to get around organisations' dysfunctional purchasing rules, which typically required board-level signoff for computers.

Heh that still exists today, though now it's usually the IT department that insists on specifying and purchasing all "computer" hardware and installing their "asset tracking" spyware on any "computers", and being in charge of what software goes on them. For automating lab setups, fortunately a raspberry pi can be hidden inside another piece of equipment, and can usually be bought as an "electronic component" without interference.

They just have a job to do. You should let them install their spyware on it. ;-)

When XP went out of service, our IT morons edict ed that all XP systems be replaced with W7 systems. Fine, except that I have a piece of equipment that requires a PCMCIA card interface and it would cost #40K to replace it. Their solution was to give me an old system and kick it off the intranet. Sneaker net is certainly no fun but having two computers on my desk is worse.

Our IT department is incompetent but from what I hear, they're one of the better ones around. Sad.