Topics

Did Tek produce a "normal" d.c. to 50 GHz scope that one could easily poke around with ? { vs. the 11801 / 11802/ 11803, CSA 803 series }

garp66
 
Edited

Did Tektronix make a 50 GHz GHz scope ??

I was looking for an older Tek GHz scope (hopefully inexpensive) to aid in design & testing of GHz pre-amps & LO's for amateur radio-astronomy. Some designs are up as high as 50 GHz, and a few down at 1.42 & 10 GHz.

I had thought that I might be able to use, (and did acquire, to repair) the 11801a & some SD-24 & SD-3x modules.
But, I have no experience with the 11801a, *yet* ...

{ Are there any GHz measurement 1180x Tek scope mentors on this forum, please ? }.

-- However, the 11801a scope *seems* too delicate { ESD damage prone, restricted signals & odd external triggering requirements } for my general poking around at these frequencies with GHz Rx & LO test set-ups ... { ? }

I do have an HP 70K / HP 70000 MMS SA + signal generator system (very happy with it) that can take me up to 40GHz, but am looking for a reasonably robust, older, inexpensive "normal" 40 to 60 GHz scope to compliment measurements in design/testing.

thank you,

Leo Bodnar
 

Yes, they do - https://www.tek.com/oscilloscope/dpo70000sx

CSA803/1180x are not "normal" scopes - they are sampling scopes and have a lot of limitations.
They only work on single volt, repetitive, low impedance signals with externally supplied trigger. Their user interface was considered mind-blowing for 1980's but is just a semi-random set of eclectic gimmicky counter-intuitive features.
Their vertical resolution is 8 bit and sampling jitter a few ps - I can't see it being useful for any good quality analogue work apart from checking that the signal "is there."
But if you know what you are getting into you should be able to eBuy CSA803 or 1180x and SD-26 20GHz head for around $500 if you don't rush.
This will get you started.
Don't get SD-30 or SD-32 from the outset.
Ironically, the limitations of CSA803/1180x are playing in your favour because you can often find an instrument that has almost never been turned on since 1980's. One of mine looks like it just came out of the factory.
Leo

cmjones01
 

On Mon, Feb 12, 2018 at 09:46 am, <@garp66> wrote:
Did Tektronix make a 50 GHz GHz scope ??
Yes, but not one that any of us are likely to be able to afford!

I was looking for an older Tek GHz scope (hopefully inexpensive) to aid in
design & testing of GHz pre-amps & LO's for amateur radio-astronomy. Some
designs are up as high as 50 GHz, and a few down at 1.42 & 10 GHz.

I had thought that I might be able to use, (and did acquire, to repair) the
11801a & some SD-24 & SD-3x modules.
But, I have no experience with the 11801a, *yet* ...
Any older multi-GHz scope is going to be a rather specialized beast, and the 11800 series (and CSA803) are no exception. Until recently, the technology for digitizing multi-GHz signals in real time barely existed, so sampling techniques, which only work on repetitive waveforms, were the only way to get such performance. As Leo has said, the 11800 series are sampling scopes. They need a repetitive waveform and a trigger signal about 35ns before the event you want to see.

I have a CSA803A on the bench and use it quite a lot. It's a quirky beast, and I'm constantly terrified of electrostatic damage to the sampling heads, but it gets useful results for me at a price I can afford. It allows me to quite happily observe multi-GHz oscillator outputs, but I usually use it for looking at fast edges in switching circuits. The SD-24 TDR head is really handy for checking cables, PCB traces and so on. Oh yes, and finding shorts in cheap SMA cables, but that's another story.

Probing high-speed (>100MHz) signals is always going to be difficult. Unless you have a really exotic active probe (and those cost serious money), it's never going to be as easy as picking up a normal low-frequency scope probe. However, there are ways to work around this. My favourite is to solder a small surface-mount resistor (about 3300 ohms often works well for me) on to the point I want to measure, then connect a short tail of subminiature 50 ohm coax to the other end of this and a nearby ground point. Bits of subminiature coax terminated in SMA connectors are cheaply available from eBay sellers in China. Though this attenuates the signal a lot, using the averaging functions on the CSA803A gets me a clean, consistent measurement.

One thing that's really useful with a sampling scope if you don't have a suitable external trigger signal is a good delay line. I've built the delay lines from a scrap 11802 into my CSA803A, and they have a trigger pickoff output. Their risetime is about 70ps, bandwidth thus about 5GHz. I think Tek made the same thing in its own box and called it the DL-1. With the delay line set up and a handful of SMA cables and surface-mount resistors, I get to reliably probe signals up to 5GHz on a small-business budget, and I'm happy about that.

Chris

Leo Bodnar
 

Chris,
You are probably thinking of DL-11? http://w140.com/tekwiki/wiki/DL-11
Another delay line trick I have used is inserting it into the circuit under test before the stage/driver you are working on but tapping the trigger off before the DL. This way observed signal is delayed but the driver remains undisturbed (assuming it can tolerate degraded risetime at the input.) This will break most systems but is useful for characterising the driver/interconnects performance.
Leo

On Tue, Feb 13, 2018 at 07:00 am, cmjones01 wrote:

I've built the delay lines from a scrap 11802 into my CSA803A, and they have a trigger pickoff
output. Their risetime is about 70ps, bandwidth thus about 5GHz. I think Tek
made the same thing in its own box and called it the DL-1.

cmjones01
 

On Tue, Feb 13, 2018 at 10:50 am, Leo Bodnar wrote:
You are probably thinking of DL-11? http://w140.com/tekwiki/wiki/DL-11
You're quite right, that's the one. Thank you for the correction.

Another delay line trick I have used is inserting it into the circuit under
test before the stage/driver you are working on but tapping the trigger off
before the DL. This way observed signal is delayed but the driver remains
undisturbed (assuming it can tolerate degraded risetime at the input.) This
will break most systems but is useful for characterising the
driver/interconnects performance.
A neat trick. Thank you for the hint - I'll remember that one!

Chris

 

On Mon, 12 Feb 2018 09:46:01 -0800, you wrote:

However, the 11801a scope *seems* too delicate { ESD damage prone & odd external triggering requirements } for my general poking around at these frequencies with my GHz Rx & LO test set-ups ... { ? }
By their nature, sampling oscilloscope heads are delicate and prone to
ESD damage. This goes double for feedthrough sampling heads which
lack a built in 50 ohm termination.

I do have an HP 70K / HP 70000 MMS SA + signal generator that can take me up to 40GHz, but am looking for a reasonably robust, older, inexpensive "normal" 40 to 60 GHz scope to compliment these frequencies in design/testing amateur RA work.
Unless you are working with untuned broadband amplifiers, I think a
sampling oscilloscope is the wrong tool and a spectrum analyzer plus
signal generator will be much more useful.

HP made some sampling based network analyzers which also operate as a
sampling oscilloscope which might be more of what you are looking for.
I show the 8753 and 8720 in my notes.