Author Topic: Which oscilloscope to look for? 300-500MHz BW  (Read 4203 times)

0 Members and 1 Guest are viewing this topic.

Offline dzsekiTopic starter

  • Frequent Contributor
  • **
  • Posts: 508
  • Country: hu
Which oscilloscope to look for? 300-500MHz BW
« on: January 28, 2019, 12:42:46 pm »
Here we are again, yet another "which oscilloscope" thread, but as you might guessed my requirements are different than that of others'...  :palm:

So, I have a strange obsession with high bandwidth analog video signals (like VGA) with pixel clock beyond standard 1080P. An oscilloscope with 300+ MHz bandwidth is desired for these applications.

For several years now my trusty companion is a HP 1720A oscilloscope, this is the base model of HP's fastest analog oscilloscope range, having 275MHz bandwidth (on paper at least), at that time this was the only scope for the purpose that I could afford. Over the years I gathered several 250MHz passive probes for it, and also got a HP 1120A 500MHz FET probe with all the dividers (which I intend to keep with any future oscilloscopes) .

When doing measurements I either use direct probing ie. having the oscilloscope as the load for the circuit of interest, or for in circuit measurement I mostly use the FET probe. If I am not measuring the video stuff, then I am doing some audio trouble shooting and other not too demanding hobby projects.

The thing is this oscilloscope was manufactured in 1977, so turning 42 this year (10 years older than me). While I don't have particular plan to ditch it, but I also can't expect to last for ever. Furthermore, as it seems it does not live up to its specs as far as bandwidth goes, since I hardly could measure rise times below 1.5ns while my pulse generator's rise time is confirmed to be 425ps with a 2.5GHz LeCroy scope. Nevertheless I can use this scope for the purpose still. The best thing I like about it is that it is completely silent, as there is no fan inside. This is important, because my workbench is located in our living room (small apartment)

The idea is not new for me to look for a new oscilloscope, in fact I've had some, but none of them stayed for too long, here is the list:

-Yokogawa DL1740: 4CH 500MHz scope with 500MS/s sampling on all channels or 1GS/s on two or less channels (20GS/s in ETS), 1MS memory. I got it as faulty (bad input preamp chip), but managed to fix it, although a few days later an other input preamp chip blew up for no reason, got it fixed, but this raised the red flag to not hold on to this scope for too long, Sold it locally with 3 more spare input chips -just in case... Functionally it was a nice scope, but had an odd cubature not fitting well in to my bench, also had a loud fan, and the bad reliabilty...

-HP 54503A: 4CH 500MHz scope, with 20MS/s sampling. This was advertised as a dead unit, but was so cheap I could not resist, and bought it for curiosity. Eventually got this fixed as well, and my general impression on this unit was generally very positive, but the low capture rate (especially with vectors or measurements on) made me sell this unit as well, oh and this one also had a loud fan.

-Currently I am playing with a GaGe CompuScope 82G scope card, what got me in this is the 2GS/s sample rate (one channel), and reasonable memory depth at 2MS. This is a nice one, and is a great supplement for the HP 1720A (some things being better, some things worse). But now I also have this for sale, although I won't be sad if there will be no taker for it, although as I've found out I'd prefer an all in one solution instead.

I was thinking about getting into the Tek 2465 line or a 485, but if I am honest these aren't a real step towards the future.

So is there any scope you can recommend that has:
- 2-4 Channels (no logic functionality required)
- Is "digital" with reasonable capture rate and possibility to extract data, etc.
- Having 300-500Mhz bandwith on repetitive signals, and single shot bandwidth of >150MHz
- reasonably quiet operation
- Can be aquired for cheap on used market (<500€)
HP 1720A scope with HP 1120A probe, EMG 12563 pulse generator, EMG 1257 function generator, EMG 1172B signal generator, MEV TR-1660C bench multimeter
 

Online tggzzz

  • Super Contributor
  • ***
  • Posts: 19280
  • Country: gb
  • Numbers, not adjectives
    • Having fun doing more, with less
Re: Which oscilloscope to look for? 300-500MHz BW
« Reply #1 on: January 28, 2019, 12:58:49 pm »
Furthermore, as it seems it does not live up to its specs as far as bandwidth goes, since I hardly could measure rise times below 1.5ns while my pulse generator's rise time is confirmed to be 425ps with a 2.5GHz LeCroy scope. Nevertheless I can use this scope for the purpose still. The best thing I like about it is that it is completely silent, as there is no fan inside. This is important, because my workbench is located in our living room (small apartment)

At these frequencies the details of the probing become important.

Don't forget that the risetimes of the probe and scope rise as sqrt(t12+t22), and that a "high impedance" *10 probe's input impedance at 200MHz is ~50ohms - which can significantly load/slug the source.

My preference is for a "resistive divider Z0" probe, since a *10 probe's input impedance is 500ohms. You can homebrew these, unlike conventional probes.

Of course you need a scope with a 50ohm input, preferably not a scope with a 1Mohm//20pF input with an added 50ohm resistor in parallel. The Tek 485 has two input attenuators, one 1Mohm//20pf, one 50ohm.
There are lies, damned lies, statistics - and ADC/DAC specs.
Glider pilot's aphorism: "there is no substitute for span". Retort: "There is a substitute: skill+imagination. But you can buy span".
Having fun doing more, with less
 


Offline dzsekiTopic starter

  • Frequent Contributor
  • **
  • Posts: 508
  • Country: hu
Re: Which oscilloscope to look for? 300-500MHz BW
« Reply #3 on: January 28, 2019, 01:19:07 pm »
At these frequencies the details of the probing become important.

Don't forget that the risetimes of the probe and scope rise as sqrt(t12+t22), and that a "high impedance" *10 probe's input impedance at 200MHz is ~50ohms - which can significantly load/slug the source.

My preference is for a "resistive divider Z0" probe, since a *10 probe's input impedance is 500ohms. You can homebrew these, unlike conventional probes.

Of course you need a scope with a 50ohm input, preferably not a scope with a 1Mohm//20pF input with an added 50ohm resistor in parallel. The Tek 485 has two input attenuators, one 1Mohm//20pf, one 50ohm.

Yes, I know about that, that’s why I pointed out that if bandwidth is concern then I mostly reach for my FET probe.
The HP 1720A also has a switchable genuine 50 Ohm input, however it employs a unified attenuator, the input capacitance is 11pF though.

As for the rise time I am also aware of the propagation of the rise times, that’s why there is a problem :) Knowing the source has 425ps rise time, but my scope showing 1.5ns rise time, doing the math backwards yields ~1.44ns rise time  for the scope, whereas it should be <1.3ns
HP 1720A scope with HP 1120A probe, EMG 12563 pulse generator, EMG 1257 function generator, EMG 1172B signal generator, MEV TR-1660C bench multimeter
 

Offline dzsekiTopic starter

  • Frequent Contributor
  • **
  • Posts: 508
  • Country: hu
HP 1720A scope with HP 1120A probe, EMG 12563 pulse generator, EMG 1257 function generator, EMG 1172B signal generator, MEV TR-1660C bench multimeter
 

Offline nctnico

  • Super Contributor
  • ***
  • Posts: 26751
  • Country: nl
    • NCT Developments
Re: Which oscilloscope to look for? 300-500MHz BW
« Reply #5 on: January 28, 2019, 10:16:26 pm »
The Tektronix TDS500 and TDS700 series are choices which fall inside your budget but expect these scopes to have problems (leaking capacitors, bad NVRAMs and bad displays). Also the NuColor color displays aren't for everyone. I get an instant headache from them. If you don't mind a fixer upper then an HP / Agilent 54815A or hp 54825A may be an option. AFAIK the first can be hacked into the second to double the samplerate. The big advantage of the HP scopes is that you don't need any external equipment (besides a BNC cable) to calibrate them and the platform is PC based so parts are easier to find.
« Last Edit: January 29, 2019, 12:01:30 am by nctnico »
There are small lies, big lies and then there is what is on the screen of your oscilloscope.
 

Offline tautech

  • Super Contributor
  • ***
  • Posts: 28136
  • Country: nz
  • Taupaki Technologies Ltd. Siglent Distributor NZ.
    • Taupaki Technologies Ltd.
Re: Which oscilloscope to look for? 300-500MHz BW
« Reply #6 on: January 28, 2019, 10:38:01 pm »
So is there any scope you can recommend that has:
- 2-4 Channels (no logic functionality required)
- Is "digital" with reasonable capture rate and possibility to extract data, etc.
- Having 300-500Mhz bandwith on repetitive signals, and single shot bandwidth of >150MHz
- reasonably quiet operation
- Can be aquired for cheap on used market (<500€)
For just a bit more you can have the latest and new with 3 yr warranty:
https://www.siglenteu.com/digital-oscilloscopes/sds2000x-e/
Avid Rabid Hobbyist
Siglent Youtube channel: https://www.youtube.com/@SiglentVideo/videos
 

Offline nctnico

  • Super Contributor
  • ***
  • Posts: 26751
  • Country: nl
    • NCT Developments
Re: Which oscilloscope to look for? 300-500MHz BW
« Reply #7 on: January 29, 2019, 12:07:12 am »
730 euro (ex VAT) isn't a bit more. It is 50% over budget! And the older high-end oscilloscopes usually have features geared towards doing accurate high frequency measurements. I have an 1GHz R&S RTM3004 on my bench but if I need to do really accurate time measurements which require a very stable trigger (ps level) then I pull out my HP 54845A.
There are small lies, big lies and then there is what is on the screen of your oscilloscope.
 
The following users thanked this post: dzseki

Offline dzsekiTopic starter

  • Frequent Contributor
  • **
  • Posts: 508
  • Country: hu
Re: Which oscilloscope to look for? 300-500MHz BW
« Reply #8 on: February 11, 2019, 09:19:59 am »
As I was thinking about it I realized that I've never bought any oscilloscope in a "known good" condition, only as-is units those I got fixed up sooner or later, or they worked right away. I think I'll stick to this habit further on, that would ease somewhat on my price constraits...
I did a bit "homework" and thought I'd share what I found. For now I focused on vintage HP/Agilent scope models as these are what I am most familiar with.

54522/42A-C: 2-4ch, 500MHz, 2GS/s on any channel, 32kpts memory, green CRT or color TFT.
pros: can be found reasonably cheap, FFT up to the full memory size
cons: difficult data retrieval (floppy, GPIB), relatively small memory, no color/intensity grading on signals

54615B/16B/16C: 2ch, 500MHz, 1-2GS/s, 5kpts memory, green CRT or color TFT
pros: maybe found at a reasonable price, relatively compact size
cons: difficult data retrieval (GPIB or RS232?), even smaller memory, FFT only on 1024 points

54810A/15A/20A/25A (infiniium): 2-4ch, 500MHz, 1-2GS/s, 32kpts memory, Windows based, color TFT
pros: color grading available, FFT up to memory size, LAN connectivity
cons: most expensive, relaitve small memory, biggest, quirks of the Windows PC background

54641A/42A (MegaZoom): 2ch, 350-500MHz, 2GS/s, 8Mpts memory, green CRT
pros: intensity grading, relatively compact size, big memory
cons: difficult data retrieval, FFT only on 2048 points

There is no clear winner, but based only on specs (not considering (knowing) any particular reliability issues etc.) I feel the 5464x is the most suitable followed by the 54522/42 leaving the entry level infiniiums to the end.
HP 1720A scope with HP 1120A probe, EMG 12563 pulse generator, EMG 1257 function generator, EMG 1172B signal generator, MEV TR-1660C bench multimeter
 

Offline Rax

  • Frequent Contributor
  • **
  • Posts: 878
  • Country: us
Re: Which oscilloscope to look for? 300-500MHz BW
« Reply #9 on: April 25, 2019, 03:07:46 am »
I have the 54522A and I love it. One of the last US-made units from HP (also, from right before Agilent re-branding), and it shows it. Solid, lab instrumentation grade materials. Nothing cheap on it, inside or outside. The 5464x series looks like today's cheaply made, plasticky scopes. Probably made in Malaysia, right after moving manufacturing our of the US.

Having said that, specs are pretty good on the 5464xs. Memory is far better than the one I have - how important is that, though? I'm sure it would absolve itself well of tasks.

For me, from my research and advice from valued friends, I am sure I made the right decision. The HP 54522A seems just about the last scope from HP's glory period, yet having modern specifications, and not costing an arm and a leg. I just had mine with calibration for less than 500 USD. It was about 14,000 USD in 1996, or just about 23,000 USD today's value. That's a hefty chunk of cash, and telling of its quality.
 
The following users thanked this post: dzseki


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf