Author Topic: DS1052E hacked to DS1102E waveform comparision  (Read 15153 times)

0 Members and 1 Guest are viewing this topic.

Offline Zyvek

  • Regular Contributor
  • *
  • Posts: 108
  • Country: us
DS1052E hacked to DS1102E waveform comparision
« on: November 24, 2010, 06:00:39 pm »
Since I'm just a basic hobbyist, this might be of more interest to me than most, but I wanted to compare my upgraded DS1052E to the other scopes in my collection. 

I just wired up a 50MHZ oscillator to compare the waveforms.  I did use the same probe (Tek P6138) on each scope, calibrated/compensated for each scope as well. The P6138 fit within the compensation range of each scope (~15pf), except the 2236.

Scope bandwidth/sample rate for each is:

Rigol 1052E/Hacked 100Mhz / 1Gs
Tek 2236 100Mhz / Analog
Tek 2340A 150Mhz / 100Ms
Tek 2445 150Mhz / Analog
Tek 2467B 400Mhz / Analog

Test oscillator was powered by an HP 3618A non-switching power supply.

I was surprised that the 2236 didn't do as well as the Rigol, given the equal bandwidth.

I was kinda hoping that the Rigol would be closer to the 2430A but of course that's not a logical thought give the 2430A's extra 50Mhz bandwidth.  I'm new to DSOs so somehow I figured the higher sampling rate of the Rigol would help.  Of course the 2430A has no easy way to save the waveform (without a GPIB equiped PC & SW) and the memory is very limited in comparison.  The 2430A seems to handle two traces without any impact vs the Rigol.  So for some finding a used 2430A might be a better deal than the Rigol (I paid ~$150 for mine)

Of course the 2445 and 2467B did better, as expected.  But I was still impressed with the Rigol overall. 

One thing that made the difference was using the P6138 probe on the Rigol vs the probes included with the 1052E.  I posted those pics in the upgrade thread here: https://www.eevblog.com/forum/index.php?topic=553.msg25021#msg25021

Apologies in advance, as I know these probably aren't very useful tests...but good nerd fun for me.

BTW, where in the Rigol setup does one set the time/date so the files saved to USB have a correct timestamp?  I couldn't find it in the manual.
« Last Edit: November 24, 2010, 06:16:34 pm by Zyvek »
-Z
 

Offline Mechatrommer

  • Super Contributor
  • ***
  • Posts: 8765
  • Country: my
  • reassessing directives...
Re: DS1052E hacked to DS1102E waveform comparision
« Reply #1 on: November 24, 2010, 10:06:05 pm »
interesting. can you post picture of the rigol probe vs the teks probe P6138? is there any possibility finding/generating different shaped 50MHz wave/source to compare? thanx for letting us know.

i'm interested in looking the difference between:
rigol + ori probe
24302236 + P6138
rigol + P6138
2467 + P6138
from the pics you posted here and in the other thread, i'm expecting the performance of 24302236 + P6138 will be in the middle between rigol + ori probe and rigol + P6138.
« Last Edit: November 26, 2010, 12:04:57 am by shafri »
if something can select, how cant it be intelligent? if something is intelligent, how cant it exist?
 

Offline jahonen

  • Super Contributor
  • ***
  • Posts: 1045
  • Country: fi
Re: DS1052E hacked to DS1102E waveform comparision
« Reply #2 on: November 24, 2010, 10:21:01 pm »
Why did you use the long ground clip for such a measurement? The long ground clip practically destroys the true performance as can be seen from my demonstration.

Probe bandwidth is usually measured with clip removed and using BNC adapter which grounds the sleeve very near the tip.

Regards,
Janne
 

Offline Mechatrommer

  • Super Contributor
  • ***
  • Posts: 8765
  • Country: my
  • reassessing directives...
Re: DS1052E hacked to DS1102E waveform comparision
« Reply #3 on: November 24, 2010, 10:22:51 pm »
P6138 in Ebay... $100 for a pair? gulp!... i'll be waiting for a while.
thanx janne, for the info as well. nice hack.
« Last Edit: November 24, 2010, 10:29:51 pm by shafri »
if something can select, how cant it be intelligent? if something is intelligent, how cant it exist?
 

Offline Zyvek

  • Regular Contributor
  • *
  • Posts: 108
  • Country: us
Re: DS1052E hacked to DS1102E waveform comparision
« Reply #4 on: November 25, 2010, 06:59:29 am »
Why did you use the long ground clip for such a measurement? The long ground clip practically destroys the true performance as can be seen from my demonstration.

Regards,
Janne

Thanks, I'm always up for learning a new tip. The Tek Probes don't have the cool spring ground tip, but I suspect this will do:

there seems to be little difference between the two?
« Last Edit: November 25, 2010, 09:54:06 am by Zyvek »
-Z
 

Offline Zyvek

  • Regular Contributor
  • *
  • Posts: 108
  • Country: us
Re: DS1052E hacked to DS1102E waveform comparision
« Reply #5 on: November 25, 2010, 07:02:04 am »
interesting. can you post picture of the rigol probe vs the teks probe P6138? is there any possibility finding/generating different shaped 50MHz wave/source to compare? thanx for letting us know.

i'm interested in looking the difference between:
rigol + ori probe
2430 + P6138
rigol + P6138
2467 + P6138
from the pics you posted here and in the other thread, i'm expecting the performance of 2430 + P6138 will be in the middle between rigol + ori probe and rigol + P6138.


Rigol Probe vs P6138 w/shortened ground tip


« Last Edit: November 25, 2010, 11:22:09 am by Zyvek »
-Z
 

Offline Zyvek

  • Regular Contributor
  • *
  • Posts: 108
  • Country: us
Re: DS1052E hacked to DS1102E waveform comparision
« Reply #6 on: November 25, 2010, 10:19:54 am »
Updated waves using the ground method suggested by jahonen:
-Z
 

Offline teixeluis

  • Contributor
  • Posts: 20
Re: DS1052E hacked to DS1102E waveform comparision
« Reply #7 on: November 25, 2010, 11:27:30 am »
Just to add that I have done a similar test, with my Rigol uncapped to 150 MHz, and the best results were obtained with the oscillator (a Rakon 45.1584 Mhz canned oscillator) connected with 1 cm length of wire to the BNC socket directly.The original Rigol probes are of very low quality, with a performance well below the rated BW.

Luis Teixeira
 

Offline mikeselectricstuff

  • Super Contributor
  • ***
  • Posts: 11876
  • Country: gb
    • Mike's Electric Stuff
Re: DS1052E hacked to DS1102E waveform comparision
« Reply #8 on: November 25, 2010, 12:13:11 pm »
P6138 in Ebay... $100 for a pair? gulp!... i'll be waiting for a while.
thanx janne, for the info as well. nice hack.

Quite aside from any performance issues, I like the old Tek probes as they are small, and have nice thin flexible cable - less prone to pulling a board off the bench when you have 4 of them attatched...
 
Youtube channel:Taking wierd stuff apart. Very apart.
Mike's Electric Stuff: High voltage, vintage electronics etc.
Day Job: Mostly LEDs
 

Offline Mechatrommer

  • Super Contributor
  • ***
  • Posts: 8765
  • Country: my
  • reassessing directives...
Re: DS1052E hacked to DS1102E waveform comparision
« Reply #9 on: November 25, 2010, 05:04:15 pm »
Updated waves using the ground method suggested by jahonen:
there is still ringing on the higher BW scope ???
if something can select, how cant it be intelligent? if something is intelligent, how cant it exist?
 

Offline Zyvek

  • Regular Contributor
  • *
  • Posts: 108
  • Country: us
Re: DS1052E hacked to DS1102E waveform comparision
« Reply #10 on: November 25, 2010, 05:24:48 pm »
Updated waves using the ground method suggested by jahonen:
there is still ringing on the higher BW scope ???


Shafri , I just now found a different 50Mhz oscillator, 3.3v SMT, what do you think?
-Z
 

Offline jahonen

  • Super Contributor
  • ***
  • Posts: 1045
  • Country: fi
Re: DS1052E hacked to DS1102E waveform comparision
« Reply #11 on: November 25, 2010, 06:42:16 pm »
The waveforms looked much better (although not perfect) to me when short ground was used. Does the oscillator have a bypass capacitor (something like 10-100nF) directly across the power supply pins? Lack of local tightly coupled bypass might cause ringing artifacts observed.

Regards,
Janne
 

Offline Zyvek

  • Regular Contributor
  • *
  • Posts: 108
  • Country: us
Re: DS1052E hacked to DS1102E waveform comparision
« Reply #12 on: November 25, 2010, 10:28:12 pm »
The waveforms looked much better (although not perfect) to me when short ground was used. Does the oscillator have a bypass capacitor (something like 10-100nF) directly across the power supply pins? Lack of local tightly coupled bypass might cause ringing artifacts observed.

Regards,
Janne

Thanks, I experimented with the cap values, which certainly squared the wave, but didn't eliminate the ringing...
-Z
 

Offline Mechatrommer

  • Super Contributor
  • ***
  • Posts: 8765
  • Country: my
  • reassessing directives...
Re: DS1052E hacked to DS1102E waveform comparision
« Reply #13 on: November 25, 2010, 11:57:04 pm »
nice! but seems short ground wont improve rigol any further compared to 24302236.
edited: i was confusing 2236 with 2430. actually i want rigol and 2236 to be competed each other. ammended my previous posts

seems the result are inconsistent between 2236 and 2430. in earlier test, the 2430 is able to detect the ring, 2236 flat. but in the second test (with short gnd) 2236 is ringing where 2430 is flat. ???
« Last Edit: November 26, 2010, 12:08:09 am by shafri »
if something can select, how cant it be intelligent? if something is intelligent, how cant it exist?
 

Offline Zyvek

  • Regular Contributor
  • *
  • Posts: 108
  • Country: us
Re: DS1052E hacked to DS1102E waveform comparision
« Reply #14 on: November 26, 2010, 04:33:00 am »
nice! but seems short ground wont improve rigol any further compared to 24302236.
edited: i was confusing 2236 with 2430. actually i want rigol and 2236 to be competed each other. ammended my previous posts

seems the result are inconsistent between 2236 and 2430. in earlier test, the 2430 is able to detect the ring, 2236 flat. but in the second test (with short gnd) 2236 is ringing where 2430 is flat. ???


I'm thinking the 2236 is not showing the detail?  Now that the signal is nicer on the 2467B I'll re-run the comparisons and post them today.
-Z
 

Offline marto

  • Newbie
  • Posts: 4
Re: DS1052E hacked to DS1102E waveform comparision
« Reply #15 on: November 26, 2010, 08:21:48 am »
I would have also like to have seen a comparison between the unhacked 50MHz limit and the 100MHz hacked signal, compared with a "true" signal. (I assume this has been done but you appear to have a pretty good range to give a ground truth) I still haven't done it to mine yet mostly because I just use if for reading serial or logic signals most of the time which makes that speed irrelevant, but maybe I should as who knows when I will need it.
 

Offline Zyvek

  • Regular Contributor
  • *
  • Posts: 108
  • Country: us
Re: DS1052E hacked to DS1102E waveform comparision
« Reply #16 on: November 26, 2010, 10:39:21 am »
Comparisons @ 50Mhz  using a Tektronix P6138 probe, 350Mhz 10MΩ 12-18pF, compensated for each scope

2236 100Mhz Analog
2430A 150Mhz/100Ms DSO
2445 150Mhz Analog
2467B 400Mhz Analog
1052E 100Mhz/2Gs

Test rig was a SMT oscillator, with a tuning cap thanks to (jahonen) Janne

I also ran a comparison on 20Mhz TTL signal generated by a BK 4051 DDS Multifunction Generator, I can post those as well if anyone wants to see 'em.

« Last Edit: November 26, 2010, 10:42:58 am by Zyvek »
-Z
 

Offline Mechatrommer

  • Super Contributor
  • ***
  • Posts: 8765
  • Country: my
  • reassessing directives...
Re: DS1052E hacked to DS1102E waveform comparision
« Reply #17 on: November 26, 2010, 01:27:14 pm »
I would have also like to have seen a comparison between the unhacked 50MHz limit and the 100MHz hacked signal, compared with a "true" signal. (I assume this has been done but you
its been proven that 100Mhz hacked is alot better than ori 50mhz. you may find somewhere in another thread here. so i think retesting here will be redundant. case study here is to compare between ori and tek probe.

@zyvek: now the only 2236 is ringing. i'm not sure its eligiblity anymore in doing the comparision, more like giving false reading it is, imo. or anyone with better idea why?
if something can select, how cant it be intelligent? if something is intelligent, how cant it exist?
 

Offline Zyvek

  • Regular Contributor
  • *
  • Posts: 108
  • Country: us
Re: DS1052E hacked to DS1102E waveform comparision
« Reply #18 on: November 26, 2010, 01:38:00 pm »

@zyvek: now the only 2236 is ringing. i'm not sure its eligiblity anymore in doing the comparision, more like giving false reading it is, imo. or anyone with better idea why?


It may be that the P6138 is just out of the compensation range for the 2236, the 2236 requires 22pf while the probe goes from 12-18. The rest of the scopes tested all require 15pf...

Check out the 20mhz test file attached as well.

-Z
 

Offline Mechatrommer

  • Super Contributor
  • ***
  • Posts: 8765
  • Country: my
  • reassessing directives...
Re: DS1052E hacked to DS1102E waveform comparision
« Reply #19 on: November 26, 2010, 01:46:00 pm »
and maybe you forgot to test with the rp220 original rigol probe to compare with. cant see how much benefit will be upgrading the probe.
if something can select, how cant it be intelligent? if something is intelligent, how cant it exist?
 

Offline Zyvek

  • Regular Contributor
  • *
  • Posts: 108
  • Country: us
Re: DS1052E hacked to DS1102E waveform comparision
« Reply #20 on: November 26, 2010, 02:01:09 pm »
and maybe you forgot to test with the rp220 original rigol probe to compare with. cant see how much benefit will be upgrading the probe.


Did you see this post:

https://www.eevblog.com/forum/index.php?topic=553.msg25227#msg25227

How is the weather by you today?
-Z
 

Offline Mechatrommer

  • Super Contributor
  • ***
  • Posts: 8765
  • Country: my
  • reassessing directives...
Re: DS1052E hacked to DS1102E waveform comparision
« Reply #21 on: November 27, 2010, 12:39:20 am »
How is the weather by you today?
rainy but hot! ;) thanx for posting the link.
if something can select, how cant it be intelligent? if something is intelligent, how cant it exist?
 

Offline jahonen

  • Super Contributor
  • ***
  • Posts: 1045
  • Country: fi
Re: DS1052E hacked to DS1102E waveform comparision
« Reply #22 on: November 28, 2010, 10:54:47 pm »
Here are some measurements from 54 MHz oscillator using a Agilent 10073C 500 MHz probe with Agilent MSO6034A compared to the DIY 1k resistor probe. I think it compares very favourable in logic applications. It works quite well especially if you have internal 50? terminator in the scope. Inline coax terminator works also quite well.

Regards,
Janne

 

Offline saturation

  • Super Contributor
  • ***
  • Posts: 4788
  • Country: us
  • Doveryai, no proveryai
    • NIST
Re: DS1052E hacked to DS1102E waveform comparision
« Reply #23 on: November 29, 2010, 12:51:01 am »
Thanks, jahonen.  Your insight in high speed techniques are great.  Its such a different beast.
Best Wishes,

 Saturation
 

Online vk6zgo

  • Super Contributor
  • ***
  • Posts: 4660
  • Country: au
Re: DS1052E hacked to DS1102E waveform comparision
« Reply #24 on: April 25, 2011, 04:19:27 pm »
 I think that a "square-ish" 50MHz waveform is a fairly savage test of a 100 MHz Oscilloscope (& a 150MHz one,for that matter)!

The frequency specs for both the Tek & Rigol scopes would normally be for the unit itself,with any probe effects being additional to

that.

The IC type "generator" will be affected by the probes,& should really be followed by a buffer stage,with a low Z output,which

could then be connected to the Oscilloscope,terminated in the same impedance.

The probe effects will then be eliminated.


VK6ZGO
 

alm

  • Guest
Re: DS1052E hacked to DS1102E waveform comparision
« Reply #25 on: April 25, 2011, 09:09:20 pm »
The frequency specs for both the Tek & Rigol scopes would normally be for the unit itself,with any probe effects being additional to that.
Last time I checked, Tek specs usually stated that the performance specs are for the scope set to 50 ohms, but are typically also valid on the 1Mohm setting with the standard probes. Agilent and Tek work hard to deliver their performance at the probe tip, since that's the only thing that counts in the end. Not so sure about Rigol.

I agree that to test just the scope, a terminated transmission line connected directly to the input is the way to go.
 

Online vk6zgo

  • Super Contributor
  • ***
  • Posts: 4660
  • Country: au
Re: DS1052E hacked to DS1102E waveform comparision
« Reply #26 on: April 26, 2011, 02:37:21 am »
Even when using the probe,a buffer would be useful,so the probe doesn't affect the output waveform from the IC.
The low Z output can be terminated directly,& the probe bridged across the termination.

Cheers,VK6ZGO
 

Offline Leo Bodnar

  • Frequent Contributor
  • **
  • Posts: 600
  • Country: gb
Re: DS1052E hacked to DS1102E waveform comparision
« Reply #27 on: April 26, 2011, 08:50:06 am »
Just a short notice - [untapped] transmission line properly terminated only at the driving end would perform just as well as terminated at both. 
There will be massive reflection off the unterminated scope 1MOhm side but it should be absorbed by the driver.

Last time I checked, Tek specs usually stated that the performance specs are for the scope set to 50 ohms, but are typically also valid on the 1Mohm setting with the standard probes. Agilent and Tek work hard to deliver their performance at the probe tip, since that's the only thing that counts in the end. Not so sure about Rigol.
I agree that to test just the scope, a terminated transmission line connected directly to the input is the way to go.

Offline scrat

  • Frequent Contributor
  • **
  • Posts: 606
  • Country: it
Re: DS1052E hacked to DS1102E waveform comparision
« Reply #28 on: April 27, 2011, 01:59:44 am »
Just a short notice - [untapped] transmission line properly terminated only at the driving end would perform just as well as terminated at both. 
There will be massive reflection off the unterminated scope 1MOhm side but it should be absorbed by the driver.

Why? Reflection should be visible (since you're looking at the high-Z side), and make the response difficult to understand...
One machine can do the work of fifty ordinary men. No machine can do the work of one extraordinary man. - Elbert Hubbard
 

Offline Leo Bodnar

  • Frequent Contributor
  • **
  • Posts: 600
  • Country: gb
Re: DS1052E hacked to DS1102E waveform comparision
« Reply #29 on: May 20, 2011, 11:29:03 pm »
Just a short notice - [untapped] transmission line properly terminated only at the driving end would perform just as well as terminated at both. 
There will be massive reflection off the unterminated scope 1MOhm side but it should be absorbed by the driver.

Why? Reflection should be visible (since you're looking at the high-Z side), and make the response difficult to understand...

Reflection of what?  If the driver has matched impedance it won't re-reflect arriving pulse/front.

Offline scrat

  • Frequent Contributor
  • **
  • Posts: 606
  • Country: it
Re: DS1052E hacked to DS1102E waveform comparision
« Reply #30 on: May 23, 2011, 07:40:32 pm »
Just a short notice - [untapped] transmission line properly terminated only at the driving end would perform just as well as terminated at both. 
There will be massive reflection off the unterminated scope 1MOhm side but it should be absorbed by the driver.

Look above, you were speaking about reflection that should be absorbed by the driver. I gave the sentence "transmission line properly terminated only at the driving end" the meaning of a line having Z characteristic impedance fed by a Z impedance source, and 1MOhm at the other end. That should cause reflections to happen.
One machine can do the work of fifty ordinary men. No machine can do the work of one extraordinary man. - Elbert Hubbard
 

Online vk6zgo

  • Super Contributor
  • ***
  • Posts: 4660
  • Country: au
Re: DS1052E hacked to DS1102E waveform comparision
« Reply #31 on: May 24, 2011, 01:19:25 am »
At the Oscilloscope end you will probably only see a higher level signal  than if it is terminated both ends.At least this has been my experience with video signals.

Of course with video signals,the signal amplitude in volts is an important specification,so an unterminated connection may cause problems to the devices further down the line.

If you monitor the signal at the generator end with another oscilloscope,you will start to see strange effects if the cable to the unterminated oscilloscope is 1/4 wavelength,(or some multiple of 1/4 wave) at one of the frequency components of the signal.


VK6ZGO
 

Offline Leo Bodnar

  • Frequent Contributor
  • **
  • Posts: 600
  • Country: gb
Re: DS1052E hacked to DS1102E waveform comparision
« Reply #32 on: May 24, 2011, 01:29:45 am »
Just a short notice - [untapped] transmission line properly terminated only at the driving end would perform just as well as terminated at both. 
There will be massive reflection off the unterminated scope 1MOhm side but it should be absorbed by the driver.

Look above, you were speaking about reflection that should be absorbed by the driver. I gave the sentence "transmission line properly terminated only at the driving end" the meaning of a line having Z characteristic impedance fed by a Z impedance source, and 1MOhm at the other end. That should cause reflections to happen.
Yes, there will be reflection going back to driver that will be absorbed in full if the Z-Z match is accurate.  The receiver won't see any reflection.

Offline scrat

  • Frequent Contributor
  • **
  • Posts: 606
  • Country: it
Re: DS1052E hacked to DS1102E waveform comparision
« Reply #33 on: May 24, 2011, 05:55:11 am »
Just a short notice - [untapped] transmission line properly terminated only at the driving end would perform just as well as terminated at both. 
There will be massive reflection off the unterminated scope 1MOhm side but it should be absorbed by the driver.

Look above, you were speaking about reflection that should be absorbed by the driver. I gave the sentence "transmission line properly terminated only at the driving end" the meaning of a line having Z characteristic impedance fed by a Z impedance source, and 1MOhm at the other end. That should cause reflections to happen.
Yes, there will be reflection going back to driver that will be absorbed in full if the Z-Z match is accurate.  The receiver won't see any reflection.

Reflections are completely damped only in the case of both ends adapted. Otherwise, high frequency stubs in transmission lines won't work, for example.
That's one way (IIRC) to estimate the distance at which a cable is broken.
One machine can do the work of fifty ordinary men. No machine can do the work of one extraordinary man. - Elbert Hubbard
 

Offline jahonen

  • Super Contributor
  • ***
  • Posts: 1045
  • Country: fi
Re: DS1052E hacked to DS1102E waveform comparision
« Reply #34 on: May 24, 2011, 06:04:59 am »
Yes it does work, like Leo described it. How the series termination (in point-to-point case) would otherwise work? One needs both ends terminated only if matching must be perfect from both directions, or the line is tapped (multipoint case).

There is 100% reflection from the mismatched load end, but since reflected and incident wave are in same phase at the end of the line, it does not cause any problem. Reflected wave then travels back to the source end, but there is no reflection from source end, since it is matched. Thus load does not see any kind of mismatch. Situation becomes completely different if signal is measured somewhere along the line, where the reflection is seen.

TDR measures reflections at the source end, where the reflection can be seen. It is consistent behavior with above description. I use same behavior to tune the series termination resistor, by adjusting the series termination resistor value so that the plateau becomes half of the final voltage (Zs+Rs=Z0).

Regards,
Janne
 

Offline scrat

  • Frequent Contributor
  • **
  • Posts: 606
  • Country: it
Re: DS1052E hacked to DS1102E waveform comparision
« Reply #35 on: May 24, 2011, 07:27:39 am »
Sorry, I was wrong!
At the return to the driver end the situation reaches its equilibrium, so no further reflections, as explained.
One machine can do the work of fifty ordinary men. No machine can do the work of one extraordinary man. - Elbert Hubbard
 

Online vk6zgo

  • Super Contributor
  • ***
  • Posts: 4660
  • Country: au
Re: DS1052E hacked to DS1102E waveform comparision
« Reply #36 on: May 24, 2011, 03:26:54 pm »
It is,of course,still best practice to terminate at the receiving end,especially if substantial power is being
transmitted through the cable,as operating at high SWR causes both voltage & current peaks at different points along the line,which may ,in a flexible coaxial cable, cause unwanted losses.

VK6ZGO

 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf