Author Topic: Need to build something to Calibrate some scopes  (Read 4270 times)

0 Members and 1 Guest are viewing this topic.

Offline ZenwizardTopic starter

  • Regular Contributor
  • *
  • Posts: 163
  • Country: us
Need to build something to Calibrate some scopes
« on: July 31, 2018, 10:57:10 pm »
Looking to Cal a few older scopes I have a 485 a 7704A and a 465 that I need to check the high frequency response on. I know the 485 is WAY out there was a whole cap and pot missing from the board when I received it. The 465 and 7704 might be in better shape. The only scope I have that is in Cal and I would trust is a Rigol 1104 100 Mhz. I need to generate some clean and fast rise square waves to calibrate the high frequency response of the verticals. My PG506 has some ringing on the front end of the pulse and I don’t have the Tunnel Diode pulsar to generate the clean edges. I am thinking about build a Schmitt Trigger Oscillator, but I don’t have a 500 Mhz scope to check it with. Also, can’t afford the Fluke 9800B. I don’t mind sending the scope in for a factory Cal but I want to make sure they can Cal it before I send it in. 

Any Ideas or chip recommendations to generate clean fast rise Square waves?
You Tube Link - Fixing broken Things
 

Offline tautech

  • Super Contributor
  • ***
  • Posts: 28142
  • Country: nz
  • Taupaki Technologies Ltd. Siglent Distributor NZ.
    • Taupaki Technologies Ltd.
Re: Need to build something to Calibrate some scopes
« Reply #1 on: July 31, 2018, 11:26:30 pm »
Displayed clean fast rise Square waves are most often the product on your connection quality and type.
It should be a quality BNC cable and terminated into 50  \$\Omega\$ either the scopes internal termination or a quality feed through termination......not a tee with a termination !

Lots of good guidance here:
https://www.eevblog.com/forum/testgear/show-us-your-square-wave/

Absolute frequency is not often required as much a a good mark time generator and the modern AWG can suit this requirement.
Avid Rabid Hobbyist
Siglent Youtube channel: https://www.youtube.com/@SiglentVideo/videos
 

Offline Muxr

  • Super Contributor
  • ***
  • Posts: 1369
  • Country: us
Re: Need to build something to Calibrate some scopes
« Reply #2 on: July 31, 2018, 11:50:03 pm »
w2aew did a video on a cheap fast pulse generator using a 74AC14 Schmitt Trigger Inverter chip. About 2ns edges.





 
The following users thanked this post: sean0118

Offline Conrad Hoffman

  • Super Contributor
  • ***
  • Posts: 1928
  • Country: us
    • The Messy Basement
Re: Need to build something to Calibrate some scopes
« Reply #3 on: August 01, 2018, 12:00:53 am »
There have been a couple extended threads on fast edge generators here, originating with the Jim Williams design.
https://www.eevblog.com/forum/projects/yet-another-fast-edge-pulse-generator/
 

Offline IconicPCB

  • Super Contributor
  • ***
  • Posts: 1527
  • Country: au
Re: Need to build something to Calibrate some scopes
« Reply #4 on: August 01, 2018, 10:27:56 pm »
There are three types of signal required to calibrate a scope.

1 a source of timing signal

2 a source of stable DC voltages

3 a source of leveled swept frequency signal

4 pulse generator with sharp edges


Timing signal used to check time base
DC voltage to check gain/linearity of vertical amplifiers ( and attenuators)
Sweeper to check 3dB bandwidth of vertical amps ( and attenuators)
Pulse generator to check transient behavior of vertical amps ( and attenuators)

If You have this gear in Your lab you need not build anything specific.

A hint:   The DC source can be any crappy old source in combination with a half decent multimeter, Your eye is good for only half a percent deviation on CRT screen.
 

Offline ZenwizardTopic starter

  • Regular Contributor
  • *
  • Posts: 163
  • Country: us
Re: Need to build something to Calibrate some scopes
« Reply #5 on: August 01, 2018, 11:41:30 pm »
I am good with most of it it is the fast edge pulse generator I need.

The scope cal equipment I have that I have confirmed is in spec are

Tektronix TG501
Tektronix SG503 no cable
Tektronix pg506 I don't trust the fast rise lots of ringing and the high amp is not working
Tektronix 191 fresh cal
Tektronix 184 Cal needs checked.
« Last Edit: August 02, 2018, 12:09:12 am by Zenwizard »
You Tube Link - Fixing broken Things
 

Offline IconicPCB

  • Super Contributor
  • ***
  • Posts: 1527
  • Country: au
Re: Need to build something to Calibrate some scopes
« Reply #6 on: August 02, 2018, 12:33:42 am »
Almost forgot....

You will need  a 50 ohm BNC feed through termination.
 

Offline ZenwizardTopic starter

  • Regular Contributor
  • *
  • Posts: 163
  • Country: us
Re: Need to build something to Calibrate some scopes
« Reply #7 on: August 02, 2018, 01:58:06 am »
Those I have in 3 Rigol ones and one Tektronix one. The tek one is really old not sure how to check it has not drifted over time. The Rigols are new.
You Tube Link - Fixing broken Things
 

Offline IconicPCB

  • Super Contributor
  • ***
  • Posts: 1527
  • Country: au
Re: Need to build something to Calibrate some scopes
« Reply #8 on: August 02, 2018, 10:52:22 am »
They are resistive terminations. Use an Ohm meter.
Make sure the BNC connector is not damaged.
 

Offline David Hess

  • Super Contributor
  • ***
  • Posts: 16548
  • Country: us
  • DavidH
Re: Need to build something to Calibrate some scopes
« Reply #9 on: August 02, 2018, 06:15:22 pm »
My PG506 has some ringing on the front end of the pulse and I don’t have the Tunnel Diode pulsar to generate the clean edges.

Are you sure about that?  You are referring to the PG506 fast rise and fall outputs, correct?  Those outputs are practically foolproof and I would trust the PG506 more than your setup and whatever you used to see any ringing.

The PG506 is fast enough (about 600 picoseconds) to do high frequency transient response calibration for the 150MHz 7704 and 100MHz 465 but you will need a tunnel diode pulser or similar for the 350MHz 485 which presents considerable difficulty.  If you just want to buy something, this is the least expensive option I am aware of.

Vertical and horizontal calibration including low and medium frequency transient response can be done with difficulty using an AC voltmeter, some attenuators, a function generator, and frequency counter.

Usually a "normalizer" is used to adjust the high impedance attenuators however a x10 probe works also if you have a relatively high voltage square wave source.  This step can be done last and does not affect other calibrations where a 50 ohm feedthrough attenuator is used.

 

Offline ZenwizardTopic starter

  • Regular Contributor
  • *
  • Posts: 163
  • Country: us
Re: Need to build something to Calibrate some scopes
« Reply #10 on: August 03, 2018, 02:45:34 am »
Here are 2 photos of the test set up You can't see it well but there is a Rigol 50 Ohm pass thru terminator on the front of the cable. This is my shortest 50 Ohm cable. If the packaging it came in is to be believed it is actually a Tektronix cable. The PG506 and scope were allowed a few minutes to warm up. This scope has not been through a cal process in a while but the voltages and ripple were checked recently both were very good. Recently is last 2 weeks. The vertical amp is a 7A26 the frame is a 7704A and the time based used in triggering was a 7B53A In the photos you can see the slightest bit of overshoot in the square wave at the beginning. I am new to very fast high speed signals so not sure is this is normal.

Just waiting for the pulser to come back in stock. I did not have the test fixtures or extenders to do the cals on the 7704 or the 7603 so I tackled the 485 first. This might have been a little far in to the deep end. I know I cant trust the 485 There was parts missing out of the R/C response network. There was also some one in there on the sweep board as well pots were swapped out to resisters. It was really strange. I am fighting with the normal struggle of initial equipment build up of the lab but learning fast and having fun.
« Last Edit: August 03, 2018, 02:52:32 am by Zenwizard »
You Tube Link - Fixing broken Things
 

Offline David Hess

  • Super Contributor
  • ***
  • Posts: 16548
  • Country: us
  • DavidH
Re: Need to build something to Calibrate some scopes
« Reply #11 on: August 03, 2018, 03:14:16 pm »
It is not possible to see what is going on at the leading edge at such a slow sweep speed.  It also looks like there might be some tilt which is a thermal balance problem and not adjustable on all 7000 mainframes.

The 7000 mainframes are more complicated because the vertical amplifiers and mainframe can be calibrated separately but this is not strictly necessary unless you want to be able to swap vertical plug-ins between mainframes without adjustments. 

It is possible to use a vertical amplifier as a 7000 mainframe standardizer by adding test points to the interface output for connecting an AC and DC voltmeter allowing all mainframe calibrations except transient response.  I have a single trace "golden" 7A16A that I used for this before I got a 7000 standardizer.  Access to the test points is feasible without an extender.

In theory transient response can be done also without the 7000 standardizer and Tektronix did this with a special extender and sampling oscilloscope on the 1GHz 7104 but this is beyond what is really needed for a 100MHz 7603 or 150MHz 7704.  If I designed my own 7000 extender, I might include this function.
 

Offline ZenwizardTopic starter

  • Regular Contributor
  • *
  • Posts: 163
  • Country: us
Re: Need to build something to Calibrate some scopes
« Reply #12 on: August 04, 2018, 12:53:58 am »
As requested pictures of faster sweep speeds. Same cable and terminator. Did allow about 20 min for the scope and PG506 to warm up. Thanks for all the help on this one.
You Tube Link - Fixing broken Things
 

Offline David Hess

  • Super Contributor
  • ***
  • Posts: 16548
  • Country: us
  • DavidH
Re: Need to build something to Calibrate some scopes
« Reply #13 on: August 04, 2018, 01:44:29 am »
The fast rise and fast fall outputs are not symmetrical.  While that looks pretty bad, you are viewing the rising edge of the fast fall output which may not be clean.  Either use the other output or change the trigger slope.

The 10mV/div setting is the "basic" sensitivity on the 7A26 where all of the high impedance input attenuators are switched out.
 

Offline ZenwizardTopic starter

  • Regular Contributor
  • *
  • Posts: 163
  • Country: us
Re: Need to build something to Calibrate some scopes
« Reply #14 on: August 04, 2018, 02:22:00 am »
I got that feeling when I was taking the pictures the input is +1  fall to 0 I need to use the -1 rise to 0 rising edge. I still need to get a pulser for the 485 but at last I can start in on the 7704 and 7603.
You Tube Link - Fixing broken Things
 

Offline tggzzz

  • Super Contributor
  • ***
  • Posts: 19281
  • Country: gb
  • Numbers, not adjectives
    • Having fun doing more, with less
Re: Need to build something to Calibrate some scopes
« Reply #15 on: August 04, 2018, 08:40:51 am »
As requested pictures of faster sweep speeds.

When part of the trace is far off the screen, there is the possibility that the amplifiers are overloaded and you might be seeing artefacts related to overload recovery. Tek scopes are better than some in that respect, but still exhibit the phenomenon.

That kind of consideration is a real problem when trying to see how fast signals settle to very near their final value. There are techniques to ameliorate such problems, but they are "not standard".
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 David Hess

  • Super Contributor
  • ***
  • Posts: 16548
  • Country: us
  • DavidH
Re: Need to build something to Calibrate some scopes
« Reply #16 on: August 04, 2018, 02:04:52 pm »
As requested pictures of faster sweep speeds.

When part of the trace is far off the screen, there is the possibility that the amplifiers are overloaded and you might be seeing artefacts related to overload recovery. Tek scopes are better than some in that respect, but still exhibit the phenomenon.

That kind of consideration is a real problem when trying to see how fast signals settle to very near their final value. There are techniques to ameliorate such problems, but they are "not standard".

That is a great point although the result shown so far does not look like that happened.  You can detect this problem sometimes by watching to see if the corner of the pulse changes shape as the position control is adjusted.

The fast rise and fast fall outputs have a variable amplitude but they are easier to use at high sensitivities if a x5 (14dB) or x10 (20dB) RF coaxial attenuator is attached.  I have always been a little dubious about relying on them at low amplitudes without an attenuator but actually they are very good even then so at 10mV/div you can get away without an attenuator.  At 1mV/div like with a 7A13, it might be a problem.
 

Offline ZenwizardTopic starter

  • Regular Contributor
  • *
  • Posts: 163
  • Country: us
Re: Need to build something to Calibrate some scopes
« Reply #17 on: August 06, 2018, 01:48:22 am »
Here are some more pictures. David was right on the money. I was using the fast fall not the fast rise edge. I miss read the front panel of the PG506  |O. That being said. Attached are some scope shots one is from my Rigol 1104Z this is new and in cal less then a year old. The others are the 7704A Vertical slot 1 Channel 1 (amp is a 7A26) The second shot is the same scope but slot B channel 2 seems to have the best looking waveform of the 4 channels (amp is also a 7A26). I am starting to think the PG506 is working correctly and this was tech error/trap fro young players.

You Tube Link - Fixing broken Things
 

Offline tggzzz

  • Super Contributor
  • ***
  • Posts: 19281
  • Country: gb
  • Numbers, not adjectives
    • Having fun doing more, with less
Re: Need to build something to Calibrate some scopes
« Reply #18 on: August 06, 2018, 06:01:07 am »
When looking at risetimes I always do a sanity check of a full cycle of a 1kHz and 1MHz signal to verify there are no low-frequency problems.

Thereafter I switch to the fastest timebase setting including any *10 switch. On my scopes this means 1ns or 2ns or 5ns.

You should also understand why Tek have the two horizontal dotted lines on the graticule; it is to aid transition time measurements. Do not change the vertical sensitivity away from the CAL setting, since that changes the bandwidth.
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 David Hess

  • Super Contributor
  • ***
  • Posts: 16548
  • Country: us
  • DavidH
Re: Need to build something to Calibrate some scopes
« Reply #19 on: August 06, 2018, 06:14:45 am »
Here are some more pictures. David was right on the money. I was using the fast fall not the fast rise edge. I miss read the front panel of the PG506  |O.

I have done that so many times with my PG506 that now I check and only do it about every third time.

Quote
Attached are some scope shots one is from my Rigol 1104Z this is new and in cal less then a year old. The others are the 7704A Vertical slot 1 Channel 1 (amp is a 7A26) The second shot is the same scope but slot B channel 2 seems to have the best looking waveform of the 4 channels (amp is also a 7A26). I am starting to think the PG506 is working correctly and this was tech error/trap fro young players.

The two vertical slots in the 7704A should perform identically with the same plug-in and there are two transient response adjustments for each inside the 7704A to allow this.  Gain matching between the slots is handled by the plug-ins which have an ingenious adjustment on the front of their control panel; push the black variable control knob in, hold it, and turn to adjust the gain of that trace.

The 7A26 has a lot of transient response adjustments for each trace so there should be no problem making the traces match.  Access to the left side of the 7A26 plug-ins may be gained by removing the 7704A's cover.

Do not change the vertical sensitivity away from the CAL setting, since that changes the bandwidth.

Tektronix used a Gilbert cell configured as a paraphase amplifier for the variable gain control so bandwidth should not vary significantly when the VAR control is used.  This paraphase amplifier is also used to calibrate the gain of each trace by pushing the VAR control in and holding it instead of popping it out.
 

Offline ZenwizardTopic starter

  • Regular Contributor
  • *
  • Posts: 163
  • Country: us
Re: Need to build something to Calibrate some scopes
« Reply #20 on: August 06, 2018, 09:50:20 pm »
Sorry about the confusion of the test setup. The 7704 has 2 7A26's installed in the verticals. Here are the shots from each of them. I let the devices warm up then i just moved the cable and changed the trigger source. If we think I can call the PG506 good I can start tuning these as if I can trust the PG506 these need a bit of a cal. Pictures follow.
You Tube Link - Fixing broken Things
 

Offline ZenwizardTopic starter

  • Regular Contributor
  • *
  • Posts: 163
  • Country: us
Re: Need to build something to Calibrate some scopes
« Reply #21 on: August 06, 2018, 09:51:07 pm »
And the Right Channel Pictures.
You Tube Link - Fixing broken Things
 

Offline rhb

  • Super Contributor
  • ***
  • Posts: 3476
  • Country: us
Re: Need to build something to Calibrate some scopes
« Reply #22 on: August 07, 2018, 12:08:29 am »
Buy one of Leo Bodnar's 40 pS rise time pulsers.  I've got one of the BNC versions and it is great.  They come with a plot of the step response of your unit. 

I'm getting ready to do a full cal of a refurbished 485, but the pulser shows it's very good as is.  But I want this tuned up for less than 1% overshoot and I plan to keep it until they carry me out in a box.
 

Offline David Hess

  • Super Contributor
  • ***
  • Posts: 16548
  • Country: us
  • DavidH
Re: Need to build something to Calibrate some scopes
« Reply #23 on: August 07, 2018, 01:32:30 am »
Sorry about the confusion of the test setup. The 7704 has 2 7A26's installed in the verticals.

I understood what you had.  Once calibrated, you should be able to exchange the 7A26s and only have to adjust their front panel gain calibration.

Quote
If we think I can call the PG506 good I can start tuning these as if I can trust the PG506 these need a bit of a cal. Pictures follow.

I think you can trust the PG506 but the only way to know is to check it against another known good oscilloscope or use one of Leo Bodnar's pulsers which I linked.

Below are a couple of photographs from my 500MHz 7904 which was a design contemporary  to the 7704A and released at the same time.

The first two photographs are a random 7A26 plug-in with my PG506 driving both traces.  The third slightly smaller trace on the right side is my 200MHz 7000 mainframe standardizer which gives some idea of the transient response of the mainframe by itself.

The third photograph is with a dual trace 400MHz 7A24 (300MHz in the 7904) vertical amplifier which has 50 ohm inputs at two different sweep speeds.  The PG506 is not really fast enough for this configuration.
 

Offline ZenwizardTopic starter

  • Regular Contributor
  • *
  • Posts: 163
  • Country: us
Re: Need to build something to Calibrate some scopes
« Reply #24 on: August 07, 2018, 01:56:01 am »
In the end I am going to get the pulser just at the moment they are not in stock. In this particular scope the one thing I do trust is the frame. I have the calibration fixture for the 7000 series frames that is in spec and working. I have also run the frame through the cal check in the service manual and the 7704 checked out great. the 7603... not so much some power supply issues in that one. And I know for sure the squair waves out of the pg 506 are good up to 100 Mhz.

Thanks for the help everyone.

You Tube Link - Fixing broken Things
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf