Calibrating a PG506 w/o Sampling System


Rick
 

Hi all,

I'm restoring my PG506 Calibration Generator (s/n B046623) after the magic smoke squirted out of the side of a tantalum cap. A few tants, supply caps,shaft couple and a replacement digital display board later and she's back up and running. The calibration procedures ends with using a sampling head to adjust out any aberrations. Tek's sampling systems seem to skip right over the 5000 and TM500 series systems. I can;t see investing in any pre-5000 or 7000 units at this stage as I'm heavily vested in the 5's. My Rigol digital scope only goes down to 5ns/div and the calibration is in the 1ns/div range. Is there something I'm missing in the sampling world? Is there a more modern way to do this? Thanks, Rick


Miguel Work
 

I have just paid 500$ for a sampling system for my 7000 scope.
-----Mensaje original-----
De: TekScopes@groups.io [mailto:TekScopes@groups.io] En nombre de Rick
Enviado el: lunes, 7 de diciembre de 2020 13:27
Para: TekScopes@groups.io
Asunto: [TekScopes] Calibrating a PG506 w/o Sampling System

Hi all,

I'm restoring my PG506 Calibration Generator (s/n B046623) after the magic smoke squirted out of the side of a tantalum cap. A few tants, supply caps,shaft couple and a replacement digital display board later and she's back up and running. The calibration procedures ends with using a sampling head to adjust out any aberrations. Tek's sampling systems seem to skip right over the 5000 and TM500 series systems. I can;t see investing in any pre-5000 or 7000 units at this stage as I'm heavily vested in the 5's. My Rigol digital scope only goes down to 5ns/div and the calibration is in the 1ns/div range. Is there something I'm missing in the sampling world? Is there a more modern way to do this? Thanks, Rick







Scanned by McAfee and confirmed virus-free.
Find out more here: https://bit.ly/2zCJMrO


Jean-Paul
 

Rick Bravo on the PG506 and repair. Great unit, I have been luck to have both newer and older versions and original manual.

From transient CAL of 2467B, I can say its a tricky and iterative process, and the correct gens and fixtures are essential.

You may get by with a 1 GHz digital scope but not the Chinese, HP, TEK or Lecroy.

The old TEK 7000 plugins are on epay but many need work, and easy to damage.

Best TEK maven on this is Dennis or Chuck Harris.

Good luck,


Jon


 

On Mon, Dec 7, 2020 at 05:17 PM, Jean-Paul wrote:


From transient CAL of 2467B, I can say its a tricky and iterative process, and
the correct gens and fixtures are essential.

You may get by with a 1 GHz digital scope but not the Chinese, HP, TEK or
Lecroy.
First of all: Unless the edge settings have been changed, it's probably not necessary to adjust them because your work hasn't influenced them. I certainly wouldn't touch them without the right equipment.
OTOH, the transient response calibration of a PG506 is a very simple adjustment, *if* you have the right equipment and perform the procedure correctly: One capacitor for the positive edge (C1000) and one for the negative edge (C940); optimize overshoot for both. That's it. Capacitor refs. are for SM "Late Model": S/N B040000 up.
However, since you're adjusting a rise/fall time <= 1 ns (that's spec, in practice usually 700 ps or better), your 'scope (as a rule of thumb) needs to have a rise time of at most 20% of that: 200 ps. That means a BW for a 'scope with Gaussian behavior of at least 1750 MHz, about 1400 MHz with many digital 'scopes.
So, using a 1 GHz BW digital 'scope won't crack it: The edge may look fine on it but probably will have serious overshoot, which you won't see on your (too slow) 'scope.
In practice, I'd consider 2.5 - 3 GHz to be the minimum BW required, taking into account that actual rise/fall time of most units is about 700 ps.

Raymond


Chuck Harris <cfharris@...>
 

I do the PG506 calibration on a tek 11801C. It reveals
all.

But, using a 7104 will only make things worse, best not
done. As Raymond says, a PG506 adjusted to a pretty
waveform on a 7104 looks like something you could spear
fish with on a 11801C.

The 7T11/7S11 pair is supposed to be adequate with the
proper sampling head, but I have never been able to get
such a pair to work reliably. It kind of drifts into
a measurement, and drifts out. I never found the problem
to be worth investigating. If someone wanted to, I am
sure that my 7T11/7S11 pair could be had for a reasonable
price... whatever that is these days.

-Chuck Harris

Raymond Domp Frank wrote:

On Mon, Dec 7, 2020 at 05:17 PM, Jean-Paul wrote:


From transient CAL of 2467B, I can say its a tricky and iterative process, and
the correct gens and fixtures are essential.

You may get by with a 1 GHz digital scope but not the Chinese, HP, TEK or
Lecroy.
First of all: Unless the edge settings have been changed, it's probably not necessary to adjust them because your work hasn't influenced them. I certainly wouldn't touch them without the right equipment.
OTOH, the transient response calibration of a PG506 is a very simple adjustment, *if* you have the right equipment and perform the procedure correctly: One capacitor for the positive edge (C1000) and one for the negative edge (C940); optimize overshoot for both. That's it. Capacitor refs. are for SM "Late Model": S/N B040000 up.
However, since you're adjusting a rise/fall time <= 1 ns (that's spec, in practice usually 700 ps or better), your 'scope (as a rule of thumb) needs to have a rise time of at most 20% of that: 200 ps. That means a BW for a 'scope with Gaussian behavior of at least 1750 MHz, about 1400 MHz with many digital 'scopes.
So, using a 1 GHz BW digital 'scope won't crack it: The edge may look fine on it but probably will have serious overshoot, which you won't see on your (too slow) 'scope.
In practice, I'd consider 2.5 - 3 GHz to be the minimum BW required, taking into account that actual rise/fall time of most units is about 700 ps.

Raymond






Jim Ford
 

Yeah, I believe that about the 7T11/7S11 pair, Chuck; I have a 7S12 and a 7S11, with a couple of S-4 sampling heads and S-51 and S-53 for triggering. PITA to get them to display anything on the 7904! At some point, I will look for an 11800 series scope like the one I had at work a few decades ago. I don't remember any issues with triggering back then. And the SD-24 sampling head got top marks for pulse fidelity from PicoSecond Pulse Labs (sold to Tek in 2014, IIRC) back then. Just got to run the purchase by the finance committee (my wife)! She can't say anything about the space it takes up; I have that covered with a 19-inch rack next to my bench....

Jim Ford

------ Original Message ------
From: "Chuck Harris" <cfharris@erols.com>
To: TekScopes@groups.io
Sent: 12/7/2020 9:32:49 AM
Subject: Re: [TekScopes] Calibrating a PG506 w/o Sampling System

I do the PG506 calibration on a tek 11801C. It reveals
all.

But, using a 7104 will only make things worse, best not
done. As Raymond says, a PG506 adjusted to a pretty
waveform on a 7104 looks like something you could spear
fish with on a 11801C.

The 7T11/7S11 pair is supposed to be adequate with the
proper sampling head, but I have never been able to get
such a pair to work reliably. It kind of drifts into
a measurement, and drifts out. I never found the problem
to be worth investigating. If someone wanted to, I am
sure that my 7T11/7S11 pair could be had for a reasonable
price... whatever that is these days.

-Chuck Harris

Raymond Domp Frank wrote:
On Mon, Dec 7, 2020 at 05:17 PM, Jean-Paul wrote:


From transient CAL of 2467B, I can say its a tricky and iterative process, and
the correct gens and fixtures are essential.

You may get by with a 1 GHz digital scope but not the Chinese, HP, TEK or
Lecroy.
First of all: Unless the edge settings have been changed, it's probably not necessary to adjust them because your work hasn't influenced them. I certainly wouldn't touch them without the right equipment.
OTOH, the transient response calibration of a PG506 is a very simple adjustment, *if* you have the right equipment and perform the procedure correctly: One capacitor for the positive edge (C1000) and one for the negative edge (C940); optimize overshoot for both. That's it. Capacitor refs. are for SM "Late Model": S/N B040000 up.
However, since you're adjusting a rise/fall time <= 1 ns (that's spec, in practice usually 700 ps or better), your 'scope (as a rule of thumb) needs to have a rise time of at most 20% of that: 200 ps. That means a BW for a 'scope with Gaussian behavior of at least 1750 MHz, about 1400 MHz with many digital 'scopes.
So, using a 1 GHz BW digital 'scope won't crack it: The edge may look fine on it but probably will have serious overshoot, which you won't see on your (too slow) 'scope.
In practice, I'd consider 2.5 - 3 GHz to be the minimum BW required, taking into account that actual rise/fall time of most units is about 700 ps.

Raymond









John Gord
 

Jim,
The HP 54120A/54121A combination also gives very good sampling performance, and is sometimes easier to find at a good price.
--John Gord

On Mon, Dec 7, 2020 at 10:40 AM, Jim Ford wrote:


Yeah, I believe that about the 7T11/7S11 pair, Chuck; I have a 7S12 and
a 7S11, with a couple of S-4 sampling heads and S-51 and S-53 for
triggering. PITA to get them to display anything on the 7904! At some
point, I will look for an 11800 series scope like the one I had at work
a few decades ago. I don't remember any issues with triggering back
then. And the SD-24 sampling head got top marks for pulse fidelity from
PicoSecond Pulse Labs (sold to Tek in 2014, IIRC) back then. Just got
to run the purchase by the finance committee (my wife)! She can't say
anything about the space it takes up; I have that covered with a 19-inch
rack next to my bench....

Jim Ford

------ Original Message ------
From: "Chuck Harris" <cfharris@erols.com>
To: TekScopes@groups.io
Sent: 12/7/2020 9:32:49 AM
Subject: Re: [TekScopes] Calibrating a PG506 w/o Sampling System

I do the PG506 calibration on a tek 11801C. It reveals
all.

But, using a 7104 will only make things worse, best not
done. As Raymond says, a PG506 adjusted to a pretty
waveform on a 7104 looks like something you could spear
fish with on a 11801C.

The 7T11/7S11 pair is supposed to be adequate with the
proper sampling head, but I have never been able to get
such a pair to work reliably. It kind of drifts into
a measurement, and drifts out. I never found the problem
to be worth investigating. If someone wanted to, I am
sure that my 7T11/7S11 pair could be had for a reasonable
price... whatever that is these days.

-Chuck Harris

Raymond Domp Frank wrote:
On Mon, Dec 7, 2020 at 05:17 PM, Jean-Paul wrote:


From transient CAL of 2467B, I can say its a tricky and iterative
process, and
the correct gens and fixtures are essential.

You may get by with a 1 GHz digital scope but not the Chinese, HP, TEK or
Lecroy.
First of all: Unless the edge settings have been changed, it's probably
not necessary to adjust them because your work hasn't influenced them. I
certainly wouldn't touch them without the right equipment.
OTOH, the transient response calibration of a PG506 is a very simple
adjustment, *if* you have the right equipment and perform the procedure
correctly: One capacitor for the positive edge (C1000) and one for the
negative edge (C940); optimize overshoot for both. That's it. Capacitor refs.
are for SM "Late Model": S/N B040000 up.
However, since you're adjusting a rise/fall time <= 1 ns (that's spec, in
practice usually 700 ps or better), your 'scope (as a rule of thumb) needs to
have a rise time of at most 20% of that: 200 ps. That means a BW for a 'scope
with Gaussian behavior of at least 1750 MHz, about 1400 MHz with many digital
'scopes.
So, using a 1 GHz BW digital 'scope won't crack it: The edge may look fine
on it but probably will have serious overshoot, which you won't see on your
(too slow) 'scope.
In practice, I'd consider 2.5 - 3 GHz to be the minimum BW required,
taking into account that actual rise/fall time of most units is about 700 ps.

Raymond









Jim Ford
 

Right, John.  I've used both a 54120A and Tek CSA803, although upwards of 25 years ago, and they are quite similar.  I do prefer Tek scopes, and in this case, the modularity of the Tek beats the HP, with its somewhat cumbersome external triggering and sampling boxes.  OTOH, one could bring the boxes close to the DUT and leave the much bulkier scope box on the bench or in a rack.  I suppose one could do the same with the much more lithe Tek sampling head(s) and (a) extender(s).  Tek for time domain, and HPAK for frequency domain!   JimSent from my Verizon, Samsung Galaxy smartphone

-------- Original message --------From: "John Gord via groups.io" <johngord=verizon.net@groups.io> Date: 12/7/20 2:56 PM (GMT-08:00) To: TekScopes@groups.io Subject: Re: [TekScopes] Calibrating a PG506 w/o Sampling System Jim,The HP 54120A/54121A combination also gives very good sampling performance, and is sometimes easier to find at a good price.--John GordOn Mon, Dec  7, 2020 at 10:40 AM, Jim Ford wrote:>> Yeah, I believe that about the 7T11/7S11 pair, Chuck; I have a 7S12 and > a 7S11, with a couple of S-4 sampling heads and S-51 and S-53 for > triggering.  PITA to get them to display anything on the 7904!  At some > point, I will look for an 11800 series scope like the one I had at work > a few decades ago.  I don't remember any issues with triggering back > then.  And the SD-24 sampling head got top marks for pulse fidelity from > PicoSecond Pulse Labs (sold to Tek in 2014, IIRC) back then.  Just got > to run the purchase by the finance committee (my wife)!  She can't say > anything about the space it takes up; I have that covered with a 19-inch > rack next to my bench....> > Jim Ford> > ------ Original Message ------> From: "Chuck Harris" <cfharris@erols.com>> To: TekScopes@groups.io> Sent: 12/7/2020 9:32:49 AM> Subject: Re: [TekScopes] Calibrating a PG506 w/o Sampling System> > >I do the PG506 calibration on a tek 11801C.  It reveals> >all.> >> >But, using a 7104 will only make things worse, best not> >done. As Raymond says, a PG506 adjusted to a pretty> >waveform on a 7104 looks like something you could spear> >fish with on a 11801C.> >> >The 7T11/7S11 pair is supposed to be adequate with the> >proper sampling head, but I have never been able to get> >such a pair to work reliably.  It kind of drifts into> >a measurement, and drifts out.  I never found the problem> >to be worth investigating.  If someone wanted to, I am> >sure that my 7T11/7S11 pair could be had for a reasonable> >price... whatever that is these days.> >> >-Chuck Harris> >> >Raymond Domp Frank wrote:> >>  On Mon, Dec  7, 2020 at 05:17 PM, Jean-Paul wrote:> >>> >>>> >>>  From transient CAL of 2467B, I can say its a tricky and iterative> process, and> >>>  the correct gens and fixtures are essential.> >>>> >>>  You may get by with a 1 GHz digital scope but not the Chinese, HP, TEK or> >>>  Lecroy.> >>>> >>> >>  First of all: Unless the edge settings have been changed, it's probably> not necessary to adjust them because your work hasn't influenced them. I> certainly wouldn't touch them without the right equipment.> >>  OTOH, the transient response calibration of a PG506 is a very simple> adjustment, *if* you have the right equipment and perform the procedure> correctly: One capacitor for the positive edge (C1000) and one for the> negative edge (C940); optimize overshoot for both. That's it. Capacitor refs.> are for SM "Late Model": S/N B040000 up.> >>  However, since you're adjusting a rise/fall time <= 1 ns (that's spec, in> practice usually 700 ps or better), your 'scope (as a rule of thumb) needs to> have a rise time of at most 20% of that: 200 ps. That means a BW for a 'scope> with Gaussian behavior of at least 1750 MHz, about 1400 MHz with many digital> 'scopes.> >>  So, using a 1 GHz BW digital 'scope won't crack it: The edge may look fine> on it but probably will have serious overshoot, which you won't see on your> (too slow) 'scope.> >>  In practice, I'd consider 2.5 - 3 GHz to be the minimum BW required,> taking into account that actual rise/fall time of most units is about 700 ps.> >>> >>  Raymond> >>> >>> >>> >>> >>> >>> >> >> >> >> >> >


Rick
 

Chuck, the Tek 11801C is intriguing. What sampling head do you use or should be used for the PG506?

Rick


Chuck Harris <cfharris@...>
 

I use SD24's... primarily because they have a TDR source built into
them.... and TDR is way cool! But the cheaper SD26's would also be fine.

Don't forget your SMA attenuators! The SD's have a limited safe range
of input voltages.... +/-3V is it. Their maximum input voltage for
measurement is +/-1V.

Static electricity will finish off an SD so quick you won't even know
it happened. Use a wrist strap.

The "C" model is the pinnacle of the species. I wouldn't waste my time
on the plain, "A", or "B"... though you might because the prices are much
cheaper.. as are the CSA models... though with them you lose half the
plugin slots, as they are reserved for level shifters, and fiber optic
receivers..

-Chuck Harris

Rick wrote:

Chuck, the Tek 11801C is intriguing. What sampling head do you use or should be used for the PG506?

Rick






Rick
 

I've been reading a bunch of the posts on the 11801C. I very intrigued. Seems like there's enough info out there that I might be able to repair one that I can get for $250. Timebase error T2523. Hmmm dare I?


Rick
 

Oops I'm off to a great start. It's code T2325. Only off by two digits, just like the lottery.


 

Error T2325 is a Timebase error, relating to the Strobe Generator, this from the 11801 Factory Diag manual to be found on KO4BB.com. There are no schematics, so this could be difficult to locate. Also, I would be wary of the SD-22 head plugged into it, as there are no shorting caps on the input SAM connectors.


Rick
 

Bob, the unit looks like is was physically damaged with the SAM connectors are are dented in. After looking at the diagnostics manual I'm losing my appetite.


 

That error is almost always due to dead batteries in the two static ram chips on the timing board. The rams are plugged into a Dallas memory hold up socket and can be removed by unplugging it. The rams can be replaced by Mouser part number 511-M48Z3570PC1. Two of them should run around $30. After installing both chips, do a few power up cycles and the error should be gone.

Regards,

Tom

On 12/8/2020 5:46 PM, Rick wrote:
Oops I'm off to a great start. It's code T2325. Only off by two digits, just like the lottery.



--
This email has been checked for viruses by Avast antivirus software.
https://www.avast.com/antivirus