Topics

7A26 transient response 5 nS all atten

Jean-Paul
 

Hello all: My workhorse 7A26 needed rework, in 7804 mainframe.
Cleaned the AC/DC/GR contacts on attn with the iso alcohol-paper strips, and exercised some scratchy controls.
Poor fast transient response on CH2, same dip and peak ~ 5 nS at ALL V/div.

Setup: Leo Bodnar pulser 40 pS>Mini Ckt HAT-15>>Mini Ckt HAT 20>>precision 50 Ohm thru>>7A26 input
Same result if 7A26 is in L or R side 7804.
Shots are: Affected 7A26, CH2, CH1, Good 7A26 Ch2 Ch1
http://crypto-museum.org/TEK/7A26/7A26.pdf

Since the aberration is constant on every V/DIV, the attenuator and 2X amp seem to be exonerated.

Before I dare probe the CH 2 amp circuit, a few questions please:

1/ Test points mentioned on schematic are NOT in PCB photos, only the parts numbers. Need a board layout with test pts.
2/ Using normal P6137 probe. placement of ground lead ?
3/ Precautions in probing around the custom TEK hybrid ICs, eg U2350, U2450?
4/ Serial: Manual has sch versions for SN B179999 < and B180000>.
My other 7A26s have this format SN, but this has no letter, 708058. I think its a late model, but see no IC date codes.
Readout PCB is 672-0051-13 Amplifier PCB is 670-2549-21 attenuators are PN200-1442-00

Hope to receive some tips from our TEK veterans!

Jon in Paris

Colin Herbert
 

Hi,
I, too, have a 7A26 with no letter at the front of the serial-number (102925). The label with this serial-number on it also identifies the plug-in as having been manufactured in Tektronix, Guernsey, which is why there is no "B" or "H" in front of the serial-number.
My manual, though originally produced in Beaverton (as I think all may have been), has a page in the "Change Information" section at the back referring to changes to Guernsey-produced 7A26s from tentative S/N 100396 and 100777.
I don't know how much use this will be to you, but I also have a source of the ICs used in the 7A26 (unless they have all been sold). They don't usually fail, but swapping them between channels (they are plug-in but be careful of bending the legs) is a pretty simple way of verifying them.
There are also a couple of resistors, one in each channel, which are at about their wattage limit, get hot and sometimes change value because of this. They are R1327 and R1329 (8K2, 0.25W. Carbon comp.) and would benefit in being replaced with resistors of higher wattage.
Colin.

-----Original Message-----
From: TekScopes@groups.io [mailto:TekScopes@groups.io] On Behalf Of Jean-Paul
Sent: 17 January 2020 12:20
To: TekScopes@groups.io
Subject: [TekScopes] 7A26 transient response 5 nS all atten

Hello all: My workhorse 7A26 needed rework, in 7804 mainframe.
Cleaned the AC/DC/GR contacts on attn with the iso alcohol-paper strips, and exercised some scratchy controls.
Poor fast transient response on CH2, same dip and peak ~ 5 nS at ALL V/div.

Setup: Leo Bodnar pulser 40 pS>Mini Ckt HAT-15>>Mini Ckt HAT 20>>precision 50 Ohm thru>>7A26 input
Same result if 7A26 is in L or R side 7804.
Shots are: Affected 7A26, CH2, CH1, Good 7A26 Ch2 Ch1
http://crypto-museum.org/TEK/7A26/7A26.pdf

Since the aberration is constant on every V/DIV, the attenuator and 2X amp seem to be exonerated.

Before I dare probe the CH 2 amp circuit, a few questions please:

1/ Test points mentioned on schematic are NOT in PCB photos, only the parts numbers. Need a board layout with test pts.
2/ Using normal P6137 probe. placement of ground lead ?
3/ Precautions in probing around the custom TEK hybrid ICs, eg U2350, U2450?
4/ Serial: Manual has sch versions for SN B179999 < and B180000>.
My other 7A26s have this format SN, but this has no letter, 708058. I think its a late model, but see no IC date codes.
Readout PCB is 672-0051-13 Amplifier PCB is 670-2549-21 attenuators are PN200-1442-00

Hope to receive some tips from our TEK veterans!

Jon in Paris

Jean-Paul
 

Colin! Many thanks, My manuals cover only B serials. Changes in Guernsey version?

All: My first reaction was calibration; page 5-7 sec 8, High Freq compensation.
But there are 10 adjustments per channel listed!
What order of adjustment? Which one might be for just the first 5-10 nS?

Mille mercis,

Jon

Albert Otten
 

Hello Jon,
My other 7A26s have this format SN, but this has no letter, 708058. I think its a late model, but see no IC date codes.
Readout PCB is 672-0051-13 Amplifier PCB is 670-2549-21 attenuators are PN200-1442-00
The Readout board is probably marked 670-2310-09. Then both board versions were introduced at B251090. You need the -01 version manual for that, Revised SEP 1985. I think the ICs have 3-digit date codes, the first indicating the year (5xx pointing to 1985). Of course the plugin itself can be newer than the ICs in it. My 713121 is from 1986.
BTW did you mean 7904 or 7104?

Albert

Chuck Harris
 

No letter, but starting with a "7" means it came from Tektronix Holland, NV,
Heerenveen, The Netherlands.

The last 5 digits of all of the serial numbers are sequential. I don't know
how they relate to the revisions of the manuals, and part changes, but I think
they are coordinated so that the revision ranges apply everywhere.

-Chuck Harris

Albert Otten wrote:

Hello Jon,
My other 7A26s have this format SN, but this has no letter, 708058. I think its a late model, but see no IC date codes.
Readout PCB is 672-0051-13 Amplifier PCB is 670-2549-21 attenuators are PN200-1442-00
The Readout board is probably marked 670-2310-09. Then both board versions were introduced at B251090. You need the -01 version manual for that, Revised SEP 1985. I think the ICs have 3-digit date codes, the first indicating the year (5xx pointing to 1985). Of course the plugin itself can be newer than the ICs in it. My 713121 is from 1986.
BTW did you mean 7904 or 7104?

Albert



Jean-Paul
 

Dear Albert and Chuck! Many thanks,

a/ Date codes of TEK ICs 413 = week 13 1984
b/ printed manuals Rev E May 1975, Rev C April 1977
c/ Bama, TekWiki, PDFs are 1972 versions.

Anyone with a link or scan of the -01 Rev SEP 1985 ?

I have 2 printed manuals (original TEK), open to an exchange if anyone has a spare of the printed late version.

to Chuck, your wisdom is appreciated re
<< calibration; page 5-7 sec 8, High Freq compensation.
...10 adjustments .order .... first 5-10 nS? >>

Bonne journée

Jon

Chuck Harris
 

Hi Jon,

I am not sure what you are asking me.

The last time I calibrated my 7A26's was so long ago,
that I certainly should do it again.. I don't use them
in a way that requires close calibration.

My recollection is they simply needed a little tweak
to smarten up the leading edge, and clean up the
chomp marks from the top.

You can get some clue as to which cap does what to
the waveform by touching it with a small metal bit...
an alignment screwdriver with a metal insert is fine.

-Chuck Harris

Jean-Paul wrote:

Dear Albert and Chuck! Many thanks,

a/ Date codes of TEK ICs 413 = week 13 1984
b/ printed manuals Rev E May 1975, Rev C April 1977
c/ Bama, TekWiki, PDFs are 1972 versions.

Anyone with a link or scan of the -01 Rev SEP 1985 ?

I have 2 printed manuals (original TEK), open to an exchange if anyone has a spare of the printed late version.

to Chuck, your wisdom is appreciated re
<< calibration; page 5-7 sec 8, High Freq compensation.
...10 adjustments .order .... first 5-10 nS? >>

Bonne journée

Jon




Albert Otten
 

Hi Jon,

The range of effect of the adjustments can more or less be guessed from RC times. For instance, C1339 (7-45 pF) in series with R1337(270 ohm) has effect in the first 3.5 ns to up to first 25 ns (taking 2*RC).
BTW: I think none of your 4 pictures shows a nice step response. Could that partly be due to the attenuators you use? Which scope type did you use? I never heard of a 7804.

Albert

Nenad Filipovic
 

Issues with 7A26 rise time/transient response are very likely associated
with aging trimmer caps going open circuit. This has been discussed already:
https://groups.io/g/TekScopes/topic/19243444#148001
My 7A26 had the same problem and I had to replace 2 of these trimmers.
After re-calibration it was good to go again.
Another possibility are the aging carbon composite resistors that tend to
drift in value over time (also discussed in the thread I posted above), but
I'd check the trimmers first.

Best Regards,
Nenad

Albert Otten
 

Hi Jon,
Anyone with a link or scan of the -01 Rev SEP 1985
You might ask Dave at ArtekManuals.com which Rev his 070-1484-01 pdf is.
I mentioned SEP 1985 since that's the one I have (also one of OCT 1987). Perhaps DEC 1984 or even APR 1984 already mentions your board versions.
I'm afraid a paper manual posted to you might never arrive because of all those strikes in Paris ;=)

Albert

John
 

My manual has a slip stating it was printed March 1981. Rev. letters vary across the sheets. There is a split of schematics & comp. locs. at S/n B180000 up.
155-0078-xx IC's appear with 01, 03, 05,07, 10 and 13 subscripts according to position and serial numbers. U2450 has four variants.
John

Jean-Paul
 

Hello all: Mille mercis pour vôtres réponses gentilles!

1/ To Albert: ...>> I think none of your 4 pictures shows a nice step response. Could that partly be due to the attenuators you use? Which scope type did you use? I never heard of a 7804.

Scope is 7904, (7804 typo) also use 7104.
Attenuators Mini Circuits HAT-15, and HAT-20, very fine.
I checked the step resp setup and scopes:
Used the Yokogawa DL1740, to check ste of Leo Bodnar pulser and attenuators, perfect!
7904 is fine, Checked Tek Cal Fix 067-0587-02 L and R
7A24 50 Ohm 500 MHz plugins, again excellent
Photos when I have more time!

Conclusion: All channels and both plugins 7A26 need transent alignment, but CH 2 of one plugin is bad.

To Albert: Although I am normally in Paris, I have a USA facility where I can receive ebay, post, etc. RE strikes in Paris, suffered since Dec 5, neither Macron/Fr Govt or unions CGT etc will really negotiate or compromise till the population suffers more. I predict a solution by Spring! The French LOVE rebellion and revolution!

To Chuck: Perfect info, I will do this next.

To Nenad Filipovic,
Sure! I am certain the culprit is a broken trim cap, and with Chuck's trimmer tapping tip, I will try to localize.
RE: ... aging carbon composite resistors

My 7A26s have NO Carbon Comp R, all are metal film, deposited, the 1/2 or 1 W are also MF. the change was probably done by the time my units were built in 1980s. (3 have SN B22... B24... B17) Very familiar with the Allen Bradley CC res, we used them in 12 kV HV modules for an avaition CRT HVPS.

(PS: To Nenad: You are in East Europe? Many connections)

To John and Albert;

ACK RE manuals, will check but I now have info to check dead trim caps and alignment. Still happy to swap with anyone that has a spare manual circa 1983 or later.

NOTES: I unearthed my 7000 plugin extenders circa 1981, (flex and rigid in TEK suitcase...) but not needed as all amplifier PCB trims and pars accessible in 7904 if left cover is removed.
Discovered the iPhone 11 with Pro Camera app is perfect for scope photography!
Out of time, will return next week.

Many thanks again to all to assist me to localize this problem and verify my setup and mainframe.

Bon Week-end à tous!

Jon in Paris

Jean-Paul
 

Hello again Chuck:
Quick Results:Tap-A-Cap: No changes. Turned every trimcap, watched and reset to original position Each one affected EXCEPT the first trim, C2336, the CULPRIT, 7-45 pF at U2350 first compensation network. RC ~ 2 nS with C2336 100 Ohm trim. Turning C2336 almost fixed response, but then it returned to the same bad transient and never reacted again, Open ckt.

C2336 is paralleded on back of PCB with C2338, fixed, 22 pF selected. I can kluge a trimcap on clips or bus wire onto the existing C2338 and avoid removing the PCB.
As a diagnostic I can prove the fault approximate the correct value and decide on the strategy to replace it.

Chuck and all others, many thanks again for your ideas and experience!

Jon in Paris

Jean-Paul
 

SOLVED: I kludged the C2336 with another trimcap 45 pF on 0.5 cm bus wire onto C2338 on back of board.
Did trim all comp R & C, very easy and iterative. ~ 22 pF Left the original trimcap onboard as its open ckt.
Transient Response is best of other channels and other plugin..
I will tweak all channels and improve this further.

http://crypto-museum.org/TEK/7A26/7A26_fix.jpg

Mille mercis a tous!

Jon in Paris

Albert Otten
 

Nice result Jon!

To John:
My manual has a slip stating it was printed March 1981. Rev. letters vary across the sheets.
There is a split of schematics & comp. locs. at S/n B180000 up.
155-0078-xx IC's appear with 01, 03, 05,07, 10 and 13 subscripts according to position and serial numbers. U2450 has four variants.
At first sight this surprised me because in OCT 1987 the is no 13. Closer inspection showed that that for the same plugin S/N range the same IC was 13 in older manuals and 10 in my recent manual. Already in SEP 1985 some numbers 13 were changed in 10. Just misprints? Also the numbering sequence is strange in time, like 7-3-0-10 for U2450. Anyway, the very last changes were at B139378.

Albert

Jean-Paul
 

Hello all:

After the 7A26 transient fix, I realized how useful these 7A26 are, so I searched my storage....2 more + 3 junkers!

2 tested 9 years ago, marked with faults, but after exercising controls,

#1 super, the best transient response.
#2 flaky BNC must fix but useable, has acceptable transient.
Both need attenuator cleaning and calibration.

THe 3 junkers are missing knobs, bad fault. I am in 7A26 nirvana.
Hours of fun to do the rework and calibrations on these marvelous plugins, at least I know the drill.

Bench festooned with 7A26...http://crypto-museum.org/TEK/7A26/7A26.jpg

Enjoy,
Jon

Albert Otten
 

Hello Jon,

Difficult to read in the picture, but for probe calibration remember that the input capacitance changed at some S/N from 20 pF to 22 pF.
I also have several 7A26s as general purpose plugins. Saves time and wear when going from one scoop to another.

Albert

Jean-Paul
 

Dear Albert: Yes wonderful to have a bunch of these versatile plugins, pluse junkers. Amazing TEK built more than 100K! My other "workhorse" is 7A24, wider BW but 50 ohm.

Pix: Here is a shot of the 2 ch transient traces on #2 plugin.
Better pix all 4 after finished.
http://crypto-museum.org/TEK/7A26/7A26_4.jpg

Cap: I am aware of the input capacitance change at B0180 from 20 >>22 pF. Due to the different input stage.
All 4 are the 22 pF later rev. Before CAL, need to find the 22 pF Capacitance normalizer, or modify another version!

Attn switch contacts: I worked on the unit with "bad BNC input" and of course its the AC/GND/DC switch, just exercising has cured the intermittent!
The CH1 attenuator has some flaky positions, so this one needs an alcohol treatment on the input attenuators.

Question: How to clean the contacts without the complex removal of the attenuators?

Without access to the contacts, to pass paper strips, could I take 100% isopropyl alcohol and just apply to the attenuator boards while exercising the switches?
One idea is to use a small syringe, place a thin plastic tube, and reach inside the plugin to the atten board surface, with plugin positioned so residue falls away from attenuator PCBs?

Await your sage advise!

Jon in Paris

Nenad Filipovic
 

Question: How to clean the contacts without the complex removal of the
attenuators?
IIRC the attenuator modules are socketed and can be easily removed. I
cleaned contacts on my 7A26 a while ago and I remember it was pretty
straightforward. Some paper strips soaked in IPA and tweezers, that should
be enough.

(PS: To Nenad: You are in East Europe? Many connections)
Yes, greetings from Belgrade, Serbia :)

Best Regards,
Nenad

Albert Otten
 

Hello Jon,

I don't have much experience with cleaning those attenuators. I think I used a squirt of Ipa when there was no access with a soaked strip of postcart.
In an old message here I saw a link to
https://www.eevblog.com/forum/reviews/tektronix-r7103-scope-teardown/ .
Scroll down to dismantled 7A26. Looks interesting at first sight, didn't read it in detail.

Albert