Date   
Re: older 1720 vectorscope service manual?

 

Anybody have a manual for one of these floating around or suggestions
> >for likely failures in this (older) unit?

I have the manual for SN B039999 and below. If you still want to go on fixing it I can scan the schematics.

/Håkan

Re: older 1720 vectorscope service manual?

Chris Elmquist
 

On Monday (01/23/2017 at 04:34AM -0500), 'Michael A. Terrell' mike.terrell@... [TekScopes] wrote:
I have a 1720, SN B029603, with a dead power supply. The schematic for later production units is nothing like the earlier power supply boards.
Indeed. I spent most of Saturday proving that :-)

The reference designators on my PCB are three digits while everything
in the later schematics are one or two digit. The layout is "close" but
not close enough to figure out which is what without a lot of reversing.

Not sure it's worth it since I was just going to make a "scope clock"
out of this unit anyway. I have other (working) true XY displays which
this is not.

I see Ignacio's offer of a 1740 manual and thank you for that. I will need
to study a little to see if that is at all close to the 1720 before I take
you up on the offer.

Chris N0JCF


-----Original Message-----
From: "Chris Elmquist chrise@... [TekScopes]" <TekScopes@...>
Sent: Jan 22, 2017 3:21 PM
To: TekScopes@...
Subject: [TekScopes] older 1720 vectorscope service manual?

I am looking for a service manual for a low serial number (B012019)
Tek 1720 VectorScope.

Online sources seem to be for S/N B060000 or later and I find that the
design on my older unit is quite different-- particularly in the power
supply where I think I'm having trouble (no light from the CRT, anode
is at about 5KV instead of 11KV and filament is at about 1V).

Anybody have a manual for one of these floating around or suggestions
for likely failures in this (older) unit?

Thanks.

Chris
--
Chris Elmquist

Michael A. Terrell
--
Chris Elmquist

Re: Tek 500-Series scope in a movie.

Michael A. Terrell
 

The movie Futureworld is running on Comet TV this month. I saw a small, rack mounted HP freq counter, and a rack mounted Tek scope, but there wasn't enough of them shown for me to identify the model numbers. The counter had labels stuck to it, along with a lot of residue of other labels. In the first part of the movie, they brag about the place costing 1.5 billion dollars, yet the props look like the junk from a '70s surplus store. :)

-----Original Message-----
From: "ditter2@... [TekScopes]" <TekScopes@...>
Sent: Jan 23, 2017 10:35 AM
To: TekScopes@...
Subject: [TekScopes] Re: Tek 500-Series scope in a movie.

I have not seen the movie yet, but intend to.

If it is a 545B, then the timing is not accurate. The 'B' model was introduced in the late 1960's (1967?), while the movie was the lead up to, and beginning of the Mercury program - 5 years earlier. A 545A would have been more accurate, although IBM was using the early 400 series - the 422, (then 453, then 475) for service of the early mainframes.

Last weekend I visited the Computer History Museum in Mountain View, CA. They have two labs showing restored running historic computers. The Digital Equipment Corporation PEP-1 lab has a running Tek 545A with CA plug-in.

Steve



Michael A. Terrell

Re: Tektronix 2440 waveform noise above 200nS timebase

Szabolcs Szigeti
 

Hi,

Are you testing it while the scope still open? My experience was that having the case off will make the scope pick up some noise.
And yes, I have the same experience on warming up, too. Doing a self-cal does reduce the noise, but that is dependent on how much the scope is warmed up. I have found considerable visible differences in performance between the cold and warm instrument. Self cal does compensate for it, but of course, there is not much point in doing a cal on a cold scope. However, doing a self cal on a warmed up unit will make bad performance while cold. But that will be ok, once it has warmed up.


Szabolcs

Re: Tektronix 2440 waveform noise above 200nS timebase

Simon Spiers
 

Szabolcs,I have tried with repet mode on, and it helps somewhat but it still looks bad. And V/div or bandwidth limit make no difference at all, As the the bodge, yes its awful, and I have a feeling thats from the factory. It seems strange that the noise does improve to a quite usable level after a few hours, and a freezer spray reveals nothing obvious.
If i feed a clean high frequency sine wave into the scope the waveform has the same noise modulated on the input. Makes no difference which channel I use.
David, thanks for the comment, the procedure was followed very carefully, and does pass the cal without problems.

Re: Tektronix 2440 waveform noise above 200nS timebase

 

On 23 Jan 2017 01:34:34 -0800, you wrote:

...

2 - there are two, slightly different adjustment procedures, depending on the serial number. Make sure you follow the proper one.
3 - make sure you turn the proper trim-pot, it is quite easy to accidentally turn an other one. I printed the page from the manual, mark the pots there with large red letters and oriented the paper the same way the scope was sitting on my bench. Also, the pot layout is different depending on the serial number!
The first time I did this, I printed the pages out to mark up and then
*still* did the calibration procedure for the wrong serial number
range. The pot layout was different but I was able to figure it out
not that doing that helped any.

Re: 468 DSO B timebase triggering

Colin Herbert
 

Are you by any chance confusing my 468 with a 7-k series? As far as I know, there isn’t a 400-series with readout. The readout feature was something that first impressed me about my 7623A after collecting a few 400-series initially.

Colin.



From: TekScopes@... [mailto:TekScopes@...]
Sent: 23 January 2017 15:52
To: TekScopes@...
Subject: [TekScopes] Re: 468 DSO B timebase triggering





It may be that the readout is causing the blank spots.It is written by the same CRT gun, so the beam can't be in 2 places at once.Try turning the readout off & see if the gaps go away.
HankC, Boston WA1HOS

[Non-text portions of this message have been removed]



_____

No virus found in this message.
Checked by AVG - www.avg.com <http://www.avg.com/email-signature>
Version: 2016.0.7998 / Virus Database: 4749/13802 - Release Date: 01/20/17



[Non-text portions of this message have been removed]

Re: TDS744A bad sampler interleave ?

Siggi
 

On Sun, 22 Jan 2017 at 18:30 beltoid97@... [TekScopes] <
TekScopes@...> wrote:



I then tried an asymetric ramp (hoping the different slopes would help
pick out misplaced segments, etc), same result (
http://www.tiikoni.com/tis/view/?id=8cc17f1 ). Note, there *does* appear
to be a second trace with a slight offset, but zooming in to the sample
level I can't confirm this, it just looks like random noise.
Huh, the doubled trace is interesting. What happens if you run your signal
generator into another channel, and GND couple CH1? What happens when you
position it up and down the screen?
I believe the position control adds an offset voltage to the input, so this
is the moral equivalent to sampling different DC levels.



I suspect the mix-up has a cycle that's a multiple of your triangle wave.
Of course, I tried varying the triangle wave frequency -- same result
always.
I think the triangle waves I've been using don't slew fast enough to be
very useful, especially if the error(s) are of the order of 10-20 samples /
10-20ns.

Last test I did (sorry, no screenshot) was a ~ 400Hz square wave, 500k
length record , 500MS/s, so I would have only a single edge in the whole
record. Zooming in to the edge showed a repeated 8-sample segment.
Ok, the MUX has an 8-sample wide interface to the memory.


I guess I'll have to take apart the machine to clean some contacts and
probe address lines... I'm still surprised none of these problems are found
by the internal self tests.
Yeah, that seems strange. Although it only manifests at certain record
lengths, so perhaps those aren't tested by the self-tests.

I hope you know of the capacitor plague that be-riddles these scopes? If
the scope hasn't been re-capped already, then you should probably make
plans to do so. There are descriptions around of how to do this, but what
often seems to happen is that the leaking electrolyte erodes traces,
turning them into opens. Even if the scope has been re-capped, unless you
know that whoever did it, did a good job in cleaning the board, you might
want to plan to clean it at least.
Another common problem I've heard mentioned with these is dicey soldering,
where sometimes retouching the soldering job on the ADC or MUXes fixes
things. The acquisition board runs pretty hot, so may thermal cycling
cracks the joints in the end.

Now, if the caps are bad on the acquisition board, it might be that one of
the MUXes is derailing in certain modes of operation because the supply
isn't holding it up. I doubt it, but it's possible.

If you look through the block diagram of one of those channels, you'll see
that any ADC can sample any channel. When the 78X scopes are pushed to
4GS/s, all the ADCs will be sampling the one enabled channel.

The interface from the ADCs to the MUX is split in two - the A/B channels.
This means that e.g. stuck bits in the ADC will manifest with a cadence of
two, though it gets complicated as the MUX also does decimation, and as far
as I know, there's no documentation on which sampling rates involve
decimation.
I think it's safe to say that problems in the ADC->MUX interface will never
manifest with a cadence larger than 2.

The MUX then stacks up samples, which are written to memory 8 at a time
(there's a 64 bit data bus out of the MUX).

If you had a stuck data bit between MUX and memory, you'd get a trouble
cadence of 1 in 8.

Hence, I think most likely this has to be an address bit, perhaps an LSB.
A stuck LSB would cause blocks of 8 samples to stomp the previous block of
8 samples on write form the MUX. On read-back, you'd only see the stomped
block.
You would then expect to see repeats every 8 samples, though I'm hard put
to fit that to the triangle wave samples you've posted pictures of.

I can't see how this'd be dependent on the record length, however.

When you debug this, you can probably get quite far by inputting the same
signal on two channels (maybe e.g. 1 & 3 to maximize symmetry), and then
comparing signals across two channels.

Mind that you MUST run a bench fan on the acquisition board if you run the
scope out of the case.

Good luck,
Siggi

Re: Upcoming Sphere Research Stuff Day

NigelP
 

Wish something like that happened in MY back-yard :(.

I'm REALLY jealous!

On the other hand my wife is probably glad it's NOT happening in my back-yard!

Nigel

Re: 468 DSO B timebase triggering

 

It may be that the readout is causing the blank spots.It is written by the same CRT gun, so the beam can't be in 2 places at once.Try turning the readout off & see if the gaps go away.
 HankC, Boston WA1HOS

[Non-text portions of this message have been removed]

Re: Tek 500-Series scope in a movie.

ditter2
 

I have not seen the movie yet, but intend to.

If it is a 545B, then the timing is not accurate. The 'B' model was introduced in the late 1960's (1967?), while the movie was the lead up to, and beginning of the Mercury program - 5 years earlier. A 545A would have been more accurate, although IBM was using the early 400 series - the 422, (then 453, then 475) for service of the early mainframes.

Last weekend I visited the Computer History Museum in Mountain View, CA. They have two labs showing restored running historic computers. The Digital Equipment Corporation PEP-1 lab has a running Tek 545A with CA plug-in.

Steve

Re: older 1720 vectorscope service manual?

EB4APL
 

I have a 1740 Waveform / Vector monitor original manual. It is free to
anybody who wants it. I suppose that with Digital TV all these monitors
has been discarded, but who knows.

Ignacio, EB4APL


El 23/01/2017 a las 10:34, 'Michael A. Terrell'
mike.terrell@... [TekScopes] escribió:

I have a 1720, SN B029603, with a dead power supply. The schematic for
later production units is nothing like the earlier power supply boards.

-----Original Message-----
From: "Chris Elmquist chrise@... [TekScopes]"
<TekScopes@...>
Sent: Jan 22, 2017 3:21 PM
To: TekScopes@...
Subject: [TekScopes] older 1720 vectorscope service manual?

I am looking for a service manual for a low serial number (B012019)
Tek 1720 VectorScope.

Online sources seem to be for S/N B060000 or later and I find that the
design on my older unit is quite different-- particularly in the power
supply where I think I'm having trouble (no light from the CRT, anode
is at about 5KV instead of 11KV and filament is at about 1V).

Anybody have a manual for one of these floating around or suggestions
for likely failures in this (older) unit?

Thanks.

Chris
--
Chris Elmquist
Michael A. Terrell


------------------------------------------------------------------------
Posted by: "Michael A. Terrell" <mike.terrell@...>
------------------------------------------------------------------------

Re: Tektronix 2440 waveform noise above 200nS timebase

 

Note that the amplitude flatness of the SG503 is only spec'ed when a particular BNC cable is used.The cable part number ( 012-0482-00) is stenciled right on the front panel, but it may not be there on earlier serial numbers.My SN is BO63557 & it has the stencil.
 HankC, Boston WA1HOS

[Non-text portions of this message have been removed]

Re: 465B for testing home electricity.

mattycurious82
 

If I may make one suggestion that I think is very much worth the effort:

Google "Dave Jones" and "how not to blow up your scope"
You should see EEVBlog episode #279. EEVBlog is a great source of info, Dave can be a bit brash at times but if you have some time he really breaks it all down nicely.

Dave has been a great source of info as well as entertainment for me for quite some time.

Re: Tektronix 2440 waveform noise above 200nS timebase

Szabolcs Szigeti
 

Hi,

I do not think there is any problem with the scope. It's normal to have this level of apparent noise on a digital scope, simply the way they work. Turning bandwidth limit down to 20MHz should verify this, as that would remove much of the noise from the analog part.

However, having said that. I'm not sure if you have the repet mode on at this sweep speed, my guess is that you do not. I tried this setup with repet on and off, and I get a similar waveform with off.

So what I think is that you see here actually the artifact from three things. The normal noise of the scope, the sampling speed and the interpolation the scope does between these points. Anything above 200ns sweep really needs repet, otherwise the sample points are so sparsely spaced that the waveform will jump all over because of the sin(x)/x interpolation. You can verify this by turning off the vectors or turning on the intensification of the sample points.

Also, you can try turning repet on and look at a low noise waveform and see how much noise you see on the signal. Repet absolutely needs a periodic input with proper trigger. Just looking at the noise will produce garbage.
Once I have some time, I'll do some experiment with my scope and try and make a video.

However, those bodge resistors look really bad. My scope has non of these, all those pots are tidy. What is the serial number of your scope? If this is really factory bodge, then maybe it was made around the time of some big revision?

Szabolcs

Re: Tektronix 2440 waveform noise above 200nS timebase

Simon Spiers
 

Hi szigszabolcs I have done the full calibration on the scope as it was failing on a few tests on the self test.
Your tips are greatly apreciated, the thing as you mention is to remember to save the calibraion mods into NVR.
Here is a video of the fault.
YouTube https://www.youtube.com/watch?v=tU2xnKK-MZ8

YouTube https://www.youtube.com/watch?v=tU2xnKK-MZ8 Watch Queue Queue Watch QueueQueue Remove all Disconnect The next video is startingstop Loading...



View on www.youtube.com https://www.youtube.com/watch?v=tU2xnKK-MZ8
Preview by Yahoo



Hi, I have experimented with ways to calibrate the 2440. Since the nvram content was lost, there was nothing to loose anyway. Initially I did not have a 250MHz



Hi, I have experimented with ways to calibrate the 2440. Since the nvram content was lost, there was nothing to loose anyway. Initially I did not have a 250MHz

Re: Tektronix 2440 waveform noise above 200nS timebase

Simon Spiers
 

Szabolcs, thank you for a great list of tips. The cooling part of this scope is a real pain when trying to adjust pots as is the aliasing technique used to set the CCD sides. I think I will remove the hybrids and clean the contacts, the noise improves with the unit warming up after an hour or so, someone suggested poor caps?
This is a video I made of the problem.
https://www.youtube.com/watch?v=tU2xnKK-MZ8 https://www.youtube.com/watch?v=tU2xnKK-MZ8





---In TekScopes@..., <szigszabolcs@...> wrote :

Hi,

I have experimented with ways to calibrate the 2440. Since the nvram content was lost, there was nothing to loose anyway. Initially I did not have a 250MHz signal generator, but had a PG502 pulse generator. Since even the first harmonics is attenuated by the scope of this frequency, I figured it may work. And it did. Later, when I managed to get an SG503 levelled sine generator, I re-did the cal, but it was basically spot on. The only difference in the process that I observed was that the stability of the PG502 is worse, so I frequently had to re-adjust the generator frequency to get a nice aliased waveform.

But any generator you use, it must of course be finely tunable to get the aliased display.

Once done a couple of times, the adjustment procedure is not that bad, but still not fun looking at the aliased waveforms for alignment and concentrating on not screwing up the steps. Good news is that you cannot break anything permanently, but a botched adjustment can make the scope completely unusable.

I followed the procedure exactly as in the service manual, and it did improve the noise and the repet mode waveform. So it is worth doing. However, be careful with a couple of things:
0 - there is a step to verify the CDD alignment. You can check if you need adjustment at all.
1 - cooling, this has been said a lot of times, use an external fan when running the scope without the case. The service manual warns about this.
2 - there are two, slightly different adjustment procedures, depending on the serial number. Make sure you follow the proper one.
3 - make sure you turn the proper trim-pot, it is quite easy to accidentally turn an other one. I printed the page from the manual, mark the pots there with large red letters and oriented the paper the same way the scope was sitting on my bench. Also, the pot layout is different depending on the serial number!
4 - follow exactly the manual. The initial setting on the CDD gains is a fiddly procedure, and if not done properly, the scope will fail the self cal. There can be as simple mistakes as forgetting to save the value after setting. Make sure you do not accidentally skip a step in the procedure. This happened to me a couple of times.
5 - on higher serial scopes, there is a separate step for adjustment of the 100ns clock, which needs a 300MHz generator. I did not have that, so I tried it with lower frequency aliased waveform, it was possible to do the alignment, but probably not to the highest precision, since the waveform was not stretched enough to be easily visible.
6- once I had the experience, it still took me about an hour of concentration to do the full calibration so make sure you have a glass of beer after this. You'll need it.

BTW. Are you looking at the noise in repet mode when you have a triggered waveform, right? My experience is repet will not give you useful results if just looking at a flat GND line or having a marginally triggered waveform. Which is kind of obvious, given the way this mode works.

Szabolcs

Re: Tektronix 2440 waveform noise above 200nS timebase

Szabolcs Szigeti
 

Hi,

I have experimented with ways to calibrate the 2440. Since the nvram content was lost, there was nothing to loose anyway. Initially I did not have a 250MHz signal generator, but had a PG502 pulse generator. Since even the first harmonics is attenuated by the scope of this frequency, I figured it may work. And it did. Later, when I managed to get an SG503 levelled sine generator, I re-did the cal, but it was basically spot on. The only difference in the process that I observed was that the stability of the PG502 is worse, so I frequently had to re-adjust the generator frequency to get a nice aliased waveform.

But any generator you use, it must of course be finely tunable to get the aliased display.

Once done a couple of times, the adjustment procedure is not that bad, but still not fun looking at the aliased waveforms for alignment and concentrating on not screwing up the steps. Good news is that you cannot break anything permanently, but a botched adjustment can make the scope completely unusable.

I followed the procedure exactly as in the service manual, and it did improve the noise and the repet mode waveform. So it is worth doing. However, be careful with a couple of things:
0 - there is a step to verify the CDD alignment. You can check if you need adjustment at all.
1 - cooling, this has been said a lot of times, use an external fan when running the scope without the case. The service manual warns about this.
2 - there are two, slightly different adjustment procedures, depending on the serial number. Make sure you follow the proper one.
3 - make sure you turn the proper trim-pot, it is quite easy to accidentally turn an other one. I printed the page from the manual, mark the pots there with large red letters and oriented the paper the same way the scope was sitting on my bench. Also, the pot layout is different depending on the serial number!
4 - follow exactly the manual. The initial setting on the CDD gains is a fiddly procedure, and if not done properly, the scope will fail the self cal. There can be as simple mistakes as forgetting to save the value after setting. Make sure you do not accidentally skip a step in the procedure. This happened to me a couple of times.
5 - on higher serial scopes, there is a separate step for adjustment of the 100ns clock, which needs a 300MHz generator. I did not have that, so I tried it with lower frequency aliased waveform, it was possible to do the alignment, but probably not to the highest precision, since the waveform was not stretched enough to be easily visible.
6- once I had the experience, it still took me about an hour of concentration to do the full calibration so make sure you have a glass of beer after this. You'll need it.

BTW. Are you looking at the noise in repet mode when you have a triggered waveform, right? My experience is repet will not give you useful results if just looking at a flat GND line or having a marginally triggered waveform. Which is kind of obvious, given the way this mode works.

Szabolcs

Re: older 1720 vectorscope service manual?

Michael A. Terrell
 

I have a 1720, SN B029603, with a dead power supply. The schematic for later production units is nothing like the earlier power supply boards.

-----Original Message-----
From: "Chris Elmquist chrise@... [TekScopes]" <TekScopes@...>
Sent: Jan 22, 2017 3:21 PM
To: TekScopes@...
Subject: [TekScopes] older 1720 vectorscope service manual?

I am looking for a service manual for a low serial number (B012019)
Tek 1720 VectorScope.

Online sources seem to be for S/N B060000 or later and I find that the
design on my older unit is quite different-- particularly in the power
supply where I think I'm having trouble (no light from the CRT, anode
is at about 5KV instead of 11KV and filament is at about 1V).

Anybody have a manual for one of these floating around or suggestions
for likely failures in this (older) unit?

Thanks.

Chris
--
Chris Elmquist

Michael A. Terrell

Re: 2465B front cover wanted

 

"benx618(g)" <benx618@...> sent me the following response since
he is having trouble logging into the Yahoo web interface.

On Sun, 22 Jan 2017 13:09:20 -0800, you wrote:

Hi David,

Since yahoo won't allow me to login at the moment I will email you.
I have researched a bit on Tek 2465 cover info to share.
You are welcome to post any useful info to the group if you wish, of course.

2465B cover note:
My understanding is that the late (indented style) cover was a reinforced redesign suitable not only for the later 2465 series (post hinged latch style) but especially for the 2467 which needs greater depth to clear the protruding CRT configuration.

2465 cover (hinged latch version) available for Trade fyi:
I have the original latching style cover (200-2742-00). I would trade for either of these needed covers:
-> 2465A/B & 2467 indented style cover (200-3199) ( 6" height)
-> 2465 option 1 DMM style (hinged latch, 7.5" height) (200-2844-00).

All for now,

Ben
On Sun, 22 Jan 2017 07:02:20 -0800, you wrote:

If you compare the 200-3199-00 and -01 versions of the front covers,
they are functionally the same so they fit the oscilloscopes the same.
As has been reported, the last digit of the part number has a much
different look to it, and if you get the light just right you can see
that the last digit was a "0" and that it was changed to a "1".
My guess is that the mold used the same -00 half of the mold to make the
inside of the -01 parts, and that the outside half of the mold was
replaced for some reason. There are some small dimensional differences
on the exterior of the cover which do not matter to the look, the fit or
the function of the plastic cover.

tom jobe...

On 1/21/2017 11:37 PM, machine guy @Mac [TekScopes] wrote:

OK, I looked again, much harder. My 2465A cover does have a part
number molded into the inside of the cover. Mine is faint, but clearly
there, PN 200-3199-01. The 1 of 01 is more pronounced than the rest
of the number. Sorry for the confusion but I just didn't see it earlier.

From: "David @DWH [TekScopes]" <TekScopes@...>
To: TekScopes@...
Sent: Saturday, January 21, 2017 8:57 PM
Subject: Re: [TekScopes] 2465B front cover wanted

I never noticed (or I forgot) that both of mine which have the large
square centered indentation have a Tektronix part number on the
inside. They are both marked 200-3199-01 instead of the 200-3199-00
you reported and they fit perfectly on my 2445B and 2440.

I wonder if the differences between the 2465B covers with the large
square centered indentation and the ones shown on Ebay are more than
cosmetic. Or maybe the Ebay ones will not fit at all and are for a
different set of instruments.