Date   

Re: Tektronix 2445 calibration issues

Szabolcs Szigeti
 

Hi,

Regarding the 2440 calibration. There are basically 4 levels of calibration
in this scope. First is the self cal, this compensates for any drift,
temperature etc. difference. This is completely automatic, and requires no
external sources. A properly maintained scope will self cal itself to be in
spec.
Second, there is the external calibration. This should only be needed if
the calibration is lost, for example due to the nvram battery discharging.
It requires some external DC levels, and a about 100kHz (?) square wave
with specific amplitude (60mV p-p, if I remember correctly). The service
manual calls for the PG506 generator, which definitely makes it easier, but
any DC source and some hacking with a square wave generator will work.
Third, there are the adjustments, which mostly deal with properly setting
up clocks for the CDD. That is a quite tedious process, requiring the SG503
levelled signal generator (but many other generators can be used, I have
even used a pulse generator instead). It mostly deals with turning the pots
inside and visually checking the alignment of the samples on the screen,
not hard, but easy to lose track in the process.
This should not be required in theory. In practice, it is useful to do
especially if there were any repairs or if the scope has not been adjusted
a long time ago. It can make a very noticable difference if these settings
are properly done. After doing this, you'll need ext and self cal,
obviously, as the process starts with clearing the nvram.
And finally there is the screen geometry adjustment, this is completely
independent of all the others, the scope provides test images and you have
to set up the screen geometry, etc. to properly display and align these
images. I suggest you do this, as long term drift can affect the screen
badly. This does not touch the normal calibration, it only involves turning
some pots to set up display geometry, so even if you do this, there is no
need to recalibrate.
Anyway, in summary, the 2440 is quite simple to recalibrate even if all
calibration is lost. Considering its age, I think this is a quite good DSO,
with some nice features.

Szabolcs

Dennis Chertkovsky <dchertkovsky@gmail.com> ezt írta (időpont: 2021. okt.
6., Sze, 23:37):

Mark,

You are correct about no batteries.
I read a little and found that the 2445 has no batteries, 2445A has a
discrete battery, and the 2445B has the dreaded NVRAM chips.

About recapping, I understand it is commonly done for these scopes, but is
it necessary to address my current issues? The scope does have a little bit
of common mode line current, but it’s a nonissue for the work I do (mostly
isolated things). I can recap if it is really necessary, but getting the
power supply out was a major bear of a task last time I did it, and I’d
rather start with the cal issues.

About resistors, maybe some have drifted due to age and hot operation, but
I’m not sure where to look; there are so many resistors. Do I just use a
thermal camera?

How is the 2440 treating you? I read that they can self calibrate, which I
would really want to have when my only direct references at home home are a
$5 microcontroller board, a few out of cal DMMs, and a very out of cal
ovenized power supply.

Sincerely,
Dennis Chertkovsky
dchertkovsky@gmail.com

On Oct 6, 2021, at 5:19 PM, Mark Vincent <orangeglowaudio@gmail.com>
wrote:

Dennis,

The plain version of this does not have the integrated battery. The B
version likely does. See if you have a battery in a plastic case on the
side board. If so, get: 667-TL5903P (Mouser part number). See if you have
surface mount electrolytics on that board. If so, replace them with radial
leaded electrolytics or tantatums. The rest of the scope should have new
electrolytics. The electrolytics should be something like Nichicon ULD and
LGR for the two main filters. Raising the amount of capacitance in the
power supply is normal by us on this group (an example is 180mfd to
330mfd). If yours has 4,7mfd tantalums in the power supply, film types
would be a better replacement. I have recapped my 2465/B and 2440 this way.
You may need to power cycle the scope a couple of times after replacing the
condensers and battery. The 2440 of mine had fail modes a couple of times
then it is right. I did put in the voltages it wanted to calibrate it.
There are some resistors that should be raised in wattage, if it is like
the 2465 series.

Mark







Re: DIY Frame for SG5010 etc ???

SCMenasian
 

If you check the schematics for the some TM50x mainframes, you will see a dot a one end of each 25 VAC and the 17.5 VAC transformer winding. These dots mark the polarity for phasing and a sufficiently sensitive ohmmeter should allow you to trace the dotted ends to the interface connector; however, you probably don't have to do this. The dotted ends all go to the 'A' side of the interface connectors. Note that not all schematics show the dots (My TM6506 does; the TM504 doesn't).

Stephen


Re: DIY Frame for SG5010 etc ???

Vintage Test
 

Ok David,

Let me know if you still need any bits. In the meantime, if anyone else would like any parts of a TM5006 - there’s virtually a whole unit, apart from the mains transformer and a few pieces of trim and the guide rails. PM me if you want any of it - the unit was totally functional and was used daily before the tx melted!

Cheers,
Mel

--
you can never have enough oscilloscopes, DMMs, valve testers or soldering irons . . .


Re: Tektronix 2445 calibration issues

Roger Evans
 

Dennis,

This is mainly a remark about DIY time mark generators, but it may have some relevance to your mixed results with the Raspberry Pi. I don't know what hardware the Pi uses to drive its GPIO pins but it is probably something like 3.3V CMOS levels. If you want clean signals into a scope for time calibration you need to match the digital output into a 50R coax line with something like a 180R and 51R divider and set the scope to 50R input impedance. Forgive me if you already know that!

Since you are clearly used to low level device programming you could get one of the Si5351a boards that sell for a few dollars on eBay and use that to generate the time marks you need with the benefits of flexibility and three 50R outputs. The Si5351a is specced to 160MHz but should provide a spot frequency of 200MHz with no difficulty. If you want frequency markers to 4GHz then you can use the AD4351 synthesiser boards but they don't go low enough in frequency to satisfy more general purpose requirements!

In the context of calibrating the 2445, you don't need to replicate the waveform of the Tek time mark generators, you just need a clean edge and always line up the same polarity edges.

Regards,

Roger


Re: DIY Frame for SG5010 etc ???

David Slipper
 

I may have tracked one down :-)

Thanks anyway.
Dave

On 07/10/2021 12:07, Vintage Test via groups.io wrote:
Me too - Scotland, just outside Glasgow. Do you want any of this stuff - I can PM you with some pics if you like?

Cheers,
Mel


Re: DIY Frame for SG5010 etc ???

Vintage Test
 

Me too - Scotland, just outside Glasgow. Do you want any of this stuff - I can PM you with some pics if you like?

Cheers,
Mel

--
you can never have enough oscilloscopes, DMMs, valve testers or soldering irons . . .


Re: DIY Frame for SG5010 etc ???

David Slipper
 

I'm in the uk.

Get BlueMail for Android<https://bluemail.me>

On 7 Oct 2021, at 08:17, "Vintage Test via groups.io<http://groups.io>" <icloud.com<mailto:mel.purcell=<a>@groups.io target=_blank>mel.purcell=icloud.com<http://icloud.com>@groups.io> wrote:

David,

I have all of the electronics, and most of the mechanics of a TM5006 available - it’s missing only it’s mains transformer. I couldn’t find one to repair it, so and waited for another TM5006 to come up at the right price ! Where are you?

Cheers,
Mel


Re: DIY Frame for SG5010 etc ???

Vintage Test
 

David,

I have all of the electronics, and most of the mechanics of a TM5006 available - it’s missing only it’s mains transformer. I couldn’t find one to repair it, so and waited for another TM5006 to come up at the right price ! Where are you?

Cheers,
Mel

--
you can never have enough oscilloscopes, DMMs, valve testers or soldering irons . . .


Re: 2901 owners, I need your help.

Tom Lee
 

I agree with Ted. The rectifier should not ever run too hot to touch. Either there is a problem with the rectifier itself, or something else is drawing excessive current. That shouldn’t be too hard to find — it will be hot, too. If you have one of those cheap IR laser-aimed thermometer pistols, you should be able to locate it without risking a finger burn, if you want to go high-tech. But you don’t need such a gizmo — old-school sniffing about works.

Tom

Sent from my iThing, so please forgive typos and brevity.

On Oct 6, 2021, at 7:44 PM, tedbmoss@cox.net wrote:

It won't work for long if you can't touch the rectifier. One or more of the TTL devices is needing replacement.





Re: DIY Frame for SG5010 etc ???

Harvey White
 

The AC windings are often put in series.  Getting one out of phase results in far less voltage than you'd want.  The average TM500 mainframe checker should have tests available to make sure that the phases are correct (so they add).

History says that some TM500 mainframes were wired improperly. It made a difference on some plugins where the AC phases were added together to make an on-module power supply.  PG505 is one, and I suspect that the voltmeters were much the same.

There's a way that you can use opto-isolators to check phasing when using a microprocessor.  Another way with the same microprocessor would be to measure peak voltages vs. time to determine phasing.

Harvey

On 10/6/2021 7:49 PM, Jeff Dutky wrote:
here is the TM500-series plug-in interface page on the TekWiki, which will tell you what voltages are needed https://w140.com/tekwiki/wiki/TM500_Series_plug-in_interfaceT

The power modules also provide NPN and PNP series pass transistors that can be (and are) used by plug-ins, so it's just a dumb power bus. Depending on what plug-ins you want to test those pass transistors may be vital.

Finally, there is some important relationship between the different 25 VAC rails that I don't fully understand. They need to have a specific phase with respect to each other (and maybe to the 17.5 VAC rail?) or Bad Things[TM] can/will happen.

-- Jeff Dutky





Re: Tektronix 2445 calibration issues

Mark Vincent
 

Dennis,

Resistors: R710, R910 https://www.mouser.com/Passive-Components/Resistors/Film-Resistors/Metal-Film-Resistors-Through-Hole/_/N-7gz41Z1yzvvqx?P=1z0vko0Z1yzbppy&Keyword=1%25+resistor&FS=True (2 in series. Pick which two parts you want) The additional meg will not hurt. R1871 279-H4P150KFDA, R1872 279-H4P180KFDA, R1853 (2) 660-MF1/2CC2213F (put in series) , R1854 71-CCF60-332K, R1880 279-H4P430KFZA, R1991 588-SM102031006FE or 588-MOX-300001006FE, R1018, R1019 approx.100,000 ohm 1W 1-5% (bleeds off filters faster. Precision not necessary), R1020 71-CCF60-133K-E3 and 71-CCF60-137K-E3 or a single 2W 270,000 ohm, R1352, R1353 to one 588-45F7R5E (original pair of 15 ohm 1/2W resistors run hot), RT1696 parallel 10,000-13,000 ohms across thermistor to speed fan up, R1220 and R1240 check and replace if out of tolerance, R125, R225 279-H4P360RFZA , R811 279-H4P330RFCA. If you want to raise the 1/8W resistors in some places on the power supply board to 1/4W, go ahead, it is your item. Some of the resistors will give problems, especially on the high voltage board to the collectors of three transistors. The others should be raised in wattage and mounted off the board. The 270,000 ohm 1/2W resistor in the power supply should be 2W. The original is stressed hard.

Electrolytics and film: The main filter 290mfd to 390mfd Nichicon LGR. See which diameter will fit and get a pair to replace them. The 47mfd decoupling throughout Nichicon ULD, The 4,7mfd tantalims to 80-R82CC4470AA30J or 505-MKS2B044701KJC00. The caps on the A5 board can be the ULD 47mfd types The 47mfd electrolytics will be 25V. The RIFA types must be replaced! These are on the power supply schematic. The mains filters C1016, C1018 594-F339X126848MDA2B, C1020, C1051 594-F339X122248MDA2B, C1065 871-B32912A5563M. The 10mfd on the fan board can be a ULD 50V type. There are no lover voltages in that series at 10mfd. The 3.3mfd C1072 in the power supply 647-ULD2G3R3MPD.

Items listed as you progress down the schematics. Items listed are ones currently in stock. This is how I listed the Mouser part numbers.

Recapping and new resistors may cure your problems. The electrolytics I listed are low ESR, high temperature and long life. The film replacements for the three 4,7mfd tantalums is advised. These tantalums are frequently bad. Film is not polarized and have lower ESR than tantalum. The resistors I listed are 1% types that have very low resistance drift with temperature, PPM/C.

Check the diodes in the power supply. Some of the Schottky diodes can get leaky. These are CR1110, CR1113-6. If bad, replace them and use compound under the TO-220 rectifier. I use MX-4 compound.

I have done these things in my 2440 and 2465/B. The 2440 is fine now. That one was redone, calibrated with the d-c and a-c voltages according to the instructions along with the other instructions for calibration. The battery was installed last. I should have done it first to get rid of the fail messages. It took a few power cycles to get them to pass. I did it over a few days, one power cycle a day then a few days wait. I doubt it takes as many days to change. I had other work and waited to power it back on when I had the time. I made sure to oil the fans in the scopes. They were dry. I did add the above compound to the heatsinks on the CCDs in the 2440. That helped cool them better. I did have a small heatsink I added to U800 in the three scopes using the same compound. I used a thicker solid wire from the screw threads to hold the heatsink down and used another nut on the wire ends to hold it down. I bent the wires around the threads as though it would be an electrical connection even though it is not for electrical but mechanical. Some here do not believe in the U800 heatsink, some do. See if the higher voltage lead to the PDA is black/dusty. If so, I recommend taking the tube out and cleaning it off. The crt removal instructions are in the manual. I find that ammonia works well. That black stuff can be conductive causing leakage paths.

I hope this has been of help to you. I know it will be some work to get it overhauled. I find the work is worth the effort and cost of new parts. Cleaning makes it appear better visually and can prevent leakage pats and heat buildup.

Mark


Re: DIY Frame for SG5010 etc ???

 

here is the TM500-series plug-in interface page on the TekWiki, which will tell you what voltages are needed https://w140.com/tekwiki/wiki/TM500_Series_plug-in_interface

The power modules also provide NPN and PNP series pass transistors that can be (and are) used by plug-ins, so it's just a dumb power bus. Depending on what plug-ins you want to test those pass transistors may be vital.

Finally, there is some important relationship between the different 25 VAC rails that I don't fully understand. They need to have a specific phase with respect to each other (and maybe to the 17.5 VAC rail?) or Bad Things[TM] can/will happen.

-- Jeff Dutky


Re: Looking for WaveStar WSTR31 Version 1.3

n4mf.sc@...
 

Hi Hakan,
That's what the info showed on the exe. I think I acquired that from Bluefeather via FTP back in 2004.

Mitch
N4MF


Re: DIY Frame for SG5010 etc ???

Jared Cabot
 

It could be done with a bit of metal work and a couple transformers if you can find them cheaply enough, but in this case DIY may end up more expensive in the end if buying all new parts.

I have designed replacement PCB's for a TM503 mainframe that can get you staeted at least.

https://www.youtube.com/watch?v=VkNRUI_QtWo


Re: Tektronix 2445 calibration issues

Dennis Chertkovsky
 

Mark,

You are correct about no batteries.
I read a little and found that the 2445 has no batteries, 2445A has a discrete battery, and the 2445B has the dreaded NVRAM chips.

About recapping, I understand it is commonly done for these scopes, but is it necessary to address my current issues? The scope does have a little bit of common mode line current, but it’s a nonissue for the work I do (mostly isolated things). I can recap if it is really necessary, but getting the power supply out was a major bear of a task last time I did it, and I’d rather start with the cal issues.

About resistors, maybe some have drifted due to age and hot operation, but I’m not sure where to look; there are so many resistors. Do I just use a thermal camera?

How is the 2440 treating you? I read that they can self calibrate, which I would really want to have when my only direct references at home home are a $5 microcontroller board, a few out of cal DMMs, and a very out of cal ovenized power supply.

Sincerely,
Dennis Chertkovsky
dchertkovsky@gmail.com

On Oct 6, 2021, at 5:19 PM, Mark Vincent <orangeglowaudio@gmail.com> wrote:

Dennis,

The plain version of this does not have the integrated battery. The B version likely does. See if you have a battery in a plastic case on the side board. If so, get: 667-TL5903P (Mouser part number). See if you have surface mount electrolytics on that board. If so, replace them with radial leaded electrolytics or tantatums. The rest of the scope should have new electrolytics. The electrolytics should be something like Nichicon ULD and LGR for the two main filters. Raising the amount of capacitance in the power supply is normal by us on this group (an example is 180mfd to 330mfd). If yours has 4,7mfd tantalums in the power supply, film types would be a better replacement. I have recapped my 2465/B and 2440 this way. You may need to power cycle the scope a couple of times after replacing the condensers and battery. The 2440 of mine had fail modes a couple of times then it is right. I did put in the voltages it wanted to calibrate it. There are some resistors that should be raised in wattage, if it is like the 2465 series.

Mark



Re: Tektronix 2445 calibration issues

Mark Vincent
 

Dennis,

The plain version of this does not have the integrated battery. The B version likely does. See if you have a battery in a plastic case on the side board. If so, get: 667-TL5903P (Mouser part number). See if you have surface mount electrolytics on that board. If so, replace them with radial leaded electrolytics or tantatums. The rest of the scope should have new electrolytics. The electrolytics should be something like Nichicon ULD and LGR for the two main filters. Raising the amount of capacitance in the power supply is normal by us on this group (an example is 180mfd to 330mfd). If yours has 4,7mfd tantalums in the power supply, film types would be a better replacement. I have recapped my 2465/B and 2440 this way. You may need to power cycle the scope a couple of times after replacing the condensers and battery. The 2440 of mine had fail modes a couple of times then it is right. I did put in the voltages it wanted to calibrate it. There are some resistors that should be raised in wattage, if it is like the 2465 series.

Mark


Tektronix 2445 calibration issues

Dennis Chertkovsky
 

Hello all,

Short questions:
Can I make my 2445 think it is calibrated without the whole cal procedure?
Why would DC balance fail?
What to fix if I am unable to calibrate at faster timebase settings?

Long story (pardon the rambling):
My 2445 initially had a weird issue yesterday when I tried to DC balance channels 1 and 2 (pulling up on both coupling levers at the same time). It did the normal screen flicker but then got stuck and showed the message “Limit 3”. Nothing I did could get it out of this, so I restarted the scope.

When it came back on, I got the dreaded “test 04 fail 10” message indicating corrupted memory, and when I bypassed it, I was greeted with a row of dots along the bottom of the screen, indicating the scope thought it was out of calibration.

I decided to try calibrate at home, and thought maybe it would be enough to just calibrate the timebase. I fashioned together a crude time mark generator using a Raspi Pico (overclocked to 200 MHz for easy math) and a DIY breadboard-BNC adapter. I wrote some code capable of generating time marks from 50 ms to 10 ns and hoped the 30 ppm crystal onboard was good enough for me. I used hardware PWM to make sure timing was as good as the crystal allowed.
Cal procedure went fine until steps 15 and 16 (50 and 10 ns time marker periods), when I simply wasn’t able to intensify the 2nd and 10th time markers and the display told me “∆ Limit”. It seemed that the timebase was running too quickly. I’ve seen this issue before, when the scope underreported the frequency of 100 MHz signals.

This rough calibration was spot on down to maybe 500 ns/div, marginal at 200 ns/div, and bad at 100 ns/div and below.
Also, it fixed the “test 04 fail 10” message, so I guess the Dallas NVRAM chips live to see another day?


What could have caused the initial DC balance issue?

Is there something that could be broken inside the scope that caused the inability to calibrate at faster timebase settings, or is it the fault of my calibration generator? I haven’t tested it with a scope at work yet, so it could have issues at higher frequencies. Additionally, at 10 ns, the peaks become more like smears than neat little marks.

Do I need to complete the entire cal procedure to get rid of the dots?

How wide are time marks supposed to be? I found myself fiddling with that setting so much that I made it a another input for my program, wide marks for coarse alignment and then finer marks for small adjustments.

And does anyone have an OCR copy of the service manual? I have the standard PDF service manual, which is an absolute pain to browse on my computer due to the weird page numbering, and it’s far too long for me to consider printing at the moment.

Thank you
Sincerely,
Dennis Chertkovsky
dchertkovsky@gmail.com

P.S.
Additionally, I checked the time mark generator against line frequency, setting trigger mode to “Line” and asking the time mark generator to make marks at a multiple of 60 Hz (3 kHz), and they moved very slowly across the screen, occasionally changing direction (line frequency varies ever so slightly, check out the UTK FNET project for more info), showing I was maybe within 0.02 Hz of line frequency, close enough.


Re: Looking for WaveStar WSTR31 Version 1.3

Andrew
 

Hey all--

I decided to make a little archive of the different WaveStar (WSTRO) 2.x versions I've found and have uploaded them to the Internet Archive.

This collection includes versions 2.5, 2.6, 2.7.4 and 2.8.1, along with release notes and manuals for each version: https://archive.org/details/WSTRO_V25-V281

I also tried to comment on what the notable differences are with each version on the archive page.

Hopefully this can help those with older scopes that are looking to run WaveStar.

Thanks


Fair market value of Tektronix 1S1, 1S2, 2B1?

Brad Thompson
 

Hello--

I have one each 1S1 and  1S2 plug-ins and one 2B1 pulser, all of which I'd like to sell. I can't locate
prices for these. The plug-ins appear to be in good condition and show no signs of
repairs or component salvage. All front-panel knobs are present. The 1S1's smaller-diameter
slightly-bent rear alignment pin will require straightening. The 2B1 is in fair cosmetic condition.

I have no way to test these and am offering them as-is and with no guarantees.

Thanks in advance for your assistance-- I'll keep estimates private so please contact me directly.

73--

Brad  AA1IP
Sent from Postbox <https://www.postbox-inc.com>


Re: 7000 series parts hulks

Michael Perkins
 

On Mon, Oct 4, 2021 at 4:15 PM Oz-in-DFW <lists@ozindfw.net> wrote:

I have a 7603, 7613, and 7633 parts hulk. All are for sale for $20 plus
whatever shipping actually costs if you can't pickup in the DFW Area. I may
need to buy a box and bubble wrap.

The plug in bays are damaged beyond hope. Teh tubes and PS's look intact
based om my casual inspection. If there is interest I'll post detailed
photos for download off my site to keep the load off the group.

I also have 7000 plugins, also pars. at least a 7A11, 7A18, and 7D14, a
test plugin of some sort and what is basically a frame with few guts.
Detailed photos later in the week. $5-$15 depending on shape.
--
Oz (in DFW) N1OZ





2941 - 2960 of 189593