Tektronix TDS3032 Forgets it’s GPIB address setting

[daveyk]

 Hello all,

My Tektronix TDS3032 Forgets it’s GPIB address setting when powered off and back on.

I do believe it’s retaining other settings. Typically, I had it set at GPIB address 2.  I’ve had to reconfigure my test software to use address 1.

I just hope this is a minor annoyance, and not the start of a more major failure.  The scopes interface module has the RS232, GPIB, and VGA option. I’m not at the scope to get the model number of the I/O module, but I think there only two versions, the other one lacking the VGA.

Right now I can’t afford to have an “oops”, so I did not take it apart yet. I wonder if that interface module has a backup battery on it? I don’t recollect the scopes having a backup battery and the scope retains its last settings. I don’t want to void the scope cert by opening it up. That shouldn’t be an issue with the I/O module.

Are schematics available for these instruments. It seems so hard to obtain schematics anymore as manufacturers are doing their best to keep most service dollars in their own pockets.

Thanks kindly to all and especially to anyone that may have service experience with these scopes.

Take care,

Dave

[Jwalling]

Does the scope keep accurate time when unplugged?
Are you running the latest firmware? (3.41)

[daveyk]

Thank you fore the reply.  I don't know that I have the latest firmware in it, I doubt it.  It did keep the address for the last year.  It just started loosing it in the last month.

I should get the latest firmware and load it on floppies (I think it takes about four) and update it, but then again, if it ain't broke, don't fix it.  When I use OpenChoice Desktop to get screen images, they are the full color images as on the screen.  I know with the newer firmware the screen captures are more singular see-thru color captures (hard to explain, they are just different).  I quite like the way they look with this scope and firmware.  I would prefer not to update.  Again, everything I use it for works perfect.  It just started reverting back to GPIB address "1" when powered up.  The time/date and settings are fine.

I am thinking it is inside the I/O module, but I saw pics of one taken apart and I don't see any batteries in it.

Dave

[daveyk]

Does the scope keep accurate time when unplugged?
Are you running the latest firmware? (3.41)

Hello again Jay,
I just checked, the time and date were incorrect.  I reset it and turned the scope off, waited a bit and turned it back on and the time a date were still correct.  So it takes a little bit for it to loose its settings.

Is it a damn Dallas DRAM?  I hates those things.  It seems to hold its calibration as far as I can tell.  Hopefully those are held in non-volitle flash ram (I2c ROM?).

For compatability reasons, I am thinking of buying a backup TDS3032 from Alltest in New Jersey.

I could try a Siglent, but I will have to write software routines to support it. Not a big deal but that does take time, so it is a deal of some sort.  Also, I can not get an answer from Siglent if you can programmically set the pulse width reference points from 50% to 10%.  Either that, or I just have to breakdown and buy a used TDS3032C/TDS3034C used.  They are still expensive used.  I may still buy the Siglent to play with; we'll see.

Dave

[Jwalling]

Does the scope keep accurate time when unplugged?
Are you running the latest firmware? (3.41)

Hello again Jay,
I just checked, the time and date were incorrect.  I reset it and turned the scope off, waited a bit and turned it back on and the time a date were still correct.  So it takes a little bit for it to loose its settings.

Is it a damn Dallas DRAM?  I hates those things.  It seems to hold its calibration as far as I can tell.  Hopefully those are held in non-volitle flash ram (I2c ROM?).

Yep, sounds like the battery is dying. And yes, another obsolete Dallas part as well. Part is a 3.3VDC DS1742W-150. They still sell the 5VDC part I think. I hate Dallas/Maxim with a passion, and have ranted about them in the past on these forums. The calibration is not stored in it, but SPC, setups, etc. are.

[tautech]

I could try a Siglent, but I will have to write software routines to support it. Not a big deal but that does take time, so it is a deal of some sort.  Also, I can not get an answer from Siglent if you can programmically set the pulse width reference points from 50% to 10%.  Either that, or I just have to breakdown and buy a used TDS3032C/TDS3034C used.  They are still expensive used.  I may still buy the Siglent to play with; we'll see.

Dave

Who in Siglent did you contact ? Jason the application support engineer in Ohio would be your best contact.

From what you've written above I'm not sure exactly what you want to do.
Can you elaborate please.
Also which model Siglent are you looking at ?

[daveyk]

Who in Siglent did you contact ? Jason the application support engineer in Ohio would be your best contact.

From what you've written above I'm not sure exactly what you want to do.
Can you elaborate please.
Also which model Siglent are you looking at ?
[/quote]

Yup Jason Chonko.  He did reply "Our scopes do not have adjustable risetime measurement thresholds at this time. We haven’t received this request very often, but I can ask about it to see if it is a feature that could be added in a future firmware revision. I wonder why there aren’t more scopes with this flexibility?"

That's interesting.  I did show him an exact picture of what I am talking about (picture attached to this message), so there shouldn't be a misunderstanding.

Anyway, back to my old, old TDS3032 with the Dallas D-RAM.  Is it socketed or soldered-in?  I ask in case I can find one.  If it is soldered in, it may not be worth the risk.

Dave

[daveyk]

"Yep, sounds like the battery is dying. And yes, another obsolete Dallas part as well. Part is a 3.3VDC DS1742W-150."

You can find them on EBAY, but they are probably as old as the one that is in it already 8-(

"SPC, setups, etc. are"

What so you mean by SPC?  Setups are of no concern.  I load setups programically through the GPIB port.

dave

[tautech]

Who in Siglent did you contact ? Jason the application support engineer in Ohio would be your best contact.

From what you've written above I'm not sure exactly what you want to do.
Can you elaborate please.
Also which model Siglent are you looking at ?

Who in Siglent did you contact ? Jason the application support engineer in Ohio would be your best contact.

From what you've written above I'm not sure exactly what you want to do.
Can you elaborate please.
Also which model Siglent are you looking at ?

Yup Jason Chonko.  He did reply "Our scopes do not have adjustable risetime measurement thresholds at this time. We haven’t received this request very often, but I can ask about it to see if it is a feature that could be added in a future firmware revision. I wonder why there aren’t more scopes with this flexibility?"

That's interesting.  I did show him an exact picture of what I am talking about (picture attached to this message), so there shouldn't be a misunderstanding.

Dave

Yes I think there is.

Ok some reference level but a you want a 50-10% setting for what ?   

You said:
Also, I can not get an answer from Siglent if you can programmically set the pulse width reference points from 50% to 10%

For measurements ? It's not clear just what you need ?

[daveyk]

If you looked at the picture I attached of my TDS3032, you can see for what I need, I measure pulse width at the 90% points on the pulse. Since it is a negative going pulse (about -300volts), the pulse width is measured from when the pulse goes down 10% and then returns back up to 10%. On a scope the means the mid-reference for pulse width readings are both at 90%. Look at the cursors in that picture. That’s measuring it manually. It is close to the automatic pulse width readings with the mid referenced set to 90%.

It’s just the way Ultrasonic pulsers are specified and measured.

Dave

[tautech]

Grabbed a couple of screenshot examples doing what I think you need.
These are from a SDS1202X-E using the standard inbuilt measurements and cursor measurement.
AWG Pulse width = 40ns.
Cursors set to 20% graticule that's not seen as brightness is set a little low.

The Time@Level is almost what you're looking for; time from Trigger level setting but measured to 50% of the rising edge.

I think I'll ask Siglent what's the use of it when an adjustable level time measurement could be more useful especially for your use case. Wonder if they'll consider changing it or adding another...............

[Jwalling]

"Yep, sounds like the battery is dying. And yes, another obsolete Dallas part as well. Part is a 3.3VDC DS1742W-150."

You can find them on EBAY, but they are probably as old as the one that is in it already 8-(

"SPC, setups, etc. are"

What so you mean by SPC?  Setups are of no concern.  I load setups programically through the GPIB port.

dave

SPC = signal path compensation. It corrects for DC offsets.

EDIT: The device is not in a socket.

[daveyk]

"Cursors set to 20% graticule that's not seen as brightness is set a little low.

The Time@Level is almost what you're looking for; time from Trigger level setting but measured to 50% of the rising edge."

Yes, I can manually use the cursors in certain situations, but you can't do that in automated testing when you just want the pulse width at the 10% down levels (that's 90% levels on a Tektronix as they think from the bottom up).

The Time@Level seems to be producing the correct result,, but that must be a coincidence considering your description of it?

Wow, what beautiful screen images the scope has!

I did get a response from Chuck that I need to study more. He mentioned a method of programmically setting the cursors at the correct levels (not sure I understand how, but I need to study his email) and then reading the time between the cursors.  It's looks like a heck of a work-around, when a simple width function with settable reference levels is what is called for.  That is why those Tek TDS3000 scopes are so expensive when new?  They are ridiculously high priced and look behind the times, but they do the job very well and last 20 years or more.

I can get another used TDS3032 as a backup for $1500 with a NIST traceable calibration and a six month warranty from Alltest in New Jersey, and I am seriously considering that right now.  I love the looks of Siglent and 200MHz BW for $400?! That's sounds insane for what all you get.  I'm pretty sure that is good enough for what I need.  I only ever use two channels.  The 20 year old TDS3032 is 2.5GS/S where the Siglent is only 1GS/S.  I'm not sure what affect that has in the long run.  I rarely ever rely on a one shot result.  I am typically running in normal free-run mode, or have the acquisition mode set for a 2 or 4 average anyway.

I do reply on P-P readings and expect/hope the channels match each other.  I know with some scopes, or most, you can have 2 or 4 channels and none of the channels agree with each other on P-P, RT, PW, etc.,. Sometimes I think calibrated scopes are a joke and get stickered each year just to keep industry (customers) and QA managers happy.   You can also have 10 Tektronix 10:1 probes, all compensated the same, and they all give you 10 different PW readings on the same scope and channel.  When you are checking a PW (at 10% down) that is supposed to be 20ns wide +/-10%, it doesn't take much of a scope probe variance to throw your reading out of tolerance. Personally, I prefer a calibrated 100:1 probe, which I think does limit the BW to ~150MHz anyway).  The company I worked for, would make their own 100:1 probe and adjust it/calibrate it on an HP Spectrum analyzer, and then base their specifications on that probe being used and a 300MHz TDS3032 scope. A lot of customers, and outside techs, figured out that they could get the sames results with a 40dB attenuator. Anyway, I definitely digressed....

I may just have to buy the Siglent $400 scope just to see if I can adapt it to my work.  A nice bright large screen sure looks nice compared to the tired old TDS3032 screens.  There's also to "too cool" factor, but it has to be able to do the job, or it is just a cool toy in the shop.

Dave

[daveyk]

"SPC = signal path compensation. It corrects for DC offsets."

Ah, yes, and that is something I always keep an eye one and I do run the compensation routine before each job, especially if setting up in a different environment (on the road).

It does appear to be holding that SPC calibration, and the date only loosed time that the scope is powered off. If it's off for 12 hours, the RTC gets 12 hours behind.  It was always loosing its GPIB address.  I can look at adding a feature to my software to reset the scope's time every time it connects to it (assuming that is programmable).

I re-configured the software for address "1" and there is no other equipment needing address "1", unless something else start loosing it memory and wanting that address. Please, please, Test Equipment Gods, do not let that happen.  I need what hair I have left.

Dave

[daveyk]

[/quote]

Yep, sounds like the battery is dying. And yes, another obsolete Dallas part as well. Part is a 3.3VDC DS1742W-150. They still sell the 5VDC part I think. I hate Dallas/Maxim with a passion, and have ranted about them in the past on these forums. The calibration is not stored in it, but SPC, setups, etc. are.
[/quote]

Hey, sorry to revisit this again. I just order one “brand new”, from China, $38.29/shipping. I am supposed to have it sometime next month  (Nov ‘18). I’ll have to post if I’m successful . It looks like a lot of work to pull the board out to replace the chip. Hopefully, I don’t break it. In September, I did buy another used TDS3032, from Alltest, in NJ, for $1500 with a NIST cert, and a six month warranty. It works perfect and matches my tired one perfectly.

I may just buy that Siglent we were talking about for what about $400/NIST cert to play with and see if I can get a software work-around for the pulse width measurement reference levels. It is probably fine for the rest of my data collection work. It is so good looking and sexy for the price - lol.

Dave

[tautech]

Hey, sorry to revisit this again. I just order one “brand new”, from China, $38.29/shipping. I am supposed to have it sometime next month  (Nov ‘18). I’ll have to post if I’m successful . It looks like a lot of work to pull the board out to replace the chip. Hopefully, I don’t break it. In September, I did buy another used TDS3032, from Alltest, in NJ, for $1500 with a NIST cert, and a six month warranty. It works perfect and matches my tired one perfectly.

I may just buy that Siglent we were talking about for what about $400/NIST cert to play with and see if I can get a software work-around for the pulse width measurement reference levels. It is probably fine for the rest of my data collection work. It is so good looking and sexy for the price - lol.

Dave

I should revisit your measurement type as I had some detailed discussion with Tech support about the Tek measurement type and there's been some FW updates lately that Siglent may have changed their T@L result.
Got some scope stuff to do this weekend so I'll check on this for you.

Edit
So after checks......nothing changed. 

[daveyk]

Hello again.  I order the battery RAM chip.  It took around 6 weeks to get.  I tore the scope down and carefully replaced it and examined the PCB.  no damaged traces (used my motorized de-soldering gun).  Job looked real good.  While it was apart, I replaced the back-light tubes ( and instrument I work on uses the same exact display ), so give some new brightness life to the display.

Put it all back together and the scope doesn't finish booting now.  The Tektronix splash screen is on the LCD, looking brighter than ever, but the a-scan does not come up nor any of the menus.  It locks up and doesn't respond to front panel controls.  Is there a button combination to do a power-on reset?

I just tried making four firmware boot floppies.  It seeks the floppy drive upon power up but never prompts to update firmware.

Now, I'm starting to panic.

Dave

[Tim5000]

Have you seen this - it may or may not help?
Looking at the part number maybe it will work for yours too...I don't have one so I cannot tell you if it will wipe something important!

TDS3034 - Suddenly always stuck at splash screen: https://forum.tek.com/viewtopic.php?t=137905

"
You could try a "hard" reboot. The process is: Start with scope powered off 1. Press and hold B Trigger button 2. Power on scope and continued to hold in B Trigger 3. Continue to hold B Trigger till unit boots or about 1 1/2 min. If the scope does not boot it will need repair.
"

Maybe worth a go, or even registering on that site and asking them (if it's still active).

[daveyk]

“"
You could try a "hard" reboot. The process is: Start with scope powered off 1. Press and hold B Trigger button 2. Power on scope and continued to hold in B Trigger 3. Continue to hold B Trigger till unit boots or about 1 1/2 min. If the scope does not boot it will need repair.
"

Maybe worth a go, or even registering on that site and asking them (if it's still active).”

Dang I wish I would have seen this earlier this afternoon. I already pulled the new chip back off and put the old chip back on.

Dave

[daveyk]

UPDATE: 27 November 2018
Replaced the BATT RAM with one I purchased out of Florida.  Ziff Sockets would not fit.  The correct pin sockets I ordered are still stuck in Texas.  I gave up and installed 24 lead carrier pins (old fashion).  I installed the "new" DS1742W-150 chip.  I partially re-assembled.  The scope booted on the first try!  The background was all read.  So I re-booted holding in the "B Trig" button to re-set it.  Got the normal blue background back.  I entered the time and date.  I ran the system tests and everything passed except the floppy drive (which is not yet attached and I don't think works anyway).  It is going through SPC now; looking good!

Now, I may Dremmel in to my old chip and try to get to the battery connections and be able to mount an external battery when this "new" batt ram decides to not work.

Dave