EEVblog Electronics Community Forum
Products => Test Equipment => Topic started by: Electro Fan on September 28, 2014, 07:11:44 pm
-
On a Tektronix 2467B there are at least a couple ways to measure voltage: #1 use the cursor knobs (which can be manually set to measure voltage), or #2 use the Measure/Volts button combination which will take an “auto” measurement and display the voltage readout but it will change the sec/div time setting (not sure why or what dictates the new sec/div time setting). Further, with method #2 if you change the sec/div time setting the voltage reading will disappear.
Net, net: is there a way to invoke an automatic voltage reading that will track (change automatically) in response to changes in the voltage inputted to the scope (and that will stay in automatic tracking mode without arbitrarily changing the sec/div time setting, but will allow the user to change the sec/div time setting without shutting off the auto voltage tracking readout)? Seems like for a top of the line product it should do this. I probably just haven't found the button/knob combination yet. Thanks
-
I was tempted to answer: what about the auto-tracking mode ?
But because I have accumulated several Tek scopes, I was not sure these series had it.
So I checked and unfortunately they don't.
Even though I reckon that the 2445/65/67 are the best portable analog scopes ever made they are lacking a few features found in their lower spec siblings like the 2245/46/47 series.
This is one of the reasons that even though I love my 2445B, I mostly use my 2247A.
The 2247 has auto-tracking on all it's auto measurements unlike the 2467B which only takes one measure (and takes ages and changes time base and attenuator settings as you noted) and even has auto-tracking cursors.
It's kind of frustrating to see that a 4X100MHz max scope like the 224x is actually more sophisticated than the flagship 2465B. Not to mention the Dallas / U800 problems that don't exists with the 224x series.
But that said, the 2465B and certainly the 2467B are in another class.
-
Thanks iDevice, That confirms what I've found. I agree the 2247A has a very thoughtful interface - hard to understand how so little of it made it into the 2467. By any chance do you know to what extent auto-tracking measurements exist on a Tektronix 485? EF
-
I agree the 2247A has a very thoughtful interface - hard to understand how so little of it made it into the 2467.
The 2465/2467 design is older than the 2245/2246/2247/2252... the CPU is a very slow MC6800, which can barely keep up with reading the knobs, let alone make fancy measurements. (The 2246 series uses an 8088.) It's too bad the 2465 series was never upgraded.
By any chance do you know to what extent auto-tracking measurements exist on a Tektronix 485?
Tek 485 has no measurements at all... it doesn't have readout.
-
Thanks iDevice, That confirms what I've found. I agree the 2247A has a very thoughtful interface - hard to understand how so little of it made it into the 2467. By any chance do you know to what extent auto-tracking measurements exist on a Tektronix 485? EF
The 485 is a somehow deviated member of the 400 series family which predates CRT readouts and automated measurements.
However that doesn't mean there isn't any measurement capability in this serie.
Any members like the 464/465/466 could be optioned with the DM44 multimeter.
This multimeter wasn't linked at all with the scope but for one notable exception: The Tek engineers had the clever idea to use the multimeter led readout and ADC to actually measure the amplitude difference of two points on the sweep generator ramp.
So by moving two highlighted points along the horizontal axis of the waveform analyzed (one point being positioned with the delay time knob and the other with the DM44 knob, the DM44 could display the duration between the two points, the calculation taking into account the time base setting by moving the decimal point and lighting the correct "range" led on the panel.
This very smart scheme allowed you to accurately measure periods, rise/fall times and even frequency by automatically displaying 1/t.
Now, my favorite model in this range is the 468 though and it's maybe the one you were thinking about.
This was one of the first hybrid scope (or maybe the first, I don't know) in a sense that it was a 465B analog scope where Tek added a digital "backpack" but again so cleverly integrated that you were able to seamlessly switch between analog and DSO.
And on this scope the DM44 option was completely converted to a DSO interface.
In analog mode you basically have a standard 465B with no readout, but when you switch on the fly to digital, you can use cursors to measure voltage and time, the resulting value being displayed on the led readout.
AND, you can even select the VT mode where both are active as the same time and coupled so that if you set one HV cursors lines crossing on the top of the waveform and the other on the bottom, you get auto-tracking voltage measurements.
The 468 was a milestone at that time and and is a masterpiece of clever engineering.
I like them so much that I have 3 of them because they are so much fun to restore/repair and even use today.
-
This multimeter wasn't linked at all with the scope but for one notable exception: The Tek engineers had the clever idea to use the multimeter led readout and ADC to actually measure the amplitude difference of two points on the sweep generator ramp.
So by moving two highlighted points along the horizontal axis of the waveform analyzed (one point being positioned with the delay time knob and the other with the DM44 knob, the DM44 could display the duration between the two points, the calculation taking into account the time base setting by moving the decimal point and lighting the correct "range" led on the panel.
This very smart scheme allowed you to accurately measure periods, rise/fall times and even frequency by automatically displaying 1/t.
The later 7000 timebases starting with the 7B85 which was introduced at the same time as the 464/DM40 and 466/DM40 do the same thing to display the delay time and delta delay time which allows the same types of measurements without reading a mechanical dial. The 2230 and 2232 use this method as well to find their sweep delay time which is displayed as part of the readout. The 2247A and I think the 2236 have an actual timer/counter for sweep delay time measurements and I think the lessor models in the 2247A series do as well.
The 7D15 universal timer/counter plug-in added this capability in a rough way to any 7000 series mainframe.
Now, my favorite model in this range is the 468 though and it's maybe the one you were thinking about.
This was one of the first hybrid scope (or maybe the first, I don't know) in a sense that it was a 465B analog scope where Tek added a digital "backpack" but again so cleverly integrated that you were able to seamlessly switch between analog and DSO.
The 7854 came out in the same year as the 468 although it uses a different method of digitizing and cannot be considered "real time" in any sense. There was an older 7704A based digital storage oscilloscope released in 1974 but it was very limited and used an external PDP computer.
This scope is quite uncommon. It appeared in the Tek catalogs for just a few years and there was no price listed for it. As far as I know they were never sold, instead Tek loaned them to important customers for evaluation. It was the first digital oscilloscope but all of the processing had to be done with a DEC PDP minicomputer which was located at the bottom of the scope cart.
- Dennis Tillman w7PF
- TekScopes@yahoogroups.com
The 468 was a milestone at that time and and is a masterpiece of clever engineering.
I like them so much that I have 3 of them because they are so much fun to restore/repair and even use today.
The 468 design lived on in the 2230 until it was replaced by the 2232.
The primary reason I went with the 2230 instead of the 468 is that it supports peak detection.
-
Thanks David, iDevice, and edavid for all the good Tek model analysis and history. Any chance you guys ever saw a scope among the 4XX or 22XX or 24XX or other series that had the vintage Tek silver masterlock-like tumbler style Time Delay Position control and also 2247A-like measurements (for voltage, rise/fall time, delay time, auto-cursors, etc.)? Thanks, EF
-
I like them so much that I have 3 of them because they are so much fun to restore/repair and even use today.
468's are neat; I have one. Sort of diverging from the original topic, but you mentioned repairing them and I haven't been able to repair mine yet - after it heats up the screen goes crazy and the neons can be seen activating, beating on it a bit or pressing on the lower left of the front panel will get it to eventually stabilize and then it'll work fine. It's hard to troubleshoot because after that point it won't happen again until it has completely cooled down, and it doesn't always happen. Since you mentioned repair have you experienced anything like this? https://www.youtube.com/watch?v=9do3yf-K25Q&list=UUgOPhIjm8qyfrQZlRGohwrQ (https://www.youtube.com/watch?v=9do3yf-K25Q&list=UUgOPhIjm8qyfrQZlRGohwrQ)
-
Thanks David, iDevice, and edavid for all the good Tek model analysis and history. Any chance you guys ever saw a scope among the 4XX or 22XX or 24XX or other series that had the vintage Tek silver masterlock-like tumbler style Time Delay Position control and also 2247A-like measurements (for voltage, rise/fall time, delay time, auto-cursors, etc.)? Thanks, EF
The Tektronix 1502 and 1503 time domain reflectometers use a similar tumbler style 10-turn potentiometer readout. I have occasionally run across oscilloscopes which had the rotary indicator replaced with a tumbler style one like you describe but while I may have seen an oscilloscope model like you describe, I do not remember where.
-
Thanks David, iDevice, and edavid for all the good Tek model analysis and history. Any chance you guys ever saw a scope among the 4XX or 22XX or 24XX or other series that had the vintage Tek silver masterlock-like tumbler style Time Delay Position control and also 2247A-like measurements (for voltage, rise/fall time, delay time, auto-cursors, etc.)? Thanks, EF
No I didn't so far, at least not in the Tek range and I guess for a reason, the only purpose of this type of knob was to be able to precisely measure the delay or delta of delays between two points to calculate timings without relying on the graticule.
But with any scope with readout, this is useless because you have even more accurate and easy to reed values on screen, so the delay or delta knobs are just rotary encoders.
Even the 4OO series scopes equipped with the DM44 were fitted with a standard large knob instead of the masterlock like one because the delay could directly be read on the led display.
The only exception to that is again the 468 because the display is never used in analog mode, so you need the knob.
In fact the 468 is actually a 465B without DM44 in analog mode.
So in a sense, the 468 would be the closest to what you want, when you switch to digital because you have both the led display and the knob.
-
I like them so much that I have 3 of them because they are so much fun to restore/repair and even use today.
468's are neat; I have one. Sort of diverging from the original topic, but you mentioned repairing them and I haven't been able to repair mine yet - after it heats up the screen goes crazy and the neons can be seen activating, beating on it a bit or pressing on the lower left of the front panel will get it to eventually stabilize and then it'll work fine. It's hard to troubleshoot because after that point it won't happen again until it has completely cooled down, and it doesn't always happen. Since you mentioned repair have you experienced anything like this? https://www.youtube.com/watch?v=9do3yf-K25Q&list=UUgOPhIjm8qyfrQZlRGohwrQ (https://www.youtube.com/watch?v=9do3yf-K25Q&list=UUgOPhIjm8qyfrQZlRGohwrQ)
Well I guess this should be easy to fix because it clearly is a power supply issue and by looking at your video it seems to be a bad solder joint or connector contact instead of a dying capacitor because it is sensible to mechanical stress.
The fact that it gets better when warming up is due to thermal expansion.
But I would check all primary PSU electrolytics anyway because a leaking cap could be the culprit by having corroded tracks and joints.
The golden rule is to check all the PSU voltages on the main board first.
Remember that all the regulated voltages are tight to the +65V (or +55V on this one, I don't remember, too many scopes ;-) so if many rails are failing at the same time it could be the +65.
For cases like you have, I use a medical hammer to trigger the problem, you know those small hammers with a soft rubber dome on each side used to knock your knee to trigger a muscular reflex, very useful.
If you have another scope, it could be used to see even small glitches in the power rails when you hammer the frame and the scope is already warm enough that the PSU is more or less stable.
When you have found the faulty rail, then you just have to trace the fault in this area.
I have seen a few cases of cracked traces, failed resistors (particularily in the +5V) on these scopes.
Good hunting...
-
I agree the 2247A has a very thoughtful interface - hard to understand how so little of it made it into the 2467.
The 2465/2467 design is older than the 2245/2246/2247/2252... the CPU is a very slow MC6800, which can barely keep up with reading the knobs, let alone make fancy measurements. (The 2246 series uses an 8088.) It's too bad the 2465 series was never upgraded.
By any chance do you know to what extent auto-tracking measurements exist on a Tektronix 485?
Tek 485 has no measurements at all... it doesn't have readout.
Actually, my 2465 sports a 6809 (or was it an 6808? I forgot) which is a slightly beefed up 6800 with some 16 bit features added. I learned to program, assembly of course, on the 6800 and 6809. Very nice little CPU.
-
Actually, my 2465 sports a 6809 (or was it an 6808? I forgot) which is a slightly beefed up 6800 with some 16 bit features added. I learned to program, assembly of course, on the 6800 and 6809. Very nice little CPU.
Sorry, it is not a 6809, it is a 6802 or 6808, which is a not beefed up at all MC6800.
-
I like them so much that I have 3 of them because they are so much fun to restore/repair and even use today.
468's are neat; I have one. Sort of diverging from the original topic, but you mentioned repairing them and I haven't been able to repair mine yet - after it heats up the screen goes crazy and the neons can be seen activating, beating on it a bit or pressing on the lower left of the front panel will get it to eventually stabilize and then it'll work fine. It's hard to troubleshoot because after that point it won't happen again until it has completely cooled down, and it doesn't always happen. Since you mentioned repair have you experienced anything like this? https://www.youtube.com/watch?v=9do3yf-K25Q&list=UUgOPhIjm8qyfrQZlRGohwrQ (https://www.youtube.com/watch?v=9do3yf-K25Q&list=UUgOPhIjm8qyfrQZlRGohwrQ)
Well I guess this should be easy to fix because it clearly is a power supply issue and by looking at your video it seems to be a bad solder joint or connector contact instead of a dying capacitor because it is sensible to mechanical stress.
The fact that it gets better when warming up is due to thermal expansion.
But I would check all primary PSU electrolytics anyway because a leaking cap could be the culprit by having corroded tracks and joints.
The golden rule is to check all the PSU voltages on the main board first.
Remember that all the regulated voltages are tight to the +65V (or +55V on this one, I don't remember, too many scopes ;-) so if many rails are failing at the same time it could be the +65.
For cases like you have, I use a medical hammer to trigger the problem, you know those small hammers with a soft rubber dome on each side used to knock your knee to trigger a muscular reflex, very useful.
If you have another scope, it could be used to see even small glitches in the power rails when you hammer the frame and the scope is already warm enough that the PSU is more or less stable.
When you have found the faulty rail, then you just have to trace the fault in this area.
I have seen a few cases of cracked traces, failed resistors (particularily in the +5V) on these scopes.
Good hunting...
I scoured pretty hard for cracks, using hammer device / cold can / microscope but never found any...but I'll look again. I think all the rails went crazy while this was happening - possibly excessive current draw rather than an open? Or maybe an open that leads to this current draw?
I have checked caps and one was on its way out, which was replaced...but it didn't fix anything.
I don't know, it's been quite a few months so I can't remember what I did, but I'll keep this info in mind when I get to working on it again, thanks :)
-
I scoured pretty hard for cracks, using hammer device / cold can / microscope but never found any...but I'll look again. I think all the rails went crazy while this was happening - possibly excessive current draw rather than an open? Or maybe an open that leads to this current draw?
Of course, it can come from a short as well, but if the whole power supply is falling at that moment, you should blow fuses. There are several secondary fuses in there.
And it should be heard as a loud humming from the transformer.
Could be interesting to monitor the digital power supply as it is completely independent, being fed by separate windings of the transformer.
If that goes down as well then it could be the transformer itself or anything in the primary.
Again, if every rails of the analog psu go down simultaneously, like I said, it could be the +65 as everything is depending from it.(except for the digital PSU of course)
I guess that if you don't see anything by examining the circuits, your only choice is to patiently trace the problem up to the source.
-
Take a very close look at that ceramic resistor network in the HV power supply. I had a 468 that had a crack in the lead on the ceramic substrate on that part. It would work until it warmed up slightly and then open. Drove me nuts tracking that one down. You might want to try tapping or pressing on it with a plastic rod.
-
Tektronix had 2 different lines with 400 series labeling.
The original, older, non-readout non-cpu scopes were in the 4xx series. Like the 465, 485, 424, 468, etc.
They had a newer TAS series, such as the TAS465, TAS475, TAS485, which support cursor measurement only. Examples attached.
-
Tektronix had 2 different lines with 400 series labeling.
The original, older, non-readout non-cpu scopes were in the 4xx series. Like the 465, 485, 424, 468, etc.
They had a newer TAS series, such as the TAS465, TAS475, TAS485, which support cursor measurement only. Examples attached.
The 468 has a CPU and supports cursor measurements of stored waveforms although it lacks CRT readout capability. It is not quite the earliest digital storage oscilloscope but it is close.
-
The 468 has a CPU and supports cursor measurements of stored waveforms although it lacks CRT readout capability. It is not quite the earliest digital storage oscilloscope but it is close.
Yep, the 468 has a CPU, (in)famous Mostek roms, System RAM, Acquisition RAM, Display RAM and lots of address decoders, I/O's, all in so lovely and highly easy to source discrete logic chips.
But the most remarkable chip is the RTW ADC, look at that chip ! (bottom left of the picture)
This is one of the first CPU based digital storage scope and why they are so respectable.
This is the kind of stuff I worked with when I started my career so I may be a bit nostalgic :)
-
Tektronix had 2 different lines with 400 series labeling.
The original, older, non-readout non-cpu scopes were in the 4xx series. Like the 465, 485, 424, 468, etc.
They had a newer TAS series, such as the TAS465, TAS475, TAS485, which support cursor measurement only. Examples attached.
The 468 has a CPU and supports cursor measurements of stored waveforms although it lacks CRT readout capability. It is not quite the earliest digital storage oscilloscope but it is close.
Yes, you are technically correct (the best kind of correct!).
I thought the 468 was a simple state machine, with no real CPU to speak of. **researches furiously** Whoops. I knew the 468 was kinda a bastard child of the series, but I thought it was simpler than it is. :-//
-
The 468 is next on my 'acquire' menu. :-+
-
I remember studying that 20 MS/s TRW flash ADC (datasheet attached) when I ran across it helping someone diagnose a digitizer problem in their 468.
The basic design of the 468 lived on in the Tektronix 2230 which replaced the TRW flash ADC with a 20 MS/s Sony sub-ranging ADC (datasheet also attached) in a 28 pin DIP intended for video applications. The design is still mostly discrete logic but uses an 8088 microprocessor instead of the 8085 in the 468, a semi-custom display controller in the form of a standard cell ASIC, and none of those cursed Mostek ROMs.
-
The 468 is next on my 'acquire' menu. :-+
Saw this, don't know if it will go for $50 or if it is fixable, but just fyi:
http://www.ebay.com/itm/Tektronix-468-100-MHz-2-Ch-Digital-Storage-Oscilloscope-Opt-2-11-Parts-Repair-/361069330323?pt=BI_Oscilloscopes&hash=item541168bf93 (http://www.ebay.com/itm/Tektronix-468-100-MHz-2-Ch-Digital-Storage-Oscilloscope-Opt-2-11-Parts-Repair-/361069330323?pt=BI_Oscilloscopes&hash=item541168bf93)
-
The 468 is next on my 'acquire' menu. :-+
Saw this, don't know if it will go for $50 or if it is fixable, but just fyi:
http://www.ebay.com/itm/Tektronix-468-100-MHz-2-Ch-Digital-Storage-Oscilloscope-Opt-2-11-Parts-Repair-/361069330323?pt=BI_Oscilloscopes&hash=item541168bf93 (http://www.ebay.com/itm/Tektronix-468-100-MHz-2-Ch-Digital-Storage-Oscilloscope-Opt-2-11-Parts-Repair-/361069330323?pt=BI_Oscilloscopes&hash=item541168bf93)
Based on the photos and described symptoms, I suspect it is repairable. It may only need to have its power supply rebuilt.
Personally I would go with a 2230 as an old combination analog and digital storage oscilloscope if only because its supports peak detection but the 468 certainly has style.
-
Hello,i'm from Czech Republic. My Tektronix 2230 had problem, 5. minuts after start, my screen not show "menu" and digital information. Work only analog. Thanks for replies.
-
Hello,i'm from Czech Republic. My Tektronix 2230 had problem, 5. minuts after start, my screen not show "menu" and digital information. Work only analog. Thanks for replies.
Welcome to the forum.
I strongly suggest you start a fresh thread in the Repair board:
https://www.eevblog.com/forum/repair/ (https://www.eevblog.com/forum/repair/)
This way it will be more easily found by those that might have similar problems.
Think carefully how you will title the thread so it is easy to find.
-
By any chance do you know to what extent auto-tracking measurements exist on a Tektronix 485?
Easy. The nearest to an auto measurement is the beam-finder button. All measurements are done by people that can count and do mental arithmetic :)
Some 4x5 scopes can have attached multimeters, which are also bodged in so they can measure time and time differences. I have a 475+dm44, and quite frankly the multimeter option is not easy to use; I prefer to use a mark one eyeball.
-
Some 4x5 scopes can have attached multimeters, which are also bodged in so they can measure time and time differences. I have a 475+dm44, and quite frankly the multimeter option is not easy to use; I prefer to use a mark one eyeball.
The DM40, DM43, and especially the DM44 significantly improve the accuracy, reliability, and ease of use of the delayed sweep delay time compared to using the mechanical dial. This is especially the case with the DM44 since it adds delta delayed sweep capability. This was not just a bodge but a deliberate design decision duplicating the capability found in the 7000 series 7B85 and 7B15 timebases.
Some later oscilloscopes like the 2236, 2247A, and I think the 2465 series in some cases took this to an extreme by measuring the delayed sweep delay time with a counter/timer.
-
Some 4x5 scopes can have attached multimeters, which are also bodged in so they can measure time and time differences. I have a 475+dm44, and quite frankly the multimeter option is not easy to use; I prefer to use a mark one eyeball.
The DM40, DM43, and especially the DM44 significantly improve the accuracy, reliability, and ease of use of the delayed sweep delay time compared to using the mechanical dial. This is especially the case with the DM44 since it adds delta delayed sweep capability. This was not just a bodge but a deliberate design decision duplicating the capability found in the 7000 series 7B85 and 7B15 timebases.
Some later oscilloscopes like the 2236, 2247A, and I think the 2465 series in some cases took this to an extreme by measuring the delayed sweep delay time with a counter/timer.
I freely acknowledge that "bodged" is unnecessarily pejorative! I'm also happy to acknowledge there might be improved resolution. However, given that the timebase accuracy is unchanged and that the DMx has additional calibration and inaccuracies, I'm unconvinced about improved accuracy.
The ease of use is a matter of personal preference; I'm not a fan, although I don't mind the 1/t facility.
-
However, given that the timebase accuracy is unchanged and that the DMx has additional calibration and inaccuracies, I'm unconvinced about improved accuracy.
The accuracy is improved by:
1. The digital readout is more accurate or at least more linear than the mechanical readout.
2. The errors contributed by the CRT and amplifier stages are removed. These were less significant when the mechanical readout was used which might explain why delta delayed capability was not used with a mechanical display; at least I do not know of any examples where this was done.
The difference is reflected in the Tektronix specifications which list different levels of horizontal accuracy depending on how the measurement is made. The difference was more significant on other oscilloscopes though; the 475 was pretty accurate anyway. Later designs took greater advantage of this.
The reason the CRT and amplifier errors are removed is that when making a delayed time or delta delayed time measurement, the CRT is used as a null device where the edges of the waveform are lined up instead of measuring the distance between them. The sweep ramp has a much better accuracy then the horizontal amplifier and CRT will support.
Slideback voltage measurements on an oscilloscope have the same advantage which is why it is trivial to make 0.1% and better voltage measurements using a differential comparator like a 7A13 on an oscilloscope which would normally only be rated for 1% accuracy or worse.
Vernier measurements benefit from the same improvement in accuracy; it is easier to line up calibrated marks than to measure the distance between them. Of course the real reason I like my vernier calipers is the same reason I like my RPN calculator; nobody will borrow them because they do not know how to use them anymore.