Tektronix New 3/4-Series Scope (June 4th!)

[nctnico]

I'm not saying it should be a cheap as the usual Chinese OEMs, but I'd pay 3-5k for a 1 GHz digital scope from Tek/Keysight/whatever that's got a decent set of options included in that price. Otherwise, I'm perfectly happy buying older gear.

A reasonable (relatively bug free) 1GHz scope from the Chinese brands will also set you back $8k to $15k.

[Brumby]

I already clarified the typo in the very next post after Terra's as a reply to his, which you quoted. If you would have scrolled another inch you would have seen it. Or you seen it, and are just bored so decided to troll me. Either way I can't believe you took time out of your day to point out word missing an "s"   

You obviously have no idea of how I conduct myself here - but please don't roll the "troll" label out so quickly.

Here is the truth of the matter:  I missed that line of your post.
Here is my response to my error:  I apologise.  You did, indeed, correct your mistake.

[porker1972]

.... and designed from the feedback of interviewing 100 scope users.

They only asked 100 people? If that number is accurate, this will be fun to watch.

I think it's hundreds:

Doesn't matter how many, it's how you ask the question and what you do with the results that matters.
Just because a majority want some feature doesn't necessarily mean it's a good idea to implement.

...and also: which customers did they ask? If it's their own customers, then it's not far off a useless exercise: they've gone through an evaluation process to buy that product and will probably buy the same brand again. If it's their competitor's customers, they might find out why the customer didn't buy their product. But even so, feature set might not be the main reason: with so many scopes now offering great performance, it's often a commodity purchase based on the best deal available at the time. Yes, past customer preference such as UI familiarity might play a part but if you're being offered a 4000-class for $5k from vendor #1, you'll have a hard time justifying the $10k-$15k equivalent from anyone else.

[porker1972]

One thing I have wanted in a scope is a selectable color palette. It appears Tek was listening when they got feedback for the 4 series. Hopefully this catches on with other scopes... more specifically in a scope I can afford 

My Lecroy LT 574 has selectable colors for UI elements and waveform traces, and I think the bigger Wavepro 900 has, too. And both scope models are over 15 years old

I think later Lecroys have selectable colors, too.

Is that really such a rare feature? 

The request for changing the colour of the traces comes from colour-blind engineers - most scopes are set with Red for Ch1 or Ch2 and Green for Ch3 or Ch4, these can't be determined. Yes, LeCroy had it some time ago, but I think it's still only on their higher-end PC-based scopes.

[Fungus]

The request for changing the colour of the traces comes from colour-blind engineers - most scopes are set with Red for Ch1 or Ch2 and Green for Ch3 or Ch4

All the 'scopes I use have Yellow, Cyan, Magenta on the first three channels - for obvious reasons, two RGB primary colors are brighter than one RGB primary color.

After that? Not much you can do. No 'scope can have eight easily distinguishable colors, not even for non-colorblind people.

[kilobyte]

And we have a price: $3,850
https://info.tek.com/www-new-gen-scope-june4-contest.html

Now whether that is base price, or optioned up we don't know 

https://info.tek.com/rs/584-WPH-840/images/en-Tektronix-New-Scope-Sweepstakes-Rules_May-2019.pdf
The Sweepstake rules from the german Tek newsletter is showing a price of 36790USD for the grand price and 24729USD for the first price

[0culus]

I'm not saying it should be a cheap as the usual Chinese OEMs, but I'd pay 3-5k for a 1 GHz digital scope from Tek/Keysight/whatever that's got a decent set of options included in that price. Otherwise, I'm perfectly happy buying older gear.

A reasonable (relatively bug free) 1GHz scope from the Chinese brands will also set you back $8k to $15k.

Sure...but I'm asking why is that so? Surely the engineering difficulties of a 1 GHz scope are pretty much solved by now. When companies have long since moved on to building ridiculously fast 50 GHz+ real time scopes at the top of their product lines, why is a lowly 1 GHz DSO so expensive still? It seems that the low end stuff that is approachable in price hasn't moved much beyond the bandwidths of old workhorses like the Tek 465/485 and the like, yet the 465/485 were NOT cheap in their day. And digital electronics in particular aren't getting any slower.

I'll give you a real-world hobbyist example: some of my interests are in FPGAs. I can clock designs at 100 MHz, and for small numbers of signals it's handy to be able to look at them on an oscilloscope for debugging. My 7904A CRO mainframe with a 7A19 amplifier and a P6201 FET probe is barely enough to make a good determination of clock integrity at 100 MHz, and that's already at least double the bandwidth of most entry level DSOs.

I guess I'm wondering ultimately, why in this world of VLSI design and such, we can't have an affordable high speed DSO that exceeds the capabilities of huge lab grade equipment of decades past.

[nctnico]

The problem is not in the hardware but the firmware of an oscilloscope has become very complicated due to all the features like decoding, measurements, statistics, etc. Nowadays oscilloscopes are more like complex data analysis systems than devices to show wiggly lines on a screen. I'm hearing good things about the Siglent SDS5000 series but that is not a cheap scope either and it seems Siglent is putting in some serious effort using beta testers.

[thm_w]

https://info.tek.com/rs/584-WPH-840/images/en-Tektronix-New-Scope-Sweepstakes-Rules_May-2019.pdf
The Sweepstake rules from the german Tek newsletter is showing a price of 36790USD for the grand price and 24729USD for the first price

Yeah its a bit odd that they'd be giving away a base model, but 24k for an optioned up model with a ton of probes/MSO probes would not be totally unreasonable. For the 5000 rigol scopes, highest end model = about 4x the price of the base model (6x in this case for tek).

Sure...but I'm asking why is that so? Surely the engineering difficulties of a 1 GHz scope are pretty much solved by now. When companies have long since moved on to building ridiculously fast 50 GHz+ real time scopes at the top of their product lines, why is a lowly 1 GHz DSO so expensive still? It seems that the low end stuff that is approachable in price hasn't moved much beyond the bandwidths of old workhorses like the Tek 465/485 and the like, yet the 465/485 were NOT cheap in their day. And digital electronics in particular aren't getting any slower.

I'll give you a real-world hobbyist example: some of my interests are in FPGAs. I can clock designs at 100 MHz, and for small numbers of signals it's handy to be able to look at them on an oscilloscope for debugging. My 7904A CRO mainframe with a 7A19 amplifier and a P6201 FET probe is barely enough to make a good determination of clock integrity at 100 MHz, and that's already at least double the bandwidth of most entry level DSOs.

I guess I'm wondering ultimately, why in this world of VLSI design and such, we can't have an affordable high speed DSO that exceeds the capabilities of huge lab grade equipment of decades past.

- They don't want to eat their market, and undercut themselves. Usually the 1GHz+ tech trickles down, and you want early adopters to pay more up front. Theoretically the Rigol MSO5000 is capable of 500MHz if not 1GHz+ in terms of hardware. Front end not sure. But no way in hell are they going to add those costs in, when they have higher end models to sell.
- Active probes, are you going to sell a 1GHz scope and throw some shitty 350MHz probes in the box? oh wait, tek MSO5000 costs $18k and includes 4 500MHz passive probes 
- Probe detect/power connections, traditionally this separates lower from higher end oscilloscopes. As soon as some low end parts start adding it, everyone will consider it (which adds $$$).
- Increased cost of circuit board dielectric (possibly).

I'm not sure what point you are trying to make with the 465, it cost $1700 in 1973 which is about $10,000 today. Can you get a better scope for $10,000 these days? yes.

[David Hess]

Surely the engineering for a 1 GHz DSO is a solved problem that could be easily spun up in lower cost models now.

Go check the cost of integrated ADCs with a minimum bandwidth of 1 GHz and a sample rate suitable for real time acquisitions.  At TI, they start at $212.  Now check the cost of integrated parts for the vertical amplifier chain and none of them will support 1 GHz operation.  They might work to 350 or 500 MHz.

So higher performance is still going to require a custom design.

I'm not saying it should be a cheap as the usual Chinese OEMs, but I'd pay 3-5k for a 1 GHz digital scope from Tek/Keysight/whatever that's got a decent set of options included in that price. Otherwise, I'm perfectly happy buying older gear.

What they could make for a lower cost is a sampling oscilloscope with a bandwidth of 4 to 8 GHz using standard surface mount construction but apparently there is not enough demand for such a thing.

[snoopy]

I'm not saying it should be a cheap as the usual Chinese OEMs, but I'd pay 3-5k for a 1 GHz digital scope from Tek/Keysight/whatever that's got a decent set of options included in that price. Otherwise, I'm perfectly happy buying older gear.

A reasonable (relatively bug free) 1GHz scope from the Chinese brands will also set you back $8k to $15k.

Sure...but I'm asking why is that so? Surely the engineering difficulties of a 1 GHz scope are pretty much solved by now. When companies have long since moved on to building ridiculously fast 50 GHz+ real time scopes at the top of their product lines, why is a lowly 1 GHz DSO so expensive still? It seems that the low end stuff that is approachable in price hasn't moved much beyond the bandwidths of old workhorses like the Tek 465/485 and the like, yet the 465/485 were NOT cheap in their day. And digital electronics in particular aren't getting any slower.

I'll give you a real-world hobbyist example: some of my interests are in FPGAs. I can clock designs at 100 MHz, and for small numbers of signals it's handy to be able to look at them on an oscilloscope for debugging. My 7904A CRO mainframe with a 7A19 amplifier and a P6201 FET probe is barely enough to make a good determination of clock integrity at 100 MHz, and that's already at least double the bandwidth of most entry level DSOs.

I guess I'm wondering ultimately, why in this world of VLSI design and such, we can't have an affordable high speed DSO that exceeds the capabilities of huge lab grade equipment of decades past.

You'll find the probes that handle the higher bandwidth are sometimes worth as much or more than the scope itself and you might say that there is nothing much inside those probes so why are they so expensive ?

And to measure a 100MHz clock accurately you can forget about anything below 1GHz bandwidth plus throw in a 1.5GHz active probe

cheers

[0culus]

https://info.tek.com/rs/584-WPH-840/images/en-Tektronix-New-Scope-Sweepstakes-Rules_May-2019.pdf
The Sweepstake rules from the german Tek newsletter is showing a price of 36790USD for the grand price and 24729USD for the first price

Yeah its a bit odd that they'd be giving away a base model, but 24k for an optioned up model with a ton of probes/MSO probes would not be totally unreasonable. For the 5000 rigol scopes, highest end model = about 4x the price of the base model (6x in this case for tek).

Sure...but I'm asking why is that so? Surely the engineering difficulties of a 1 GHz scope are pretty much solved by now. When companies have long since moved on to building ridiculously fast 50 GHz+ real time scopes at the top of their product lines, why is a lowly 1 GHz DSO so expensive still? It seems that the low end stuff that is approachable in price hasn't moved much beyond the bandwidths of old workhorses like the Tek 465/485 and the like, yet the 465/485 were NOT cheap in their day. And digital electronics in particular aren't getting any slower.

I'll give you a real-world hobbyist example: some of my interests are in FPGAs. I can clock designs at 100 MHz, and for small numbers of signals it's handy to be able to look at them on an oscilloscope for debugging. My 7904A CRO mainframe with a 7A19 amplifier and a P6201 FET probe is barely enough to make a good determination of clock integrity at 100 MHz, and that's already at least double the bandwidth of most entry level DSOs.

I guess I'm wondering ultimately, why in this world of VLSI design and such, we can't have an affordable high speed DSO that exceeds the capabilities of huge lab grade equipment of decades past.

- They don't want to eat their market, and undercut themselves. Usually the 1GHz+ tech trickles down, and you want early adopters to pay more up front. Theoretically the Rigol MSO5000 is capable of 500MHz if not 1GHz+ in terms of hardware. Front end not sure. But no way in hell are they going to add those costs in, when they have higher end models to sell.
- Active probes, are you going to sell a 1GHz scope and throw some shitty 350MHz probes in the box? oh wait, tek MSO5000 costs $18k and includes 4 500MHz passive probes 
- Probe detect/power connections, traditionally this separates lower from higher end oscilloscopes. As soon as some low end parts start adding it, everyone will consider it (which adds $$$).
- Increased cost of circuit board dielectric (possibly).

I'm not sure what point you are trying to make with the 465, it cost $1700 in 1973 which is about $10,000 today. Can you get a better scope for $10,000 these days? yes.

I guess the 485 is more applicable.

The problem is not in the hardware but the firmware of an oscilloscope has become very complicated due to all the features like decoding, measurements, statistics, etc. Nowadays oscilloscopes are more like complex data analysis systems than devices to show wiggly lines on a screen. I'm hearing good things about the Siglent SDS5000 series but that is not a cheap scope either and it seems Siglent is putting in some serious effort using beta testers.

Fair enough.

Surely the engineering for a 1 GHz DSO is a solved problem that could be easily spun up in lower cost models now.

Go check the cost of integrated ADCs with a minimum bandwidth of 1 GHz and a sample rate suitable for real time acquisitions.  At TI, they start at $212.  Now check the cost of integrated parts for the vertical amplifier chain and none of them will support 1 GHz operation.  They might work to 350 or 500 MHz.

So higher performance is still going to require a custom design.

I'm not saying it should be a cheap as the usual Chinese OEMs, but I'd pay 3-5k for a 1 GHz digital scope from Tek/Keysight/whatever that's got a decent set of options included in that price. Otherwise, I'm perfectly happy buying older gear.

What they could make for a lower cost is a sampling oscilloscope with a bandwidth of 4 to 8 GHz using standard surface mount construction but apparently there is not enough demand for such a thing.

Good points. I am a bit surprised this stuff is still so expensive. I guess high performance ADCs and related parts are still niche enough that economies of scale don't kick in. And if custom designs are still needed, that blows the hypothetical low price 1 GHz DSO out of the water.

I'm not saying it should be a cheap as the usual Chinese OEMs, but I'd pay 3-5k for a 1 GHz digital scope from Tek/Keysight/whatever that's got a decent set of options included in that price. Otherwise, I'm perfectly happy buying older gear.

A reasonable (relatively bug free) 1GHz scope from the Chinese brands will also set you back $8k to $15k.

Sure...but I'm asking why is that so? Surely the engineering difficulties of a 1 GHz scope are pretty much solved by now. When companies have long since moved on to building ridiculously fast 50 GHz+ real time scopes at the top of their product lines, why is a lowly 1 GHz DSO so expensive still? It seems that the low end stuff that is approachable in price hasn't moved much beyond the bandwidths of old workhorses like the Tek 465/485 and the like, yet the 465/485 were NOT cheap in their day. And digital electronics in particular aren't getting any slower.

I'll give you a real-world hobbyist example: some of my interests are in FPGAs. I can clock designs at 100 MHz, and for small numbers of signals it's handy to be able to look at them on an oscilloscope for debugging. My 7904A CRO mainframe with a 7A19 amplifier and a P6201 FET probe is barely enough to make a good determination of clock integrity at 100 MHz, and that's already at least double the bandwidth of most entry level DSOs.

I guess I'm wondering ultimately, why in this world of VLSI design and such, we can't have an affordable high speed DSO that exceeds the capabilities of huge lab grade equipment of decades past.

You'll find the probes that handle the higher bandwidth are sometimes worth as much or more than the scope itself and you might say that there is nothing much inside those probes so why are they so expensive ?

And to measure a 100MHz clock accurately you can forget about anything below 1GHz bandwidth plus throw in a 1.5GHz active probe

cheers

Well yeah, making a good probe isn't easy. They are all compromises in some way. My highest frequency active probes are my P6201 FET probes, which are favorably matched with the 7A19 amplifier in my 7904A, as well as with my 2465B. I also have a P6056 passive probe, theoretically good up to 3.5 GHz at a penalty of impedance, but haven't had chance to try it out yet. I got all of them for a fraction of what they cost new.

I guess for me, it will be a waiting game to see how the high performance name brand 'scopes of today (last decade or so) age. In the meantime, you'll find me using my Tek boat anchors. Hoping to find a nice 7104 to go with my 7904A. 

 

[David Hess]

Low-z passive probes would be cheap if there was enough demand for them.

[EEVblog]

https://info.tek.com/rs/584-WPH-840/images/en-Tektronix-New-Scope-Sweepstakes-Rules_May-2019.pdf
The Sweepstake rules from the german Tek newsletter is showing a price of 36790USD for the grand price and 24729USD for the first price

Yeah its a bit odd that they'd be giving away a base model, but 24k for an optioned up model with a ton of probes/MSO probes would not be totally unreasonable. For the 5000 rigol scopes, highest end model = about 4x the price of the base model (6x in this case for tek).

I think it's way to close to make the Series 4 viable against the Series 5. $3700 sounds more on the mark. Remember, they market this as suitable for "almost any budget", for me that's gotta be sub $5k for a decent optioned model.

[TheDefpom]

I did a little video on this, just has some of my opinions in it, and references this thread.

🔴 #561 New Tektronix Oscilloscope - Tektronix 4 Series MSO ?

https://youtu.be/C2-9QUDrDuQ

[Mr Nutts]

The request for changing the colour of the traces comes from colour-blind engineers - most scopes are set with Red for Ch1 or Ch2 and Green for Ch3 or Ch4, these can't be determined.

I understand 

Yes, LeCroy had it some time ago, but I think it's still only on their higher-end PC-based scopes.

I think these are also the only ones that were actually developed by Lecroy

I also think anything below the WS3000 is designed by someone else (Siglent?) 

I don't know if you can change colors on a Siglent scope

[tautech]

I don't know if you can change colors on a Siglent scope

None of the current low value models, dunno about the SDS5000X but will have a squiz at one when arrives next week..........should RTFM I guess. 

[nctnico]

The request for changing the colour of the traces comes from colour-blind engineers - most scopes are set with Red for Ch1 or Ch2 and Green for Ch3 or Ch4, these can't be determined.

I'm not quite sure about that. On the color versions of the Tektronix TDS500/TDS600/TDS700 series you can also change all the colors. It is not a new feature for Tektronix but it may be they have re-introduced it.

[snoopy]

https://info.tek.com/rs/584-WPH-840/images/en-Tektronix-New-Scope-Sweepstakes-Rules_May-2019.pdf
The Sweepstake rules from the german Tek newsletter is showing a price of 36790USD for the grand price and 24729USD for the first price

Yeah its a bit odd that they'd be giving away a base model, but 24k for an optioned up model with a ton of probes/MSO probes would not be totally unreasonable. For the 5000 rigol scopes, highest end model = about 4x the price of the base model (6x in this case for tek).

I think it's way to close to make the Series 4 viable against the Series 5. $3700 sounds more on the mark. Remember, they market this as suitable for "almost any budget", for me that's gotta be sub $5k for a decent optioned model.

It will probably be the MSO5 with a smaller screen, lower sampling rate and 200MHz, 350MHz, 500MHz and 1GHz bandwidth options. Why reinvent the wheel ?

[simone.pignatti]

here is some magic:

MSO46 4-BW-1500
Oscilloscope 6 channels 1.5GHZ

MSO46 4-BW-1000
Oscilloscope 6 channels 1.0GHZ

MSO46 4-BW-500
Oscilloscope 6 channels 500MHZ

MSO46 4-BW-350
Oscilloscope 6 channels 350MHZ

MSO46 4-BW-200
Oscilloscope 6 channels 200MHZ

MSO44 4-BW-1500
Oscilloscope 4 channels 1.5GHZ

MSO44 4-BW-1000
Oscilloscope 4 channels 1.0GHZ

MSO44 4-BW-500
Oscilloscope 4 channels 500MHZ

MSO44 4-BW-350
Oscilloscope 4 channels 350MHZ

MSO44 4-BW-200
Oscilloscope 4 channels 200MHZ

 

[JPortici]

No isolated channels, huh ? (six channels is cool, though)

[EEVblog]

So no more 2 channel model? Makes sense.
The price for the 1.5GHz is going to go sky high.
Will be intyeresting to see if all models are 1.5GHz, or they have two different front ends to get the cost down.

[David Hess]

So no more 2 channel model? Makes sense.

It especially makes sense because if triggering is done digitally, then an entire digitizer has to be included for the external trigger input anyway and if the vertical signal conditioning uses an integrated solution, it makes little sense to have special version for a trigger input.  At that point, all of the hardware for a complete vertical input is included anyway so it might as well be a full vertical input.  I am only surprised that this never led to three channel DSOs.

[Bud]

So guessing from the nomenclature, the scopes are bandwidth software upgradeable. Will keep some folks on this forum busy 

[LapTop006]

I wonder if it's 6analog + 16digital or the same as the 5 series.

I was hoping for 8, so far the only way I've run out of channels on my HPAK has been rail monitoring and even my simple board has more than six rails.