Tektronix TBS1000B

[echen1024]

Tektronix has just revamped their lowest cost scopes. To be honest, not ally hat impressive. Sure, a 2ch HW counter and 34 measurements might make better than the DS1052E, but in the comparison they neglect the fact that the 1000E has long memory, and, for a starting cost of $550, go with the DS1074Z. And seriously? STILL 2.5k memory depth? I recommend Tek add at least 1-2M memory, as well as some STANDARD rudimentary serial tries and maybe decides. That, at around $450-500 would be quite a deal.

http://tektronix.gcsmarket.com/TBS1000B/?s=TBS1000BO14ENTEST&UID=3449521

[Hydrawerk]

STILL 2.5k memory depth?

It is probably still based on that old TDS220 ASICs or something like that. Hey, 2500 points of memory for a scope released in 2014 is crap.

[Hydrawerk]

http://www.tek.com/oscilloscope/tbs1000b-digital-storage-oscilloscope

[Rigby]

STILL 2.5k memory depth?

It is probably still based on that old TDS220 ASICs or something like that. Hey, 2500 points of memory for a scope released in 2014 is crap.

Yeah, but how often do you need a lot more than that?  I don't.  I mean, I'll admit, I'm a newb, but when I do serial captures I don't really use much memory.  I'm not doing a full 60,000 waveform record or anything, that's for sure, I just need a few milliseconds worth of I2C or serial data or whatever.

How often do you need the huge deep segmented memory stuff, really?  (I'm honestly asking.)

[EEVblog]

Ready to step up to the best-in-class performance
of the TBS1000B Series?

What, the class of 1995?

[mikeselectricstuff]

$1190 for a 100MHz scope with no memory to speak of - is this an early April Fool or are they just taking the  piss? 

[hans]

Wait 550$ is for 50MHz. The 70MHz is a whopping 890$.

For that money you can get a DS1074Z-S, which has 4 ch, also 1GSa/s (I bet both scopes have shared ADC's), roughly 10000x (how many orders....) the amount of memory, 30k waveforms/second + intensity graded display, serial decoding, and a 2 channel 25MHz AWG.

Tektronix, wow.. just wow.. They must have had the laziest product development meetings in the "entry level scope"-project. "Just sell the same shit again with a new UI".


Why memory depth matters IMHO... Single shot mode; capture a burst of data and view it to your hearts content.
You can capture the whole picture at once, and zoom into every detail you need. Imagine a burst packet that's too long to capture at the smallest detail you require. Then you have to pan to approximately where you want to look, single shot that area, analyze it, etc. That only works if the trigger is bang on and the signal timing stays identical, otherwise your signal jitters all over the place.
It makes life so much easier.

[EEVblog]

How often do you need the huge deep segmented memory stuff, really?  (I'm honestly asking.)

Once you feel the power of it, you use it every day and expect to have the ability to zoom into the waveform to see high frequency detail on every capture.
This is why scopes with switchable short memory (to improve the speed) annoy me, as Murphy says you'll always have it set to short memory when you need long memory.
2.5K was adequate back in 1995 when the big innovation was always-on "real time" mode which the Tek 200 series pioneered, but since then it has become a running joke.
This scope is not designed for the general market, it is designed for those corporations who have procedures in place for that scope, and just want to buy another one with no fuss. Or for the educational market who get huge discounts and once again have documentation in place for these old scopes.
Only a fool would buy it for any other reason.

[mikeselectricstuff]

STILL 2.5k memory depth?

It is probably still based on that old TDS220 ASICs or something like that. Hey, 2500 points of memory for a scope released in 2014 is crap.

Yeah, but how often do you need a lot more than that?

Almost all the time. Which is why every new scope (apart from this Tek joke) for the last 10 years or so has a substantial amount of memory.

[AutomationDude]

How often do you need the huge deep segmented memory stuff, really?  (I'm honestly asking.)

Once you feel the power of it, you use it every day and expect to have the ability to zoom into the waveform to see high frequency detail on every capture.
This is why scopes with switchable short memory (to improve the speed) annoy me, as Murphy says you'll always have it set to short memory when you need long memory.
2.5K was adequate back in 1995 when the big innovation was always-on "real time" mode which the Tek 200 series pioneered, but since then it has become a running joke.
This scope is not designed for the general market, it is designed for those corporations who have procedures in place for that scope, and just want to buy another one with no fuss. Or for the educational market who get huge discounts and once again have documentation in place for these old scopes.
Only a fool would buy it for any other reason.

Couldn't have said it better, myself

Event the DPO3000 series of scopes (which you'll pay thousands of dollars for) still use the base 2.5k of memory.  They are simply relying on their name at this point in order to move product in the lower end of the market.

[w2aew]

Couldn't have said it better, myself

Event the DPO3000 series of scopes (which you'll pay thousands of dollars for) still use the base 2.5k of memory.  They are simply relying on their name at this point in order to move product in the lower end of the market.

Not true.  The DPO3000 and MSO3000 come standard with 5M of memory on each channel.  The new MDO3000 has 10M standard on each channel.

[echen1024]

Couldn't have said it better, myself

Event the DPO3000 series of scopes (which you'll pay thousands of dollars for) still use the base 2.5k of memory.  They are simply relying on their name at this point in order to move product in the lower end of the market.

Not true.  The DPO3000 and MSO3000 come standard with 5M of memory on each channel.  The new MDO3000 has 10M standard on each channel.

I think you mean the TDS3000. But that thing is old.

[Rigby]

memory is so cheap.  once it leaves the ASIC entirely we'll hopefully start seeing good amounts of memory independent of the ASIC used in a design.  The ASIC will be capture and processing, and not storage.

[w2aew]

Couldn't have said it better, myself

Event the DPO3000 series of scopes (which you'll pay thousands of dollars for) still use the base 2.5k of memory.  They are simply relying on their name at this point in order to move product in the lower end of the market.

Not true.  The DPO3000 and MSO3000 come standard with 5M of memory on each channel.  The new MDO3000 has 10M standard on each channel.

I think you mean the TDS3000. But that thing is old.

Even the TDS3000, which IS getting old, is 10k memory. I think the intention was was the much newer DPO/MSO2000 and 3000 series were intended to replace the TDS3000, but customer demand for the TDS3000 has kept them in the portfolio.

[echen1024]

Couldn't have said it better, myself

Event the DPO3000 series of scopes (which you'll pay thousands of dollars for) still use the base 2.5k of memory.  They are simply relying on their name at this point in order to move product in the lower end of the market.

Not true.  The DPO3000 and MSO3000 come standard with 5M of memory on each channel.  The new MDO3000 has 10M standard on each channel.

I think you mean the TDS3000. But that thing is old.

Even the TDS3000, which IS getting old, is 10k memory. I think the intention was was the much newer DPO/MSO2000 and 3000 series were intended to replace the TDS3000, but customer demand for the TDS3000 has kept them in the portfolio.

Same reason why the TDS1000 series is still produced. I was introduced to oscilloscopes on that thing at 5 years old browsing Fry's with my dad. I just LOVED all the knobs.

[AutomationDude]

Couldn't have said it better, myself

Event the DPO3000 series of scopes (which you'll pay thousands of dollars for) still use the base 2.5k of memory.  They are simply relying on their name at this point in order to move product in the lower end of the market.

Not true.  The DPO3000 and MSO3000 come standard with 5M of memory on each channel.  The new MDO3000 has 10M standard on each channel.

You're right, I was thinking of the TDS3000 series of scopes... always get 'em confused (and they have 10k, not 2k5 points). 

Thanks 

[Hydrawerk]

Here I made a first impression video.

Note the funny dual counter with no traces displayed.

[tom66]

I expect they keep with the old design because they have an ASIC/frontend chip which does all that and they've made it for so long it's very cheap. The DS1074Z gets a ton of performance from a  ~$100 FPGA and $80 ADC but I'm guessing Tek probably spend less than a fifth of that on the ASIC.

Perhaps they'll sell a lot of these instruments to the educational market because of teachers that want you to "read the divs off directly" (IOW, they want you to use a scope in the most basic way possible with minimal assist from the features like deep memory, serial decode)... we had older Hameg scopes which were mixed analog-digital and I confess to frequently using the measurement facility even though I was completely capable of reading it directly, it's just easier.

The comparison chart picks the worst case for the other scopes and the best case for the Tek. The Rigol does 1Mpt of memory in all but the fastest sample rate. Who cares about more measurements if the Rigol ones are better (if they are?) The Rigol does have a "zoom function" - both dual timebase and run/stop zoom. It also samples up to 1GSa/s unless you use both channels, though the Tek does not halve sample rate with two channels, I suppose that is a benefit. And the Rigol does have limit test and a pass/fail mode too!

Who segments the market by offering 50,70,100,150 and 200 MHz models? Seriously! 50, 100 and 200 at most! I'd say drop the 50MHz and keep the 70MHz model as the base as most modern scopes are starting from that point. It just makes more sense to have less models, less inventory overhead, less cost to support.

[EEVblog]

I expect they keep with the old design because they have an ASIC/frontend chip which does all that and they've made it for so long it's very cheap. The DS1074Z gets a ton of performance from a  ~$100 FPGA and $80 ADC but I'm guessing Tek probably spend less than a fifth of that on the ASIC.

Perhaps they can't even make the ASIC more, but still have a warehouse full of millions of them?

[AutomationDude]

Couldn't have said it better, myself

Event the DPO3000 series of scopes (which you'll pay thousands of dollars for) still use the base 2.5k of memory.  They are simply relying on their name at this point in order to move product in the lower end of the market.

Not true.  The DPO3000 and MSO3000 come standard with 5M of memory on each channel.  The new MDO3000 has 10M standard on each channel.

I think you mean the TDS3000. But that thing is old.

Yeah, that's the one I meant... It's so old now (as I'm getting to be), blah, blah blah LOL

[han]

i'm sure 80% will go to school / university that didn't care for performance.
And they buy because they had the same brand and type with the existing scope.
Here, they even still buy analog scope for for one odd reason, the analog scope had a vertical/horizontal cal pot...

[Hydrawerk]

Note the dim Run/Stop LED.
I like the big multipurpose knob.

[echen1024]

Do CH1 and CH2 light up?

[Hydrawerk]

Do CH1 and CH2 light up?

No, they do not light up.
It is visible in my video. http://youtu.be/p1ysgtBNIs0

[echen1024]

Do CH1 and CH2 light up?

No, they do not light up.
It is visible in my video. http://youtu.be/p1ysgtBNIs0

Yes,  I see now. Even my old Siglent had blinkenlights on the channels!