EEVblog Electronics Community Forum

Products => Test Equipment => Topic started by: Chipguy on April 09, 2014, 07:19:18 pm

Title: Tektronix is going to ditch all the low memory TDS models
Post by: Chipguy on April 09, 2014, 07:19:18 pm
Hey guys,

I sold some of my Tektronix oscilloscopes recently. While researching some details I also downloaded some manuals from the Tek site which can only be accessed when you are logged in.

So yesterday I got a call from Tektronix asking me about my recent activities and if they can "support" me in any way.
Well, I told them what I think of their entry value scopes and that anyone would be a fool not to buy a Rigol scope for either less than half the price or at least double the features for the same price.

So he told me that they are going to ditch all the 2.5K and 10K memory scopes soon (oh they had scopes with 10K memory ?)
and replace them with "something more modern"  :clap: WOOHOO
I told him "That's about time" but he could not say when these new models will become available.

He basically said that there will be some major changes in the lower price segment.
OK then, I am very curious when we are going to see something new from Tektronix.

Maybe they can make it back into the market. I don't think their entry models sell very good at the moment.

Cheers
Chipguy (@Chippolev on twitter)


Title: Re: Tektronix is going to ditch all the low memory TDS models
Post by: Rigby on April 09, 2014, 07:43:11 pm
Ooooh.
Title: Re: Tektronix is going to ditch all the low memory TDS models
Post by: krenzo on April 09, 2014, 08:56:59 pm
Oh boy, I can't wait for them to start selling rebadged Rigol 1074s for double the cost!
Title: Re: Tektronix is going to ditch all the low memory TDS models
Post by: Wim_L on April 09, 2014, 09:19:28 pm
So he told me that they are going to ditch all the 2.5K and 10K memory scopes soon (oh they had scopes with 10K memory ?)

Yep, the TDS3000 series has 10k per channel. Those are pretty good scopes (up to 500 MHz, 4 channels, quite responsive), and probably the highest performance ones they have that can run on batteries (not entirely sure about that, I'm not following their new developments all that closely). The memory is indeed quite small by current standards, but still a good scope for analog work.
Title: Re: Tektronix is going to ditch all the low memory TDS models
Post by: Chipguy on April 09, 2014, 09:26:09 pm
Oh boy, I can't wait for them to start selling rebadged Rigol 1074s for double the cost!

Haha, I haven't thought of that: Some rebadged stuff. Gonna be interesting,   |O or :-DD
Title: Re: Tektronix is going to ditch all the low memory TDS models
Post by: NiHaoMike on April 09, 2014, 09:49:34 pm
Interesting that there's a "road test" of a 2.5k memory scope.
http://www.element14.com/community/roadTests/1282 (http://www.element14.com/community/roadTests/1282)
Title: Re: Tektronix is going to ditch all the low memory TDS models
Post by: Hydrawerk on April 09, 2014, 10:49:10 pm
Yep, the TDS3000 series has 10k per channel. Those are pretty good scopes (up to 500 MHz, 4 channels, quite responsive), and probably the highest performance ones they have that can run on batteries

DSO6000 (up to 1GHz) can also run on batteries. http://www.home.agilent.com/en/pc-1000003091%3Aepsg%3Apgr/infiniivision-6000-series-oscilloscope?nid=-536902766.0&cc=US&lc=eng (http://www.home.agilent.com/en/pc-1000003091%3Aepsg%3Apgr/infiniivision-6000-series-oscilloscope?nid=-536902766.0&cc=US&lc=eng)

http://www.home.agilent.com/en/pd-1081755-pn-DSO6012/battery-power-option-for-all-6000-series-models?nid=-33523.635257.00&cc=US&lc=eng (http://www.home.agilent.com/en/pd-1081755-pn-DSO6012/battery-power-option-for-all-6000-series-models?nid=-33523.635257.00&cc=US&lc=eng)

But it is not so compact as Tektronix TDS3000.
Title: Re: Tektronix is going to ditch all the low memory TDS models
Post by: Hydrawerk on April 09, 2014, 10:50:14 pm
Oh boy, I can't wait for them to start selling rebadged Rigol 1074s for double the cost!
Hmm, are they gonna sell some rebadges scope by Siglent, as BK Precision and LeCroy do???
Title: Re: Tektronix is going to ditch all the low memory TDS models
Post by: David Hess on April 09, 2014, 11:38:14 pm
Besides the obvious functional things you can do with large acquisition memories like long high sample rate pretrigger records and single shot acquisitions, what other advantages do they have?

I still work in the 1k to 4k range and except for the above two cases.  I have always found large acquisition memories to just slow things down compared to a short one and delayed acquisition.  I would trade high waveform acquisition rates for long record lengths although the later allows the former in instruments that support segmentation.
Title: Re: Tektronix is going to ditch all the low memory TDS models
Post by: Chipguy on April 10, 2014, 05:41:06 am
I still work in the 1k to 4k range and except for the above two cases.  I have always found large acquisition memories to just slow things down compared to a short one and delayed acquisition.  I would trade high waveform acquisition rates for long record lengths although the later allows the former in instruments that support segmentation.

This is not the only issue the TDS scopes have. They also have a low waveform update rate, like 50 Hz or something.
There are many occasions where you set the memory depth to a low value in order to get the update rate (which you can forget with the TDS models) but there are also other occasions, escpecially when you check digital signals, where you really need to scroll through a lot of waveform because your trigger event is far away from the position you want to check.

So in any respects the TDS models really becoming useless nowadays.
They neither give you low memory but 50K or so update rate nor they do give you a lot of memory to go through.

We hat 7 Tek scopes at work, only 2 are remaining. The rest is now Rigol.
Title: Re: Tektronix is going to ditch all the low memory TDS models
Post by: EEVblog on April 10, 2014, 07:23:09 am
Besides the obvious functional things you can do with large acquisition memories like long high sample rate pretrigger records and single shot acquisitions, what other advantages do they have?

My latest video uploading now provides one answer to that.
Title: Re: Tektronix is going to ditch all the low memory TDS models
Post by: EEVblog on April 10, 2014, 07:28:26 am
I do hope this isn't just the sales drone spinning a yarn.
Tek are at least 5 years behind (some would say 10) on entry level scope technology.
Title: Re: Tektronix is going to ditch all the low memory TDS models
Post by: Kjelt on April 10, 2014, 07:33:54 am
My latest video uploading now provides one answer to that.
You are slowly changing from an EE to a marketing pro  :)
Title: Re: Tektronix is going to ditch all the low memory TDS models
Post by: KedasProbe on April 10, 2014, 08:42:15 am
Maybe they will take a new approach and introduce some low end 'PCI-E' scopes.
like http://www.alazartech.com/octnews.htm (http://www.alazartech.com/octnews.htm)
Maybe some complete integrated small black box PC with the same technology, just connect screen, keyboard and mouse.
Except for the mechanical design they don't have to change much from what they have.

edit: price reference 130Euro for the PC parts, now internally add the PCIe scope part.
http://www.minipc.de/catalog/il/1795 (http://www.minipc.de/catalog/il/1795)
A motherboard 90Euro http://www.minipc.de/catalog/il/1620 (http://www.minipc.de/catalog/il/1620) again just use the PCI-e port to plug in existing designs. (the USB speed bottleneck of USB scopes can so easily be solved this way)
Title: Re: Tektronix is going to ditch all the low memory TDS models
Post by: tszaboo on April 10, 2014, 10:03:52 am
Oh thank god. Really, I've met so many TDS2000 series scope, and I hate them all. The engineer who decided to use the screen as the sampling memory was a genius indeed, but I hate him, because every single company looks at the sample rate, and buys this instead of a proper one.
Title: Re: Tektronix is going to ditch all the low memory TDS models
Post by: EEVblog on April 10, 2014, 11:55:12 am
More competition for Rigol and Agilent can only be a good thing. The low end is well covered but the mid range 2000 series is basically Rigol or Agilent or total crap.

Yes, Tek are now very competitive in the Agilent 3000X price range with the MDO3000, but they start at over $3K
Title: Re: Tektronix is going to ditch all the low memory TDS models
Post by: Chipguy on April 15, 2014, 06:30:58 pm
Well Tektronix sort of "delivered" today, from the looks.
There is now the TBS1000 entry scope series.

And they got a whopping 2.5k memory for each channel, whooop deeee doooo.  :-DD
http://www.tek.com/tbs1000-tbs1000b-and-tbs1000b-edu-digital-oscilloscopes (http://www.tek.com/tbs1000-tbs1000b-and-tbs1000b-edu-digital-oscilloscopes)

So going through a bunch of I2C or SPI packages and see if there is some collision in your bus.... oops... not enough memory, can't capture even 3 bytes and see if your SCK rings or something.

It's ideal for educational stuff I believe. That's why they made a dedicated version.
For what they are they are still too expensive. You can get much better value for money from Rigol.

http://www.rigolna.com/products/digital-oscilloscopes/ds1000Z/ds1074z-s/ (http://www.rigolna.com/products/digital-oscilloscopes/ds1000Z/ds1074z-s/) USD $818and 4 channels
The Tek 1072B only has 2 channels and costs USD $890

I don't want to see Tektronix gone of the market, but I am afraid they may get in trouble in the future....


Title: Re: Tektronix is going to ditch all the low memory TDS models
Post by: Rigby on April 15, 2014, 06:39:38 pm
this was discussed previously, and is in fact the origin of this thread; the 2.5k models are rumored to be going away, finally.

what gets me is that memory is freaking cheap, so cheap that one could reliably include several GB of RAM into a scope for < $50 after markup.  I'd like to see a scope as a logic analyzer be able to capture several dozens of minutes on 2-4 channels, if the need ever came around.

if a scope has an 8-bit ADC, then each point is 1 byte of data.  12 megapoints = 12 megabytes of data.  4GB of RAM can be purchased for less than $50.  Why do we not have scopes with 4 gigapoints of memory?  Is that too much to be useful?  Where is the line between enough memory and too much?  Is that line the same for everyone?
Title: Re: Tektronix is going to ditch all the low memory TDS models
Post by: GiskardReventlov on April 15, 2014, 06:42:00 pm
I'm definitely not an expert and not in the market but if I was going to spend ~$800-1000 u.s on a scope and my choices were tek or rigol. I'd spend a little more for tek. How old is the ds1074 now?
Title: Re: Tektronix is going to ditch all the low memory TDS models
Post by: Rigby on April 15, 2014, 06:43:22 pm
How old is the ds1074 now?

Less than a year, I think.

I agree with you about the brand, but 2.5k?  I don't need it ATM but I'm told I will and that lack of enough RAM will be terrible.
Title: Re: Tektronix is going to ditch all the low memory TDS models
Post by: jlmoon on April 15, 2014, 06:52:54 pm
Now if Agilent would just turn up the input volume on that 3000x series scope.. it would be very close to perfect!   :scared:
Title: Re: Tektronix is going to ditch all the low memory TDS models
Post by: free_electron on April 15, 2014, 06:55:16 pm
Well Tektronix sort of "delivered" today, from the looks.
There is now the TBS1000 entry scope series.

And they got a whopping 2.5k memory for each channel, whooop deeee doooo.  :-DD
http://www.tek.com/tbs1000-tbs1000b-and-tbs1000b-edu-digital-oscilloscopes (http://www.tek.com/tbs1000-tbs1000b-and-tbs1000b-edu-digital-oscilloscopes)

:palm:

i get those calls too from tektronix. whenever you download a manual or software behind the registration page you get phone calls.

they must be getting really desperate flogging their kit...
Title: Re: Tektronix is going to ditch all the low memory TDS models
Post by: Kryan9 on April 15, 2014, 07:13:13 pm
this was discussed previously, and is in fact the origin of this thread; the 2.5k models are rumored to be going away, finally.

what gets me is that memory is freaking cheap, so cheap that one could reliably include several GB of RAM into a scope for < $50 after markup.  I'd like to see a scope as a logic analyzer be able to capture several dozens of minutes on 2-4 channels, if the need ever came around.

if a scope has an 8-bit ADC, then each point is 1 byte of data.  12 megapoints = 12 megabytes of data.  4GB of RAM can be purchased for less than $50.  Why do we not have scopes with 4 gigapoints of memory?  Is that too much to be useful?  Where is the line between enough memory and too much?  Is that line the same for everyone?

If you're going to add memory, you have to have a memory interface, which isn't really as cheap as compared to just adding a DDR3 SDRAM chip alone and calling it a day. Assuming the front-end is FPGA-based, this can mean that you end up significantly increasing the size of the FPGA because you need the memory controller logic (nowadays it's included in modern FPGA's), as well as the bus logic that provides read and write access control so the ADC's can have their data input to the RAM, as well as letting the main processor read data out to be displayed and do some number crunching, PLUS semi-significant on-chip or possibly off-chip buffers with almost zero latency since SDRAM has the column access times and periods where the display needs to be updated with some data which can add up. This all adds up to more dev time and bigger FPGA's which increase exponentially in price.

For OScopes which have higher-performance ASIC's the redesign isn't anywhere near this simple anyways, so it's not much of an option if you want a TDS1104B with 4 GB of sample memory that doesn't cost $5K+.
Title: Re: Tektronix is going to ditch all the low memory TDS models
Post by: pickle9000 on April 15, 2014, 07:19:58 pm
One thing interesting about the TBS1000B-EDU is the ability to edit and show images and text to the user, as well as being able to save information to a PC from the scope. Could be real handy for providing service information to a tech or on a manned production line. I doesn't sound really tricky to implement and could be added to any scope as a base feature.
Title: Re: Tektronix is going to ditch all the low memory TDS models
Post by: bronson on April 15, 2014, 08:14:37 pm
Anyone who's excited about the new scopes, I gotta ask:

Were you excited about the Keithley 2100?

Because I was.  Ughhhhhhhh....  What a heartbreaker.
Title: Re: Tektronix is going to ditch all the low memory TDS models
Post by: echen1024 on April 16, 2014, 01:16:07 pm
Tektronix seems to have old, patient people working there.

But I think they are just using the same ASIC with everything inside including the 2.5k memory, and have been for the past 10 years.
Title: Re: Tektronix is going to ditch all the low memory TDS models
Post by: David Hess on April 16, 2014, 02:22:29 pm
Tektronix seems to have old, patient people working there.

But I think they are just using the same ASIC with everything inside including the 2.5k memory, and have been for the past 10 years.

It is a big step from using the relatively small but fast embedded memory in an ASIC or programmable logic for the acquisition record to using an external fast and wide memory bus capable of supporting the real time sample rate.  That step is only made larger by most oscilloscope applications not benefiting from the long record length that external acquisition memory allows.  The extra long record length also does not help with the waveform acquisition rate because it still has to be processed.

Ever notice how some DSOs have much lower maximum sample rate when the record length is extended beyond a specific length?  Now you know why.

While long records are indispensable for specific applications, those are a minority of DSO applications.  I am suspicious that marketing is driving a "record length war" in the sense that processor manufacturers had a "MHz war".  I would rather have better graded index displays, faster waveform acquisition rates, and lower latency among other deprecated features.
Title: Re: Tektronix is going to ditch all the low memory TDS models
Post by: w2aew on April 16, 2014, 02:40:13 pm
Tektronix seems to have old, patient people working there.

But I think they are just using the same ASIC with everything inside including the 2.5k memory, and have been for the past 10 years.

I believe that Tek intended the DPO/MSO2000/3000 series to replace the bench TDS line, but customer demand for them has kept them in the catalog.  I just got the 2014 catalog, and they're still in there.
Title: Re: Tektronix is going to ditch all the low memory TDS models
Post by: David Hess on April 16, 2014, 03:04:19 pm
Tektronix seems to have old, patient people working there.

But I think they are just using the same ASIC with everything inside including the 2.5k memory, and have been for the past 10 years.

I believe that Tek intended the DPO/MSO2000/3000 series to replace the bench TDS line, but customer demand for them has kept them in the catalog.  I just got the 2014 catalog, and they're still in there.

Their market segmentation has always confused me but I do not see them replacing portable battery powered operation of the TDS series nor the isolated channels of the TPS2000 unless there was low demand and maybe not even then.  Those two separate features would seem to require two series of oscilloscopes which are not any of the current DPO/MSO series.
Title: Re: Tektronix is going to ditch all the low memory TDS models
Post by: Chipguy on April 16, 2014, 04:10:14 pm
Tektronix seems to have old, patient people working there.
But I think they are just using the same ASIC with everything inside including the 2.5k memory, and have been for the past 10 years.

I can tell for a fact that this ASIC is around since at least 1996.
At work and home I had several TDS340 and TDS340A. Out of curiousity I opened one a long time ago and found that Tektronix was working with Maxim. There was a chip branded "MAXTEK" which contained all the guts including the RAM. I think the ADC's were something special and seperate IC's.

For the projects I work on that small amount of memory is mostly useless, so I have only one TDS340 left which will go on sale soon.

It is very dissapointing that Tektronix is not taking the technology any further. When that guy told me that they are going to replace all the TDS models I hoped they would introduce something that can easily keep up with Rigol or Agilent. At that pricepoint this is rather pathetic.
Come on Tektronix, release something for a reasonable price thats really shakes the market. The technology is all there.
Title: Re: Tektronix is going to ditch all the low memory TDS models
Post by: ConKbot on April 17, 2014, 11:19:35 am
no more 2.5k pt memory depth scopes?


(http://i.imgur.com/sSDlDT2.jpg)
Title: Re: Tektronix is going to ditch all the low memory TDS models
Post by: David Hess on April 17, 2014, 03:49:50 pm
Tektronix seems to have old, patient people working there.
But I think they are just using the same ASIC with everything inside including the 2.5k memory, and have been for the past 10 years.

I can tell for a fact that this ASIC is around since at least 1996.
At work and home I had several TDS340 and TDS340A. Out of curiousity I opened one a long time ago and found that Tektronix was working with Maxim. There was a chip branded "MAXTEK" which contained all the guts including the RAM. I think the ADC's were something special and seperate IC's.

Tektronix sold their semiconductor fabrication facilities to Maxim back around 1994.  Maxim continued to produce their existing ASICs for some time (including the 234-0408-20 which replaced the venerable Tektronix 155-0022-00 which I believe was the longest produced Tektronix IC) and there is speculation that quality of the Maxim produced parts suffered significantly.
Title: Re: Tektronix is going to ditch all the low memory TDS models
Post by: rf-design on April 17, 2014, 04:00:13 pm
Their market segmentation has always confused me but I do not see them replacing portable battery powered operation of the TDS series nor the isolated channels of the TPS2000 unless there was low demand and maybe not even then.  Those two separate features would seem to require two series of oscilloscopes which are not any of the current DPO/MSO series.

TPS is possible eaten up by Fluke. Is there a specific feature which sells against a 200MHz Fluke? Now they have a 4CH and a 2CH+DMM handheld.
Title: Re: Tektronix is going to ditch all the low memory TDS models
Post by: rf-design on April 17, 2014, 04:11:50 pm
Tektronix seems to have old, patient people working there.

But I think they are just using the same ASIC with everything inside including the 2.5k memory, and have been for the past 10 years.

It is a big step from using the relatively small but fast embedded memory in an ASIC or programmable logic for the acquisition record to using an external fast and wide memory bus capable of supporting the real time sample rate.  That step is only made larger by most oscilloscope applications not benefiting from the long record length that external acquisition memory allows.  The extra long record length also does not help with the waveform acquisition rate because it still has to be processed.

Ever notice how some DSOs have much lower maximum sample rate when the record length is extended beyond a specific length?  Now you know why.

While long records are indispensable for specific applications, those are a minority of DSO applications.  I am suspicious that marketing is driving a "record length war" in the sense that processor manufacturers had a "MHz war".  I would rather have better graded index displays, faster waveform acquisition rates, and lower latency among other deprecated features.

With the PCIe 3.0 one or two x4 lanes could shift the deep memory interface from FPGA/ASIC to PC. Cheap memory come wide and slow relative to an ADC beside graphic card memory but PCIe rise in speed and avaibility.
Title: Re: Tektronix is going to ditch all the low memory TDS models
Post by: echen1024 on April 17, 2014, 04:25:36 pm
Why not implement GDDR5 memory? That should be quite fast.
Title: Re: Tektronix is going to ditch all the low memory TDS models
Post by: rf-design on April 17, 2014, 05:46:32 pm
Why not implement GDDR5 memory? That should be quite fast.

Yes, it is 20% faster than PCIe 3.0 per wire/wire pair but the memory controller with DMA and all that stuff would better fit into an expensive NRE ASIC instead of an FPGA. So it is a matter of difference of  the return for the NRE if you already plan to built on a PC platform.
Title: Re: Tektronix is going to ditch all the low memory TDS models
Post by: GiskardReventlov on April 17, 2014, 05:59:15 pm
The only question now is who'll do the first review/teardown of the TBS1104 (4-chan 100MHz)?

But maybe there needs to be a shootout between Rigol* 1074 and TBS1104 first?

*I might go with Tektronix just because there's no ambiguity to saying tektronix where as it's ree-gull or rye-goal

Title: Re: Tektronix is going to ditch all the low memory TDS models
Post by: free_electron on April 17, 2014, 06:22:26 pm
Anyone who's excited about the new scopes, I gotta ask:

Were you excited about the Keithley 2100?

Because I was.  Ughhhhhhhh....  What a heartbreaker.
excited ? really ? its not even made by keithley ... relabeled wingpangpong
Title: Re: Tektronix is going to ditch all the low memory TDS models
Post by: David Hess on April 17, 2014, 09:05:45 pm
Tektronix seems to have old, patient people working there.

But I think they are just using the same ASIC with everything inside including the 2.5k memory, and have been for the past 10 years.

It is a big step from using the relatively small but fast embedded memory in an ASIC or programmable logic for the acquisition record to using an external fast and wide memory bus capable of supporting the real time sample rate.  That step is only made larger by most oscilloscope applications not benefiting from the long record length that external acquisition memory allows.  The extra long record length also does not help with the waveform acquisition rate because it still has to be processed.

Ever notice how some DSOs have much lower maximum sample rate when the record length is extended beyond a specific length?  Now you know why.

While long records are indispensable for specific applications, those are a minority of DSO applications.  I am suspicious that marketing is driving a "record length war" in the sense that processor manufacturers had a "MHz war".  I would rather have better graded index displays, faster waveform acquisition rates, and lower latency among other deprecated features.

With the PCIe 3.0 one or two x4 lanes could shift the deep memory interface from FPGA/ASIC to PC. Cheap memory come wide and slow relative to an ADC beside graphic card memory but PCIe rise in speed and avaibility.

I have never had a good experience dealing with real time data transfers which cross from the PC peripheral interface to system memory and pushing the throughput just makes it worse.  Going through the PCIe bridge to the memory controller to the system memory and expecting real time operation to support sustained sample rates strikes me as difficult to impossible.  At best, everything in that chain would need to be qualified and the cost would be higher than just interfacing a wide memory controller to the FPGA or ASIC directly.

Last time I checked, common PCIe bridges did *not* support a throughput equal to their interface rate.
Title: Re: Tektronix is going to ditch all the low memory TDS models
Post by: Chipguy on April 17, 2014, 10:38:00 pm
Today I had another session with one of these old 2.5K Tektronix scopes. (TDS340A)
I had to check an OPAMP which is able to deliver a few hundred milliamps at the output.
However those amps are usually also pretty fast, in this case GBW is 240 MHz.

So I wanted to check if there are any oscillations on the output.
Due to the lack of memory I can not see any oscillations while I have one full cycle on the screen (1 KHz).
The sampling frequency in this case was something like 500 KSa/s.
So oscillations above 250 KHz will not be captured.

And guess what, the circuit does have an oscillation at around 84 MHz.
I was only able to see it by setting the timebase to a low value and playing around with the trigger level.
Thats useless. You relly can not see stuff like that on that scope.

Title: Re: Tektronix is going to ditch all the low memory TDS models
Post by: bronson on April 17, 2014, 10:48:24 pm
excited ? really ? its not even made by keithley ... relabeled wingpangpong

That's exactly my point.  It sure looked good in the prerelease pamphlets.  I had the space allocated on my credit card and everything.

Then I saw the onebundone LCD and....  oh, the disappointment.
Title: Re: Tektronix is going to ditch all the low memory TDS models
Post by: David Hess on April 18, 2014, 02:36:42 am
Today I had another session with one of these old 2.5K Tektronix scopes. (TDS340A)
I had to check an OPAMP which is able to deliver a few hundred milliamps at the output.
However those amps are usually also pretty fast, in this case GBW is 240 MHz.

So I wanted to check if there are any oscillations on the output.
Due to the lack of memory I can not see any oscillations while I have one full cycle on the screen (1 KHz).
The sampling frequency in this case was something like 500 KSa/s.
So oscillations above 250 KHz will not be captured.

And guess what, the circuit does have an oscillation at around 84 MHz.
I was only able to see it by setting the timebase to a low value and playing around with the trigger level.
Thats useless. You relly can not see stuff like that on that scope.

Isn't the TDS340A record length only 1000 points?  I thought the display was 1/2 of the total record length at 50 points per division with 10 divisions.  In that case, 1 cycle at 1 kHz (1 millisecond) across 10 divisions (500 points) would be 500 kSamples/second while the needed sample rate would require a record length of like 500k points or so.  Such long record lengths are a relatively recent development in inexpensive DSOs.

The TDS340A does not have variable persistence but its inconvenient vector or dot accumulate mode should have caught this and peak detect acquisition mode certainly would have as that is what it is designed for.

The DSOs I regularly use are older than your TDS340A but even though they support a longer record length, I usually use 1000 points and rely on peak detect mode to capture such things.
Title: Re: Tektronix is going to ditch all the low memory TDS models
Post by: Chipguy on April 18, 2014, 08:23:07 am
DavidHess: The TDS340A has 2.5K points, which is awful.
When you see a single period of a 1KHz waveform the scope samples it with something like 50 kHz.
Even an old analog scope would show you the high frequency oscillation with the beam getting suspiciously wide and fuzzy,
looking kinda out of focus.  ???
That would make you notice something is going on.

The TDS340A however gives you the impression that everything is fine   :-BROKE
I really have to get rid of that. What a ripper  :--

Those engineers at Tektronix are sleeping :=\
Title: Re: Tektronix is going to ditch all the low memory TDS models
Post by: Thomas on April 18, 2014, 09:38:25 am
Quote
Then I saw the onebundone LCD and....  oh, the disappointment.
You're talking about the 2110, right?
That's the one with the horrible LCD:
(http://www.nortelcoelectronics.no/keithley_2110_front-jpg?pid=Native-ContentFile-File&r_n_d=49328_&adjust=1&x=308&from=0)

2100 is a totally different meter:
(http://www.nortelcoelectronics.no/keithley_2100-jpg?pid=Native-ContentFile-File&r_n_d=49328_&adjust=1&x=308&from=0)
Title: Re: Tektronix is going to ditch all the low memory TDS models
Post by: Chipguy on April 18, 2014, 12:57:52 pm
The Keithley 2110 has indeed the poorest display I have ever seen on a good brand bench multimeter.
What a waste of space in that window, it could be so much better.
Next worst are the older HP bench multimeter with LCD I would say.
Title: Re: Tektronix is going to ditch all the low memory TDS models
Post by: David Hess on April 18, 2014, 02:03:02 pm
DavidHess: The TDS340A has 2.5K points, which is awful.
When you see a single period of a 1KHz waveform the scope samples it with something like 50 kHz.

The operating manual, service manual, 1998 catalog, and the math say that the TDS340A record length is 1000 points but I will settle for "short".

Quote
Even an old analog scope would show you the high frequency oscillation with the beam getting suspiciously wide and fuzzy,
looking kinda out of focus.  ???
That would make you notice something is going on.

The TDS340A however gives you the impression that everything is fine   :-BROKE
I really have to get rid of that. What a ripper  :--

An analog oscilloscope certainly would but so would this oscilloscope with peak detection, envelope mode, or accumulate turned on.  My *older* DSOs have the same limitations and they could have detected the oscillation as well assuming it was larger than the input noise.

Send it to me.  I can put it to good use.  It would make a find upgrade for my 2440 which that series replaced despite having a shorter record length. :)

Quote
Those engineers at Tektronix are sleeping :=\

Yes, shame on them for not including a longer record length in a low end real time DSO from 1998.
Title: Re: Tektronix is going to ditch all the low memory TDS models
Post by: Chipguy on April 18, 2014, 02:27:27 pm
Quote
The operating manual, service manual, 1998 catalog, and the math say that the TDS340A record length is 1000 points but I will settle for "short".

Errr right, even only 1000. Seems I mixed that up with the TDS2000 series.

Quote
Yes, shame on them for not including a longer record length in a low end real time DSO from 1998.
Well it was ok back then.

I am moaning about them not taking that any further 16 years later.

When you see the rest of the product politics from Tek it's clear that they are just crawling technical progress.
They just give you minor innovations in order to maximize the profit by selling you another scope, soon, with some more monor innovations.

However we are living in a time where new competitors are easily keeping up easily.
They (Rigol etc.) eventually will become the technology leaders if Tek keeps progressing like this.
Title: Re: Tektronix is going to ditch all the low memory TDS models
Post by: SeanB on April 18, 2014, 02:31:06 pm
Probably as well they are running low on stock of the special ASIC's used in these scopes, and the cost of having a fab make them with an old process is going to be expensive, compared to the scope factory cost, so likely they will soon obsolete them and keep the remaining stock for service spares until they are all gone out of warranty and lifetime support period, when the whole line and spares stock will be dropped.
Title: Re: Tektronix is going to ditch all the low memory TDS models
Post by: rf-design on April 18, 2014, 02:40:44 pm
Probably as well they are running low on stock of the special ASIC's used in these scopes, and the cost of having a fab make them with an old process is going to be expensive, compared to the scope factory cost, so likely they will soon obsolete them and keep the remaining stock for service spares until they are all gone out of warranty and lifetime support period, when the whole line and spares stock will be dropped.

Yes, you hit the nail. Everyotherthingwise would be irrational.
Title: Re: Tektronix is going to ditch all the low memory TDS models
Post by: David Hess on April 18, 2014, 03:26:26 pm
Quote
Yes, shame on them for not including a longer record length in a low end real time DSO from 1998.
Well it was ok back then.

I am moaning about them not taking that any further 16 years later.

When you see the rest of the product politics from Tek it's clear that they are just crawling technical progress.
They just give you minor innovations in order to maximize the profit by selling you another scope, soon, with some more monor innovations.

However we are living in a time where new competitors are easily keeping up easily.
They (Rigol etc.) eventually will become the technology leaders if Tek keeps progressing like this.

Rigol is more like a marketing leader.  Tektronix has not been a technology leader since about the time Danaher bought them.

While long record lengths are nice and have become less expensive, they are irrelevant to most DSO applications including this one.

Rigol and others are going to make use of long record lengths in their marketing to distinguish themselves from Tektronix whether this feature is important or not.  This comes from a company who could not get peak detection right and deliberately conflated it (they lied) with envelope detection in their DS1000E(D) and DS1000B series oscilloscopes.
Title: Re: Tektronix is going to ditch all the low memory TDS models
Post by: w2aew on April 18, 2014, 07:20:54 pm
....  Tektronix has not been a technology leader since about the time Danaher bought them....

I beg to differ a little on this point.  Sure, I'm biased a little, but this statement isn't really true.  Tek has many areas where they are simply unmatched.  The high speed Arbitrary Waveform Generators feature 50GS/s sample rate at 10b resolution - no one even comes close.  The RSA series of realtime signal analyzers feature 165MHz of realtime bandwidth, with realtime triggering features unmatched by anyone else.  The next generation of realtime scope technology, called asynchronous time interleaving, will give Tek 70GHz of bandwidth at 200GS/s - look for products based on this later this year.  Even after more than two years since its introduction, no one else offers anything that competes with the MDO4000 mixed domain oscilloscope, a mixed signal scope that includes a 3/6GHz time correlated spectrum analyzer.  The story continues in many other instrument categories.  Sure, Tektronix offerings at the entry level aren't the lowest price point, but then again, the entry level hobby market never has been a focus for Tektronix (unfortunately).  The only reason that the older technology scopes such as the TDS series are still in the catalog is because of customer demand for them.  Otherwise, they would have been replaced by the DPO/MSO series long ago.
Title: Re: Tektronix is going to ditch all the low memory TDS models
Post by: Hydrawerk on April 19, 2014, 07:17:01 pm
Tektronix is still better than LeCroy with their crap WaveAce Siglent scopes.
Title: Re: Tektronix is going to ditch all the low memory TDS models
Post by: echen1024 on April 19, 2014, 07:24:48 pm
Tektronix is still better than LeCroy with their crap WaveAce Siglent scopes.
We have those at our local Fry's Electronics. They are always locking up and the salespeople have no idea what to do. So every other week, I go in and reboot the oscilloscopes.
Title: Re: Tektronix is going to ditch all the low memory TDS models
Post by: David Hess on April 19, 2014, 07:36:29 pm
Tektronix is still better than LeCroy with their crap WaveAce Siglent scopes.
We have those at our local Fry's Electronics. They are always locking up and the salespeople have no idea what to do. So every other week, I go in and reboot the oscilloscopes.

Unfortunately I have moved states away from Fry's Electronics or I would stop by occasionally and try to evaluate the various oscilloscopes with my handy portable pocket sized battery powered fast transition signal source.
Title: Re: Tektronix is going to ditch all the low memory TDS models
Post by: AndyC_772 on April 19, 2014, 07:52:10 pm
Tektronix is still better than LeCroy with their crap WaveAce Siglent scopes.
We have those at our local Fry's Electronics. They are always locking up and the salespeople have no idea what to do. So every other week, I go in and reboot the oscilloscopes.

Not a good idea. If a piece of kit is prone to locking up, potential buyers need to know. It's important information.
Title: Re: Tektronix is going to ditch all the low memory TDS models
Post by: David Hess on April 19, 2014, 08:09:07 pm
Tektronix is still better than LeCroy with their crap WaveAce Siglent scopes.
We have those at our local Fry's Electronics. They are always locking up and the salespeople have no idea what to do. So every other week, I go in and reboot the oscilloscopes.

Not a good idea. If a piece of kit is prone to locking up, potential buyers need to know. It's important information.

Too bad marketing does not see it that way.

"Now with the reliability and performance of Microsoft Windows!"
Title: Re: Tektronix is going to ditch all the low memory TDS models
Post by: linux-works on April 20, 2014, 07:17:14 pm
Tektronix is still better than LeCroy with their crap WaveAce Siglent scopes.
We have those at our local Fry's Electronics. They are always locking up and the salespeople have no idea what to do. So every other week, I go in and reboot the oscilloscopes.

reboot the scopes?

do you have to hit control-alt-1khz or something?

lol ;)

Title: Re: Tektronix is going to ditch all the low memory TDS models
Post by: echen1024 on April 20, 2014, 10:06:26 pm
Tektronix is still better than LeCroy with their crap WaveAce Siglent scopes.
We have those at our local Fry's Electronics. They are always locking up and the salespeople have no idea what to do. So every other week, I go in and reboot the oscilloscopes.

reboot the scopes?

do you have to hit control-alt-1khz or something?

lol ;)
Yea. I keep pestering them to sell me the display Fluke 287, but so far, no luck.