EEVblog Electronics Community Forum
Products => Test Equipment => Topic started by: awesomechapro on December 05, 2024, 08:16:38 am
-
Hi all,
I’m tossing up whether I should get an older Tektronix TDS3012B or a newer Hantek DSO2D10 oscilloscope. As the Tektronix is probably built better but the Hantek is more up to date. Which one would you go for?
-
What are the prices for both?
-
The Tektronix TDS3012B has 1.25 Gs/s per channel vs 1 Gs/s shared on the Hantek DSO2D10, but the Hantek has more memory depth. Don't buy the Hantek new, there are better alternatives at this price range (see the Siglent/Rigol threads).
-
What are you going to use it for?
As mentioned, the Hantek has a lot more record memory, and it also has protocol decoding. Those may be useful/important to you.
From what I've read about the Hantek, most users were pretty disappointed with it relating to reliability/stability. Do a few searches here.
I don't know your budget, but maybe compare the price and specs of the low end Siglent and Rigol models. You'll get something much more useful immediately, and with a longer useful life.
-
I would definitely go for the TDS3012B; it's an older model, but the performance is commensurate with the Tektronix brand.
Oscilloscopes are like "knives" for hardware engineers. The ease of use of the UI is the sharpness of the knife.
If it is not easy to use, you will spend minutes trying to find the mode you need. The UI may not work smoothly.
Just because something is cheap doesn't mean it's good, as evidenced by the widespread use of Tek TBS for educational purposes.
As for Hantek, you can ask anyone who has used it, but I would choose Tek based on performance, ease of use of the UI, and hackability.
(I am a Tek/hp believer. understand that i am biased.)
-
the performance is commensurate with the Tektronix brand.
It was when the unit was new, many moons ago. Who knows what it is now.
And even if it is still factory fresh, the spec sheet performance of this model, whilst good of its era, is significantly below that of any of the current, popular low-end models from either Siglent or Rigol. The Tek has very short memory depth and only 9-bits of resolution. I don't know what you're expecting to pay for the Tek but based on average values I see on auction sites, if you're prepared and able to spend 100USD more, then I think that the Rigol 802 (whilst flawed in some aspects) would be a much more useful tool for most of the common use-cases of a contemporary hobbyist, and represent a better long-term investment.
imho.
-
Hi all,
I’m tossing up whether I should get an older Tektronix TDS3012B or a newer Hantek DSO2D10 oscilloscope. As the Tektronix is probably built better but the Hantek is more up to date. Which one would you go for?
We cannot advise you without price information.
This question is only about the price and your budget.
P.S. I have a Hantek DSO2D10 which I bought for $130 and I am happy with it, but I know there are other opinions out there.
I wouldn't pay more than $200 for it though.
-
had the hantek d15 , was not impressed by it sure memory ... but
i would go to the TEK right away ... abd some bw hack can be made ...
protocol decoding, save money and get some logic analyzer, you can find some at very low prices ... Kingst LA xxxx series, and some ds logic
low price is not always an excuse, save money get good stuff if possible, and in the long term it will pay back
-
They are both around $300 AUD (around $200 USD)
I have been looking at the Siglent and Rigol options but I only have the budget for the really entry level models which would probably have similar specs
-
What do you want to use the oscilloscope for? The TDS3012B has an extremely short memory so if you want to capture longer events with some detail, you'll run out quickly.
IMHO neither is a good option nowadays. How about spending US $50 on a cheap pocket DSO from Ali-express and save money to buy one of the entry level Rigol or Siglent DSOs later on?
-
They are both around $300 AUD (around $200 USD)
I have been looking at the Siglent and Rigol options but I only have the budget for the really entry level models which would probably have similar specs
$200 is a bit on the high end.
The Tek is technically 25 years old, afaics. It has a floppy disk! ;)
I don't know.
The Hantek is okay for ~$150, but for $200+ I'd rather save up for a Siglent/Rigol.
You can get the Hantek cheaper though, right?
https://www.aliexpress.com/item/1005006741286094.html (https://www.aliexpress.com/item/1005006741286094.html)
-
Save
Get Siglent or Rigol, maybe some xmass lower prices ?? and add some free bw upgrades loll new models are out and cost effectives ...
for an oldie like the Tek you have to be sure it pass self tests and yes you have the "battery" problem who could happen, and yes it's old
Hantek, meh works but cheaaaap get the 2d15 ??? so far i don't hear about many problems on them ... you have a thread about them here
for all machines ... be sure to get the latest fw updates os some ol floppies for the Tek loll had problems to get some :palm:
-
The Tektronix TDS3012B has 1.25 Gs/s per channel vs 1 Gs/s shared on the Hantek DSO2D10, but the Hantek has more memory depth. Don't buy the Hantek new, there are better alternatives at this price range (see the Siglent/Rigol threads).
Samples/second is an irrelevant metric: the only metric that matters is bandwidth.
Examples:
- 35 years ago I was using HP scopes to examine 800ps risetimes. It was 25MS/s
- I have a 200ps/1.7GHz 1970 scope that uses BC107 (fT=350GHz) transistors in the signal path
- I have a scope with <100ps risetimes and a 37kS/s sampling rate
-
Hi all,
I’m tossing up whether I should get an older Tektronix TDS3012B or a newer Hantek DSO2D10 oscilloscope. As the Tektronix is probably built better but the Hantek is more up to date. Which one would you go for?
Provided the scope works, either will be better than no scope. Beyond that we would need to know what you need to use it for.
The cheaper scope would be better if it enables you to buy other necessary equipment.
Don't forget to ensure you have the right class of probe for your use cases. Probes aren't cheap, and the wrong class can ruin your day. FFI, see the references at https://entertaininghacks.wordpress.com/library-2/scope-probe-reference-material/ (https://entertaininghacks.wordpress.com/library-2/scope-probe-reference-material/)
Example: the Tek IsoVu probes have a base price of £10700. I doubt they are what you need! https://www.tek.com/en/products/oscilloscopes/oscilloscope-probes/isovu-isolated-probes (https://www.tek.com/en/products/oscilloscopes/oscilloscope-probes/isovu-isolated-probes)
-
The Tektronix TDS3012B has 1.25 Gs/s per channel vs 1 Gs/s shared on the Hantek DSO2D10, but the Hantek has more memory depth. Don't buy the Hantek new, there are better alternatives at this price range (see the Siglent/Rigol threads).
Samples/second is an irrelevant metric: the only metric that matters is bandwidth.
Examples:
- 35 years ago I was using HP scopes to examine 800ps risetimes. It was 25MS/s
- I have a 200ps/1.7GHz 1970 scope that uses BC107 (fT=350GHz) transistors in the signal path
- I have a scope with <100ps risetimes and a 37kS/s sampling rate
Would you, please, stop confusing people with useless comments of how sampling scopes sample at slow sample speeds..
They do not. They have equivalent sampling rate defined by sample aperture of individual sample in reconstruction process. They do not violate Nyquist, but cleverly reconstruct signal shape from thousands of separate trigger events.
And people are talking about Real Time sampling scope for which realtime sampling rate IS important.
You know, like ANY average scope made in the last 20 years...
And in the last more than 10 years, scopes mostly don't even provide ETS function anymore..
-
The Tektronix TDS3012B has 1.25 Gs/s per channel vs 1 Gs/s shared on the Hantek DSO2D10, but the Hantek has more memory depth. Don't buy the Hantek new, there are better alternatives at this price range (see the Siglent/Rigol threads).
Samples/second is an irrelevant metric: the only metric that matters is bandwidth.
Examples:
- 35 years ago I was using HP scopes to examine 800ps risetimes. It was 25MS/s
- I have a 200ps/1.7GHz 1970 scope that uses BC107 (fT=350GHz) transistors in the signal path
- I have a scope with <100ps risetimes and a 37kS/s sampling rate
Would you, please, stop confusing people with useless comments of how sampling scopes sample at slow sample speeds..
They do not. They have equivalent sampling rate defined by sample aperture of individual sample in reconstruction process. They do not violate Nyquist, but cleverly reconstruct signal shape from thousands of separate trigger events.
And people are talking about Real Time sampling scope for which realtime sampling rate IS important.
You know, like ANY average scope made in the last 20 years...
And in the last more than 10 years, scopes mostly don't even provide ETS function anymore..
Please read and understand what I wrote in context, before having a knee-jerk response.
I used those examples to illustrate why a scope's sample/second metric is meaningless. In particular comparing 1.25GS/s with 1.0GS/s is not useful: they are both 100MHz scopes!
The only metric that matters is the bandwidth.
-
The Tektronix TDS3012B has 1.25 Gs/s per channel vs 1 Gs/s shared on the Hantek DSO2D10, but the Hantek has more memory depth. Don't buy the Hantek new, there are better alternatives at this price range (see the Siglent/Rigol threads).
Samples/second is an irrelevant metric: the only metric that matters is bandwidth.
Examples:
- 35 years ago I was using HP scopes to examine 800ps risetimes. It was 25MS/s
- I have a 200ps/1.7GHz 1970 scope that uses BC107 (fT=350GHz) transistors in the signal path
- I have a scope with <100ps risetimes and a 37kS/s sampling rate
Would you, please, stop confusing people with useless comments of how sampling scopes sample at slow sample speeds..
They do not. They have equivalent sampling rate defined by sample aperture of individual sample in reconstruction process. They do not violate Nyquist, but cleverly reconstruct signal shape from thousands of separate trigger events.
And people are talking about Real Time sampling scope for which realtime sampling rate IS important.
You know, like ANY average scope made in the last 20 years...
And in the last more than 10 years, scopes mostly don't even provide ETS function anymore..
Please read and understand what I wrote in context, before having a knee-jerk response.
I used those examples to illustrate why a scope's sample/second metric is meaningless. In particular comparing 1.25GS/s with 1.0GS/s is not useful: they are both 100MHz scopes!
The only metric that matters is the bandwidth.
But you are wrong.
You cannot have a scope with a 100MHz BW with realtime sampling of 120MS/s.
It violates Nyquist and won't show you actual signal, but a downconverted version.
-
The Tektronix TDS3012B has 1.25 Gs/s per channel vs 1 Gs/s shared on the Hantek DSO2D10, but the Hantek has more memory depth. Don't buy the Hantek new, there are better alternatives at this price range (see the Siglent/Rigol threads).
Samples/second is an irrelevant metric: the only metric that matters is bandwidth.
Examples:
- 35 years ago I was using HP scopes to examine 800ps risetimes. It was 25MS/s
- I have a 200ps/1.7GHz 1970 scope that uses BC107 (fT=350GHz) transistors in the signal path
- I have a scope with <100ps risetimes and a 37kS/s sampling rate
Would you, please, stop confusing people with useless comments of how sampling scopes sample at slow sample speeds..
They do not. They have equivalent sampling rate defined by sample aperture of individual sample in reconstruction process. They do not violate Nyquist, but cleverly reconstruct signal shape from thousands of separate trigger events.
And people are talking about Real Time sampling scope for which realtime sampling rate IS important.
You know, like ANY average scope made in the last 20 years...
And in the last more than 10 years, scopes mostly don't even provide ETS function anymore..
Please read and understand what I wrote in context, before having a knee-jerk response.
I used those examples to illustrate why a scope's sample/second metric is meaningless. In particular comparing 1.25GS/s with 1.0GS/s is not useful: they are both 100MHz scopes!
The only metric that matters is the bandwidth.
But you are wrong.
You cannot have a scope with a 100MHz BW with realtime sampling of 120MS/s.
It violates Nyquist and won't show you actual signal, but a downconverted version.
Has that been suggested anywhere in this thread?
If not, why are you raising a strawman argument?
-
The Tektronix TDS3012B has 1.25 Gs/s per channel vs 1 Gs/s shared on the Hantek DSO2D10, but the Hantek has more memory depth. Don't buy the Hantek new, there are better alternatives at this price range (see the Siglent/Rigol threads).
Samples/second is an irrelevant metric: the only metric that matters is bandwidth.
Examples:
- 35 years ago I was using HP scopes to examine 800ps risetimes. It was 25MS/s
- I have a 200ps/1.7GHz 1970 scope that uses BC107 (fT=350GHz) transistors in the signal path
- I have a scope with <100ps risetimes and a 37kS/s sampling rate
Would you, please, stop confusing people with useless comments of how sampling scopes sample at slow sample speeds..
They do not. They have equivalent sampling rate defined by sample aperture of individual sample in reconstruction process. They do not violate Nyquist, but cleverly reconstruct signal shape from thousands of separate trigger events.
And people are talking about Real Time sampling scope for which realtime sampling rate IS important.
You know, like ANY average scope made in the last 20 years...
And in the last more than 10 years, scopes mostly don't even provide ETS function anymore..
Please read and understand what I wrote in context, before having a knee-jerk response.
I used those examples to illustrate why a scope's sample/second metric is meaningless. In particular comparing 1.25GS/s with 1.0GS/s is not useful: they are both 100MHz scopes!
The only metric that matters is the bandwidth.
But you are wrong.
You cannot have a scope with a 100MHz BW with realtime sampling of 120MS/s.
It violates Nyquist and won't show you actual signal, but a downconverted version.
Has that been suggested anywhere in this thread?
If not, why are you raising a strawman argument?
Problem is that it is you that have a knee jerk response without knowing whole discussion.
Yes, exactly THAT was discussed.
A scope with claimed 250 MHz BW and sampling of 312.5 MS/sec.
Do you understand now?
Of course that if you sample 100MHz BW with 500MS/s, 1GS/sec or 1.25GS/sec, that in that case it does not matter.
It is a 100MHz BW scope. And your statement means that. I agree.
But that was not discussed, but mathematically insufficient sampling...
And than you jump in with "sampling rate does not matter..." argument. :-//
Of course I reacted.
-
It took 12-bit to get me away from analog scopes. I've used 8/9 bit scopes at work for years and they didn't move me enough to get one, no matter how fast or expensive. The memory depth on the older scopes is too small. Plus, the FFT on 8/9 bit scopes seems near to worthless for any real world problem. Depending on what you're doing, # of channels and bandwidth, even the lower end Siglent SDS800 series could make you quite happy for not much $$.
-
- I have a 200ps/1.7GHz 1970 scope that uses BC107 (fT=350GHz) transistors in the signal path
That's a pretty fast transistor, even today ;)
BTW for DSO use, Sample Rate should only be discussed along with Bandwidth as they are both limiting DSO functions that are inter-related. There's a new DSO (won't mention brand as it's likely to start another flame war) that when "enabled" and using all 4 channels has an effective Channel Sampling Rate which violates it's Channel Nyquist Bandwidth!!
Best
-
It took 12-bit to get me away from analog scopes. I've used 8/9 bit scopes at work for years and they didn't move me enough to get one, no matter how fast or expensive. The memory depth on the older scopes is too small. Plus, the FFT on 8/9 bit scopes seems near to worthless for any real world problem. Depending on what you're doing, # of channels and bandwidth, even the lower end Siglent SDS800 series could make you quite happy for not much $$.
Had a similar experience long ago. When we were allowed in the lab (our value was perceived at creating more company $ developing systems, circuits, chips rather than in the lab testing such), not too impressed by the DSOs back then and always reached for a Tek Analog Scope.
When we semi-retired we needed a scope for our home lab and acquired a pair of familiar Tek 2565Bs which we repaired/restored. Later we studied this forum and decided on a DSO for our lab and selected the SDS2000X+, which even tho 8 bit ADC based was quite impressed (it has 10 bit mode) with the performance/features wrt to cost.
Now with the 12 bit SDS800X HD, this moves to another level of performance/features vs $, of course our "appetite" was created with the SDS2000X+ and we are getting hungrier :-+
Best
-
- I have a 200ps/1.7GHz 1970 scope that uses BC107 (fT=350GHz) transistors in the signal path
That's a pretty fast transistor, even today ;)
Oh... picky picky picky.
30 years ago I worked on a project demonstrating data comms in the 60GHz ISM band. Filled a room, and was deafening.
Now there are many chips for short-range distance measurement, especially for cars.
-
It took 12-bit to get me away from analog scopes. I've used 8/9 bit scopes at work for years and they didn't move me enough to get one, no matter how fast or expensive. The memory depth on the older scopes is too small. Plus, the FFT on 8/9 bit scopes seems near to worthless for any real world problem. Depending on what you're doing, # of channels and bandwidth, even the lower end Siglent SDS800 series could make you quite happy for not much $$.
Yup, the "use case" is critical information.
I started on 1GHz 6-bit boat anchor DSOs at work. OK for some purposes.
I quite like my 10MHz 14bit Digilent Analog Discovery. The added AWG and simple digital pattern generation/capture make it easy to implement, say, N-path filters and Tayloe mixers. Having the decent post-processing makes it easy to demonstrate their surprisingly interesting characteristics.
-
- I have a 200ps/1.7GHz 1970 scope that uses BC107 (fT=350GHz) transistors in the signal path
That's a pretty fast transistor, even today ;)
Yes, I was quite shocked too. Having still worked with the TO18 BC107/1087/109 in the seventies of last century, I've always thought they had 200 MHz transition frequency, and that only at relatively high collector currents between 15 and 25 mA. Had I known about the secret 350 GHz of these transistors, I would still have been baffled how a relatively big leaded component can be that fast...
-
- I have a 200ps/1.7GHz 1970 scope that uses BC107 (fT=350GHz) transistors in the signal path
That's a pretty fast transistor, even today ;)
Yes, I was quite shocked too. Having still worked with the TO18 BC107/1087/109 in the seventies of last century, I've always thought they had 200 MHz transition frequency, and that only at relatively high collector currents between 15 and 25 mA. Had I known about the secret 350 GHz of these transistors, I would still have been baffled how a relatively big leaded component can be that fast...
BC107/108/109 and plastic TO-82 counterparts BC182/3/4..