Hi Mike,
Reading the specs it seems like quantization ENOB is => 8bits and vertical resolution is 12 bits so I am not sure what direction I am going to go at this point. Not sure what I would be getting for the money spent and I am thinking that I am going to just stay with what I currently have for the time being.
Reading the specs it seems like quantization ENOB is => 8bits and vertical resolution is 12 bits so I am not sure what direction I am going to go at this point. Not sure what I would be getting for the money spent and I am thinking that I am going to just stay with what I currently have for the time being.
Agree on the ENOB spec. A bit fuzzy. I may still spring for either a 1000 or 4000 series 4 channel. The main difference I see is the 4000 offers dual A/D's for twice the sampling rate. I wonder too if the 4000 series gets a better graded A/D than the 1000 offering more real bits. eevblog has done some videos on one of them. Another reason for me to get another scope is 4 channels. When I bought mine a couple years ago I thought 2 channels, sure plenty for what I want to do. Wrong... Several times I'd be doing something and be thinking if only one more channel. If I get a 4 channel, I'll have 6 if I use both scopes. Plenty for sure this time. One more advantage for me for the 1000/4000 is I like real buttons/knobs. I've used several pc interfaces that simulate controls and always find them less than satisfying. I wish the rigol had a knob for each channel gain. I'm always changing the gain forgetting to switch channels. But cost and real estate means it is not going to happen. I'll post here if I pull the trigger with my thoughts.
Maybe another to look at, the Siglent SDS2104x Plus which has 10 bit quantization, and is reasonably priced at around $1400 USD,
or the SDS2104x HD at 12 bits for $3300 which is a bit of a budget buster.
The SDS2104x Plus is a 10 bit, 4 channel 100MHz scope which should be a significant step up from my TBS1052B which does have separate sensitivity controls for each of its two channels, but is otherwise rather unremarkable and not very good for analog stuff.
Note that they are currently offering a bandwidth upgrade to the next step if you purchase one of the Plus series scopes at TE equipment (US)
or the SDS2104x HD at 12 bits for $3300 which is a bit of a budget buster.
The SDS2104x Plus is a 10 bit, 4 channel 100MHz scope which should be a significant step up from my TBS1052B which does have separate sensitivity controls for each of its two channels, but is otherwise rather unremarkable and not very good for analog stuff.
Note that they are currently offering a bandwidth upgrade to the next step if you purchase one of the Plus series scopes at TE equipment (US)
Last edited:
Kevin, just backtracking to the Picoscope, I have a NZ friend who has a 2204 which piqued my interest, and then a thread on Audiokarma ( https://audiokarma.org/forums/index.php?threads/heyboer-williamson-bode-plots.875702/ ) added a bit more insight on the 2206B being able to achieve Bode stability plotting out to 1MHz, and possibly below 2Hz, for valve audio amp assessment. I had found the typical soundcard route to be excellent, but having to resort to manual spot frequency assessment using a sig gen and normal scope was a pain for >96kHz, where the important GNFB stability performance was often embedded. Was that area of interest on your radar?
Ciao, Tim
Ciao, Tim
It was, but primarily planned for general troubleshooting and to reduce the clutter on my work bench. Blode plotting would be a nice feature to have since I have been design discrete op-amps for a couple of years now.
I use an RTX-6001 for frequency response, THD, noise and SMPTE and other IMD measurements. For quick and dirty measurements I use my Amber 3501A.
I use an RTX-6001 for frequency response, THD, noise and SMPTE and other IMD measurements. For quick and dirty measurements I use my Amber 3501A.
Yeh, the RTX-6001/Soundcard style tool is excellent for rapid benchtop testing. Afaik, even a doubling of that upper frequency limit of 90-96kHz is unlikely with that technology even though 384kHz sampling is being seen. So for the application specific Bode Plot capability a different pathway like the Picoscope 2206B/4224A/524x with FRA software seems like the only way forward, but there are some forum comments about the low frequency end that indicate some learning curve is needed.
The ability to power the Pico from a laptop would provide the signal isolation that I presently get using a soundcard and laptop, but maybe you are after safety isolation as well.
The 1Meg inputs on the 2206B/4224A should allow a generic 100:1 probe, but whether such a probe could achieve flat response well out past 100kHz is a concern, and the 20MHz bandwidth seems too low for your need. The 524X uses 200kohm inputs, which makes it a bit more specific to the Pico.
The ability to power the Pico from a laptop would provide the signal isolation that I presently get using a soundcard and laptop, but maybe you are after safety isolation as well.
The 1Meg inputs on the 2206B/4224A should allow a generic 100:1 probe, but whether such a probe could achieve flat response well out past 100kHz is a concern, and the 20MHz bandwidth seems too low for your need. The 524X uses 200kohm inputs, which makes it a bit more specific to the Pico.
Yeah I was looking at 5242D and 5243D which are 60MHz and 100MHz scopes respectively. The one thing that they do that the other devices I am considering don't is the Bode plot option. (They offer high resolution quantization over most of the frequency range I would use one which is another reason I was interested.)
According to Pico the 5242D and 5243D have standard 1M ohm input impedance and can be used with most standard 1X/10X probes.
The software is very unintuitive and I need to find a manual for the software before I would commit to this device. I also will need to figure out how to deal with common mode noise from the Nuc based instrumentation server - it won't be used with anything else. Using a high isolation medical isolation transformer helped a lot with the other instrumentation I use with that computer so perhaps it is enough.
Edit: I found the software manual so I will read through that.
According to Pico the 5242D and 5243D have standard 1M ohm input impedance and can be used with most standard 1X/10X probes.
The software is very unintuitive and I need to find a manual for the software before I would commit to this device. I also will need to figure out how to deal with common mode noise from the Nuc based instrumentation server - it won't be used with anything else. Using a high isolation medical isolation transformer helped a lot with the other instrumentation I use with that computer so perhaps it is enough.
Edit: I found the software manual so I will read through that.
Last edited:
The Siglent 1104-XE with their arb generator can also do Bode plots. here is a hack to enable using different Arb's with the Siglent bode feature. Bode plots with a scope or audio setup will be limited in the dynamic range if thats an issue. Thats where network analyzers excel. I use an Omicron Bode for that when needed myself. For the high frequency end the NanoVNA might be excellent. You would need to make some adapters and no real need for the extended HF range (5 GHz on the one I have). Definately cheaper than any other option above 100 KHz.
The digital scopes like the Siglent do take some practice and a learning curve to get utility and they have features (like waveform and spectrum on the same display) that are remarkable once you master the tricks. I have not fired up my 7854 in months, and only use it for exceptional tests.
The digital scopes like the Siglent do take some practice and a learning curve to get utility and they have features (like waveform and spectrum on the same display) that are remarkable once you master the tricks. I have not fired up my 7854 in months, and only use it for exceptional tests.
The Bode plot from the Pico 2206B (8-bit) for valve amp application appears to be sufficient to allow identification of gain and phase margins in an amp with circa 20dB feedback, and uses a 3rd party FRA software (still being supported but for 32-bit and up to Win10), but is limited to 1MHz and appears to be about the same cost as the 1104-XE.
It's interesting how different people work and think.
I think PicoScope actually has one of the most intuitive interfaces out there.
In fact, it's the main reason why we got one at work, because the lab guys found all the other scopes way to confusing.
(plus being able to save your measurements directly on a server was a big improvement).
To unfortunate that there regular 4 channel version of the 2000 series isn't being sold anymore 🙁
I've been experimenting more with the software, but of course in demo mode only since I don't own one, but it seems fairly logical.
I know this thread has gone stale, but I just saw it so thought I would add my thoughts.
I have a 5244B that I picked up used on ebay at a 75% discount, and at that price I thought it was a good deal. At the full price there are a ton of other interesting options.
Mine is from the prior generation of these scopes; since it is only usb2 I would expect it is slower than the D models that are usb3. But it is already pretty fast. A large part of my motivation for buying it was that I was looking for a silent scope with a large screen - it didn't take much research to determine that picoscopes were probably my best bet. I already had a 2204a that I had been using for a handfull of years and really liked, so this was a low-risk decision for me. I like a lot about it and usually use in in 12-bit mode so I get the higher resolution but can still see any high-frequency garbage that may indicate instabilities. Having Bode plots to 20 MHz with essentially zero setup time is great as well. Finally, I sometimes like to do more detailed analysis of signals and the picoscope software lets me save data in matlab/octave format; since I am already on the computer I have Octave open and can suck in the data and have at it.
The newer pico software is becoming pretty good, but still has the fatal flaw of not yet including user-defined keyboard shortcuts for all of the settings. But if you ever use non-Windows machines, I would not recommend any picoscope. They supposedly are trying to fix that with the new software, but reports on eevblog seem to indicate they are not successful so far.
If your very small signals are audio frequencies, then a deep-memory 8-bit scope with a true 500 uV/div setting may not be too far behind the 5000D which only have 2 mV/div at which don't get the full resolution. Most modern scopes have resolution enhancement and/or digital lowpass filtering, and these can be used to add resolution at lower frequencies where you want it. Each bit costs a factor of 4 in sample rate of course, so for example a 2 GS/s 8-bit scope gives you 11 bits at 31 MS/s; if this were in a 500 uV/div setting the 5000D would need 13 bits at 2 mV/div to be equivalent. And zooming the vertical scale in the picoscope software is wonky as you don't get divisions at nice intervals in general, so you might actually prefer the non-pico solution anyway.
Another option for low-level signals is a low-noise preamp, of course.
Happy scope hunting!
jason
I have a 5244B that I picked up used on ebay at a 75% discount, and at that price I thought it was a good deal. At the full price there are a ton of other interesting options.
Mine is from the prior generation of these scopes; since it is only usb2 I would expect it is slower than the D models that are usb3. But it is already pretty fast. A large part of my motivation for buying it was that I was looking for a silent scope with a large screen - it didn't take much research to determine that picoscopes were probably my best bet. I already had a 2204a that I had been using for a handfull of years and really liked, so this was a low-risk decision for me. I like a lot about it and usually use in in 12-bit mode so I get the higher resolution but can still see any high-frequency garbage that may indicate instabilities. Having Bode plots to 20 MHz with essentially zero setup time is great as well. Finally, I sometimes like to do more detailed analysis of signals and the picoscope software lets me save data in matlab/octave format; since I am already on the computer I have Octave open and can suck in the data and have at it.
The newer pico software is becoming pretty good, but still has the fatal flaw of not yet including user-defined keyboard shortcuts for all of the settings. But if you ever use non-Windows machines, I would not recommend any picoscope. They supposedly are trying to fix that with the new software, but reports on eevblog seem to indicate they are not successful so far.
If your very small signals are audio frequencies, then a deep-memory 8-bit scope with a true 500 uV/div setting may not be too far behind the 5000D which only have 2 mV/div at which don't get the full resolution. Most modern scopes have resolution enhancement and/or digital lowpass filtering, and these can be used to add resolution at lower frequencies where you want it. Each bit costs a factor of 4 in sample rate of course, so for example a 2 GS/s 8-bit scope gives you 11 bits at 31 MS/s; if this were in a 500 uV/div setting the 5000D would need 13 bits at 2 mV/div to be equivalent. And zooming the vertical scale in the picoscope software is wonky as you don't get divisions at nice intervals in general, so you might actually prefer the non-pico solution anyway.
Another option for low-level signals is a low-noise preamp, of course.
Happy scope hunting!
jason
@jasonRF
I really like the new version 7 software.
Especially zooming-in, setting triggers, loading saved data is so nice.
The zooming function also can be a bit tricky for those who are very use to the time/division setting.
The graph numbers are leading, not the settings, which can be really confusing at first.
Also the math function is FINALLY useful compared to the total garbage from the majority of most scopes.
You can just plugin any formula you would like, combining any input etc.
Their (thermal) loggers are also really nice.
The logging software also works with the scopes I think?
For audio I only use a scope for just the standard stuff, all the more accurate measurements I just do with a good USB audio device these days. 24 bits with SNR of better than 112dB is pretty easy these days!
It's too unfortunate that the 100Mhz versions are so pricey, I need the bandwidth for Class-D amps and some other things 🙁
I really like the new version 7 software.
Especially zooming-in, setting triggers, loading saved data is so nice.
The zooming function also can be a bit tricky for those who are very use to the time/division setting.
The graph numbers are leading, not the settings, which can be really confusing at first.
Also the math function is FINALLY useful compared to the total garbage from the majority of most scopes.
You can just plugin any formula you would like, combining any input etc.
Their (thermal) loggers are also really nice.
The logging software also works with the scopes I think?
For audio I only use a scope for just the standard stuff, all the more accurate measurements I just do with a good USB audio device these days. 24 bits with SNR of better than 112dB is pretty easy these days!
It's too unfortunate that the 100Mhz versions are so pricey, I need the bandwidth for Class-D amps and some other things 🙁
Agree about the picoscope math channels, since they are quite powerful. And using a keyboard to type them in is very convenient. Even then, I still find there are things I need something like Octave to do (eg cross-spectrum analysis).
I Also agree about the price of 100+ MHz models. Compared to Tek scopes they are competitive, but the last couple generations of Rigol/Siglent scopes make Pico look pretty expensive for hobbyists. Business customers probably don’t care as much.
When I was shopping for a picoscope on ebay, i basically was holding out until I found a good deal on a 2-channel 100+ MHz model in my budget, and was expecting to end up with a 2000 or (hopefully) 3000 series. I just lucked out by noticing the 5244b from a seller that accepted my offer that was way below the buy-it-now price.
Cheers
Jason
I Also agree about the price of 100+ MHz models. Compared to Tek scopes they are competitive, but the last couple generations of Rigol/Siglent scopes make Pico look pretty expensive for hobbyists. Business customers probably don’t care as much.
When I was shopping for a picoscope on ebay, i basically was holding out until I found a good deal on a 2-channel 100+ MHz model in my budget, and was expecting to end up with a 2000 or (hopefully) 3000 series. I just lucked out by noticing the 5244b from a seller that accepted my offer that was way below the buy-it-now price.
Cheers
Jason
I'm having a Picoscope 5203 starting back from 2010 and still using it! Changed the fan a couple of times because it was noisy over time but the rest is working properly. Didn't had the time or money to ship it back to Picotech to calibrate it ... yet. Since it's sending data over USB , I'm also using a Intona USB isolator for it upgraded with a linear PSU. The signals are sent to my desk monitor which is far bigger than any other scope I have lying around. In fact, this Picoscope was the main reason why I don't use scopes that are showing signals on a small screen. I guess I'm suffering of claustrophobia... 🙂 I also have a Tek which is configured to send signals using local network to my monitor. If you're going this road then there's no coming back!
Not sure why anyone would think a week old thread was stale.. @jasonRF 🤣 I am glad however that you replied as there are many thought provoking comments, the pre-amp comment - I have a 20dB/40dB low noise pre-amplifier for uses like this, and use an RTX-6001 for spectrum analysis, noise measurements and frequency response. The bode plot function looks interesting since I have been dabbling with bad discrete op-amp design for a couple of years now, and need measurement capability to 20MHz or so. The other reason is space.
I have been experimenting with the Pico scope 7 software, it seems to work pretty well. Demo mode obviously is not reflective
I have not made the decision and check frequently for better deals on eBay, so far not really seeing the right deal.
I have been experimenting with the Pico scope 7 software, it seems to work pretty well. Demo mode obviously is not reflective
I have not made the decision and check frequently for better deals on eBay, so far not really seeing the right deal.
Yeah - most picoscope offerings on ebay are super overpriced.
Regarding high-frequency Bode plotting, the FRA4Picoscope app is very nice and easy to use, and since I bought the 5244b last year whenever I need Bode plotting I almost always fire it up. But I also have been able to write python scripts to do Bode plotting from ~100kHz to 20+ MHz using a lowly Owon vds1022 scope and a tinySA as a signal source (which technically can generate as low as 10 kHz but it is pretty ugly that low), and while it doesn't have the exact same performance as the 5244b it works quite well. If you are comfortable with that kind of thing then it opens up more scope options since most will at least have scpi, as long as you have a signal source that you also can control via scpi or some kind of api. If you want a nice professional Bode plotting tool then this approach doesn't make sense unless you want to invest a lot of time to make one yourself.
Good luck,
jason
Regarding high-frequency Bode plotting, the FRA4Picoscope app is very nice and easy to use, and since I bought the 5244b last year whenever I need Bode plotting I almost always fire it up. But I also have been able to write python scripts to do Bode plotting from ~100kHz to 20+ MHz using a lowly Owon vds1022 scope and a tinySA as a signal source (which technically can generate as low as 10 kHz but it is pretty ugly that low), and while it doesn't have the exact same performance as the 5244b it works quite well. If you are comfortable with that kind of thing then it opens up more scope options since most will at least have scpi, as long as you have a signal source that you also can control via scpi or some kind of api. If you want a nice professional Bode plotting tool then this approach doesn't make sense unless you want to invest a lot of time to make one yourself.
Good luck,
jason
So I pulled the trigger tonight and ordered a Pico Scope 5242D directly from Pico Technology. I did not see anything on eBay that was sufficiently attractive to warrant a purchase, the 5242 and 5243 on offer there were older models and there has been significant improvement and I like the idea of a warranty and knowing that the thing should be fully functional when I get it. That said these are expensive enough I pondered the decision for an excessive period of time.. LOL
I will report back once I've got it and have something worthwhile to say.
I will report back once I've got it and have something worthwhile to say.
Well I like it, and quite a lot.
Internal generator to input A at 1MHz, and 16 bit quantization.
Installed FRA4, this is a bode plot of the scope itself. I think this is pretty reasonable performance for 10MHz measurement bandwidth. Running at 12 bit resolution.
Internal generator to input A at 1MHz, and 16 bit quantization.
Installed FRA4, this is a bode plot of the scope itself. I think this is pretty reasonable performance for 10MHz measurement bandwidth. Running at 12 bit resolution.