Author Topic: Is this oscilloscope perhaps the best for the beginner?  (Read 4560 times)

0 Members and 1 Guest are viewing this topic.

Offline Johnny B Good

  • Frequent Contributor
  • **
  • Posts: 354
  • Country: gb
Re: Is this oscilloscope perhaps the best for the beginner?
« Reply #50 on: March 27, 2020, 10:28:44 pm »
I have got to get that SDS1104X-E - that’s really nice.
Yes it's the Peak Hold that cleans up all the rubbish down on the noise floor when it's generally of zero interest compared to the fundamental and all its harmonics.

If there's other FFT's you wanna see please ask.

 That snippet of info (Peak Hold) was the major clue as to how you'd managed to 'cleanse' the spectrum of all signs of the sideband pedestals: just submerge them in a sea of raised noise floor artefacts - nice trick!  ;)

 I had to have another read of the user guide before I could usefully experiment further with the suggested acquisition modes and FFT settings listed in the math section of the guide before I was able to approximate your results.

 Initially, I'd thought that perhaps the differences had been due to my use of a cheap 'n' cheerful Feeltech FY6600 signal generator compared to the more expensive Siglent unit you'd employed. Now that I've managed to closely approximate your FFT result, you can hardly see the difference other than for the even harmonics present in the cheap generator's output versus the lack of same in your spectral plots for the Siglent generator which probably just means they're at a much lower level, hidden within that raised up noise floor along with all the sideband noise pedestals.

 I've attached screen captures to show my own version of this test using the aforementioned FY6600 as the test source. The rise and fall times for square edged waveforms is limited to 8ns but there is a little work around to reduce them to just 4ns if the 8ns isn't quite fast enough for the task in hand. AFAICR, for a square wave output on CH1 configure a duplicate square wave on CH2 and use it to 100% amplitude modulate CH1. [Edit] I tried that but couldn't get it to work - there had been a trick to speed up the rise and fall times, I've just simply forgotten the details.  :(

 You might find all of these now tucked out of sight sideband noise pedestals distracting but I don't. Indeed, I found their absence from the plot rather disconcerting to the point where I then had to figure out exactly what settings you'd used (and therefore which ones to avoid) to get such a distorted plot.

 Since you so kindly offered, could you post another FFT plot using more normal settings to capture the more interesting portions such as the actual noise floor and sideband pedestals generated by the Siglent generator? I'd like to compare them to the plots I'm getting from this dirt cheap Feeltech signal generator of mine.

 Also, could you forego the display of the signal waveform being shared in split screen mode? At the 50μs per division timebase setting you're obliged to use with the FFT display, a 1MHz square wave trace provides rather limited information to the point where subtracts rather more from the display than it adds. It's far better to run the FFT function in exclusive mode when capturing screenshots.

 If you feel the need to offer a picture of the waveform being analysed, I think you may as well set the timebase appropriately and capture it as a separate wave trace before readjusting the timebase and switching back to the FFT screen for further examination and analysis.

JBG
« Last Edit: April 02, 2020, 08:27:43 pm by Johnny B Good »
 

Offline tautech

  • Super Contributor
  • ***
  • Posts: 18762
  • Country: nz
  • Taupaki Technologies Ltd. NZ Siglent Distributor
    • Taupaki Technologies Ltd.
Re: Is this oscilloscope perhaps the best for the beginner?
« Reply #51 on: March 28, 2020, 09:50:49 am »
If we are to compare AWG signal purity this needs a different approach and it matters how many bit they each might be. A slow triangle or ramp should distinguish the steps just with a clean sharp trace however if we are to compare sinewaves the setup matters as we are not only looking for harmonics but sidebands and phase noise too.

To do all this with affordable equipment will only get you so far as we are down near the noise floor and as these instruments are digital the clock stability of each also matters unless they have a common reference clock.

As FFT results are dependant on timebase settings and memory depth we always have a tradeoff between these two and the frequency of the sampled signal so to keep as many data points on the display to get the best trace that results from interpolation reconstruction. Further as we examine signals closely with lower Span and dBm/ settings the noise floor also drops so we must keep chasing it with lower and lower Reference settings.

Anyways, some screenshots for study, again from SDG6022X 1MHz sine @ 0dBm into 50 ohm external termination.
Avid Rabid Hobbyist
 
The following users thanked this post: Johnny B Good

Offline digitron

  • Newbie
  • Posts: 2
  • Country: us
Re: Is this oscilloscope perhaps the best for the beginner?
« Reply #52 on: March 29, 2020, 03:39:28 pm »
 

Offline MyHeadHz

  • Regular Contributor
  • *
  • Posts: 163
  • Country: us
Re: Is this oscilloscope perhaps the best for the beginner?
« Reply #53 on: March 29, 2020, 04:28:17 pm »
If you can afford the Rigol DS1054Z it does offer quite a bit more useful functionality.

... especially once you update it with hacked firmware.    :D  (check the usual northern BOLTR channel for details)
 

Offline tautech

  • Super Contributor
  • ***
  • Posts: 18762
  • Country: nz
  • Taupaki Technologies Ltd. NZ Siglent Distributor
    • Taupaki Technologies Ltd.
Re: Is this oscilloscope perhaps the best for the beginner?
« Reply #54 on: March 29, 2020, 07:25:00 pm »
Is this one better than the SDS1104X-E?

https://www.tequipment.net/Rigol/DS1202Z-E/Digital-Oscilloscopes/?b=y&v=7906
Welcome to the forum.

It's a 2 channel DSO and introduced as competition to SDS1202X-E.
Avid Rabid Hobbyist
 
The following users thanked this post: digitron

Offline wizard69

  • Frequent Contributor
  • **
  • Posts: 322
  • Country: us
Re: Is this oscilloscope perhaps the best for the beginner?
« Reply #55 on: March 30, 2020, 02:17:19 am »
 The question of what is cheap is an interesting one.   I'm looking at this from the perspective of somebody entering engineering school, and for many in that situation spending $50 on a scope might be a poor diversion of money.   It is easy to spend a $1000 a year just on books and a laptop is a modern day requirement.   So this is most defiantly dependent on ones financial condition going into school.    In the end after some experience with a cheap scope and with the schools lab equipment, experience will guide the buyer in the right direction.

As for funky used scopes that can be a problem but on the other hand it is also an opportunity to learn how to repair and work with such equipment.   It also prepares you for industry because you don't often have access to the newest and best devices once you leave school.   This is often the biggest issue with graduates, that is their inflexibility when it comes to tools of the trade.

By the way I'm not trying to dismiss the value of a good scope, what I'm trying to point out is that every situation is different.    Even out in industry a scope might not be the first thing you reach for depending upon what your job is.   For the first few years I did a lot of calibration work on process controls and frankly a good DVM was a requirement.   For a student I would certainly recommend a good, high end, handheld DVM that is book bag compatible; if they have money to throw at instrumentation.  In the end almost any electronics related form of employment will make use of a good DVM, that might not be a case for a scope.    I'm just not convinced that getting a scope is the first thing I would suggest for somebody entering engineering/technical school.   Yes they will eventually want one, but I'd flesh out that work bench with other things before the scope.    Frankly I can't see a scope being hugely important in the first year anyways.

Far better to look for a cheap or even free unit until your collection of bench tools is fleshed out.
::)
Define cheap.
And learn on some old clunker that mightn't even be reliable ?  :scared:

That's where I started and sure don't miss needing to fix a scope before you can do anything with it. Been there done that and got the award for it. Never again.

My mentor, a Dr of EE once said of all the tools available to him, if he had to only have one it would be a scope and for a good while I've pondered on this and now know how right he was such has been the advancements in DSO's in the last few years.
 

Offline gcewing

  • Regular Contributor
  • *
  • Posts: 66
  • Country: nz
Re: Is this oscilloscope perhaps the best for the beginner?
« Reply #56 on: March 30, 2020, 03:21:08 am »
I've recently acquired a Siglent SDS1202X-E too. I've only used it for one major project to date, but so far I'm very happy with it. 200MHz, 1GS/s, 2 channels, 14Mpts.

I could also have got 100MHz 4 channels for about the same price, but I decided I'd rather have more bandwidth than more channels. Also I got dedicated knobs for each channel, which I also like.

At NZ$700 it wasn't the cheapest thing I could have got, but it seems very good value for that price and should last me a long while.
 
The following users thanked this post: tautech, Johnny B Good

Offline Johnny B Good

  • Frequent Contributor
  • **
  • Posts: 354
  • Country: gb
Re: Is this oscilloscope perhaps the best for the beginner?
« Reply #57 on: April 02, 2020, 09:42:31 pm »
@tautech

 Thanks for those screenshots. They were rather interesting (in the sense that I had a little difficulty trying to emulate them  :) ). Anyway, I did my best and here are my four best plots. One problem I hit straight away was the -24dB ref limit on the 1dB per division setting. I had to use the 5dB setting (2dB was still too limiting) in order to show the noise floor (the interesting bit  ;) ).

 It's quite obvious that the FFT function built into the SDS1104X-E is an upgrade over the SDS1202X-E's FFT function. I suppose it's only fair seeing as how they reduced its per channel control knobs ration to a quarter of the 1202's allocation of two per channel :( (and still they charge double for the privilege of having four channels on tap >:().

 Considering the front panel real estate that dedicated control knobs would consume, I suppose it's more or less inevitable (at least at the 'affordable price' end of the 'scope market). However, when such multiplexing of control knobs is applied to a two channel DSO such as that Rigol DS1202Z-E that got mentioned earlier, it just cheapens it (and, imo, that Rigol was butt ugly cheap enough to begin with).

 With a four channel 'scope, it's a case of "Needs Must" but with a two channel 'scope, it just looks like a case of "Bean Counteritus" having its wicked way with the UI and design aesthetic. Still, in view of the likelihood that a beginner's next 'scope will likely be a four channel offering cursed with this shortage of control knobs, that cheap Rigol may be not such a bad way to start after all (saves the prospective purchaser becoming spoilt by the luxury of dedicated control knobs  ::) ).

JBG

 
The following users thanked this post: tautech

Offline tautech

  • Super Contributor
  • ***
  • Posts: 18762
  • Country: nz
  • Taupaki Technologies Ltd. NZ Siglent Distributor
    • Taupaki Technologies Ltd.
Re: Is this oscilloscope perhaps the best for the beginner?
« Reply #58 on: April 02, 2020, 10:07:53 pm »
@tautech

 Thanks for those screenshots. They were rather interesting (in the sense that I had a little difficulty trying to emulate them  :) ). Anyway, I did my best and here are my four best plots. One problem I hit straight away was the -24dB ref limit on the 1dB per division setting. I had to use the 5dB setting (2dB was still too limiting) in order to show the noise floor (the interesting bit  ;) ).
Good work and glad to see you're getting to grips with using FFT.  :-+

 
Quote
It's quite obvious that the FFT function built into the SDS1104X-E is an upgrade over the SDS1202X-E's FFT function. I suppose it's only fair seeing as how they reduced its per channel control knobs ration to a quarter of the 1202's allocation of two per channel :( (and still they charge double for the privilege of having four channels on tap >:().
Yes and no !
While X and X-E models share a similar UI that originated from a major FW upgrade in SDS2000 and was implemented in the later SDS1002X models it has continued to evolve as X-E models have substantially more processing power so can comfortably support 1Mpts FFT. Markers of different types and table displays also have been added, none of which would've been possible in the lesser processing power in earlier X models.

Quote
Considering the front panel real estate that dedicated control knobs would consume, I suppose it's more or less inevitable (at least at the 'affordable price' end of the 'scope market). However, when such multiplexing of control knobs is applied to a two channel DSO such as that Rigol DS1202Z-E that got mentioned earlier, it just cheapens it (and, imo, that Rigol was butt ugly cheap enough to begin with).
Multiplexed controls you either love or hate it appears and once I was never a fan of them either however once forced to get used to them you do quite quickly as I discovered when lucky enough to beta test SDS1104X-E.

Quote
With a four channel 'scope, it's a case of "Needs Must" but with a two channel 'scope, it just looks like a case of "Bean Counteritus" having its wicked way with the UI and design aesthetic. Still, in view of the likelihood that a beginner's next 'scope will likely be a four channel offering cursed with this shortage of control knobs, that cheap Rigol may be not such a bad way to start after all (saves the prospective purchaser becoming spoilt by the luxury of dedicated control knobs  ::) ).
It's only market RRP and BOM cost driven.  ;)

Avid Rabid Hobbyist
 

Offline nigelwright7557

  • Frequent Contributor
  • **
  • Posts: 269
  • Country: gb
    • Murton-Pike Systems
Re: Is this oscilloscope perhaps the best for the beginner?
« Reply #59 on: April 03, 2020, 12:51:16 am »
It depends what you want to use the scope for.
If its high speed digital stuff or RF then you are going to have to speed some decent money on a scope.
If its just audio frequencies then even a cheap £20 ebay USB PC scope would do the job.
PCBCAD51/PCBCAD360/PCBCAD720 PCB design software https://www.murtonpikesystems.co.uk
 

Offline Johnny B Good

  • Frequent Contributor
  • **
  • Posts: 354
  • Country: gb
Re: Is this oscilloscope perhaps the best for the beginner?
« Reply #60 on: April 04, 2020, 03:50:36 am »
I've recently acquired a Siglent SDS1202X-E too. I've only used it for one major project to date, but so far I'm very happy with it. 200MHz, 1GS/s, 2 channels, 14Mpts.

I could also have got 100MHz 4 channels for about the same price, but I decided I'd rather have more bandwidth than more channels. Also I got dedicated knobs for each channel, which I also like.

At NZ$700 it wasn't the cheapest thing I could have got, but it seems very good value for that price and should last me a long while.

 Often the bandwidth initially offered, can be hacked upwards, depending on the sampling rates available to all four channels in simultaneous use. A sampling rate of 512MSa/s can comfortably support a 200MHz bandwidth. If that 100MHz 4 channels DSO you'd been considering had a dual 1GSa/s per DAC on its acquisition board like the Siglent 1104X-E, the only impediment to doubling the bandwidth up would be the unavailability of a hacking software package for that model of DSO and the possible need to make a minor hardware modification.

 Hacking your DSO to a wider bandwidth can range from an 'easy peasy' firmware only upgrade to 'a little bit tricky' firmware plus some soldering depending on the make and model involved. If you already have any experience of hacking devices, this may offer a way to get a bigger bang for your buck if you choose your initial purchase with this in mind. In this case, you've chosen so wisely, Siglent have already pre-hacked it for you.  :)

 If you researched this purchase as carefully as I had some eighteen months back, I'm sure you'll be aware of the missing caps issue in the channel acquisition boards that had afflicted the early production run. By the time I ordered mine nearly eighteen months ago, this problem had been fixed in the later versions that were being sold so I've never had to take mine apart.

 Like yourself, I've been extremely pleased with what my £365 (delivered from a UK agent) had bought me. Going on currency exchange rates, you got yours for about the equivalent of £330. The UK agent has only dropped the price by just a fiver which is reassuring - it implies the 2nd hand market isn't awash with used 1202X-E's suggesting that most of their wise owners haven't felt any need to replace it with something better.  :)

 Anyway, I'm about to inflict yet another set of FFT plots upon this thread which I think may be of some interest, if not to your good self, then at least to the OP and any others who may be monitoring the activity here for helpful hints in purchasing a DSO (perhaps like me, only their 2nd 'scope in 40 plus years  ::) ).

 These are power rail noise spectra of the 5.18v output from the 1.3A rated 7805 sized buck converter (a cheap mini 360, iirc, from Banggood) that I'd elected to use over the more prosaic 7805 analogue regulator choice. I discovered that its switching frequency is just over 1.2MHz when the output current is above the threshold that takes it out of 'Hiccup' economy standby mode which is guaranteed to be true in this case (300mA or so once the OCXO is up to temperature).

 I'd wanted to solder an SMA-F socket to the board to properly examine the ripple and noise before completely assembling it but lacked suitable cables and adapters so, after making do with the 'scope probe tip and spring clip ground contact to get a more accurate picture, I boxed it all up to "Suck it and see" how well it would work.

 Earlier this week, the required kit arrived so I pulled it apart to fit an SMA-F socket with 1000nF ceramic cap in series (shunted with a 1K resistor to allow me to measure the DC voltage with a DMM with an error of just -0.01%) with a short co-ax run to the Vcc test point, allowing me to reassemble it and monitor voltage, noise and ripple any time I wanted.

 The FFT is set to show dBV rms meaning -20dB corresponds to 100mV rms, -40dB to 10mV and so on. The half meter of SMA-M to BNC male connecting cable is terminated with a 50 ohm thru-line terminator at the scope. Since the test point impedance is close to zero ohms and directly driving the noise and ripple into the cable, there's no need to apply any 6dB correction factor when interpreting the voltage levels.

 The first six images show ever wider frequency spans starting with a 1MHz close up view of the noise floor. The second encompasses a 2MHz span to reveal the 1.2MHz switching component just poking up above the background noise (the 1MHz spike is part of the noise coming from the logic ICs as is virtually all of the rest of the spikes visible in the remaining images). Hopefully, the additional filtered feed from that point to the OCXO and GPS module (and cmos RRO opamp) will keep most of that digital hash suppressed in this more critical area - I may shift the co-ax connection to the OCXO's +5v pin later on to check this out.

 The final two images demonstrate the effect of applying the 20MHz bandwidth limit as per the now out of date standard, previously used for such noise and ripple traces on equipment now long since past its BBD. With the almost universal use of dc-dc converters in modern equipment which can put out high levels of switching noise transients that can extend into the low hundreds of MHz in the case of dc-dc converters with poor component layout and design, the use of the 20MHz B/W limit is no longer appropriate.

JBG


 

Offline digitron

  • Newbie
  • Posts: 2
  • Country: us
Re: Is this oscilloscope perhaps the best for the beginner?
« Reply #61 on: April 04, 2020, 11:18:36 am »
I just want something to troubleshoot arcade boards, logic, gates, memory, cpu. Mostly 5v stuff...I'm nowhere near you guys on this stuff and I'm not sure if I should get a logic probe or use that money towards a cheap scope? Thx
« Last Edit: April 04, 2020, 11:27:49 am by digitron »
 

Offline Johnny B Good

  • Frequent Contributor
  • **
  • Posts: 354
  • Country: gb
Re: Is this oscilloscope perhaps the best for the beginner?
« Reply #62 on: April 04, 2020, 02:49:41 pm »
I just want something to troubleshoot arcade boards, logic, gates, memory, cpu. Mostly 5v stuff...I'm nowhere near you guys on this stuff and I'm not sure if I should get a logic probe or use that money towards a cheap scope? Thx

 It looks like you're describing vintage kit (5v logic gates etc) which suggests you may be better served with a logic analyser rather than (just) a DSO alone. Although it seems unlikely that '80s and '90s arcade games will be using anything other than the ubiquitous 7805 voltage regulators to provide the 5v Vcc rail voltages, who's to say that such kit couldn't have been repaired or revamped with drop in dc-dc converter replacements (same footprint but no need to bolt them onto a heatsink)?

 Ceramic caps can fail, causing excess noise and ripple which could introduce inexplicable faults in the logic processes you're trying to probe (and even analogue voltage regulators can become oscillators if their decoupling caps go open circuit or high resistance). That cheap Rigol and a separate (also cheap) logic analyser might be a more cost effective investment than a relatively pricey 'all in one' MSO solution.

 Since you will be initially testing the voltage rails for noise and ripple caused by faulty components to avoid going on a wild goose chase with your logic analyser, the effects you're looking for are unlikely to be as subtle as the effects of the dc-dc converter noise and ripple submerged in a sea of logic gate power rail noise as displayed in my own FFT plots. Indeed, they'll likely be apparent enough in the waveform traces without resorting to the FFT function and that being the case, you could buy yourself a 2nd hand worn and battered 100 or 200 MHz B/W DSO with only one working channel to cover such initial examination of the voltage rails, leaving more money in the budget for a decent LA.  ;)

 I have a memory of 'decent' 2nd hand LAs costing upwards of a thousand dollars being discussed in the FY6600 thread last year so did a quick search for "cheap logic analysers" and found this short lived thread from just over 4 years ago:

https://www.eevblog.com/forum/testgear/budget-logic-analyzer-which-one-do-you-recommend/msg854725/#msg854725

 Here's a less older thread discussion (from only two years ago):

https://hackaday.com/2018/01/03/spite-thrift-and-the-virtues-of-an-affordable-logic-analyzer/

 And, here's an example of just how cheap you can go!  ::)

https://www.amazon.co.uk/Hobby-Components-24MHz-Analyser-1-1-16/dp/B00DAYAREW

 If it is vintage arcade kit from the '80s and '90s, that last item might even prove sufficient (at least for your initial requirements). However, only you know your needs so, if you haven't already come across any of those examples above in your own researches, take another long hard look at what's out there. I'm not in the market for a LA so that's as far as I'm going (at least right now ;) ).

 All things considered, a decent MSO could prove to be your best option after all - it depends on your requirements (would just looking at the traffic on an eight bit data bus suffice or will you need to capture 16 bit addresses as well?).

JBG
 
The following users thanked this post: digitron

Offline wizard69

  • Frequent Contributor
  • **
  • Posts: 322
  • Country: us
Re: Is this oscilloscope perhaps the best for the beginner?
« Reply #63 on: April 04, 2020, 04:23:41 pm »
IN a case like this I strongly suggest holding off on your buy until you get a better understanding of the hardware you are working on.    You will want to understand the clock rates of various components so that you can better spec a scope.   Also the mix of analog to digital is important.    Also I wouldn't be surprised at all if some vintage hardware comes in 3.3 volt variants.

As for a logic probe, that might not be a bad investment and frankly is something you can build yourself.   Logic probes can be low cost and remain useful over time.   I wouldn't however suggest their purchase over a good DVM or some other pieces of primary test equipment.    Most people would consider a scope a primary piece of test equipment so there is some conflict there   

If this is a long term investment it might pay to take what ever bandwidth you decide is about right and double it.   The other issue that you will have to consider is what sort of serial port debugging that a potential new scope supports.   While I'm not really involved in arcade systems I have to think that the use of serial ports of various types would be common even in some vintage equipment.

In a nut shell this highlights the problem of suggesting hardware to somebody when you are not highly familiar with the equipment they will be working on.   It isn't easy to say to somebody that XYZ model is all you need when it might not even come close.   In the case here; I could see SPI (or other serial tech) being used to interconnect boards in an arcade system, I just have no idea if I'm right.

I just want something to troubleshoot arcade boards, logic, gates, memory, cpu. Mostly 5v stuff...I'm nowhere near you guys on this stuff and I'm not sure if I should get a logic probe or use that money towards a cheap scope? Thx
 
The following users thanked this post: Johnny B Good, digitron


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf