Understanding Entry Level Scopes? (Limitations & Uses)

[Fungus]

99% of what a 'scope is normally used for is looking at the shape of waves on screen. Bandwidth is the only parameter that counts for that and the cheapo Rigol has more bandwidth then just about anything below $1000 or so.

Really that should be the end of the argument, right there. Everything else is secondary.

In that case: you can buy used 500MHz to 1GHz DSOs (Tektronix TDS500 or TDS700 series for example) oscilloscopes for less than $1000.

That would be true - if they were on the list of options being argued over.

[Muxr]

99% of what a 'scope is normally used for is looking at the shape of waves on screen. Bandwidth is the only parameter that counts for that and the cheapo Rigol has more bandwidth then just about anything below $1000 or so.

Really that should be the end of the argument, right there. Everything else is secondary.

In that case: you can buy used 500MHz to 1GHz DSOs (Tektronix TDS500 or TDS700 series for example) oscilloscopes for less than $1000.

I dunno that it's a terrible idea per se. Memory depth is pretty low on those though like 15Kpts? compared to Rigol's 24Mpts. Also those scopes are old now. And for your first scope you probably want something that may not be on its last legs. They also don't have the reparability of the old analog Tek scopes, but that's a whole other topic.

[Wuerstchenhund]

In that case: you can buy used 500MHz to 1GHz DSOs (Tektronix TDS500 or TDS700 series for example) oscilloscopes for less than $1000.

I dunno that it's a terrible idea per se. Memory depth is pretty low on those though like 15Kpts? compared to Rigol's 24Mpts. Also those scopes are old now. And for your first scope you probably want something that may not be on its last legs.

2nd hand is not limited to Tek. I've seen quite a few LeCroy WaveRunner2 LTs going for less than $1k, which come with bandwidths of 350Mhz to 1GHz, 1GSa/s+ sample rates (on each channel, i.e. no half-channel degradation) and with up to 8M memory, plus a ton of math and analysis functions. They're not even overly old (2001 to 2004) and thanks to Iwatsu's (the company who built them) excellent build quality are generall good for many years to come.

[nctnico]

99% of what a 'scope is normally used for is looking at the shape of waves on screen. Bandwidth is the only parameter that counts for that and the cheapo Rigol has more bandwidth then just about anything below $1000 or so.

Really that should be the end of the argument, right there. Everything else is secondary.

In that case: you can buy used 500MHz to 1GHz DSOs (Tektronix TDS500 or TDS700 series for example) oscilloscopes for less than $1000.

I dunno that it's a terrible idea per se. Memory depth is pretty low on those though like 15Kpts? compared to Rigol's 24Mpts. Also those scopes are old now. And for your first scope you probably want something that may not be on its last legs. They also don't have the reparability of the old analog Tek scopes, but that's a whole other topic.

Regarding the TDS500/700 series: The memory depth is more like 1Mpts on the older models! And FFT (up to 32kpts IIRC) with averaging of the results so you at least you can get a stable FFT result. But there are newer oscilloscopes out there for less than $1000 and many nice features except protocol decoding.

[pascal_sweden]

Not all LeCroy / Iwatsu scopes are great though.

Take for example the LeCroy Wavejet Touch 354:

The menu consumes a considerable amount of the screen real estate, and it takes that space even when the menu is hidden:
https://youtu.be/uDBuUI5GZ7Q?t=965

Placement of the buttons is not done wisely:
https://youtu.be/uDBuUI5GZ7Q?t=1392

The update rate is very slow when making adjustments:
https://youtu.be/uDBuUI5GZ7Q?t=1589

The intensity grading is really poor:
https://youtu.be/uDBuUI5GZ7Q?t=1626

There is no decoding for I2C, SPI and UART:
https://youtu.be/uDBuUI5GZ7Q?t=1834

It almost defeats the purpose of having a 5000 USD whizbang scope!
https://youtu.be/uDBuUI5GZ7Q?t=1870

[Muxr]

99% of what a 'scope is normally used for is looking at the shape of waves on screen. Bandwidth is the only parameter that counts for that and the cheapo Rigol has more bandwidth then just about anything below $1000 or so.

Really that should be the end of the argument, right there. Everything else is secondary.

In that case: you can buy used 500MHz to 1GHz DSOs (Tektronix TDS500 or TDS700 series for example) oscilloscopes for less than $1000.

I dunno that it's a terrible idea per se. Memory depth is pretty low on those though like 15Kpts? compared to Rigol's 24Mpts. Also those scopes are old now. And for your first scope you probably want something that may not be on its last legs. They also don't have the reparability of the old analog Tek scopes, but that's a whole other topic.

Regarding the TDS500/700 series: The memory depth is more like 1Mpts on the older models! And FFT (up to 32kpts IIRC) with averaging of the results so you at least you can get a stable FFT result. But there are newer oscilloscopes out there for less than $1000 and many nice features except protocol decoding.

Yeah, that's not bad.. Rigol's FFT is pretty weak. I like Wuerstchenhund's idea of looking at those LeCroy (Iwatsu) scopes, didn't realise they can be found for ~$1000.

[Wuerstchenhund]

Not all LeCroy / Iwatsu scopes are great though.

Take for example the LeCroy Wavejet Touch 354:

The menu consumes a considerable amount of the screen real estate, and it takes that space even when the menu is hidden:
https://youtu.be/uDBuUI5GZ7Q?t=965

Placement of the buttons is not done wisely:
https://youtu.be/uDBuUI5GZ7Q?t=1392

The update rate is very slow when making adjustments:
https://youtu.be/uDBuUI5GZ7Q?t=1589

The intensity grading is really poor:
https://youtu.be/uDBuUI5GZ7Q?t=1626

There is no decoding for I2C, SPI and UART:
https://youtu.be/uDBuUI5GZ7Q?t=1834

Yeah well, it's an upper entry-level scope based on an Iwatsu design from around 2003, so yes, it's no racer. On the other hand it's exactly what certain large industrial customers want, which is why it is still in portfolio. Customers that want a robust, reliable and simple to use decent portable scope that is backed by great service.

And now just for fun, compare the price of the 500Mhz WaveJet Touch and say a 500MHz Keysight DSOX3054T. 

[David Hess]

Out of interest, would you recommend an analog scope as a good complementary scope to a (relatively) cheap DSO?

Would you think horse and buggy is a good complementary to a modern car? 

There's no point in getting an analog scope if you have a half-decent DSO, unless you're into antiques or get one for free or a few bucks as a curiosity item.

A better comparison would be an old manual transmission VW bug with carburetor and points which is viable for basic transportation and eminently maintainable versus a modern vehicle which occasionally suffers firmware crashes and mysterious check engine notifications requiring dealer service.

With care a working 100 MHz dual timebase Tektronix can be found in the US for $50 or a 2230 DSO for $100 but in Australia it will be better to go with a new cheapo Chinese DSO.

That was not a deliberate rhyme; that is my story and I am sticking to it.

[Wuerstchenhund]

Out of interest, would you recommend an analog scope as a good complementary scope to a (relatively) cheap DSO?

Would you think horse and buggy is a good complementary to a modern car? 

A better comparison would be an old manual transmission VW bug with carburetor and points which is viable for basic transportation and eminently maintainable versus a modern vehicle which occasionally suffers firmware crashes and mysterious check engine notifications requiring dealer service.

Well, my favorite comparison is still that the analog scope is the equivalent of measuring voltage with a light bulb 

[marmad]

The biggest limitation of the Rigol DS1054Z compared to its bigger DS2000 cousin is bandwidth and sampling rate.

Have you used both DSOs? The DS2000 is about tenfold faster than the DS1000Z (it has a whole separate FPGA for doing post-processing). As someone that has owned the DS2000 for 4 years and then used the DS1000Z for several months, I found it almost unbearable to use for anything other than the simplest waveform viewing. Turn on virtually any extra functions (especially when 4-channels are running) and the response of the DSO drops to a smail's crawl... it's terrible. Unless you need 4-channels almost exclusively (and most people don't: the majority of oscilloscopes produced have been 2 channels), I'd rather own a fast, 2-channel DSO that actually does all of the things it's supposed to do - and does them well
(EDIT: except FFT - Rigoi's uniformly suck at that).

I can give you some examples of what I mean. Rigol's DS1054Z is loaded with features.. and while the main timebase function of the scope generally performs well.. the serial decoding is implemented poorly. Instead of decoding the actual signal the decoder only decodes from the screen's frame buffer. So if the signal isn't the best and your scale isn't adjusted for maximum resolution you will get decoding errors. That's just one example.

Ok, now I'm fairly certain you've never used the DS2000. Yet another big difference you didn't mention (aside from speed): the DS2000's decoding works really well (I use it all the time). It's fast and it decodes the entire memory, not just display memory. It also decodes while in segmented memory, which brings me to another huge difference between the two DSOs: segmented memory. The DS1000Z's implementation is a useless joke (it's basically the same crappy implementation as the old DS1000E series). The DS2000, OTOH, has real segmented memory capture with timestamps and analysis. They are worlds apart.

Almost anyone that has used a DS2000 would then find it (as I did) almost painful to use a DS1000Z. They are not "cousins" - the DS2000 is an adult; the DS1000Z is a child.

[Gyro]

Well, my favorite comparison is still that the analog scope is the equivalent of measuring voltage with a light bulb 

Yes, we are aware. Been there, done that one to death. 

[Muxr]

The biggest limitation of the Rigol DS1054Z compared to its bigger DS2000 cousin is bandwidth and sampling rate.

Have you used both DSOs? The DS2000 is about tenfold faster than the DS1000Z (it has a whole separate FPGA for doing post-processing). As someone that has owned the DS2000 for 4 years and then used the DS1000Z for several months, I found it almost unbearable to use for anything other than the simplest waveform viewing. Turn on virtually any extra functions (especially when 4-channels are running) and the response of the DSO drops to a smail's crawl... it's terrible. Unless you need 4-channels almost exclusively (and most people don't: the majority of oscilloscopes produced have been 2 channels), I'd rather own a fast, 2-channel DSO that actually does all of the things it's supposed to do - and does them well
(EDIT: except FFT - Rigoi's uniformly suck at that).

I can give you some examples of what I mean. Rigol's DS1054Z is loaded with features.. and while the main timebase function of the scope generally performs well.. the serial decoding is implemented poorly. Instead of decoding the actual signal the decoder only decodes from the screen's frame buffer. So if the signal isn't the best and your scale isn't adjusted for maximum resolution you will get decoding errors. That's just one example.

Ok, now I'm fairly certain you've never used the DS2000. Yet another big difference you didn't mention (aside from speed): the DS2000's decoding works really well (I use it all the time). It's fast and it decodes the entire memory, not just display memory. It also decodes while in segmented memory, which brings me to another huge difference between the two DSOs: segmented memory. The DS1000Z's implementation is a useless joke (it's basically the same crappy implementation as the old DS1000E series). The DS2000, OTOH, has real segmented memory capture with timestamps and analysis. They are worlds apart.

Almost anyone that has used a DS2000 would then find it (as I did) almost painful to use a DS1000Z. They are not "cousins" - the DS2000 is an adult; the DS1000Z is a child.

I have a DS2000 and I don't remember the serial decoding being done in memory. Also I remember DS2000 being pretty sluggish. I've been thinking of upgrading the firmware on it, so I will give it a spin. If you're right that's pretty nice.

[marmad]

I have a DS2000 and I don't remember the serial decoding being done in memory.

It's something they fully implemented a few years ago already.

Also I remember DS2000 being pretty sluggish.

Compared to what? Surely you don't mean compared to the DS1000Z - it's not even close.

[Muxr]

I have a DS2000 and I don't remember the serial decoding being done in memory.

It's something they fully implemented a few years ago already.

Also I remember DS2000 being pretty sluggish.

Compared to what? Surely you don't mean compared to the DS1000Z - it's not even close.

Just sluggish, as in not very pleasant to use in some menus, and even just basic things have input lag, although in general it's a good scope don't get me wrong, certainly usable and for the money great. My old HP 54610B is much more responsive than it (not even comparing it to my RTM 1054). I haven't used the DS1k.

[Wuerstchenhund]

Well, my favorite comparison is still that the analog scope is the equivalent of measuring voltage with a light bulb 

Yes, we are aware. Been there, done that one to death. 

Great 

[marmad]

My old HP 54610B is much more responsive than it (not even comparing it to my RTM 1054).

Well, the fact that your HP is old is not really relevant; just because electronics get less expensive and some new or improved features (e.g. the DS2000 has 14,000x more memory than the 54610B), it doesn't mean all features are improved.

I haven't used the DS1k.

Trust me: if you're complaining about the 'sluggishness' of the DS2k, you won't want to   

[Muxr]

Trust me: if you're complaining about the 'sluggishness' of the DS2k, you won't want to   

Duly noted 

[Muxr]

Well, my favorite comparison is still that the analog scope is the equivalent of measuring voltage with a light bulb 

That intensity grading though   

[David Hess]

Well, my favorite comparison is still that the analog scope is the equivalent of measuring voltage with a light bulb 

It is more like a 2-dimentional magic eye tube with index grading for the 3rd dimension.

As far as measuring voltage, my analog 7603 has no trouble making accurate voltage measurements to 4 digits ... using slideback techniques and an auxiliary voltmeter which is effectively built in.   A normal voltmeter by itself cannot measure voltages over wide bandwidths but combining the two instruments allows something which neither can do alone and a DSO by itself cannot do either.  How often do I actually do this?  Occasionally but almost never where the extra accuracy over a DSO is needed.

If someone needed an inexpensive entry level oscilloscope with high bandwidth, then I would recommend an older analog oscilloscope but in this case given that it is Australia where all the old American HP and Tektronix gear got poisoned by the native wildlife and bandwidth is not a consideration, even I say go with something cheap which means Chinese.

While I have not gotten to evaluate one personally, my recommendation for a step up from the cheapo Chinese stuff would be one of the R&S HMO1202 series DSOs.  I really like what I have seen about how they operate and how they are built.  I wonder though if this should apply to all of the Hameg descended R&S DSOs.

[rstofer]

Out of interest, would you recommend an analog scope as a good complementary scope to a (relatively) cheap DSO?

Would you think horse and buggy is a good complementary to a modern car? 

There's no point in getting an analog scope if you have a half-decent DSO, unless you're into antiques or get one for free or a few bucks as a curiosity item.

Or at this point - would it just better to spend the combined money on a slightly better DSO like the DS2000 series?

Hardly. Get a starter scope like the mentioned Rigol DS1054z (although I'd rather get a GW Instek GDS-1054B which is in the same price range as a more mature alternative, though), and later down the road when you find yourself having outgrown you scope then you might want to start a look at a more advanced big brand scope anyways.

OP, you're in the wrong forum/web site if you think that anybody, other than me, will recommend an analog scope.  OTOH, my well worn, $200, Tektronix 485 has a bandwidth of 350 MHz, far higher than what I would be willing to pay for a digital scope.

Every thing I have done for the last 12 years since I retired has been with that 485.  There was nothing I needed to do that I couldn't do with that scope and, really, there still isn't.  But, my, the digitals sure look pretty - and all those numbers!  Remember this:  Everything you see on a digital scope is reconstituted after some arbitrary number of samples are taken and the data manipulated by a C programmer.  Do the sample hit the important bits?  Beats me!  The programmers can't even get the GUI right, why do I believe their calculations?

The downside to analog scopes is that, generally, they don't have a one-shot mode.  That is the ONLY advantage to the digital scope.  Oh, sure, the measurements are neat but we've been doing measurements for generations before digital came along.  Somehow, we got it done...  But single shot is truly useful when trying to decode an SPI transaction.

For the low frequencies involved with embedded projects or robotics, there really is no need for anything beyond 100 MHz and my 350 MHz is overkill.  But I always liked the Tek 4xx series scopes and I just bought what I wanted.  I remember back to the days with the 455 was king and mainframes really were 'big iron'.  Most of the field service guys had the 60 MHz 455.

I wouldn't recommend buying an analog scope as a backup to the 1054Z until you get a lot farther along and have bought everything else.  It just isn't that important.  OTOH, the old Tektronix scopes were very high quality and very reliable.  Not this Chinese crap!

[David Hess]

OP, you're in the wrong forum/web site if you think that anybody, other than me, will recommend an analog scope.  OTOH, my well worn, $200, Tektronix 485 has a bandwidth of 350 MHz, far higher than what I would be willing to pay for a digital scope.

I can say the same thing about my first 2230 which I picked up for $60 which is far less than I was willing to pay for a modern 100 MHz DSO.  When I decided that I needed 4 channels, I picked up another 2230 for $80.

The downside to analog scopes is that, generally, they don't have a one-shot mode.  That is the ONLY advantage to the digital scope.  Oh, sure, the measurements are neat but we've been doing measurements for generations before digital came along.  Somehow, we got it done...  But single shot is truly useful when trying to decode an SPI transaction.

My 2230s are also DSOs so they have no problems with single shot acquisitions.

But I recognize that people in the US have a big advantage in accessing vintage test equipment at low prices and many users are not technically oriented enough at least initially to maintain an old instrument no matter how much they might learn from the experience.

The photo below showing warm up drift was taken after I picked up a 2232 to keep my 2230s company.  The displays are configured to show a 2k (2048 point) record length on a 1k (1024) point wide display.  Yes, they cannot decode SPI or make automatic measurements but the cursor measurements work fine and 2 GS/s equivalent time sampling is completely adequate for 100 MHz signal integrity testing which often comes down to looking for something weird.

[rstofer]

Besides, they LOOK like oscilloscopes!

[Neganur]

The 1054Z is an 'entry level' scope.  That means it is a good 'first' scope, probably not the 'last' scope.  Clearly, it would be out of place in a very high tech lab unless it was the designated 'beater' scope.  But it certainly meets my needs!

For the sake of argument, a person from Keysight called my msox3104A an entry-level scope.

That stung a bit.

I think the rigol is excellent for what it costs.

[Berni]

For the sake of argument, a person from Keysight called my msox3104A an entry-level scope.

That stung a bit.

I think the rigol is excellent for what it costs.

Well to be fair the x2000/x3000 scopes are technically the "crappiest" scopes that Keysight makes right now. Not that they are bad scopes (I use one regularly and its nice) and they can certainly beat cheap Rigols in areas, The x6000 is what is considered mid-range at Keysight (Tho what you mostly get is more bandwith pretty much, not a lot in the way of features).

[vk6zgo]

Out of interest, would you recommend an analog scope as a good complementary scope to a (relatively) cheap DSO?

Would you think horse and buggy is a good complementary to a modern car? 

There's no point in getting an analog scope if you have a half-decent DSO, unless you're into antiques or get one for free or a few bucks as a curiosity item.

A better comparison would be an old manual transmission VW bug with carburetor and points which is viable for basic transportation and eminently maintainable versus a modern vehicle which occasionally suffers firmware crashes and mysterious check engine notifications requiring dealer service.

With care a working 100 MHz dual timebase Tektronix can be found in the US for $50 or a 2230 DSO for $100 but in Australia it will be better to go with a new cheapo Chinese DSO.

That was not a deliberate rhyme; that is my story and I am sticking to it.

   

A DS1054Z at US$400,at current exchange rates equates to $A526,before freight!
The halcyon days of the "mighty Oz"coming in at 1.3 x the US $ are long gone.

Buying a DS1054Z or similar from an Australian agent won't leave you much change from $A700.
Chinese with freight isn't that much better.

Unfortunately local eBay sellers see these sort of prices & think they can sell their old 100Mhz  analog Teks for $A300 plus!
"Tell 'em they're dreaming!!"