Trying to pick an oscilloscope

[Pcrandall]

Hello everybody. FNG here. Hopefully I’m in the right forum. I’ve been reading eevblog on and off for years but finally decided to sign up.

Anyway, I’m having trouble trying to figure out which oscilloscope would be appropriate for me. I’m doing development on the TI c2000 launchpad with the 28069 chip running at 90MHz. I’ll be running a prescaler on it, so I don’t know if I’ll need to see the full 90MHz. I’m gonna be developing switch mode supplies and 3 phase inverters. Since I’m doing 3 phase, I figured a 4 channel would be good but that’s where I’m stuck. I’ve been looking at the Rigol 1054, but I’m not sure if the front end bandwidth would be high enough at 100MHz(hacked). I’ve been looking at the Siglent SDS1104X-E at 200MHz, but that only has 14M memory depth. Is that enough? I’ve also been looking at the Rigol 1202Z-E that’s on sale for $299 right now. And that has a 24M memory depth but only 2 channels. Rigid MSO5000 series would be ultimate, but that would break me.

I just don’t know what to do. My budget is $300 - $600ish. Maybe a little more. Just wondering what the eevblog hive mind has to say.

I appreciate the help.

[rstofer]

I doubt that any pin on a uC will toggle at 90 MHz.  The internal shared bus usually prevents that kind of thing.  I suspect 10 MHz is more reasonable and this is easy to see on a 100 MHz scope.

Then too, you really only need to see that the waveform looks reasonable, you would capture and decode it with a logic analyzer and 200 MHz LAs aren't overly expensive.

The SDS1104X-E, unlocked to 200 MHz, would be the choice today.  I have the DS1054Z unlocked to 100 MHz and it hasn't limited me yet.  I do have an old Tek 485 350 MHz 2 channel scope if I ever need it.  That seems unlikely!

None of your power related experiments are going to even come close to taxing a 100 MHz scope.  You will need 10x or 100x scope probes.   Probably 100x.  It is best if these are not switch selectable.  There's nothing like hammering the input stage with a signal that should have been viewed at 10x but the switch was at 1x.  Check with a DMM before probing!

Consider differential probes for SMPS projects (or so I have been told, I don't do this kind of work).

99.9+% of the time all you want to see is what you have on the screen.  It doesn't take anywhere near 14M points to fill a screen.  Personally, I have never had a reason to concern myself with record length.  Pick a User Manual and read up on the subject.  It wasn't that long ago that K sized record lengths were an advanced feature.  I don't think analog scopes even have that as a concept and we went to the Moon with those.  Yes, there are some storage scopes around...

From the DS1054Z User Manual

Analog channel: standard 12M pts (single-channel), 6M pts (dual-channel), 3M pts (3/4-channel); optional 24 Mpts (single-channel), 12 Mpts (dual-channel), 6 Mpts (3/4-channel)

So, unlocked, the DS1054Z has the highlighted memory size.

From the Siglent Data Sheet

SIGLENT’s new SDS1000X-E Super Phosphor Oscilloscopes feature two
channel and four channel models. The two channel model is available with
a 200 MHz analog bandwidth, a single ADC with 1 GSa/s maximum sample
rate, and a single memory module with 14 Mpts of sample memory. The
four channel scope is available in 100 and 200 MHz models and incorporates
two 1 GSa/s ADCs and two 14 Mpts memory modules. When all channels
are enabled, each channel has sample rate of 500 MSa/s and a standard
record length of 7 Mpts. When only a single channel per ADC is active, the
maximum sample rate is 1 GSa/s and the maximum record length is 14 Mpts.

You can have 2 channels running at 14 Mpts (one per ADC) or two running in a single ADC at 7 Mpts.  So, this is somewhat larger, with multiple channels, than the DS1054Z due to the parallel ADCs and memories.  For a single channel, the Rigol wins.  It's a toss up!

This is not a specification I would spend any time on.

[cypher007]

I’ve not got a great deal of scope experience but I bought one of these a couple of months back for £190.

https://www.ebay.co.uk/itm/291115611819

Believe it can be unlocked to 200mhz.

[bdunham7]

Do you have any experience with either oscilloscopes or the circuits you mentioned?  Do you know what exactly you would use it to look at?  Examining logic circuits running at 90MHz (if there are any on the board at that level is another question) requires a probe that costs more than your budget all on it's own.  Also, you might want to consider an MSO with a logic analyser section.  To safely work on your power inverter you may need multiple HV isolated probes of one sort or another--just having 4 channels isn't enough.  The Siglent SD1104X-E seems to be the go-to entry level scope these days--I have one--but any of the ones you mentioned will work well for you.  Choosing one is just the beginning of your worries.  And why are you worried about 14M memory depth vs 24M or whatever?  It's highly unlikely you will find that to be any sort of limitation.

[Pcrandall]

Thank you for the advice. I do have extensive experience working with high voltage inverters and VFDs. I’m a commercial/industrial refrigeration technician. I’ve installed/set up and diagnosed countless 460 three phase vfds. I completely understand the circuit and how it works and what signals I’m looking for. I doubt I’ll be messing around with the 90MHz signal directly. I’ll be running a prescaler to get the carrier wave to 20KHz. I was into electronics years ago and only used an old analog scope and I don’t even know what it was.  I’m not worried about the 14M depth. I guess I don’t understand that aspect of the scope. Just wanted to make sure 14M was enough. In Dave’s tear down video, he kinda himmed and hawed about that aspect.  I just want to be able to accurately see the carrier and fundamental waves so I can fine tune the coding.

I’m stuck between the Siglent 1104 or the Rigol 1054. I’m leaning more towards the Siglent.

[bdunham7]

Thank you for the advice. I do have extensive experience working with high voltage inverters and VFDs. I’m a commercial/industrial refrigeration technician. I’ve installed/set up and diagnosed countless 460 three phase vfds. I completely understand the circuit and how it works and what signals I’m looking for. I doubt I’ll be messing around with the 90MHz signal directly. I’ll be running a prescaler to get the carrier wave to 20KHz. I was into electronics years ago and only used an old analog scope and I don’t even know what it was.  I’m not worried about the 14M depth. I guess I don’t understand that aspect of the scope. Just wanted to make sure 14M was enough. In Dave’s tear down video, he kinda himmed and hawed about that aspect.  I just want to be able to accurately see the carrier and fundamental waves so I can fine tune the coding.

I’m stuck between the Siglent 1104 or the Rigol 1054. I’m leaning more towards the Siglent.

Just get the Siglent, then.  I've had both.  There are aspects you haven't asked about yet that will likely matter later on where the Siglent is far better. 

Just curious--if you have tested and worked on HV VFDs and such, what instrumentation did you use in that case?  Something like a Fluke 190 series?

[tggzzz]

Firstly the signal's frequency does not indicate the required scope frequency response. What matters in the signal's risetime. If you doubt that, consider a 100ns wide 1pps (i.e. 1Hz) signal - if you think that the 1Hz means you will only need a 5Hz scope, you will be disappointed.

Secondly, if you are probing anywhere "near" the mains, then you need to understand what might happen if you attach a probe's sheild to an "unfortunate" node. Never ever be tempted to "float the scope" by disconnecting the scope's protective mains earth.

Finally, if you are putting probes any where near an off-line SMPS, then you need to be very careful with the type of scope probes.

Putting the second and third points together, you should seriously consider HV differential probes. You can try to synthese the differential aspect using a scope's "invert ch2 and add" function, but that has significant high frequency differential mode problems.

Understand the basic theory and practice and safety hints at https://entertaininghacks.wordpress.com/library-2/scope-probe-reference-material/

[Pcrandall]

Just get the Siglent, then.  I've had both.  There are aspects you haven't asked about yet that will likely matter later on where the Siglent is far better. 

Just curious--if you have tested and worked on HV VFDs and such, what instrumentation did you use in that case?  Something like a Fluke 190 series?

What am I not asking about? I am leaning towards the Siglent, but what do you like about it better? I understand the bandwidth and sample rate and all that. I was only curious about the memory depth.

We always used regular fluke multimeters to check the three phase frequency. We did have some kind of oscilloscope that we busted out once in a while. It was some kind of fluke. We didn’t use it much. Pretty much stayed in the office because it was so expensive.

I’m still designing the boards for the PFC, DC bus, and inverter section. I won’t be probing any high voltage right now. When the time comes, I’ll get something me differential probes and I’ll have my fluke multimeter. For right now, I’ll be working only on the development board to fine tune the coding. I want to see the signals I’m creating and sending to the gate drivers.

[Pcrandall]

Tggzzz

Thank you for the link. I haven’t seen that. There’s a lot of good stuff in there.

I’m having a hard time trying to read what you wrote. The syntax makes it kinda hard to read. But I think I know what you’re saying. The first is about rise time, which I understand. The second was about being careful probing near the mains and don’t ever float the scope. I would never do that.

When I get everything squared away in kicad, I’ll post the board and schematic. I’m pretty much working with Texas Instruments and Infineon chips. I kinda like their circuit designs. I’m working on an interleaved PFC and try to incorporate ZVT. The DC bus will be at 400VDC and then the inverter will drop it down to 200VAC three phase. The inverter side will have TI interlocking gate drivers with Infineon threnchstop IGBTs. It’ll have a 20KHz carrier frequency with a 400Hz fundamental. There’s obviously more to it, but that’s the general gist of things.

But like I said in my post above, For right now I’m only going to be probing the logic and signal side of things. I’m initially only going to be working on the C2000 lauchpad to make sure I have the proper signals feeding the gate drivers.

[james_s]

If you want to poke at line voltage VFDs then a differential probe is a must, don't try to cheat with various hacks. If you need to cut corners to meet the budget, buy a cheaper scope and get at least one HV differential probe. The highest frequencies you're likely to find in a VFD outside of the core of the microcontroller are in the tens of kHz, *maybe* a few MHz for something like a SPI or I2C interface to the operator controls. Any of the 4 channel Rigol, Siglent, etc scopes will meet your needs.

It's easy to get paralyzed by indecision, so don't worry about it too much. Buying a scope that costs a few hundred bucks is not a major life decision, take good care of it and you'll be able to sell it for a substantial portion of the purchase price if you change your mind or find that it's not cutting it. Not too long ago I saw a used DS1054z sell on ebay for a bit more than the cost of a brand new one.    They're so cheap to begin with that they don't depreciate much as long as they're not abused.

[Weston]

If you are doing power electronics stuff you should strongly consider getting a MSO with logic probes, its really useful to be able to probe the control signals without using up analog channels, and its nice to be able to see all of the control signals at once regardless of what nodes you are probing.

[bdunham7]

What am I not asking about? I am leaning towards the Siglent, but what do you like about it better? I understand the bandwidth and sample rate and all that. I was only curious about the memory depth.

We always used regular fluke multimeters to check the three phase frequency. We did have some kind of oscilloscope that we busted out once in a while. It was some kind of fluke. We didn’t use it much. Pretty much stayed in the office because it was so expensive.

I’m still designing the boards for the PFC, DC bus, and inverter section. I won’t be probing any high voltage right now. When the time comes, I’ll get something me differential probes and I’ll have my fluke multimeter. For right now, I’ll be working only on the development board to fine tune the coding. I want to see the signals I’m creating and sending to the gate drivers.

You should read some of the many other threads on this subject and if past experience holds, there will be plenty of input on this thread.  The 'which o-scope' topic seems to come around every few months.  Just a few quick items that you haven't asked about--the Siglent has a spectrum analyzer function that is actually usable at a basic level, whereas the Rigol is essentially joke quality.  The Rigol has some minor but annoying issues regarding how the traces are drawn on the screen whereas the Siglent seems to be fairly accurate in how it represents things.  This is a fairly advanced topic and one you likely won't need to worry about.  The Siglent also has the capability of adding 16 digital channels with an a adapter, so it is an MSO--although the optional add-on costs nearly as much as the scope.  And, if you get a Siglent function generator, you can do frequency response analysis on circuits, even passive ones, fairly directly. 

There's more, but I'm sure someone else will chime in.  B/t/w, those expensive Fluke 4-isolated-channel color scopemeters are pretty much the gold standard for things like VFDs and they are available on eBay fairly reasonably--but probably still out of your budget. 

[Pcrandall]

I apologize for starting this thread. I know there’s a lot of people who ask about which oscilloscope to get. I believe when people ask this question, they’re just trying to get other people’s opinions to see what scope would work for their particular project.

I’ve decided that the Siglent SDS1104X-E is gonna be on my bench pretty soon. The Bode plot function is a game changer. That will certainly help me develop my power supply section.

I called Saelig today and talked to John in tech support for almost 45 minutes about anything and everything.  Great guy and very knowledgeable. He didn’t pressure me on anything. He gave me great professional advice and gave me some ideas on how to approach my project. What’s even better is that they are local. They’re only like 15 minutes away from me. I can go there anytime.

The only problem for me is Transat has it on sale right now for $465. John couldn’t match that. The only thing he could take off is the eevblog coupon for 5%. That’ll just cover some of the taxes for me. So I really don’t know if I should buy the cheaper one from Transcat($465), or buy from Saelig($499) with the great knowledge and support and potentially start building a good relationship with them.

I thank all you guys for the great help and advice. Please forgive me for the redundant questions. I’m still learning.

[rstofer]

I would go with Saelig, the difference in cost wouldn't be a consideration.  Other people are after the lowest possible price.  I think a comfortable working relationship is worth a few bucks.

I'm sure either supplier will do a credible job so it just comes down to flipping a coin.

[bdunham7]

I apologize for starting this thread. I know there’s a lot of people who ask about which oscilloscope to get. I believe when people ask this question, they’re just trying to get other people’s opinions to see what scope would work for their particular project.

I’ve decided that the Siglent SDS1104X-E is gonna be on my bench pretty soon. The Bode plot function is a game changer. That will certainly help me develop my power supply section.

I called Saelig today and talked to John in tech support for almost 45 minutes about anything and everything.  Great guy and very knowledgeable. He didn’t pressure me on anything. He gave me great professional advice and gave me some ideas on how to approach my project. What’s even better is that they are local. They’re only like 15 minutes away from me. I can go there anytime.

The only problem for me is Transat has it on sale right now for $465. John couldn’t match that. The only thing he could take off is the eevblog coupon for 5%. That’ll just cover some of the taxes for me. So I really don’t know if I should buy the cheaper one from Transcat($465), or buy from Saelig($499) with the great knowledge and support and potentially start building a good relationship with them.

I thank all you guys for the great help and advice. Please forgive me for the redundant questions. I’m still learning.

I like to save buck as much as anyone, but I would certainly be willing to pay a little extra to have an actual store within driving distance.  Electronics stores have been drying up since I was a kid. And the difference is $9 + possibly illegally evading sales tax if Transat doesn't collect it.  Go buy it so you can start asking questions about how it works. And start saving up for the signal generator.

[james_s]

On the other hand online shopping has been a godsend for me. I hate going to stores, dealing with traffic, parking, sales people, standing in line, etc. I remember some stores I'd actively look at the floor and make a beeline for the products I was interested in and just quietly hope that I don't get approached by a sales droid. I'd much rather read online reviews and peruse forums like this and then order the item I want online than set foot in a physical store. About the only thing I want to pick out and buy in a physical store is produce and other non-processed foods.

[tautech]

Just get the Siglent, then.  I've had both.  There are aspects you haven't asked about yet that will likely matter later on where the Siglent is far better. 

Just curious--if you have tested and worked on HV VFDs and such, what instrumentation did you use in that case?  Something like a Fluke 190 series?

What am I not asking about? I am leaning towards the Siglent, but what do you like about it better? I understand the bandwidth and sample rate and all that. I was only curious about the memory depth.

We always used regular fluke multimeters to check the three phase frequency. We did have some kind of oscilloscope that we busted out once in a while. It was some kind of fluke. We didn’t use it much. Pretty much stayed in the office because it was so expensive.

I’m still designing the boards for the PFC, DC bus, and inverter section. I won’t be probing any high voltage right now. When the time comes, I’ll get something me differential probes and I’ll have my fluke multimeter. For right now, I’ll be working only on the development board to fine tune the coding. I want to see the signals I’m creating and sending to the gate drivers.

You seem to have a reasonable handle on where you're heading and the risks of venturing into measurements on the hot side.
About the SDS1104X-E mem depth, as you know it uses two 1GSa/s ADC's that each have 14 Mpts of memory support which is shared when you use 2 channels on the same ADC. In summary a total of 28 Mpts is available to be shared over the 4 channels with 14 Mpts available on 2 channels if on separate ADC's.

Looking forward to when you might need to venture onto the hot side, I'd recommend 100:1 probes for simple measurements where ground loops won't be an issue and differential measurements when ground loops will be an issue but like any 3 phase measurements the cost of 3 differential probes can be more than the scope. 

Siglent do a 2ch ISFE option that might well suit your needs .....basic as it is however you can buy two of them for the cost of a single differential probe.
https://siglentna.com/product/isfe-isolated-front-end/

Further discussion on ISFE for a chap needing similar capability starts here:
https://www.eevblog.com/forum/testgear/siglent-sds1204x-e-released-for-domestic-markets-in-china/msg2444394/#msg2444394