Looking for a replacement general purpose oscilloscope for home use

[nctnico]

With the new firmware installed, the autocal behaves the same.   

You can turn off the autocal but it is not sticky.  Once the scope has been power cycled, it will be enabled.   They will store all of the settings as defaults on power down but this function is not included.   My guess is customers would bitch about the inaccuracy.  They may be some INI or registry setting that would overwrite how the autocal works but I have never seen anything published for it.

In the end Lecroy scopes like these aren't made to be used as daily drivers. These are to be considered data acquisition & signal analyses systems in 1 box for which accuracy and repeatability of measurements matters most. Sure you can use them as an oscilloscope up to some point but there are limits due to the relatively simple architecture (completely seperated acquisition & processing logic) Lecroy used.

[moerm]

What's to complain about, you low-life who dared to (AFAIK) never buy TME worth millions of $ from them, and who, to make things worse, dared to "infest" many with a "I like LeCroy" virus?!

I'm an example of your "victims". I always took you to be someone who again and again, intentionally or not, showed a LeCroy scope to be something nice and desirable.

   When I bought the first one, there really was nothing brand new with a decent BW and in my price range.  I was familiar with the brand and bought a used one.  That scope is has many features that you would only find in a high end scope still today.  Then again, there are things a cheap modern scope can do that this scope can't.   Consider math functions were important for physics.  Modern scopes have a lot of digital decoding for example.   

Similar here plus "the targeted at research" was attractive to me. But frankly, if someone like you isn't worth (to them) to be treated well I have only a GFY! for them.

We don't appear to do much in the states and guessing the high end market dried up.  Companies needed to find some way to be profitable.   As they shift their focus to higher volume, lower end test equipment and rebranding, hard to say how it will work out. 

To the former I don't comment as that would risk to get (or be taken by some as) "political".
Re the latter, frankly, I think they missed that train. Siglent nowadays is a very acceptable alternative and I strongly doubt that your (or my) region will be able to compete with chinese wages and costs.

Sad, I really liked LeCroy until I found out what a__holes they can be (generally are towards "small fish"?).

On a more serious note: situations like this one are appalling to me. How a largish corporation dares to clearly not give a flying f_ck about a grown up man/customer, no matter whether direct or indirect, who (AFAIK) never uttered anything in any way unfair about them.

One more reason for us to hope and wait for somewhat equivalent alternatives from Siglent and possibly others. And @Siglent, pay attention! There's an important lesson to learn here. Stay the way you are, continue to listen to customers and to treat them well and fair!

A few years ago, a member here had provided details on how to defeat the licensing with their older X-Stream software.   There were some comments about LeCroy not letting that slide as they still used that software with some of their current products.   And they make money on those licenses.   I suspect at least some of their reluctance to remain more open was due to this little episode.

They won't like how (certainly not only) I react to that. "I don't give a flying f_ck about your needs and wants" is an attitude not reserved for LeCroy ...

It's simple: Once I own something I can do - and do - with it whatever I please and I happen to like "looking" at code (incl. binary) and I also happen to have a not insignificant amount of expertise in crypto.
As long as I don't sell what I discover they can KMA.

I hope, you somehow get ahold of the firmware you need.

Not unexpected the 5 GHz probes are not listed as an accessory for 2 GHz scopes as the 2.5 GHz active probes cover this range.

Turn it whichever way you want, incl. - probably correctly - assuming that I'm sometimes stupid.
But that won't change the fact that I won't consider Siglent active probes a day earlier then when they have a "probe" section in their "product" main menu.
Although I like Siglent a lot, I really do.

[tautech]

Not unexpected the 5 GHz probes are not listed as an accessory for 2 GHz scopes as the 2.5 GHz active probes cover this range.

Turn it whichever way you want, incl. - probably correctly - assuming that I'm sometimes stupid.

You and me both ! 

But that won't change the fact that I won't consider Siglent active probes a day earlier then when they have a "probe" section in their "product" main menu.

Yep there are some inconsistencies across their websites, the US have an Accessories header but none of their other sites seem to. 
Yet we can drill down and see any accessory is in such a list and then find other products that might not be listed on a product page.
Have a squiz here in the US site:
https://siglentna.com/products/accessories/

Although I like Siglent a lot, I really do.

They've been a darn good company to deal with for our now 11 years with them, not perfect in any way but darn good regardless.

[joeqsmith]

With the new firmware installed, the autocal behaves the same.   

You can turn off the autocal but it is not sticky.  Once the scope has been power cycled, it will be enabled.   They will store all of the settings as defaults on power down but this function is not included.   My guess is customers would bitch about the inaccuracy.  They may be some INI or registry setting that would overwrite how the autocal works but I have never seen anything published for it.

I'm torn on this one.
Like you I wish for total control and anything not required for a quick measurement need have the functionality to be set to OFF and remain this way after reboot.

The more serious user of course knows an instrument needs to get to temp equilibrium before taking precise measurements but to have LeCroy holding their hand seems a little OTT just so they won't get complaints of system performance when the scope is still cold.

Do you have a User definable Default to change this behaviour with a single button press immediately after boot ?

It's a scope and 8-bits at that.  Precise measurements may be a bit of a stretch under the best case.      

As for the single button, not that I am aware of.   You  maneuver to the menu, select a tab, disable the autocal.   Typically, turn the scope on.  Maybe after 10 minutes you set it manually, assuming you remember.   I'm not sure how you could have your cake and eat it too.  I can understand why they chose to make it work this way but I can't understand why they don't offer other options.   It really only comes into play when looking at slow signals and forgetting to disable that feature after power up.   If it had some intelligent mode where if the temperature has changed and there are no trigger events for some time, manually trigger.  Of course then I would be pissed I lost my data I captured on the screen.      

In the end Lecroy scopes like these aren't made to be used as daily drivers. These are to be considered data acquisition & signal analyses systems in 1 box for which accuracy and repeatability of measurements matters most. Sure you can use them as an oscilloscope up to some point but there are limits due to the relatively simple architecture (completely seperated acquisition & processing logic) Lecroy used.

Interesting opinion.  While I consider my WaveMaster to be a bit too specialized for my daily use, not so with the WaveBlunder.   Consider my first scope was an old tube type Dumont.  I doubt it had a MHz BW.   The first brand new scope I  purchased for home was a Hitachi analog with a Z80.  It had some digital readouts on the screen, 4-channel, and a 100MHz BW.  My first real use of a DSO was a large Tektronix mainframe where I worked.  All scopes.  Wouldn't have a use for any of them today.   

os·​cil·​lo·​scope ä-ˈsi-lə-ˌskōp
ə-
: an instrument in which the variations in a fluctuating electrical quantity appear temporarily as a visible wave form on a display screen

Now we could split that down from there but I consider any DSO to include data acquisition & signal analyses.   I suspect the limits you mention are  not something I have a use for, or it is my lack of owning a modern scope making me oblivious to them. 

[nctnico]

The most common oscilloscopes have a tight coupling between the acquisition and generating the visible traces (rendering). This allows for dealing with deep memory in a faster way, doing hardware decoding and rendering acquisitions into traces much faster. The way Keysight has constructed their Megazoom Asic based scopes is the complete opposite of how Lecroy oscilloscopes are setup. You'd say they are both oscilloscopes but they are different like a Philips and Torx screwdriver are different.

[tautech]

With the new firmware installed, the autocal behaves the same.   

You can turn off the autocal but it is not sticky.  Once the scope has been power cycled, it will be enabled.   They will store all of the settings as defaults on power down but this function is not included.   My guess is customers would bitch about the inaccuracy.  They may be some INI or registry setting that would overwrite how the autocal works but I have never seen anything published for it.

I'm torn on this one.
Like you I wish for total control and anything not required for a quick measurement need have the functionality to be set to OFF and remain this way after reboot.

The more serious user of course knows an instrument needs to get to temp equilibrium before taking precise measurements but to have LeCroy holding their hand seems a little OTT just so they won't get complaints of system performance when the scope is still cold.

Do you have a User definable Default to change this behaviour with a single button press immediately after boot ?

It's a scope and 8-bits at that.  Precise measurements may be a bit of a stretch under the best case.      

As for the single button, not that I am aware of.   You  maneuver to the menu, select a tab, disable the autocal.   Typically, turn the scope on.  Maybe after 10 minutes you set it manually, assuming you remember.   I'm not sure how you could have your cake and eat it too.  I can understand why they chose to make it work this way but I can't understand why they don't offer other options.   It really only comes into play when looking at slow signals and forgetting to disable that feature after power up.   If it had some intelligent mode where if the temperature has changed and there are no trigger events for some time, manually trigger.  Of course then I would be pissed I lost my data I captured on the screen.      

How much time does your autocal consume ?

I dug out my SDS6204A of which I was darn sure I could disable the QuickCal but now it seems not. 
@~10m a 2s QuickCal is performed.

[alm]

Not unexpected the 5 GHz probes are not listed as an accessory for 2 GHz scopes as the 2.5 GHz active probes cover this range.

Is the 2.5 GHz figure probe-only bandwidth or bandwidth at the probe tip with a supported scope? If the former, then connecting a probe with 2.5 GHz -3dB bandwidth to a scope with 2 GHz -3 dB bandwidth won't give the full 2 GHz bandwidth, and there is a valid argument for using a higher bandwidth probe to maximize the bandwidth with that scope. For example look at this Agilent 1156/7/8A datasheet where they specify recommend a 1.5 GHz probe for a 1 GHz scope, a 2.5 GHz probe for a 1.5 GHz scope and a 4 GHz probe for a 2.25 GHz scope.

[tautech]

Not unexpected the 5 GHz probes are not listed as an accessory for 2 GHz scopes as the 2.5 GHz active probes cover this range.

Is the 2.5 GHz figure probe-only bandwidth or bandwidth at the probe tip with a supported scope? If the former, then connecting a probe with 2.5 GHz -3dB bandwidth to a scope with 2 GHz -3 dB bandwidth won't give the full 2 GHz bandwidth, and there is a valid argument for using a higher bandwidth probe to maximize the bandwidth with that scope.

Good question yet as always RTFM.
P17
https://siglentna.com/wp-content/uploads/dlm_uploads/2021/11/SAP1000_SAP2500_UserManual-EN01A.pdf

[alm]

Good question yet as always RTFM.
P17
https://siglentna.com/wp-content/uploads/dlm_uploads/2021/11/SAP1000_SAP2500_UserManual-EN01A.pdf

Unless the F****** manual is misleading. Can you explain, particularly for the SAP1000, how a 1 GHz bandwidth probe on a 1 GHz scope can have a 1 GHz bandwidth at the probe tip, while it should be -6 dB down at 1 GHz by the definition of bandwidth being the -3 dB point? Either the probe-only bandwidth is overly conservative (the minimum guaranteed bandwidth must be well above 1 GHz), or the bandwidth at the probe tip is way optimistic. If the "1 GHz" probe is actually guaranteed to be > 1.5 GHz probe-only, that would be useful information when using the probe with a scope with a wider bandwidth.

[joeqsmith]

How much time does your autocal consume ?

I dug out my SDS6204A of which I was darn sure I could disable the QuickCal but now it seems not. 
@~10m a 2s QuickCal is performed.

...  Autocal on this scope takes maybe 1 - 3 seconds.  Basically a life time.

So there I am, waiting for my illusive event and the scope's temperature/time/whatever has changed and it wants to autocal but is stuck waiting for the trigger, just like I am.   Then it happens, the elusive trigger that both the autocal and me have been waiting on.  The scope starts the autocal.  A second event happens immediately and it is still running the autocal. 

The most common oscilloscopes have a tight coupling between the acquisition and generating the visible traces (rendering). This allows for dealing with deep memory in a faster way, doing hardware decoding and rendering acquisitions into traces much faster. The way Keysight has constructed their Megazoom Asic based scopes is the complete opposite of how Lecroy oscilloscopes are setup. You'd say they are both oscilloscopes but they are different like a Philips and Torx screwdriver are different.

At least its not cars this time.   But yes, I understand your point except I really have nothing to gauge it on.   In other words,  my perspective would be based on 15+ year old tech.  I have no clue as to what LeCroy, Keysight, or any other scope manufactures offer today.   

One page 1, I showed using the WaveBlunder to decode the SPI bus on a radio used in a cheap multi-meter. 
https://www.eevblog.com/forum/testgear/looking-for-a-replacement-general-purpose-oscilloscope-for-home-use/msg5442182/#msg5442182

That was a pretty clean environment.  Here I showed it sniffing the communications on my bikes, while running.  For my use cases around the house, it's been good enough for most tasks.  I consider it good general purpose scope.  It's just not mechanically very robust and a few other problems.   Maybe a new mid range scope would be better. 
   
https://youtu.be/h9pixcy3GOU

[tautech]

Good question yet as always RTFM.
P17
https://siglentna.com/wp-content/uploads/dlm_uploads/2021/11/SAP1000_SAP2500_UserManual-EN01A.pdf

Unless the F****** manual is misleading. Can you explain, particularly for the SAP1000, how a 1 GHz bandwidth probe on a 1 GHz scope can have a 1 GHz bandwidth at the probe tip, while it should be -6 dB down at 1 GHz by the definition of bandwidth being the -3 dB point? Either the probe-only bandwidth is overly conservative (the minimum guaranteed bandwidth must be well above 1 GHz), or the bandwidth at the probe tip is way optimistic. If the "1 GHz" probe is actually guaranteed to be > 1.5 GHz probe-only, that would be useful information when using the probe with a scope with a wider bandwidth.

Maybe you missed this bit from the Warranted characteristics table:
Bandwidth (with scope) 1 GHz（SDS5104X） 2 GHz（SDS6204A）

[alm]

Maybe you missed this bit from the Warranted characteristics table:
Bandwidth (with scope) 1 GHz（SDS5104X） 2 GHz（SDS6204A）

If you sell scopes, then I think it's important to know that if you have a 1 GHz BW signal generator feeding into a 1 GHz BW probe into a 1 GHz BW scope, then the guaranteed frequency response is -9 dB at 1 GHz, so the -3 dB bandwidth will be substantially less than 1 GHz (about 500 MHz for a classic 1st order Gaussian response). To guarantee -3 dB at 1 GHz all components need to have substantially more than 1 GHz of guaranteed bandwidth. Exactly how much more depends on the shape of the frequency response of each component. You are basically adding up the curves (in dB, multiplying in linear voltage units). If you prefer time domain, then you can also talk about the geometric sum of rise times, but I couldn't find a rise time specification for this probe.

[tautech]

Maybe you missed this bit from the Warranted characteristics table:
Bandwidth (with scope) 1 GHz（SDS5104X） 2 GHz（SDS6204A）

If you sell scopes, then I think it's important to know that if you have a 1 GHz BW signal generator feeding into a 1 GHz BW probe into a 1 GHz BW scope, then the guaranteed frequency response is -9 dB at 1 GHz, so the -3 dB bandwidth will be substantially less than 1 GHz (about 500 MHz for a classic 1st order Gaussian response). To guarantee -3 dB at 1 GHz all components need to have substantially more than 1 GHz of guaranteed bandwidth. Exactly how much more depends on the shape of the frequency response of each component. You are basically adding up the curves (in dB, multiplying in linear voltage units). If you prefer time domain, then you can also talk about the geometric sum of rise times, but I couldn't find a rise time specification for this probe.

You can find some further info on SAP1000 from this post about probe testing:
https://www.eevblog.com/forum/testgear/probe-into-probes-whats-up/msg4678768/#msg4678768

The measured rise times are all there for you to do some simple maths. 
Not datasheet rise times but actual real measurements !

[Performa01]

I think it's important to know that if you have a 1 GHz BW signal generator feeding into a 1 GHz BW probe into a 1 GHz BW scope, then the guaranteed frequency response is -9 dB at 1 GHz, so the -3 dB bandwidth will be substantially less than 1 GHz (about 500 MHz for a classic 1st order Gaussian response).

Textbook doctrins are far too simplistic to reflect the real world. They don't help much when it comes to practical applications.

It would be more helpful to know that neither a (not bandwidth-limited) wideband oscilloscope frontend nor any oscilloscope probe will actually have a first order gaussian response and any serious levelled signal generator will have an amplitude flatness of <1 dB (0.5 dB if you don't exploit its full requency range).

Back in early 2019, I've measured the system bandwidth of an SAP1000 prototype in combination with a pre-production unit of the SDS5104X (Siglents first GHz-DSO). That resulted in ~1.11 GHz:

https://www.eevblog.com/forum/testgear/at-last-siglent_s-sds5054x-touchscreen/msg2165488/#msg2165488

The bandwidth of the DSO frontend with direct coax connection was only very slightly better at 1.17 GHz and I need to stress that the frequency flatness of my pre-production SDS5000X was not optimal yet and the frontend for the final devices has undergone some redesign.

https://www.eevblog.com/forum/testgear/at-last-siglent_s-sds5054x-touchscreen/msg2430978/#msg2430978

[alm]

Textbook doctrins are far too simplistic to reflect the real world. They don't help much when it comes to practical applications.

Are you saying that in the real world, a probe with 1 Ghz -3 dB bandwidth with a scope with 1 GHz -3dB bandwidth will have 1 GHz system bandwidth? Because that's what the warranted specifications say. Any SAP1000 with a probe-only bandwidth of 1.01 Ghz is considered in spec according to the manual, as is any SDS5104X with 1.01 Ghz bandwidth. Is the roll-off is that sharp that these two together will have at least 1 GHz system bandwidth together? Or does Siglent do DSP correction for these probes?

Back in early 2019, I've measured the system bandwidth of an SAP1000 prototype in combination with a pre-production unit of the SDS5104X (Siglents first GHz-DSO). That resulted in ~1.11 GHz:

https://www.eevblog.com/forum/testgear/at-last-siglent_s-sds5054x-touchscreen/msg2165488/#msg2165488

The bandwidth of the DSO frontend with direct coax connection was only very slightly better at 1.17 GHz and I need to stress that the frequency flatness of my pre-production SDS5000X was not optimal yet and the frontend for the final devices has undergone some redesign.

https://www.eevblog.com/forum/testgear/at-last-siglent_s-sds5054x-touchscreen/msg2430978/#msg2430978

So clearly that particular probe and scope are performing better than advertised. Certainly nice for you as a user, but not something one can consistently rely on, otherwise obviously the manufacturer would have put it in the datasheet.

[KungFuJosh]

So clearly that particular probe and scope are performing better than advertised. Certainly nice for you as a user, but not something one can consistently rely on, otherwise obviously the manufacturer would have put it in the datasheet.

It's actually typical of Siglent to modestly rate their toys.

[2N3055]

I am confused as to what what datasheet/manual you are using.
In manual I just downloaded, Siglent clearly states:

SAP1000
Bandwidth (probe only) >1 GHz
Bandwidth (with scope) 1 GHz（SDS5104X)
 
SAP2500
Bandwidth (probe only) >2.5 GHz
Bandwidth (with scope) 2 GHz（SDS6204A）

So unlike what you are saying, it clearly specifies system BW. And states that probe BW is larger to guarantee system BW.

Like Performa I tested too, but with 1GHz SDS6104H12 and SDS3104xHD.
BW is 1GHz combined.

Same is with their differential active probes.
Their 5GHz SAP5000D is
Bandwidth (Probe only) >5 GHz
Bandwidth (with Oscilloscope) 4GHz（SDS7404A）

I didn't test it myself (yet) but Siglent is, if anything, very conservative with BW specifications.
I will test if I have a chance, but I expect it to match BW specs without problems.
My biggest problem is generating test signal of sufficient amplitude flatness up to 5GHz at this moment.
 

[Performa01]

Are you saying that in the real world, a probe with 1 Ghz -3 dB bandwidth with a scope with 1 GHz -3dB bandwidth will have 1 GHz system bandwidth? Because that's what the warranted specifications say. Any SAP1000 with a probe-only bandwidth of 1.01 Ghz is considered in spec according to the manual, as is any SDS5104X with 1.01 Ghz bandwidth. Is the roll-off is that sharp that these two together will have at least 1 GHz system bandwidth together? Or does Siglent do DSP correction for these probes?

The textbook statement was a general one, because especially passive high impedance probes usually do not behave like a low-pass at all, let alone a first order Gaussian one.

Siglent list their active probes as accessory for certain oscilloscopes and then a 1 GHz probe actually means that you get at least 1 GHz system-bandwidth with that probe in conjunction with the scope(s) it belongs to. And that’s not only true for active probes. I guess this is pretty common - try to measure the system bandwidth of your Keysight or Tek oscilloscope and see how the textbook theory applies (not!) – if you use the industry standard test method, that is.

If on the other hand, you take a random probe, just because it has a high bandwidth rating and looks sexy, and use it with a DSO were the HF-compensation of the probe doesn’t fit, you might actually end up with a performance that is worse than with matching probes that have a lower bandwidth rating.

So clearly that particular probe and scope are performing better than advertised. Certainly nice for you as a user, but not something one can consistently rely on, otherwise obviously the manufacturer would have put it in the datasheet.

Up to this point I’ve never seen any serious T&M-gear that would not exceed its specifications. For lower end devices, this is usually even more pronounced. Have you tried to calculate what bandwidth the SAP1000 would have to have according to the textbook, in order to achieve a system bandwidth of 1.11 GHz when the bandwidth of the DSO itself is 1.17 GHz? You should definitely try this exercise 😉

[alm]

If on the other hand, you take a random probe, just because it has a high bandwidth rating and looks sexy, and use it with a DSO were the HF-compensation of the probe doesn’t fit, you might actually end up with a performance that is worse than with matching probes that have a lower bandwidth rating.

I have certainly used active FET probes on instruments like spectrum analyzers and scopes they may not have been designed for. Both higher and lower bandwidth than the probe. Their output are a standard 50 Ohm, after all. Nothing brand-specific about them other than power supply. Does Siglent do HF compensation (DSP correction) on their 50 Ohm probes in the 1-2 GHz range? I know other brands do it for the high bandwidth 20 GHz+ probes, but I wouldn't expect this to be necessary for a 1 GHz probe.

Up to this point I’ve never seen any serious T&M-gear hat would not exceed its specifications. For lower end devices, this is usually even more pronounced. Have you tried to calculate what bandwidth the SAP1000 would have to have according to the textbook, in order to achieve a system bandwidth of 1.11 GHz when the bandwidth of the DSO itself is 1.17 GHz? You should definitely try this exercise 😉

I'm not complaining about the instrument exceeding its published specifications. That is pretty normal indeed. But the specifications are not consistent with each other. To meet the 1 GHz bandwidth at the probe tip, both scope and probe need to substantially exceed (the definition of substantial depends on the steepness of their response curves) 1 GHz bandwidth on their own. So what if I send the SAP1000 to a cal lab. They probably won't have the exact Siglent scope it was made for, so they might use a network analyzer, or a wider bandwidth Siglent scope, to test the bandwidth. They measure the SAP1000 bandwidth to be slightly above 1 GHz as stated in the data sheet, and declare it meeting its specifications. Yet this may not be enough to guarantee it meeting the 1 GHz at the probe tip spec. I expect an "A-brand" to have their specifications in order, and not written by an intern who failed the signal processing course.

[nctnico]

So clearly that particular probe and scope are performing better than advertised. Certainly nice for you as a user, but not something one can consistently rely on, otherwise obviously the manufacturer would have put it in the datasheet.

It's actually typical of Siglent to modestly rate their toys.

It doesn't matter. Warranted specs are the specification you base the certainty of the measurement results on. If you buy 10 oscilloscopes to use in a production line, you aren't going to measure each of them AND/OR trust they won't get worse over time and temperature due to component temperate, drift and aging.

[pdenisowski]

Up to this point I’ve never seen any serious T&M-gear that would not exceed its specifications. For lower end devices, this is usually even more pronounced.

Agree completely.

[KungFuJosh]

It's actually typical of Siglent to modestly rate their toys.

It doesn't matter. Warranted specs are the specification you base the certainty of the measurement results on. If you buy 10 oscilloscopes to use in a production line, you aren't going to measure each of them AND/OR trust they won't get worse over time and temperature due to component temperate, drift and aging.

Yes, you're right, it doesn't matter because the datasheet clearly states that the probes are independently over spec'd, and meet the stated bandwidth connected to the scopes. 🤷