The unlocked Rigols do not perform like good 100 MHz oscilloscopes should
That's complete rubbish. An unlocked Rigol behaves exactly like a DS1104Z (which is a good oscilloscope).
just apparently no one really knows .. but still thanks for the answers
in life everything is not easy and everything has its causes and consequences
How _should_ a "good" 100MHz scope perform, then?Measure on the main window, not on the zoom !
I'm wondering how that matters to me, that Rigol measures screen data, along with there only being 7 bits shown on the screen. I guess I should be able to figure that out.
This thread has lost the plot.
What is unknown is if the DS1104Z 100MHz model and unhacked DS1054Z 50MHz model display the same behavior.
But seriously, stop being a smartass and accept that the DS1054Z has been out for years, and that nearly everyone who has one has hacked it, and that there are no reports of damage (or malfunction) due to the hack.
Really. I though I had got insane when I tried Leo Bodnar's pulse generator.
I own both a Siglent SDS1202X-E and a Rigol DS1074Z. The Siglent did the right thing. The Rigol still added some interpolation in dots mode.
Scope is solid.
Really. I though I had got insane when I tried Leo Bodnar's pulse generator.
The Rigol still added some interpolation in dots mode.
Most probably sin(x)/x was on when testing, or some other acquisition settings were different between the two 'scopes.
What is unknown is if the DS1104Z 100MHz model and unhacked DS1054Z 50MHz model display the same behavior.No it isn't.
You know the hardware is identical, right?
How _should_ a "good" 100MHz scope perform, then?Measure on the main window, not on the zoom !
I'm wondering how that matters to me, that Rigol measures screen data, along with there only being 7 bits shown on the screen. I guess I should be able to figure that out.
What, someone doesn't think the zoom window is telling the truth? Or is telling less truth than the unzoomed screen? Other than being compressed vertically?
You know the identical hardware is being used ever so slightly differently in those two models, right?
Sorry if that all sounds a bit like ... it is a bit of that, but also a bit of my desire to see facts and good explanations instead of the typical internet opinions and assumptions. There is too much latter all over the place.
You know the identical hardware is being used ever so slightly differently in those two models, right?
Sorry if that all sounds a bit like ... it is a bit of that, but also a bit of my desire to see facts and good explanations instead of the typical internet opinions and assumptions. There is too much latter all over the place.Aren't you making the same baseless claim, and really adding nothing of what you want to the discussion?
I'm not at all interested in the 1054 vs 1104 debate. Who cares? All that matters is the absolute performance of the instrument, so we don't need to compare the 1104 unicorn, almost nobody has one anyway.
What is unknown is if the DS1104Z 100MHz model and unhacked DS1054Z 50MHz model display the same behavior.
No it isn't.
You know the hardware is identical, right?
Most probably sin(x)/x was on when testing, or some other acquisition settings were different between the two 'scopes.
sin(x)/x is my guess, too. Turn it off if you don't like it.
Borjam's claim that "dots should be samples" is just an unfounded assumption on his part.
PS: Who looks at pulse generators in dot mode anyway? You can't calculate rise times from dots.
If the hacks are so common as some claim, and reverting the hack also claimed to be easy, and those Rigol scopes as popular as some others claim, it should be trivial to get quite a few more results to prove both cases one way or another. (Well, I didn't look what kind of measurement David Hess was looking for, maybe those are not as trivial.)
I'm also trying to understand exactly what David is explaining above, so I provided my interpretation of what a good trace would look like, and the region where the Rigol is not good...
If referring that first part you quoted above, then with that particular part I was not expecting Fungus to prove his previous sentence, I was pointing out the mistake that he is showing only relying on identical hardware -idea, leaving out the different usage. Thus, some of the assumptions, based only on the identical hardware, may not necessarily be valid. (On my claim of different usage: If the hardware is identical, it must be used differently; this is pure logical deduction, otherwise the models would behave exactly the same way with bandwidth as is, and no bandwidth hacks would be needed. I think Fungus knows all that, but just didn't consider it in this context.) (Lets see if I can modify the earlier message to clarify a bit... Done.)
Also, indeed I am not trying to add any Rigol facts to discussion, I'm merely pointing that if there is an argument going, the parties should try to provide more info (not necessarily instantly), especially if the other side has some measurements or other provable facts (potentially) showing otherwise.
Otherwise things end easily like in my mind: one side refers to earlier measurements showing a potential difference, another side replies with "nonsense, identical hardware", and I get to think "what kind of answer is that, the 'identical hardware' doesn't even consider different usage of the hardware, which could explain something, and even if different usage was considered and somehow ending with idea that the scopes should behave exactly the same, the different measurement results (contradiction of another assumption) should still be explained somehow, even if found out to be just user error."
You might not be interested on it [1054 vs. 1104] (and maybe not many others, either), but since it is out there and fungus replied, I tried to point on the weaknesses on Fungus' arguments, in the hopes that something more solid might appear. If nothing happens, no big deal.
Also note that David Hess' question is not a comparison between 1104 and 1054, but comparison between hacked 1054 and both unhacked models. Getting results just from the unhacked 1054 would already be an improvement to the situation. And he has noted the possibility of that (minor) deficiency on that "absolute performance" of the hacked 1054, so that question should then be of interest to you, too, though not necessarily a major interest.
That looks like a different problem if it is one. The change in slope from negative to positive indicates that if there was an overload, the amplifier has recovered by the time you indicate. I kept a copy of the photograph where I first noticed it shown below.
It would be practically impossible for improper transient response calibration to produce that response (it should be wavy instead of straight) and it is way too fast for it to be a thermal tail. It does look like the result of overload due to exceeding the maximum slew rate of a linear amplifier driving a transistor into cutoff or saturation.
I did not think much of it until I saw a second example and a user took measurements of bandwidth versus vertical sensitivity which showed something was going on. The same problem could cause both behaviors.
A hacked 100MHz DS1054Z should be compared to a 100MHz DS1104Z. Interpreting this should be much easier since it is a comparison of supposedly identically performing DSOs. If the 100MHz DS1104Z shows signs of overload from the fast edge, then maybe they do perform identically but that indicates nothing good about the DS1104Z.
I had posted some images of mercury switch generated edge - will be challenging to find.
If referring that first part you quoted above, then with that particular part I was not expecting Fungus to prove his previous sentence, I was pointing out the mistake that he is showing only relying on identical hardware -idea, leaving out the different usage. Thus, some of the assumptions, based only on the identical hardware, may not necessarily be valid. (On my claim of different usage: If the hardware is identical, it must be used differently; this is pure logical deduction, otherwise the models would behave exactly the same way with bandwidth as is, and no bandwidth hacks would be needed. I think Fungus knows all that, but just didn't consider it in this context.) (Lets see if I can modify the earlier message to clarify a bit... Done.)But I do not agree that the hardware is being used differently at all. Generally, there are simply limits placed in software to stop a parameter setting from going beyond a certain constant value. The hack likely just changes those constants from one value to another. For example, software limit the sweep speed setting. I see no way to deduce that a 1054, a 1054 hacked, and an 1104 ALL set to the identical user settings would be 'being used differently.' That seems nonsensical, but perhaps you could fabricate an example for discussion to provoke some open thought?
Wasn't hard. First was a charged capacitor and fast switch. I do not recall what I was doing on the second, but probably something similar with different triggering. What should I be looking for?
We know the design is the same or at least I assume it is. We do not know if Rigol grades or selects them.
A hacked 100MHz DS1054Z should be compared to a 100MHz DS1104Z.
I also took a reading at 250MSa/s. Funny enough, the input divider relay is now actuated between 250 and 245mV/div. The difference of the rise time between the two settings now is neglible but something ugly goes on with the sin(X)/x trace reconstruction. It becomes obvious that even in dot mode (not only vector display), the 'scope uses this correction function (or something else...