ฝัง
- เผยแพร่เมื่อ 15 ต.ค. 2024
- Episode 804
I always assumed I had a 100MHz scope. Maybe it worse, maybe it's better...
I re-measured with a clean sine wave. I get 180MHz at 3db point for two channels. with all four channels on it drops to 100MHz for all four.
Rigol: www.banggood.c...
Hack it: www.instructab...
Be a Patron:
/ imsaiguy
I wondered about the "fix up" for the Rigol, never go around to doing my DS4024, will give it a try some day, I guess it's out of warranty by now. Never really wanted to mess around fearing I would brick the thing. The other week I was comparing two Rubidium oscillators that at 100Mhz (after a 10 clock multiplier) thinking, be cool if I could get the Rigol to display at 500Mhz I would add another 5x multiplier to the testing. I should look to for the fixing data. Thanks for the video, possibly will bump into each other at excess or Anchor some day.
You have to consider that you use square signal at imput. To preserve squere signal integrity, you need bandwidth at least 5 * first harmonic @ 100MHZ its 500 MHZ of bandwidth... This is my rule of thumb.
You are correct. I used the generator since it is small and would fit on screen. I have repeated the tests with a sine wave and get about the same results. Since the first harmonic of the square wave is the sine wave, it was a quick test of the bandwidth.
Thanks for showing the achilles heel of these modern DSO scopes. I have a Siglent 1104X-E, an old Tektronix 784A, and an older yet Tektronix 2230 (analog). I do a lot of high power audio amplifier work, so I do not really approach the limits of any of these oscilloscopes, but, I still prefer my Tek 2230 analog scope for some reason. I can't place my finger on any one thing, but perhaps it is because I was schooled with analog and it is just what I like to use. This happens to be a topic where there are many differing views, from many different perspectives, BTW! Regardless, thanks for the insightful video. Fred
Yes, analog scopes have a different feel. I've gotten used to the digital ones and like them. I still have a Tek 336 that is both analog and sort of digital (line digitizer). much better than persistent phosphor types.
@@IMSAIGuy I have a Tektronix 465M. I like the old scopes but it's nice to be able to grab a one shot signal and freeze it on the screen. I think they had ones that would do that back in the day but they were expensive IIRC. Bought a Rigol 1054z, hacked it, I have no regrets. :)
I like Analog scopes more because of "the feeling". But for serious work, DSO's is the best choice always.
Would love if a manufacturer just simulated the feel of old scopes. And no. I don't like Tektronix version of it.
I think you would get better results if you use the TinySA as a generator since it produces sine waves. I have used it to measure the bandwidth of my Tektronix 465B which is rated at 100 MHz. With the TinySA, it came out as 130 MHz which is reasonable since Tektronix was conservative with their specs. Also , note that dB is dB. What that means is that 3 dB is 3dB whether referred to voltage or power. Thus 3 dB down in power is 1/2. 3 dB down in voltage is .707 which, since it is voltage, is also 1/2 power. Bandwidth is usually defined as 1/2 power which is why you use 0.707 as the factor when applying voltage to measure the bandwidth of an oscilloscope.
The TinySA is ok on low output, but some harmonics. Don't use the high output it is square wave. I repeated the test with my HP8921 (very clean sine wave). I measured 180 MHz with two channels and 100 MHz with all four.
@@IMSAIGuy WOW! That's a great scope!
What signal generator are you using ?
Bandwidth is measured -3dB (voltage) of a sine wave.
The input in the test is probably from an function generator which can't provide the clean sine wave.
The sine wave was from my HP8921. very clean sinewave.
So the scope is displaying an distorted signal.
Maybe an additional instrument is required to find out the reason of the "slow rise time square wave" instead of sine wave.
@@paulcohen1555 opps! sorry I was using a square wave. forgot, I had to re-watch the video. I'll have to repeat with my 8921
I assume it will test the same, at max bandwidth the scope is throwing away all the higher harmonics and leaving the fundamental of the square wave (the leftover sinewave).
@@IMSAIGuy The scope is certainly ATTENUATING the harmonics, but they will still be present and provide extra power on top of the fundamental and provide error. Out of band responses for the input circuitry will also throw it off. Scope input BW measurements should be performed with a clean, stable sinewave.
Hi! Nice video. Could you please generate a TH-cam video on how you "Hacked" your Rigol SA1074Z Scope. I tried to follow your suggested "Instructable" link, but a number of procedures were not clear to me. For example, where did the Telnet part fit into the scope set up? A lot of us would most probably like your special interpretation as to how to Hack your scope! Thanks.
It has been too long. here is a site: www.n5dux.com/ham/riglol/
You just need to type in the information and it gives you a hacked key. You just type it into your oscilloscope and that's it!
What happens when all 4 channels are in use at the same time?
I think the same hardware is used for 50Mhz - 200Mhz versions but the badged 200 is a two channel only; the 1202z unit.
two channels I'm getting 180 MHz 4 channels only 100 MHz
With 1 GSa/s capability, I thought the aliasing wouldn't occur until about 500 MHz. I have an old Tek TDS784A scope that also has 1 GSa/s. I'll have to try the same thing. I think I'll use a sine function instead of the square wave though.
If he turns on a seconds channel that 1GSa/s sampling rate is shared between the two channels, so it's only 500MSa/s. Although in the vid I only see channel 3 enabled.
Is the output of that signal generator leveled ? If it isn't then it's a case of Heisenberg's uncertainty principle. note. I'm sure it's not leveled in any way but that's just feeling not actually looking at the boards specs :-)
I've checked the level.
just tested my DS1102ze not hacked -. rated at 100mhz I was able to get clean sine with internal hardware counter up to 215mhz
Maybe you can turn off the interpolation explicitly on the rigol also and benefit from the vector-mode if there are only a few points. This should make more visible whats going on by conntecting the dots with straight lines instead of showing what looks nices (Everyone likes sinusoidials :-))
Does the "breaking" have anything to do with the screens refresh rate?
no
Hi, thanks. Is there a way of testing and verifying the bandwidth up to 500 MHz without fancy expensive equipment? I don't have a function generator that goes that hi.
you can measure the rise time, but you will need a very fast edge:
th-cam.com/video/iImBK_st5ww/w-d-xo.htmlsi=GfdaOwziodeqIDM1
th-cam.com/video/QCq1REgpBCA/w-d-xo.htmlsi=4USaPHd0c6lIbSTJ
th-cam.com/video/WutBUVs7_z4/w-d-xo.htmlsi=z3SF3c4qa9mxYIgt
@IMSAIGuy thank you for the reply. I watched the video in the link, a few times. Sounds like I am gone have to learn some more before I can obsorb all that there was in the videos.
www.tek.com/en/documents/primer/evaluating-oscilloscopes
Most all digital oscilloscopes do a sinx(x) - Interpolation on the trace. This may help to get a more realistic trace, if you have enough sample-points. If there are only a few sample-points availabe, they fool you because they show sinusoidial traces instead of the real sample values. On my Siglent, this feature can be turned off explicitly. I set the default to disabled.
It looks like your Rigol turns this "feature" off when you switch to point-display.
huh 70% it is -3.09db or -6.18db for a power.
Activate the frequency counter ;-)
I did off camera. It only worked up to a certain point (forgot where).
@@IMSAIGuy Press Measure -> Counter to select CH1-CHx
@@richterd21 Yes, I use it all the time, the counter only works up to a 100mhz then stopped working at the higher frequencies.
I just copied from tek website, better then me trying to explain(www.tek.com/support/faqs/how-bandwidth-related-rise-time-oscilloscopes)
Historically, oscilloscope frequency response tended to approximately follow the rule: Bandwidth x risetime = 0.35. This corresponds to a 1- or 2-pole filter roll-off in the frequency domain. Today, at the high end, most real-time digital oscilloscopes more closely follow this rule:
Bandwidth x rise time = 0.45.
This corresponds to a much steeper frequency roll-off above the specified bandwidth. The steeper roll-off is more desirable in digital oscilloscopes that oversample by 4x, 3x, or even less because it prevents aliasing by eliminating any signal above the Nyquist frequency (1/2 the sample rate - the minimum sample rate required for accurate signal representation).
So if you have a fast pulse gen like leo bodnars 40 pS gen you can calculate bandwidth from risetime, a standard risetime measurement is in every scope nowadays.
The link is not working. I think it must be a rise time measurement. I will be doing a future video on this. I first need to get the equipment to product a very fast edge. It's coming
@IMSAI Guy Here is the link working, sorry about that 🤗 www.tek.com/support/faqs/how-bandwidth-related-rise-time-oscilloscopes
Its just a reference to the piece of text i did put below.
Leo bodnar's 10 mhz/40 pS risetime pulser is fast enough to do this test, not so cheap but a nice piece of engineering with software to adjust the output level, low noise etc
It comes with its own individual test report of performance.
www.leobodnar.com/shop/index.php?main_page=product_info&cPath=124&products_id=295
40 pS equals to 25 ghz so thats several magnitudes over a 1 ghz scope but if you want even more accurate/higher he also has a 30 pS pulser but only with sma connectors and twice the price 😉
If your not familiar with him: have a look at his GPS DO , its a really well developed piece of kit, i use it to reference all my gear to 10 mhz.
Do you have a good used Tektronix with 500mhtz or better? Can't believe you don't, being a lab guy. Thanks for the look.
Haven't found one for a cheap price. Was a 1 GHz Lecroy for $400 but it was too big! I don't have much room. I've been thinking of upgrading my Rigol to a better model.
@@IMSAIGuy Some of the 1GHz-labeled Lecroys require a special adapter to interleave 2 channels into one in order to get the rated 1GHz, otherwise they will only sample at a useful rate up to 500MHz. So do be aware of that if you do get one. They are big. I have an old 500MHz Lecroy 7200A (the one from the 1990s, not the newer 7200a) which is gigantic (and loud), but nice to use. :-)
Most scopes does 500 millihertz just fine, atleast in DC coupling mode.
@@larslindgren3846 🤣