2 жыл бұрын
2 жыл бұрын
2 жыл бұрын
3 жыл бұрын
4 жыл бұрын
4 жыл бұрын
4 жыл бұрын
4 жыл бұрын
4 жыл бұрын
5 жыл бұрын
5 жыл бұрын
5 жыл бұрын
5 жыл бұрын
Пікірлер
@TrimeshSZ 12 күн бұрын
Be careful, you are starting on the slippery slope to timenuttery...
@citizenrich 17 күн бұрын
Great experiment, thanks.
@LongnoseRob 17 күн бұрын
Nice comparison!
@steubens7 3 ай бұрын
oooh that reminds me, got a laptop that can actually run scopehal now ... need to check it out
@LongnoseRob 3 ай бұрын
Would be good to also compare instant Ramen with fresh Ramen 😂
@AndrewZonenberg 3 ай бұрын
Yeah this was a bit of a rush job given the calendar constraints, I had some other projects on this lab bench until Saturday afternoon. So not enough time to run as many samples as I wanted.
@timeltdme4355 3 ай бұрын
i find the lack of serial decode trigger quite a repeating pattern in Siglent, did you compare this scope to something from R&S ?
@simonbaxter8001 5 ай бұрын
Not sure why Siglent can't expand the software functionality of their lower SDS200X Plus range to include eye diagrams, etc. In todays digital comms world, I don't know why they think adding things like CAN, CA FD, Flexray data decoders is ok at the low end, but actual signal integrity measuring features (which is what the scope is actually for) of the same signals is not even included unless you spend 3x or 4x the price!
@AP-nb7in 6 ай бұрын
Thanks Andrew for this review & sharing your BIG BIG 🙂experience, great. Question : I got a Siglent 2504Xplus (500MHz, 2GS/s). I'm doing "research work" (private) that moved to "high speed" design PCBs and the effect of passive compoments. MSO/ debugging & so on, are not in my "scope", I only focus on time-domain analysis. With my SDS2504Xplus, I'm stuck and face its limits. You mention in the video this 6000A scope misses features vs alternatives from Keysight&other. No brainer, anyone is aware of that 🙂 But, for a private "researcher" LOL, , that can extend the bill to get this SDS6204A at max (12kEur VATincl.) => do you think it's a bargain or not ? vs alternative on the market ... regarding pure time-domain analysis : rise time / eye-diagram / etc... Any reply is welcome 🙂
@wolpumba4099 10 ай бұрын
This is so cool. Can I play with this software without having an oscilloscope? I only have ADALM Pluto, SDRPlay RSP1 and Saleae Logic.
@AndrewZonenberg 10 ай бұрын
We don't have any SDR support yet, but I'd like to add it. The Saleae Logic is also on the wishlist but not currently supported. But you can load data from a lot of different interchange formats (e.g. CSV, raw I/Q, VCD, binary waveform files from most scope vendors) and then use ngscopeclient for protocol decoding. So for example, you could capture activity on a SPI flash chip with the Saleae then use ngscopeclient to decode up to a list of memory addresses and read/write transactions, filter for accesses to a specific address range, etc.
@wolpumba4099 10 ай бұрын
*Summary of the Video* *Introduction* - Review and demonstration of a Bit Error Rate Tester (BERT). - Introduction to NG Scope Client for remote control and signal processing. *Features Explored* - Equalization settings for optimizing link performance. - Real-time Bit Error Rate (BER) and eye pattern measurements. - Use of BERT as either source or sink in tests. *Experiments Conducted* - Testing at various data rates, including 14 gig and 25 gig. - Use of different test patterns like PRBS 7 and PRBS 9. - Experiment with analog emphasis on the transmit side. *Software Interface* - GUI has some rough edges and quirks. - Custom pattern settings affect all ports, not clearly indicated in UI. - Complete remote control possible via API. *NG Scope Client* - Allows for better control and visualization. - Can read out real-time BER measurements at any point in the eye. - Open-source and easy to integrate with new instruments. *Final Thoughts* - Positive about the hardware, but software needs improvement. - Looking forward to more experiments with the BERT. - Invites viewer feedback in the comments. *Positive Findings* 1. *Hardware Quality:* The reviewer is impressed with the BERT hardware. 2. *API Flexibility:* The API allows for complete remote control of the instrument. 3. *NG Scope Client:* This tool enhances the user experience by allowing remote control and signal processing. 4. *Equalization Parameters:* The BERT allows for fine-tuning of equalization settings. 5. *Real-Time Measurements:* The software provides real-time Bit Error Rate (BER) and eye pattern measurements. *Negative Findings* 1. *Software Quirks:* The reviewer finds some rough edges in the software, particularly in the GUI. 2. *User Interface:* The UI could be more intuitive, especially when dealing with custom patterns. 3. *Configuration Resets:* Changing the clock frequency causes the BERT to reset, affecting the outputs. 4. *Documentation:* Lack of clear documentation or guidance on some features. Overall, the reviewer is positive about the hardware but suggests that the software could use some improvements.
@xDevscom_EE 10 ай бұрын
Small box that can! Thank you for sharing, can't wait to see what high speed results you'd get with it. Need to upgrade scopes to 30 GHz now? :)
@XtianApi 10 ай бұрын
I love these videos. I am just a hobbyist with a siglent sds1204x-e, but I watched this for fun, lol. Something is wrong with me
@manfred4283 Жыл бұрын
I would like to try this Software. Any plans to make installation a bit easier like providing deb package or even a PPA? Does it support Siglent waveform files for offline analysis?
@AndrewZonenberg Жыл бұрын
I'm not sure what the Siglent binary waveform file format is. We support Teledyne LeCroy .trc, Tektronix .wfm, and Keysight/Agilent (and some Rigol) .bin. We also support CSV, VCD, raw I/Q, and a few other generic interchange formats. We have experimental deb-format packaging set up by a Debian maintainer, but it's not built automatically during CI builds. Still not what I'd consider release grade yet.
@manfred4283 Жыл бұрын
Thanks for the feedback. Tektronix wfm sounds great as well. Might be an alternative to the Tekscope software we use at work some day. Siglent is able to save csv files. So this should be an option if the bin files are not yet supported. Great project anyway, i'm looking forward to a beta that is ready for endusers. Thanks for your work!
@theelmonk Жыл бұрын
@@manfred4283 When installing applications like this with many dependencies I usually find problems with some of them - perhaps the author has installed more recent packages, or used some latecomer in the shifting sands that is Python. Sometimes it's fixable, sometimes it requires more knowledge than I've got, But this is not one of those, it's very competently built. I followed the build instructions (on Debian 11) with minor variations to suit my preferred directory layout and it all came together without a hitch. I'm hoping working with live data will be an equally pleasant experience. The only suggestion I can add is that it would be great if there were some example saved files to permit experimenting with the interface before tackling the hardware. Perhaps there are and I haven't found them yet.
@karkomass Жыл бұрын
Thanks for the great master class! This content is pure gold. Looking forward to watch additional videos maybe on your open source probes.
@oly463 2 жыл бұрын
It's looking good , thanks for sharing
@oliverkeller2342 2 жыл бұрын
nice review! did your unit also exhibit a strong whining noise, emitted by the fan? its frequency is continuously adapting every few seconds which is annoying in an otherwise silent office
@AndrewZonenberg 2 жыл бұрын
I can't comment on fan noise as my lab is quite loud (about 65 dBA ambient) thanks to all of the servers and other lab equipment. Any sound this scope made was completely drowned out. I actually had to do fairly extensive denoising on the raw audio for it to not sound like I was vacuuming the whole video.
@nanda2902 2 жыл бұрын
Hi Andrew, I'm currently preparing sample for cross section viewing in a Scanning Electron Microscope (SEM). I need a really good and clean cut and sand paper isn't clean enough.. Can you please advise a cleaner method of cutting? Thank you in advance
@AndrewZonenberg 2 жыл бұрын
Hi, Generally whatever you use for the first cut is going to cause a lot of damage. A low speed diamond wire saw or similar will help, but there will still be cracking and possibly some heat effects in the immediate vicinity of the cut. This is why you start by making the initial cut a bit back from the plane of interest, then use increasingly finer abrasives and fairly slow polishing speeds to get closer to your final imaging plane with increasingly higher surface quality. What grades of sandpaper are you using? I usually start around P500 and run out to P10000 before switching to the diamond polishing paste. For low magnification optical microscopy (stereo scope) even around P5000 would likely be enough. For optical microscopy I usually run slightly submicron, say 500 or 250nm diamond. For SEM, depending on magnification, you might want to go all the way to colloidal silica with particle sizes in the tens of nm.
@johannglaser 2 жыл бұрын
Around 16:36, it seems that the vertical resolution of the signal is quite low. Esp. around the extremes, the quantization is clearly visible. Which vertical setting did you use at the scope? 1V/div or even coarser?
@AndrewZonenberg 2 жыл бұрын
It was 73 mV/div. I'm not sure how the fine gain works in this scope since I didn't do a full teardown of the frontend, but the excessive quantization suggests that it's probably *actually* 100 mV/div in hardware and just rescaled for display (so you lose ~1/4 of your total ADC codes offscreen). Additionally, the PCIe test used a probe with 10:1 attenuation so the actual input signal in hardware was 7.3 mV/div (likely actually 10 mV/div).
@johannglaser 2 жыл бұрын
@@AndrewZonenberg Thanks for the details! I've taken the liberty and analyzed an image. I found around 10.7 pixels per LSB, or about 6.5mV per LSB. With an 8-bit ADC this would be approx. 1.65V full scale or close to 200mV/div (assuming 8 div vertically as seen at 6:29). So, considering my inaccuracies, perhaps the scope's lowest true setting is 200mV/div (with your 10:1 probe, or 20mV/div with a 1:1 probe), and all smaller settings are software zoom.
@johannglaser 2 жыл бұрын
Apart from this "observation": Really interesting review. I liked especially when you pointed out the shortcomings, because I learned more on how to use a powerful scope. :-)
@oscarpeters1351 2 жыл бұрын
nice video, I am still unsure which scope i should choose. the new Siglent SDS2104X HD or something else like a Rohde & Schwarz or Tekronix, what do you think of the user interface of the Siglent scope? do you have any suggestions what a good choice might be for an 350MHz 4 channel scope?
@AndrewZonenberg 2 жыл бұрын
I was promised a 2104X+ HD test unit in a week or two so I can see how I like it, and may do a review there as well. I expect it to be mostly the same as the 2104X+ except with the 12-bit ADC instead of 8. The UI is quite nice - smooth and responsive and generally reminds me a lot of a LeCroy scope in terms of general industrial design, except the pop-up menus are on the side rather than the bottom. I wish they supported multiple grids though - this is a feature that most higher end scopes have, and they have the screen real estate to make use of it even on the lower end models like the 2104X+. But some entry level scopes from the big names don't have that capability either.
@oscarpeters1351 2 жыл бұрын
@@AndrewZonenberg thanks, i will wait for the video to pop up on your channel :)
@Ender_Wiggin 2 жыл бұрын
Fuckin baller lab ! i love it
@Ender_Wiggin 2 жыл бұрын
Fucking hell ya this is what i have been looking for. Bring eye diagrams to the masses! Now I need a good inexpensive scope that works with this
@tigerwang202 2 жыл бұрын
awesome !great job👍
@elliot8786 2 жыл бұрын
Thank you, Andrew, for doing this video. It is very interesting and important stuff. From me I wanted to recommend this book from Tektronix: "ABCs of Probes". It is basics on simple probes. Thank you a lot again for doing this work !
@kosmic5695 2 жыл бұрын
I never heard the term "R-C" probes before. They are normally called passive probe and are sold with virtually every oscilloscope still today. Not sure why you are are saying they were more adapted to vacuum tubes and now deprecated ? Also, about the S21 measurement of the passive probe I would double check my results. I did some test (mostly scalar) with a noise source and got excellent results (< +-1dBm) with a Lecroy pp005 up to 500Mhz on a 2GHz Scope. That being said, actives probes were still outperforming the passive one.
@AndrewZonenberg 2 жыл бұрын
I call them R-C divider probes to avoid confusion with transmission line probes, which are also passive (just a resistor). Yes, they're the most common kind of passive probe, but they're not the only kind. They're optimized for extremely low DC loading at the expense of high input capacitance, which is great for really low speed signals with weak drive (a la vacuum tubes) but not the best choice for fast switching signals driven by CMOS push-pull drivers with a fair bit of output current. Both probes have their place but I think for most modern transistorized circuitry a resistive probe is the better passive probe for ~90% of measurements.
@AndrewZonenberg 2 жыл бұрын
As far as flatness goes, I did another test to illustrate my point. Tone at nominal 0 dBm was generated with a Siglent SSG5060X-V, passed through through six feet of RG-188 cable, and to common port of a Mini-Circuits ZFRSC-42-S+ resistive splitter. No level correction was performed to compensate for loss in this cable. Port 1 of splitter goes to thru line on a LeCroy PCF200 test fixture with the far end terminated to 50 ohms, port 2 goes through a Mini-Circuits FL086-24SM+ cable to channel 2 of a LeCroy WaveRunner 8404M-MS oscilloscope in 50 ohm mode. Channel 1 has a PP022 passive probe with spring ground, measuring across the transmission line. Power at port 1 and 2 was measured with the WR8K-SPECTRUM option looking at the highest peak. In this setup, CH1-CH2 shows S21 of the probe (ignoring the <0.3 dB of loss in the FL086-24SM+ cable and the PCF200 fixture) 1 MHz: CH1 = -2 dBm, CH2 = -0.2 dBm, S21 = -1.8 dB 50 MHz: CH1 = -9.5 dBm, CH2 = -1.4 dBm, S21 = -8.1 dB 100 MHz: CH1 = -6.4 dBm, CH2 = -2.3 dBm, S21 = -4.1 dB 200 MHz: CH1 = -12.4 dBm, CH2 = -5.6 dBm, S21 = -6.8 dB 300 MHz: CH1 = -7.7 dBm, CH2 = -7.2 dBm, S21 = -0.5 dB 400 MHz: CH1 = -7.3 dBm, CH2 = -8.9 dBm, S21 = +1.5 dB 500 MHz: CH1 = -13.2 dBm, CH2 = -10.5 dBm, S21 = -2.7 dB
@DrTune 2 жыл бұрын
That was fascinating! thanx
@andrewpeck3450 2 жыл бұрын
Great video, thank you!
@tablatronix 2 жыл бұрын
awesome
@LongnoseRob 2 жыл бұрын
Nice walkthrough on sample-preparation,
@roseelectronics4582 2 жыл бұрын
Quite powerful, impressive! Can you upload more videos of FX-100?
@grahamsutherland1106 3 жыл бұрын
This looks really useful for quickly iterating through design and diagnostic ideas, without spending time and money on hardware revisions! Also shows just how demanding PAM4 can be at these rates.
@212Elevators 3 жыл бұрын
Do you plan to do a full system test, including going into depth about Kidde's interconnection relay (which connects the home alarms to it)?
@AndrewZonenberg 3 жыл бұрын
It was a pretty simple setup. There's a box above the pull station that has a 24V power supply for the horn/strobe and one of the Kidde relay systems. The relay module has a 9V power output which goes to one leg of the pull box, the other leg goes to the home alarm interconnect line so when the box is pulled they all sound and the relay closes. At that point it's just a matter of wiring 24V to one side of the relay and the horn/strobe power to the other side and it's done. I test it monthly.
@212Elevators 3 жыл бұрын
@@AndrewZonenberg Alright here's the other question: If you activate a detector, does the alarm circuit latch (stay on until reset) or is it momentary (resets on its own when the smoke clears)?
@AndrewZonenberg 3 жыл бұрын
@@212Elevators Everything is momentary electrically, however the pull station has a mechanical latch holding the switch closed until you open the box with the hex key to reset.
@212Elevators 3 жыл бұрын
@@AndrewZonenberg Got it. Then finally, the panel sounds like it's a Simplex 4005 or something. What exactly is it?
@AndrewZonenberg 3 жыл бұрын
@@212Elevators No panel, just point to point wiring. Kidde home detectors, Gentex Commander 3, Kidde relay module, Honeywell (I forget the model) pull box, and a mains to 24V power supply. Long term I'm looking at putting a FM-200 or Novec 1230 system in the lab. That will require putting in an actual panel but I need to add smoke dampers to the exhaust fan in the ceiling and some rubber seals to the door - and come up with about $20K in cash - first.
@yannickf.2382 3 жыл бұрын
Sorry but the sound is terrible, thanks
@ChrisSmith-jr6bx 3 жыл бұрын
You have all that expensive equipment. Why not invest in a decent microphone for us to listen to. Can't listen to the rest. Sorry
@AndrewZonenberg 3 жыл бұрын
I've been working on improving the sound setup for a while. Check out some of my most recent videos and you should find it's greatly improved. The high ambient noise level in the lab makes recording there a bit of a challenge. I may re-record some of my older content at some point with better sound quality once I'm satisfied with things.
@decee1157 3 жыл бұрын
That lab is just gorgeous. Huge fan of the HMC8043 Wish i had that setup. Out of everything you got i only got the stickyvise
@gsuberland 3 жыл бұрын
That UI is really polished. Great demo.
@camaronruss9476 3 жыл бұрын
Omg this video randomly played and I thought it was my system
@archiebrown8120 2 жыл бұрын
How those arent fire alarms they are fire buzzers
@dallasrailfanner3870 2 жыл бұрын
@@archiebrown8120 bro you heard fire alarm smoke detectors and fire alarms, not buzzers
@stevenbacon3878 3 жыл бұрын
Very impressive!
@smunaut 3 жыл бұрын
I wish I had any scopes with ref clk inputs, let alone two of them 😅
@AndrewZonenberg 3 жыл бұрын
Only the LeCroy (and the other LeCroy off the left side of the shot that you can't see) is mine, the MSO6 is on loan from a friend to support development. You can still use the sync feature without a shared refclk but there will be more timebase jitter and you might see drift in longer acquisitions due to slightly different sample rates. The only hard requirement is that the primary instrument have a trigger output. You don't even need an external trigger input on a secondary scope, although if you don't have one you obviously waste a channel with the trigger cascade.
@funtechu 3 жыл бұрын
This is neat!
@xDevscom_EE 3 жыл бұрын
Yay! Looking forward for older Tek's support :)
@codingmarco 3 жыл бұрын
Porting is really easy for glscopeclient, it took me about a weekend to get a basic port of GPIB support and an HP scope from the 90s (I haven't created a pull request though since I did it quite quick-and-dirty)
@victorpreuss 3 жыл бұрын
amazing tool! congratulations
@eljeffy 3 жыл бұрын
one time some kid in like 1st grade pulled the fire alarm at lunch and the teacher hit him with a spatula
@xDevscom_EE 4 жыл бұрын
Nice, sound is better this time! :)
@Exit911YT 4 жыл бұрын
How do you interconnect Fire Alarms and Home Smoke Detectors together?
@AndrewZonenberg 4 жыл бұрын
Standard home smoke detectors go off when they detect smoke (or the test button is pushed) and output a 9V DC signal onto the red interconnect line (relative to mains neutral). Some more advanced models support digital data modulated onto this 9V bus, such as to indicate carbon monoxide conditions etc, but a 9V DC signal always will be interpreted as a fire condition. Kidde sells a relay module which takes in 120VAC and outputs 9V DC plus SPST relay contacts, which I use to actuate the horn/strobe if the alarm system is triggered. The 9V DC output from the module and the interconnect line go to the pull station, which puts 9V on the alarm signal when activated. As of now there's no battery backup for the pull station or horn/strobe, so if grid power is lost these will be nonfunctional, but the other detectors/alarms are on battery and will work fine during a power outage.
@chrizlax 4 жыл бұрын
Really nice snapping window management, and the small text is neat too. Too many applications waste all my monitor space without any increase in usability, yours looks really efficient and fast to get the graphs on the screen you want. I liked the minimal config as well, e.g for the PLL, no need to set anything but the expected frequency. Is autoscaling planned for the stats plots? (e.g BER) Seems it would save you some time?
@AndrewZonenberg 4 жыл бұрын
BER plots are actually hard coded right now to I think 0 to 1e-12 on the Y axis. Pretty much all of the stats plots already autoscale on the Y axis, but since the X axis is shared among plots in a group it's a bit trickier. When you have plots with incompatible X axis units in one group (say volts and picoseconds) it draws against the underlying value without regard to units. Which can lead to some display issues. As far as X axis autoscaling, I try to use that very carefully because it's easy to end up annoying the user. Eye patterns and BER bathtub curves both take over the entire waveform group's X axis and force it to be +/- 1 UI from center. I haven't used it elsewhere yet.
@reps 4 жыл бұрын
Is all processing happening on PC? Impressive software either way!!
@AndrewZonenberg 4 жыл бұрын
Yes. All of the protocol decoding, clock recovery, etc is my own open source code running on the PC. For this demo I'm pulling waveform data from a Teledyne LeCroy WaveRunner 8104-MS. The native data format is essentially raw 8-bit ADC samples plus some metadata for sample rate, trigger offset, etc. Everything past that is my code. Libscopehal also supports Rigol, R&S, Siglent, and Agilent scopes to some degree, although the LeCroy driver is the most well developed and tested since that's what I have in my lab. Nobody's offered up a Tek to test on yet, and we've never tested the Agilent driver on a modern Keysight scope but it might work.
@tablatronix 4 жыл бұрын
Why do you not use UV curable soldermask ? Curious
@AndrewZonenberg 4 жыл бұрын
Honestly? I don't have any on hand. I keep meaning to pick some up and experiment. The epoxy method works quite well from a mechanical/electrical perspective, just hard to get an exact color match with the tints I have.
@tablatronix 4 жыл бұрын
@@AndrewZonenberg Yeah it seems if you were deep in a mod or repair, you do not want to remove the board move it to heat, maybe you have probes attached and maybe a mess all over desk ( maybe lol ), uv flashlight and blast it in place, I have some UV green I have not tested it yet.
@KeithMakank3 4 жыл бұрын
go dr z!
@smunaut 4 жыл бұрын
Nice videos ! Thanks for sharing those tips / technique. Audio is a bit on the quiet side though, so might be worth boosting a bit :p
@AndrewZonenberg 4 жыл бұрын
I actually boosted it as far as I could without clipping *and* applied some range compression to make the narration even louder. The dynamic range is massive between the quiet ambient, me talking, and the louder sounds like the power tools and compressed gas. If I don't mind clipping the spray sounds I could probably add another few dB of gain, though.
@AndrewZonenberg 4 жыл бұрын
Sorry about the black screen during the intro. There was supposed to be a long shot of me talking at the lab bench and somehow that got deleted without me noticing... Guess this is what happens when I edit video at 3AM. Also at 5:20 I meant the bus changes from layers 3 to 4, not 2 to 3.