ฝัง
- เผยแพร่เมื่อ 28 มิ.ย. 2024
- In this episode Shahriar reviews the Analog Discovery Pro 3000 Series, particular the ADP3450 model. The instrument offers a combination of mixed signal oscilloscope and dual-channel arbitrary waveform generator. The WaveForms free software can be used to perform a wide range of measurements. You can find more information about this product at:
digilent.com/shop/analog-disc...
The review is organized at follows:
00:00 - Introductions
00:35 - Instrument overview, functions, specifications and design
04:53 - Detailed teardown and architecture overview
11:17 - WaveForms software, oscilloscope capabilities, catching glitches, trigger functions
19:18 - Mixed-Signal measurements, logic analyzer, cross-domain triggering
22:19 - Serial decode, digital trigger, correlated measurements
24:03 - Waveform generation, synchronized signals, modulations, sweeps, FFT, 3D spectrogram
28:46 - Network analyzer functionality, measuring linearity and frequency response of DUT
33:42 - Impedance analyzer functionality, measuring complex impedance across frequency
36:29 - Additional capabilities, tracer, scripting, etc.
37:39 - Concluding remarks
www.TheSignalPath.com
/ thesignalpath
www.Patreon.com/TheSignalPath - วิทยาศาสตร์และเทคโนโลยี
Great, very thorough and fair review, Shahriar! Do any viewers have any questions that we might be able to answer about the ADP3450?
Great work on this system - especially the software. Are there plans for a higher-end unit? 100Mhz? 200? 500?
A 500Mhz/14bit box with that software would be a great tool for power electronics designers. I understand that is a big leap on the hardware side though - especially at 14bits.
Has any testing been done with the ADP3x50 as a Frequency Response Analyzer in SMPS analysis (e.g. like an Omicron Bode 100)? Can the Network Analyzer feature in WaveForms filter all the switching artifacts etc. or is it only meant for linear circuits?
Could this be used with a laptop (or using the wifi dongle) isolated from the mains? For example, measuring line-referenced circuits.
@@Factory400 we're working on pushing it further... in a few days, we'll be doing a limited release of the ADP5250, which is closer to what you're wanting. Full specs soon, keep an eye out!
@Jeremy L In many cases, yes! Just give it the old "college try". Ain't nothin' to it but to do it, baby!
NIIIIIIICE ive been using the analog discovery 2 for a few years and this looks like they finally answered some of my wishes. great review.
That software looks fantastic, hope they bring out a model with a teeny bit more bandwidth
The bandwidth and real time memory depth is woeful for the price :-(
The software is the same as for the Analog Discovery at a quarter of the price.
The software seems amazingly smooth. Very well put video. Thanks.
I also like how the Waveforms app runs on Windows Mac and Linux including ARM.
Great review! I have and worked with the AD2 (Analog Discovery 2) and well, the software support is the same its just limited in other HW ways. I have to say, I find it quite strange that room for a shield was not foreseen from the beginning - on the AD2 of course there was no space, so there was no point in even trying, but here, the device is so large, a shielding box should have been foreseen from the beginning. Also considering the price tag, you would expect a bit more attention to such things. Also the 50R input termination should not have been a major price adder, and again space was available.
Hey! So, the kind of funny story is that these designs are fairly new to our team. The good news is we realized the error, fixed it, and got the updated versions out en masse! Pobody's nerfect ;)
@@DigilentInc There is always room for an Analog Discovery Pro 2 :D I guess if you collect some more feedback from various users, it would make sense to make an updated version in a bit of time.
@@DigilentInc Kudos for reacting and admitting!
I totally agree. Placing shields over the channels is a well-established standard on scopes. Even on much lower-priced brands such as Rigol. This is essential, and the 14 bit resolution of the ADP only accents this necessity further. I'm quite surprised by the dual-braid GND connection too, which is a little unorthodox.
@@DigilentInc Would you please be able to share some cross-talk metrics of the scope channels?
I have no idea what most of this is but I will learn thank you for the review!
Excellent review - Thanks.
VIEWERS OF SHAHRIAR,
who are you and why are you here?
I'm an electrical engineer from Germany and since I'm mostly spending my time at Excel and PowerPoint, I enjoy the hands-on videos that display a deep understanding of high frequency circuits. It's like second hand lab time and I love it 😃
Been here for a very long time. Shahriar's videos are what made me start to understand the 'voodoo' that is RF in depth. I'm an EE but my day job is writing software. I'm still rooting for this channel to pass 100K subs soon - it's one of the best on the platform.
Web designer/dev with love of electronics on the side. I watch just to hopefully gain some knowledge! Cheers.
Technician, but I like to pretend I'm an EE on the side!
mechatronics-turned-EE, lots to learn.
I'm Shahriar's mother. I check all his videos for accuracy.
:)
Very nice review! I have the Analog Discovery 2 and the Digital Discovery from Digilent and their Waveforms SW is quite nice (available also for linux). I also like the company policy with a strong focus on academic support.
nice presentation, thanks
7:45... It looks like on the left hand side.... they shifted the middle Relays to the left to make room for the mounting screw.
Better software analysis than some of my standalone instruments 😑. I really enjoyed that walkthrough. We are truly living in a golden age of instrumentation.
From the data sheet it seems the AD9648 runs both channels at 125MSPS at the same time, so you should still get the full sample rate with all channels turned on.
Limit the resolution by 2-4 bits and increase the bandwidth by 2-4 and I'm in 🙂
btw: review as good as always 👍
For a moment you really get me going!! I thought you have a Video Signal Decoder... I have tons of security video around me in the 900 Mhz band but I can not see any... would you recommend any device to see what is been broadcast??
The scope test board (eBay item #0023) seems to be this one: Agilent N2918B Eval Board. As of June 2022 it seems not be available anywhere.
Excellent review. Have a comparison with the Picoscope 5000 series for example? Thanks
Very nicely done. Can you provide any information on the mixed signal testing board? I am learning to use the ADP2450 as well as some other test equipment and the board would be ideal as a learning tool . It looks like you got it from ebay, but I am having difficulty in finding it.
Finally Shahriar reviews a test device that I don't have to sell my house to have it!
Hello Shahriar, great and informative video and testing procedure as usual! Thank you for that fantastic channel. Could you please share the dev board you are using for this video? I'd be interested to get something similar. many thanks!
Wow - some really fantastic usability features in the software and the 14bit dynamic range is very nice. I don't have any application at the moment where a 50Mhz bandwidth would not be deadly though.
Exactly - we know that spec-for-spec, the numbers don't match with other benchtop scopes. HOWEVER, we also know that you're getting a more flexible, holistic solution as well. We like to say "Good enough for most things". We need to work on our marketing, though... ;)
is it 14 bits ADC over 2v peak to peak or 4v peak to peak input voltage so what would be the lsb of the oscilloscope in mv?
I recently bought this scope and was disappointed by the extremely limited input buffer in live trace mode. Not sure why that limitation exists when record mode offers so much more memory.
It would be interesting (at least to me!) to have you review one of the inexpensive nanoVNA devices that can be had for $50-$150. I just ordered one for myself since I've decided to get back on HF as an amateur radio operator after many years. The nanoVNA seemed like a great excuse to fill an "antenna analyzer" need vs. some very specific solution. My needs (now) are mostly in the sub GHz range, and these device seem accessible to very many people. Thankfully, I'm not trying to debug mm wave 5G systems with one, though!
I'd love to see a RF focused product in their Discovery line-up with things like spectrum analyzer, vector network analyzer, and etc. Any thoughts on what you'd include for folks learning RF?
Oh wow, I hit send on this before hitting minute 30 -- impressive!
Seams like a pretty thought through product.
It might be a very good choice for someone starting with electronics and searching for an all in one instrument.
But it also seems handy for doing some automated testing of a small scale production environment.
Yeah, we're excited about that aspect of it, for sure!
Can someone speak about the differences between AD Pro, Picoscope and NI Virtual Bench? Seems like all of these have models in the similar class of product and I am not sure which one would be the best value for money.
@16:45 seems like the two signals are slightly out of sync by about 2us. Is that the Scope or the source
where can I get a metal pointer like the one you use? is it steel or aluminum
Looks like a scriber, Aluminium handle with hard metal Tip, used to mark metals in a Workshop..
The software is just mind blowing! I never liked pc based oscilloscopes, but I'd buy this one in a jiffy (if I had the money, lol).
For now I've to live with my hantek DSO5102P.
Analog Discovery and Analog Discovery 2 both enjoy the same software. AD2 is a quarter of the price.
@@nezbrun872 i think I'd rather try to grab their digital discovery since what I actually need now is a decent logic analyzer. My first choice is ds logic plus as that one is quite affordable, but digilent makes better product, I think.
Nice ! 😎
Excellent review as always, I would like to see a direct competitor to this guy please....cheers.
Picoscopes come to mind.
You should have a look into Cleverscope
Great video. Thank you so much. Can you let me know what test board you are using ? I’m thinking about buying adp5250 and would like to follow along using that
It is an old eval board I picked up from eBay. There are a lot like it. Search for Oscilloscope Training Boards. Or evaluation boards.
An excellent demo of the Waveforms software.
I just followed along running your same tests on an Analog Discovery 2 (I happen to have the same N2918A eval/demo board). There's nothing in these demos that you can't do on an AD2! There are some minor differences regarding the sample rate and memory depth (125MHz vs 100MHz, & AD2 has half the memory depth), and the SFDR isn't quite as good, but otherwise everything ran perfectly well.
There's one benefit with the AD2: you can power the demo board from it as the AD2's power supplies go to +/-5V.
The only things I can't figure out are with the software:
1. How do you get taller Logic channels?
2. How do you get the decoded I2C values to show up as a "bar chart"?
3. How do you rotate the 3D Spectrogram?
I have looked at the Digilent instruments before and they possibly would be a good idea for a student lab setup in undergrad education - but they are an overkill for our application and would probably not survive long in a real student environment either. For a typical undergrad engineering lab application probably a 2 MHz analog bandwidth and 8 bit sampling would be enough - in a breadboard-type of environment you can hardly reach higher signal frequencies. For this particular instrument I also dislike the use of a proprietary scripting language (from what I could see) - everyone else i migrating to python now and for good reasons.
I had a student working on a similar project with enough capabilities for this type of environment and we got some quite good results, but of course we were undercritical and the project ended a little more than halfway. It was based on a single STM32 microcontroller and some jelly-bean opamps - total cost could perhaps be as low as 50 USD, low enough for students to afford it themselves, or to give it away/restock failed hardware quickly and cheaply.
Hi Uwe! If you're a fan of Python, we totally agree with using it with WaveForms SDK!
Cool instrument. I have a Digilent Electronics Explorer which is a scope, dmm, wavegen, and LA built into a breadboard, its extremely handy. The waveforms software is so powerful, even if my hardware is a bit less capable. How do you think this new piece compares to the Nat Inst. Virtual Bench? Could they have fit a DMM and some power supplies to it or are they protecting the market of the Virtual Bench, now that NI owns Digilent?
I have one of the first Virtual Benches. I love it. Programmable power supply, MSO, function generator, DMM all in one is useful for unit/component testing electronic hardware. Quickly.
Man, i reckon more work went into software dev for this than the hardware dev. Very impressive.
Hey can you or have you done a video on magnatrons and why they go bad? Like from a microwave?
OMG!!! Reviewing a 50MHz scope??? That's like DC for you isn't it?
Indeed - normally he requires >50Mhz bw from his multimeter
@@DrTune LOL!!!
... and his temperature controlled soldering iron probably uses PWM at 26GHz!
I have ADALM2000 sw is simpler but for quick measurement useful as well.
what is that EBAY #0023 board that you are using? Is it some custom board or is it available for sale?
It’s just a board I picked up from eBay. There are many of these types of circuits available.
@@Thesignalpath When you say 'these types' what do you mean? Are these specifically made for testing? I'd be really interested to find out more.
What is EBay #00023? Is it like the Rigol DK-DS6000 oscilloscope test/demo board?
looks much like an Agilent evaluation kit N2918A
It's the one just before EBay #00024.
Capable tool for a reasonable price. Nicely reviewed.
YMMV, but it seems rather expensive to me, considering an Analog Discovery 2 is a quarter of the price with the same software.
I was looking for an MSO earlier in the month and I seriously considered this. In the end I bought a digital discovery with the full probe complement and a Rigol MSO5000 with the full LA probe and leads. The Rigol probe was almost $400 but it is quite long (1m ?) and looks to be very well made along with lots of grabbers and flying leads. I'm very interested to hear your take on this new pro version of the Analog Discovery nevertheless. Cheers.
Your choice of the MSO5000 is a better one IMHO. I'm not sure why you need the Digital Discovery if you have the Rigol's LA probe kits?
(Spoiler: I have the Digital Discovery and an MSO5000 with LA pod too, but I bought the DD quite some time before the Rigol).
@@nezbrun872 Thanks for the response. I bought the DD because it can act as a pattern generator and for Waveforms and the Scripting. I also have an AD and lots of other Digilent gear/FPGA boards. pmods, etc. so this just added to the menagerie.
7:30 Sorry I think the AD9648 is dual-channel, therefore each of the 4 channels has the full 125 MS/s and the FPGA has a total of 500 MS/s to process.
sadly not many logic analyzers seem to be able to easily combine multiple bits into an integer and plot that as a waveform...
This software can do this.
Their software is best. Keysight should learn something from them!
I have legacy version of it!
👍
I am a little confused how they did not think about isolation between 4 14 bit channels, but no harm no foul i guess
Ya, they did the same mistake in the original analog discovery 2. A lot of cross talk between channels!
Probably some silly mistake like forgetting the mask when sending to the PCB fab.
@@raviteza8 That's inexcusable.
What is competition for it?
I guess Picoscope and RedPitaya. If you can call it competition :P
And Cleverscope
I'm seeing this at USD1800 In UK. Any better deals to be had? Inc vat of course.
There's a 2 channel model for $800, both very much in-line with entry level oscilloscope pricing.
I've generally found that unless you are an accredited US-based educational institution or student, their prices are way beyond what most of would be willing to spend.
Fancy 14-bits :)
I would have loved to have a variable ADC like in the picoscope, selectable between 8,12 and 14 bits for example. The Software is so rich and better than Picoscope. It's a pity, it just goes to 50 MHz and only 125 MS/s. They could have easily chosen a better ADC and gone upto 200MHz.
The design considerations to get up to a few hundred MHz probably would drive the cost way up. For $1,300 USD I'm of the opinion that the hardware and software featured here is great for general purpose low speed analog and digital use.
When scopes like the 1054Z can be had for $350 new, the 14 bit depth, dual arb, dual triggers, and software suite have to make up that already massive price difference. Quadrupling the analog bandwidth would put this into an entirely different market category.
@@kevina.4036 The Picoscope 5000 series has a variable ADC with selectable bitrate. Just by simply selecting a better ADC, and keeping the total data through put the same, I m sure a higher BW and sample rate and can be achieved at the same price. A lot of scopes do that already, for example the new Rigol 5000 series.
Everything gets a bit more complicated at higher frequencies, especially the cost of the probes.
Yes this is the one thing stopping me from buying it.
Do you really need WiFi in a device thst has ethernet? You can always plug it into your 1/2.5/5/10-GBit LAN and use the Lab's WiFi, which is probably much better and much recently updated, than any WiFi module you will put into a device as an afterthought.
You can use that device in the field too, like debugging some stage audio equipment, when you sit with your laptop at the mixer station...
Hmm, is the USB port is USB 2? Having 250 Msps * 14 bps = 3.5 Gbit/s of samples to transfer assuming no compression.
That's not how this device works; it's not a dumb ADC that dumps a firehose of data to the host computer. The FPGA handles the triggering/windowing on the hardware side so the only data that needs to be transmitted to the host computer is the 32k of waveform memory, times the update rate (probably 20-30 times per second, just a guess), times the number of active channels. The specs explicitly say they are USB 2.0 ports.
This device almost makes it up the hill and just as it is about to crest, it decides to give up. These hw specs are so underwhelming. A lot of Chinese devices have been coming out with a lot better specs for a lot less for years now. I don't know what the limiting factor is here. That is the only thing preventing me from pulling the trigger.
Was this video produced in collaboration with or sponsored by Digilent, or any other third party?
No.
900 euros for a student equipment :))
The price is WAY too high. For that price there are much more capable units. You can get a 5000 series Rigol Scope for that price with 8GS and all the other features they offer but better. You can actually get the Rigold 5000 AND have $400 left over. Those $80 Hantek USB scopes have all the same features at 1/14 th the price.
Do the Rigol include a waveform generator?
True, but let me tell you, the Waveforms software is truly a pleasure to use - so powerful and flexible and well-executed. I use it with my Analog Discovery 2 connected to my silent MacBook Pro M1 and the display is beautiful and speed of executing a task is great, easily repeatable/recallable/exportable - and I can hear a pin drop. I have a Rigol 1054z as well, but I prefer to use my AD2 always.
tums down for this oscilloscope , not super exciting as you used to say, sorry Shahriar .
Первыйнах.