TSP

Рет қаралды 15,459

Күн бұрын

Пікірлер: 43

@garbagebag185 9 ай бұрын

At 43:45 it appears that the same noise waveform is added to both the I and the Q channel simultaneously, making a gaussian distribution in only one axis. The way which makes a circular gaussian distribution to do it is to have two sources of independent noise being separately added to the I and the Q channel.

@Thesignalpath 9 ай бұрын

Yes, it appears the noise is fully correlated in I/Q, which is incorrect.

@P.Barros 9 ай бұрын

Masterclass on how a particular RF design can ruin the signal-to noise ratio in a RF signal generator, and how the theory perfectly matches the measurements made. Just brilliant Shahriar !. I wonder why they decided to place the attenuator so early in the RF chain and not right at the output as most of other vendors do.

@2ftg 9 ай бұрын

It can then be a cheaper one that does not need to handle as much power. Additionally, it will drift less from heating due to less dissipation. Still not the best idea.

@P.Barros 9 ай бұрын

@@2ftg Yes, those could be some of the reasons. They need to handle +15 dBm, we just need the data sheet of those attenuator chips.

@andymouse 9 ай бұрын

Its very well explained at 12:45.

@ctid107 9 ай бұрын

I've always regarded a lot of RF stuff as something of a black art, however your explanations are so good that I experience quite a few "now i get it" moments in all your videos. Thanks

@Xsiondu 9 ай бұрын

It is absolutely sorcery! But we stand on the shoulders of giants who left us very detailed books full of theory and pictures.

@boy442985682 9 ай бұрын

Beautifully informative as always. In an industry where a lot of this knowledge is gained via experience over many years, or kept under lock and key by senior engineers or academics for job security, or just because it’s hard to find time to learn some of this until you need it, This channel continues to make me a better young engineer in a world where we’re going need more young engineers very quickly. Thanks again!

@PROWEB7 6 ай бұрын

WAW, very nice vidéo, thank you, I don't understand why you don't have more subscribers! I'm sorry about that, your videos are wonderful

@trickyrat483 9 ай бұрын

Damn, you've left them a lot to think about. I'll wait for version 2.0. :)

@MIsam-fv9kb 9 ай бұрын

Great Video Dr.Shahriar I really liked the detailed Analysis. High quality content as usual.

@oriole8789 9 ай бұрын

Very nice compact instrument, thanks for the detailed review! Very nice to see thought put into fixing the main connector, if needed. The UI seems well-designed and hopefully they can address various glitches that you've found. I find that at least in my own work life, whenever I identify a large number of product/service issues for customers, it often overwhelms them, even though I try to clearly mark priorities as well. I end up checking back months later, only to find some of the most glaring issues still present. I understand the mechanisms for how things like that can happen, but it's still very odd to see that "from the outside". This type of product would require high effort from the software side for sure. Thanks again for the video.

@AdrianRumpold 9 ай бұрын

NRQ6-POOCH - nice touch! 🐈‍⬛

@non-human3072 9 ай бұрын

Awesome, thanks for all the videos

@pyrokinetikrlz 9 ай бұрын

very informative!

@preste3 9 ай бұрын

As soon as you said individual licenses, I lost interest in the product. 😢

@AaronHe-z6s 8 ай бұрын

On the RF board, we see the silkscreen -- "Transcom". So what is the relationship between "Sanko" and "Transcom"? Thanks a lot！

@der.Schtefan 9 ай бұрын

First thing I do on all Android devices is to activate developer mode and disable ALL animations. They are just taking away precious time, especially the keyboard sliding animation.

@indeliblecanopus 9 ай бұрын

I'm curious, do these RF instruments (with range below 6Ghz) always use Rogers materials for rf section or it can get away with lower performance materials to satisfy the specs?

@andymouse 9 ай бұрын

Yeah good old FR4!

@asdfzxcv6268 9 ай бұрын

Why would you need anything better than FR-4 here? No planar filter or other fancy structures requiring precision, no long transmission lines to cause high losses. The only RF things are the short pieces of transmission lines, not a problem to get good enough impedance matching on FR-4.

@andymouse 9 ай бұрын

:)@@asdfzxcv6268

@asdfzxcv6268 9 ай бұрын

I've got a question about filters. It looks like there are filters directly connected in series. It's not the first time I see two or three filters connected directly one after another (sometimes planar) in the teardown videos. Normally filters have their designed characteristics when they are terminated with Z0 but if a filter is connected to another one, it can see Z0 only in pass-band of it's neighbor. I wonder if there are some special methods of designing such multi stage filters to avoid complete mess and spurious responses in stop bands? Or it's just trial and errors tuning to get proper stop band regardless of the mismatches?

@aliemlek 9 ай бұрын

👍👍👌👌👏👏

@nraynaud 9 ай бұрын

nice video. Have you determined what phone they used as a base for the instrument? A few years back I tried to find a phone PCB to use as a base for a product, the same way, but not speaking Chinese I couldn't find anything afordable.

@drewwollin3462 9 ай бұрын

Very helpful. I have a couple of vector signal generators and had been a bit disappointed with the spurious signals and not-so-great phase noise. However, based on this review, they are not so bad. Drew VK4ZXI

@cosmolittle1395 9 ай бұрын

Signal Hound make a 6GHz vector signal generator at quite a low price (VSG60A). Could I request a tear down and review of this to compare with the GH60?

@tommihommi1 9 ай бұрын

nice video flexing your amazing spectrum/signal analyzer collection :D Could have done most of the measurements using just one of them, but that would be boring

@FullyArticulate 9 ай бұрын

Excellent video, as usual (especially the signal path analysis)... Except for the end "it just needs a bit of work on the software to really make it competitive". It's so easy to throw stones at efforts to make affordable useful products, especially when testing them with $100k+ Keysight and R&S equipment. It looks like this is about $10k, although I have no idea what the software licenses cost. The closest competitor I can find is maybe the FieldFox, and that's going to run about $30k for similar capabilities. So, it seems, out of the gate, it's about 1/3rd the cost for a competitive set of features, albeit with flaws. Cost is a critical component of competitiveness, otherwise, nothing will "compete" with the 6 or 7 figure priced units that few can afford, and keeps new entrants out of the market from even trying to innovate in this space.

@Thesignalpath 9 ай бұрын

FieldFox is an analyzer, not a generator. The instrument is also being characterized using the various instruments, it is not being compared to them since this is a generator and analyzers are used to evaluate its performance.

@carlosdanielmendozaperez6907 9 ай бұрын

If I were a vendor and Shahriar makes this suggestions to me , I'd be very happy, plus, software corrections are not expensive to implement, and the critics were very polite and constructive.

@FullyArticulate 9 ай бұрын

@@carlosdanielmendozaperez6907I completely agree that his suggestions are always very polite and constructive. However, the line about it being "competitive" is what bothered me. It's far too easy to compare $10k units against $100k units and say they aren't "competitive". Shaming vendors for not being as good as something at 10-100x the price does no one any good. If you can afford the $100k R&S VSG, have at it. I can't, and am willing to sacrifice noise floor or capability or some bugs so that I can actually buy something. That doesn't mean the cheaper unit isn't "competitive", just that it has compromises.

@Thesignalpath 9 ай бұрын

Compromises on noise floor, phase noise or harmonics performance are a normal part of making purchases based on cost. Errors in the firmware, bugs and incorrect signal generation has nothing to do with cost - it must be corrected. This is true for a $10K instrument as it is for a $100K instrument. Again, the instrument is being compared with other instruments in its own class - not against something which costs $100K more.

@jackm_ 9 ай бұрын

The way this thing is made, it looks like an hack: hack together a phone, a bigger battery, the actual hardware and some electronics to connect everything together. Write an app that works on the phone's android version. God forbit it connects to the internet, or it will have to be updated from time to time and we all know how android changes between security patches so that your app magically doesn't work again.

@prutser67 9 ай бұрын

And if your device does not work and you want to replace a part, the number is lasered off.😢

@vincei4252 9 ай бұрын

@@prutser67 That's a line in the sand.