If you were talking to someone whose organization is considering Keysight Oscilloscopes, what would you say?
How would you rate it and why? Any other tips or advice?
It's a very nice product. Whatever feature that I was looking for, it was within my budget. It was affordable. I can upgrade it later, if necessary. Plus, if something with an initial option was going out of my budget, they offered to give me something like a refurbished model. This made it worthwhile because the other option I explored was way beyond my budget, and I simply could not get approval for that additional amount of money. Always reach out to the area manager or application engineer to show your interest. They are able to hook you up with the right technical application engineer, who can understand what you are trying to do, then recommend what are the best possible solutions and give you a demo for you to feel comfortable. It is also good for them to show you the features. While it's more intuitive than other scopes, there is a little longer learning curve. I do consider it quick to learn. It is targeted towards embedded systems and controller cards, like USB and PCIe.
My advice would be to get the demos because everyone's application tends to be different. Sometimes it's the fixturing, or sometimes you realize some things in particular that make a difference for your needs. It's hard, just in theory, to know. There's very much a learning curve. It's hard for me to talk about it in detail because I have used the tool for over 20 years, so as it evolves, I evolve too. Since the very beginning of time I've used many different solutions for measuring, starting off with HP then finally going to Keysight. It's always been a reference measurement method. The solution is not field upgradable, typically. You need to buy a new one to go faster. This product is a ten out of ten. The DCA-X is the industry standard. It's hard to get better.
My advice would be to evaluate the Keysight product. In our company we have been using the Keysight equipment for at least 15 years. We use many different versions so I do not remember specific model numbers but we use a lot of the scopes, as well as PPD. Technology advances in this field so we are always in a learning curve because there is always new technology, new products, new features come out. I would rate this product at eight out of ten because of the features, ease of use, and the support.
This is a very good product to do your basic signal integrity tests and check your signals in the systems. I would recommend using this product. It took us 30 to 40 minutes to get acclimated to the solution. The user interface is good and neat, with not much clutter. It's straightforward to use, so anybody can quickly understand the product and use it. For what we need, we get that with the features on the current design. We are not using it to support connect devices nor are we planning to going forward.
The functionality is excellent, but the reliability is poor. The application space that the solution is targeting is PC components: hard disk drives. We are not using this product to support connected devices.
Study what you're looking at. Analyze. See if the tool is applicable to what you are trying to measure. You have to do your own research and see if the solution matches the application you're looking for and what the requirements are. Of course, there's always a price budget so that's something that you have to look at. But I would highly recommend this product. It's a good product performance-wise. I've used the product for more than seven years. I've used it at different companies as well, not just my current company. We used sampling oscilloscopes and real-time scopes to make eye diagram measurements. You're taking signals from your device on the test and looking at what it looks like; how clean the signal generator is from the transmitter. In the past I've also used it to look at the receiver eye too where you can actually drive a signal or a backplane and then look at it on the receiver end, using the scopes. In terms of a learning curve, when I started my career with this product it took about a year to get really familiar with the tools. After that, it was complete repetition that makes you an expert. You keep learning more things about it as time goes on with the same product and there are new features with different specs. It took at least a year to get familiar with the product but every time you learn new things, it's under six months or so. We have no plans to use it to support IoT. It's more for networking switches. The current solution we have in the lab is upgradable. Keysight has done a good job there where they have given us a solution where they can actually upgrade it. At the moment what we have is the latest upgrade. All it required was buying the license and upgrading the scope.
Keysight is number one and top of the line. While there is a learning curve, they do provide training. Keysight does a three-day training onsite. We are not using this product to connect with support connected devices, like IoT development, and are not planning to going forward.
The learning curve is not that steep. If you are used to some of the measurements beforehand, it is easy and straightforward. The application space that we are using the solution to target is high-speed serial.
It is a good tool to use. If they had a debug feature, it would be helpful. We use the tool to measure some waveforms, which we normally give to our customers for use in data center applications. We are not using this product to support connected devices, like IoT development, nor do we have plans to do so going forward.
Make sure you know what speeds you are trying to test. Ask them for a demo. Ask them for some correlation data, some other products that they have measured. Ask for good pricing. Don't go for the highest-bandwidth scope if you don't need it because it's going to cost you more and it really doesn't make any difference in your measurements. You need to see what you are trying to test and buy accordingly. In terms of a learning curve, I've been using smaller scopes for ten to 15 years so it is not an issue for me. As long as you have a basic understanding of what you are trying to see, what you expect when you try to make measurements, then you can learn the knobs pretty fast. You can go really advanced, but you have to understand what you are trying to do, what changes you are trying to make on the scope. Otherwise, you'll get garbage; garbage in garbage out. If you don't have the right inputs you will see something and you will think it's wrong. It's equal parts them doing the right stuff, but also us trying to use the equipment correctly. If you know how to design, or what you expect to see on the scope, then mostly if your design works properly on your product, you should be seeing the right things. Then, the ramp-up time is not too long. As for the solution being field-upgradable, there are some things that we can switch out. I'd give the product a seven out of ten. They can make it a little bit more user-friendly through the GUI. Otherwise, it's fine.
Moving forward to the next-generation of design, the support from testing is very important because you could end up in a situation that you are working on something where there is no testing instrument available. If you can work with a company that works on the leading edge, provide them with feedback and insight, that will be helpful. You will be able to start early engagement, using a solution before anybody else has a final product. My advice would be to build a very good relationship with a testing company. The learning curve depends. In the old days it was a matter of getting used to a hardware interface. These days everything is Windows-based. Sometimes there is a learning curve there but it really depends on the user, in my opinion. We are facing a paradigm shift here. The traditional, analog Z signal is very good and easy to implement, but when you go to higher data rates, it's not sufficient because your fundamental Nyquist frequencies are going up quite a bit. We are facing a lot of losses in the media, including the PCB packaging. What we need to do is maintain the same Nyquist frequencies, but the tradeoff is that we have to go to a different way of doing things. Analog Z would not be sufficient and we have to go to the PAN4 or PAN-N types of structures of the signals. We need to switch and it's happening now already. I would rate this product at ten out of ten. The reliability and quality have always been there over the years that I've been using Keysight. Of course, other brands are pretty good too. It's just that I have had more experience dealing with Keysight for 20-something years.
We are doing high-speed copper cable measurement at about 112 gigabit per second requiring bandwidth at about 40 to 50 gigahertz. The majority of our products are connected devices. We are currently using it for 56G ANSI, 56G PAM-4, and 112G PAM-4. These are our three main signals. We are utilizing the full bandwidth of the product. Keysight equipment is very popular.
Create a shortlist and have a demo with each of them. Try each one. Once you do some trials, experiments, you will know what you want to choose. There are several things we look at: One is the accuracy, the second is the reliability of the equipment, and the third is how easy it is to use. There is no learning curve because I've been using it for so long and it's pretty simple. The solution is field-upgradable. We may add more functionality to the PLTS. I would rate the product at eight out of ten. It's pretty easy to use, pretty reliable. It doesn't break, so it can work in a very harsh environment. But for a lot of additional functionality, we need to pay extra. So that is where there is room for improvement.
Be aware that there are other solutions out there that are cheaper. And if you're willing to put in the extra work to make up for the disparity between the products, I think it can be done. There was a learning curve, but only in so far as I needed to be more familiar with signal integrity design in general. They do offer a lot of resources available to budding SI engineers. As soon as you get the basics down, you get the gist of how to use their equipment. Our solution is field-upgradable. We plan on upgrading it, probably within the next five to ten years. I would rank the product at seven out of ten. It does everything I need it to do, but it is really expensive and I think there are alternatives out there that I could choose that would be cheaper. But we already have the product.
I would definitely recommend trying this product. It's a really reliable and high-performance machine. In terms of a learning curve on the equipment, I had some training on the VNA before we purchased it, so for me training was done within a week or so. It was not really hard. But for colleagues, it took them some time to learn it. Upgrades are always welcome. Apart from VNA, we need to purchase all the fixtures, other equipment to connect to VNA. They are very expensive tools, and that's part of upgrading what we are doing in the lab. The solution, performance-wise is a ten out of ten. Price-wise, it's a six out of ten.
I would certainly recommend that colleagues use the same solution for their application. I don't remember at the moment which model we're using, but it's the one with the highest bandwidth for the VNA. There is a little bit of a learning curve but it's not too difficult because there is good support. I had some previous experience and it took me a few days to start and then, to get a little bit more familiar, took a few weeks. In terms of field-upgradability, there a possibility by changing the software to add new features and we want to take advantage of that. For example, I know that the VNA can be also a spectrum analyzer, so we want to try to take advantage of that without buying another instrument. It will enable us to use it for two purposes.
Review case studies from the manufacturers' data sheets and compare their price and their performance. In the high-speed area there are a lot of data that need to be considered, such as your dynamic range for the equipment and the noise level. We need very high accuracy, so it's better to compare before buying from any supplier. I have been using this product for at least ten years. Already in my third year of my PhD I was using Keysight, although I don't think it was called Keysight at the time. In terms of the model I'm using currently, I don't remember the type and specific model number but I know it's a 50 gigahertz VNA, from 300 kilohertz to 50 gigahertz. I would rate this product at eight out of ten because the equipment is very expensive, it's very high-end. But it is very accurate equipment. Keysight is a leader in this area.
The application space the solution targets is usually telecommunications, like 4G and 5G. We are measuring RX and TX signals.
I suggest looking over their use case site. We are using it to connect IoT to our products. We're measuring high-speed signals, like USB, PCI Plus, HDMI, DDR, memories, power rails, etc. The application space the solution targets is embedded systems. For example, an embedded PC for industry automation, including wellness in strength machines or bicycles.
If you want to use it manually and automate it, it's easy. However, for automation, if the documentation was better, it would decrease your time using it. At first, you need to read the documentation, or if you want to search for any particular thing, it takes lot of time. So, you need to invest a lot of money to start this. If someone has some experience with Keysight, then it would be easier. We are measuring signals mostly related to switching, like networking. The type of signals that we are measuring are signalling integrity. I don't think it's being used for connection devices. I have seen past developments, and it's mostly used for taking measurements on the board to check how it works.
We always are sharing this solution with our peers and colleagues. The solution is targeting high-speed serial applications. For high-speed signal designs, you need to get the eye diagram beautiful. The learning curve of the solution depends on the quality of the material. Some things take a day or two to get acclimated, and others take a year or two. I am working on some larger devices, so I don't plan on using it for IoT. Though some of our department using it for IoT, I don't have any details.
I'm researching spectrum analyzers. Any recommendations for a solution for a 6g testing solution?