Altium
Next generation electronics design solutions
In a special interview with Australian electronics design consultancy Gryphon Systems Engineering, Technical Director, Peter Stephens explains how his company is using Altium’s Innovation Station to overcome the complexities of custom electronics design…
 |
 |
 |
Envision (E): Tell us a little about Gryphon System and what you do.
Gryphon Systems Engineering (GSE): We’re an electronics design consultancy focusing on customized system integration. We do a lot of complex measurement and control type contracts. And we have to be agile to make money, especially right now.
We design specialized bridging technology, and use a lot of high speed data acquisition cards and post processing on those designs. The work can be quite niche and we do a lot of contract work for hi-tech companies.
E: Your choice of design environment has been an important decision for you. Why did you choose Altium’s Innovation Station?
GSE: Our approach with each contract is to start without preconceived notions of what the best solution will be. We like to start each project like a blank sheet. And if something already exists, we will integrate it, if not we will build it.
In the past, this held a lot of risk. It would almost always spin out. You think it will be done in a month and it turns out to be eight weeks. And that hurts. We’re a small business, so it’s important we do projects when we say we are going to do them.
When we heard about Altium and the buzz around the Innovation Station, it sounded too good to be true, because it really is a paradigm change. Of course, the IC and FPGA companies also have their development platforms as options to consider but these options have always been too fragmented to get real with results with.
E: So how long has Gryphon been using the Innovation Station?
GSE: Since March of this year. It’s a compelling approach and Altium has made a very bold step. In some senses, it’s quite a confronting step for a firm such as Gryphon to take.
But it has cut Altium some slack. We know it is so different to everything else. This really is a new approach and this outweighs the small niggles which might exist.
The Altium Innovation Station has worked really well for us. We are able to generate a design for the FPGA within Altium Designer, test that on the NanoBoard, and then get the board fabricated, and we then use that time to develop the application layer software. We can still test some of this in the NanoBoard while waiting for the fabricated board.
It has really worked out because of the breadth of peripherals which are there. This really matches up with the controls we need. Which means if we want to roll the board out we can use design components which are on the NanoBoard, we can test those, and change or modify as we need. But by not moving too far from the NanoBoard, it really reigns in the design risk and makes it more manageable. And it means that when we say we can deliver in four weeks, we can because the risk is a lot lower.
E: What projects have you been designing using the Innovation Station?
GSE: One is FPGA based,
the other is working off an
The FPGA project is giving us the biggest bang for buck from the product. It is because we almost don’t have to code in VHDL at all. It actually allows us to easily pull the components together and test them on the NanoBoard. The Innovation Station is actually a good learning tool to get into FPGAs, because you can learn about FPGAs and still be able to deliver real world products at the same time.
E: Take us through your design processes, and what you’ve been experiencing?
GSE: For standard PCBs, life has got a lot quicker. The big thing has been that I’m spending less time checking, because of the way Altium links schematic and board designs. I don’t spend as much time checking my design because Altium Designer will actually flag you when you are making a mistake and help you correct it. When starting a new design with the right design rules, this easily can reduce development time by about 30%. And the functions are very intuitive, so there were no steep learning curves.
We are saving in the overall development of the systems. One area, which is huge, is where there is a design component in the NanoBoard that we can lift straightaway. We have seen this improve by 50-60%. We are able to easily add components, for things like a LCD display which are already on the NanoBoard.
When it comes to adding the software, we’re seeing huge savings. For new designs, it’s about 30%, because we spend more time on the design and validation up front, rather than on a writing and debugging drivers.
 |
|
 |
|
 |
E: You’ve also been using the Software Platform Builder feature a lot.
GSE: It is just a knock-out feature. When you design a board, you end up in a situation where you have a lot of peripherals and you have to develop a driver for each of them. Then you have to validate them, which is not a trivial task. That is where Software Platform Builder gives us a huge head start. It basically generates the software platform for you.
But there are two kinds of benefits that stem from using it. There is the low level part, which is more like boiler-plate code where you are basically getting and redoing what is very standard, much like using an SPI library. But the Platform Builder actually allows you different levels in which you can access the SPI library. So you have an SPI port from which you can communicate in a raw form, and the platform builder actually provides you with a driver which lets you use that SPI port in a more packaged fashion, so using higher level function calls for talking to specific devices. So that makes it easier to write code straight away because it actually means that we don’t have to write the drivers ourselves anymore, we are actually just testing the drivers. We have skipped the need to code the low level drivers.
The second benefit, which is even more powerful, is that the Software Platform Builder has a POSIX real-time library. It means we can actually use what’s in the Software Builder for an effective threading mechanism and not have to go away and buy an RTOS [real-time operating system] separately. This can be an expensive thing to buy and because we only have small product-runs, this can save us a lot of money.
E: Tell us a little about using Altium’s unified design environment.
GSE: When things are bundled together it is usually because they lack functionality as individual bits. That way they look better when together. We’ve found that with Altium Designer, you actually get a fully functional software environment.
Altium’s unified design environment also helps us to be vendor independent, which is important to us. It means that we aren’t going to be held up on FPGA components from any particular brand. It also means we can develop with a vertical path. We can start with a mid-range device. This way we give ourselves a lot of room, so if the complexity goes up, we can upgrade the device or pull back if we need to meet a certain price point.
The reality is that you usually have to choose tools which can lock you into hardware in indirect ways. But because Altium Designer is vendor independent, it means in an exceptionally power sensitive application, we can use an FPGA that has the best power profile, rather than having to worry about the fact that the design and software is too intricately embedded in a specific brand or model of processor.
E: You’ve been using the OpenBus feature to overcome some fears, we hear!
GSE: The exciting but scary part of Altium Designer, which is actually now the most important part, is the OpenBus feature. When you look at it as presented in Altium’s videos it’s insane! It seems over-the-top, unreal! It makes the process look trivial when you know you have books and huge pages of VHDL code for implementing anything in the FPGA.
Then suddenly, you’ve got all these wishbone compliant components you can just drag and drop and then map and address, and basic memory mapping is handled for you in just one step. It is very scary! You ask yourself, am I going to buy the product for this feature, or am I only going to use it every once in a while?
It’s quite funny because you never see FPGA design in those terms. It is truly amazing. It really is, because it means that if the design suddenly requires another three serial ports… no problem. You just pull the serial ports into the OpenBus document, then drag and drop the drivers, using the software platform, to get them running. Once you’ve done that, all you need to do is test it on the NanoBoard. Basically, you can code in ten minutes; it’s that easy.”
E: How long does it take to get a prototype up and running using the Innovation Station?
GSE: It is not unreasonable to get a prototype working within a week on the NanoBoard and that means something you can take to a customer and get that ‘wow’ effect going. To validate the design as an engineer, you can do that inside a day because you know whether or not you are heading in the right direction.
We believe in test early, fail early. You eliminate things that are going to hold you up early in the piece.
And with the NanoBoard, you can change the peripherals around. It is really quite good, because in the future, you don’t know what peripherals will be prominent. And it is important that the development tool has reasonable life-cycle. With the NanoBoard, the peripherals are not bolted on the board, which is different to other vendor boards. This means it can be expanded and the product will have a longer life.
E: How would you wrap up this interview?
GSE: I’m genuinely excited about this product, it’s not everyday you find a product that changes the paradigm.
For more information on Gryphon Systems Engineering and how it uses Altium Designer, please read the customer success story or visit the website www.gryphonsystems.com.au