2.4GHz ISM Band/BLE Analyser and Tester Solutions

[MixedSignalSystems]

Hi All,

I’m doing some work for a fabless semiconductor company that is designing ICs with a 2.4GHz RF interface.

They are going to use a proprietary protocol but it will be broadly similar to BLE. Low power, fairly low bandwidth, similar radio characteristics (modulation scheme, etc.).

They are currently exploring hardware evaluation and production test solutions for the PHY.

Beyond the usual suspects, e.g. Rohde & Schwarz, Anritsu, Teradyne, etc. are there any interesting analysers, sniffers, etc. that we should take a look at?

They will want to look at PHY performance with poor quality clocks, drift, packet corruption, etc. Everything that you might want to look at if you were to fully evaluate and characterise a RF PHY (and higher stack layers, if possible) block.

I have no idea what kind of capital budget these guys have but they are a sizeable fabless semiconductor company so they will have some money to spend.

I’d be grateful for any recommendations. Often, we fail to look past the big guys (Nobody ever got fired for buying IBM!) but I bet that there are interesting companies out there that I haven’t heard of yet.

Many thanks,
Kieran

[cdev]

People have gotten fired for lacking imagination when they are up against imaginative people.

At the end of the day its all about creativity.

[cdev]

"proprietary protocol" is shorthand for insecure and generally nobody should use something like that.

Even Bluetooth had its problems, I'd expect proprietary anything to be even worse.

Only idiots would trust a proprietary protocol for anything. Its like closed source software. Its likely to be insecure.

[Marsupilami]

Hey Kieran,

I work for a large-ish T&M company on RF semi test systems. Here's my somewhat biased take on this.

Every semi maker/designer wants to get their chip into the next iPhone, Galaxy or car infotainment system, etc. The customer, the maker of such devices will have very stringent requirements for the test equipment used for V&V and production test of the chips. It's mostly about legal and financial responsibility. If the chip datasheet says x dBm for a certain quantity and it turns out that in some weird corner case it's x+y because of incorrect verification they will want to sue your and the test equipment provider's pants off. That's very hard to do when the T&M company is 2 German college graduates and their cousin with the gross yearly profit of 3 potatoes. For this reason they avoid accepting such solutions even when it looks technically sound.
This doesn't of course mean that in earlier phases of product development such "cheaper" equipment couldn't be used but if you'll have to buy the Keysight box for V&V anyway you are creating an extra problem of having to correlate your early dev measurements to your V&V equipment and at that point using the same, expensive instrument throughout the whole process starts making more financial sense. This is why a handful of big brands have hegemony.
In certain high volume situations it might worth for the semi company to invest in at least partially building the test solution to save on equipment cost. In that case the product of a smaller, less known equipment maker can be used as a starting point, but then there has to be willingness, and resources to basically do V&V on the test equipment itself before or while working on the actual semiconductor device development. Few big ones ever can or want to do that.

[Marsupilami]

Only idiots would trust a proprietary protocol for anything. Its like closed source software. Its likely to be insecure.

Maybe standardized ones don't cover the specific latency and bandwidth requirements. Maybe security is not at all a requirement. Maybe they plan to implement robust security features in higher levels of the protocol stack. What was that about people getting fired for lacking imagination?

[MixedSignalSystems]

This doesn't of course mean that in earlier phases of product development such "cheaper" equipment couldn't be used but if you'll have to buy the Keysight box for V&V anyway you are creating an extra problem of having to correlate your early dev measurements to your V&V equipment and at that point using the same, expensive instrument throughout the whole process starts making more financial sense.

Thanks for your reply. I appreciate everything that you are saying. In particular, the likelihood of just delaying the inevitable is non-zero in this application so I'll certainly keep that in mind.

[cdev]

If you don't use their phones you cant buy what they are selling, whatever it is, even if they give you one!

[patstew]

I work for RF Creations, we make PHY testers and protocol analyzers for bluetooth classic and LE, and also for proprietary bluetooth extensions (down to phy level) for very large chip vendors. We've also done bespoke test equipment for fully proprietary schemes. For a 2.4 GHz system we would be able to do that as a software update on one of our existing hardware platforms, the amount of work would depend on how 'broadly similar' your scheme is. You can contact us at enquiry@rfcreations.com if you're interested.