Oscilloscope bandwidth vs sample rate


#1

What would be good parameters for an oscilloscope to use with the P1 and Photon? I’ll looking into getting an oscilloscope to find any noise or interference issues.

I was looking at the picoTech 2000 series with 20MHz, because of cost. I dont really want to spend more than $275

Any help will be appreciated.


#2

20MHz should be good for pretty much anything around a core/photon/electron.

The picoscopes are a good entry ramp, but you might be amazed at what Tektronics hardware you can get on ebay for the same money (speaking as someone that owns both picoscopes and Tek’s.)


#3

I took a look and found several are the tectonics better? I’m all about the quality functionality over size. Would u recommend the 2225 model?


#4

You do need to do your homework. This site seems to have some opinions:http://www.usedoscilloscope.org/used-tektronix-oscilloscope/2200-series/

For that price you would get a great front end and solid analog performance, but none of the cute digital capabilities you will get with the picoscope software. Personally, I think both have their place, and an analog scope will lie to you a lot less often than a digital one.
You have to factor in your tolerance for risk and whether you just want a tool or would like a solid starter scope.

The picoscope takes less space on a desk (if you dont count the computer) too.

Sorry I can’t give you a black or white answer, like most interesting problems this is kind of a grey area :slight_smile:


#5

My favorite combi:


#6

I realize that “you need to do your homework” sounds harsh - it wasn’t meant that way.

You asked for advice, and it is rude of me to say “do your homework” - what I meant was that it appears not all Tek 2xxx 'scopes are equivalent. I have 4 Teks here of different vintages (two of which were bought on ebay) and one picoscope.

I should have said, that I’m not an expert by any means on the nuances of Tek scopes by model number, but I do think that a scope is only as good as it’s front end, and to me a good(*) Tek is very hard to beat.

Digital scopes can lie to you in a number of ways, not the least of which is undersampling the input so that the trace you see on the screen doesn’t reflect the real signal. However, having a display instantly tell you the pulse width in uS can be worth it’s weight in gold when you are debugging a digital system.

If you do buy a scope on ebay, make sure that the screen is still bright and not burnt-in.

Over time, I would recommend getting both a “real” scope and a picoscope, but which one to get first is a difficult decision, and I fear I’m not making it any easier for you. Sorry.

(*) Good, does not necessarily mean new in this context.


#7

Thanks for the advice, I have a better understanding on the two different types. I’ll do some more reading to figure out what I want to do with it and pick the more appropriate one. I guess my initially thought was to use the oscilloscope to help determine any EMI or noise on the custom PCB to help with FCC part 15 verification.


#8

Hi @wesner0019

I just want to point out that your goals of only spending $275 and finding EMI or noise to help with FCC part 15 verification are fundamentally in conflict. You might get some useful information from a 20MHz bandwidth scope for EMI but not a whole lot!

That said, a 20 MHz scope is great for debugging i2c or SPI or serial problems and anything related to the peripherals around a core or photon.


#9

After some more research I came to that conclusion. Would you by chance know what kind of equipment would be needed for verification to determine not huge issues exist before sending to a lab?


#10

While you might be able to explore some things with a high-speed oscilloscope, you really need a spectrum analyzer with a good RF front end. These instruments are very expensive relative to your budget–a cheap spectrum analyzer might be $30k.

It would probably be cheaper to find a lab that will let you come and do some pre-testing before you go for real testing. Somewhere in the $3,000 to $10,000 range is my experience.


#11

I found this within a search. I know this is a far cry from a lab grade unit, but could this be good enough to detect any glaring emissions? Has anyone had experience with one of these before?


#12

Maybe this kind of thing instead:


#13

I just glanced quickly, but it looked looked that device only goes up to 1.7Ghz.

I would recommend BladeRF. Paired up with GNU Radio, it can act as a very capable spectrum analyzer for only $650. http://nuand.com/

I have worked in the wireless test and measurement world for nearly ten years, and I am amazed at what these new SDR platforms can do. One can create entire LTE nodeB emulators for less than $2k. They are just fantastic, I have used the BladeRF and would easily recommend it.

However, to @bko’s point, you may still want to talk to a lab. RF is very tricky and unless you know precisely what you are looking for and have all the right equipment to measure it (including isolation equipment/chambers), it is pretty difficult.


#14

They speak the truth, if you’re worried about emissions compliance, you need a spectrum analyser and those are a different beast entirely.

I agree that some of the new SDR equipment, and even some of the butchered DTV dongles can give remarkable results for very little money, you should plan for a couple or pre-scan sessions at your nearest EMC test facility (I’m a personal fan of NWEMC here in the US, if you’re lucky enough to live close to one - but plenty of other options are available) - pre-scans where you are not paying for any certificate can be remarkably cost effective.

So a combination of SDR tools in your lab, plus pre-scans would be my recommendation - if your budget does not stretch to commercial-grade test equipment in your lab.

But, you should still buy a 'scope, if you’re serious about making/supporting anything.


#15

Thanks for the great tips. I am planning on building up my electronics tooling arsenal as the funds become available to do so.


#16

I just looked into NWEMC and one facility is 10 miles from my house! I live in the Minneapolis area.


#17

The NWEMC folks up in Brooklyn Park are great, you will not be disappointed.


#18

@eely22 I am looking at the FCC limits for an unintentional radiator for a class B device. Using the below, the testing can be done between 9kHz to >960MHz, How much higher than 960MHz does test go for?

http://www.ecfr.gov/cgi-bin/text-idx?SID=4b7ab9921ce319cdec7c254953ee2ba7&mc=true&node=pt47.1.15&rgn=div5#se47.1.15_1109

Also do you know off hand, if when the P1 is placed on a custom PCB with a switching LiPo power path charger and a few general I/O and an I2C OLDE screen would need FCC verification or does the P1 FCC certification meet the boards needs? My guess is that the PCB with the P1 will still need to be verified for FCC Part 15 Class B Unintentional radiator. I’ve read so many different things on-line that everything seems to be conflicting.


#19

Anything with WiFi is going to be an intentional radiator.


#20

I have worked in wireless test and measurement, but not on regulatory testing. I must confess upfront that I am not an expert on FCC regulations!

You’re best bet is to contact the lab and talk to them about your specific project. They will know the regulations and what you need to test for, and they should quote you for free.