4 GS/s Oscilloscope with averaging and VERY fast retrigger.

[frogg]

I'm not sure what you think the differences are between a high speed digitizer with FPGA on-board and a modern FPGA dev board, but uh........ok, mate

[Marco]

I'm not sure what you think the differences are between a high speed digitizer with FPGA on-board and a modern FPGA dev board, but uh........ok, mate

One works out of the box, the other doesn't.

What exactly do you have in mind? ADC12J4000EVM + TSW14J10EVM + TEC0330 and some hope and prayer you can make it all work together?

[nctnico]

I'm not sure what you think the differences are between a high speed digitizer with FPGA on-board and a modern FPGA dev board, but uh........ok, mate

One works out of the box, the other doesn't.

What exactly do you have in mind? ADC12J4000EVM + TSW14J10EVM + TEC0330 and some hope and prayer you can make it all work together?

Agreed. For the OP's purpose the analog front-end design will matter a lot too and there maybe be impedance requirements as well. It is not trivial to get a high frequency signal into an ADC. Besides that getting from the demo applications which come with a dev board to something that is actually useful in a production environment is going to take a serious amount of work.

[rsjsouza]

What exactly do you have in mind? ADC12J4000EVM + TSW14J10EVM + TEC0330 and some hope and prayer you can make it all work together?

I love that. Quite on point.

[moore]

Dear all,

I'm working on an experiment which measures the response of a system to a 2-10 us long pulse sequence. The response is very deep in the noise and requires a lot of averaging of the response signal. The system recovers instantaneously after the pulse sequence, and thus we would like to repeat the pulse sequence as soon as possible. I'm struggling to find a signal averaging oscilloscope that's able to respond faster than 1 microsecond (FastFlight-2 signal averaging oscilloscope, not available anymore) after the end of a trace. We can't miss triggers, as a missed trigger means missed data, missed averages, and therefore longer to collect enough averages to obtain meaningful data. Does anyone have an idea for how to average a million or so repetitive segments of analog signal without losing data?

Best,
E

We have a similar acquisition need, mass spectrometry with high rep rate lasers to get high dynamic range after averaging.  Gave up on the FastFlights years ago, have tried the Acquiris, they are ok but expensive. Currently using Gage Applied, see product line here:
http://www.gage-applied.com/  (click tabs for 8 or 12 bit digitizers which have the 4-6Gs/s rates)

They have worked hard on minimizing rearm time and keeping data flowing for high rep acquisitions.   Also I've been impressed with how their ENOB stays adequate at high bandwidth.   They have some ok software for plug n play, but likely you have to use their API and do some programming to get exactly what you want.  They do have firmware for fast averaging and other functions. 

All in all a solid company with good support that seems to be not well known outside the exotic scientific applications community.  They are not Keysight or Ametek/Ortec size but they punch above their weight class.

[frogg]

the OP is asking for an integrated product rather than building something. Scopes have the front end required and triggering performance

Hi Someone,

I'm sure we are just talking about the same thing. It is extremely common in industry to pair what is essentially an FPGA development board with a high speed digitizer front end. There are usually a few cards or even a single integrated card in some cases that fit into a backplane, like NI FlexRIO. Sometimes the terminology varies by manufacturer, sometimes it's a "development board". Sometimes it's just called a "high speed digitizer."

This is extremely common for use in EW and radar systems and high speed optics.

The FPGA synthesis toolchain is very simple and easy to use, especially if the necessary logic is straightforward like the OP states.

Yes, there is some minor assembly required, but it is really trivial.

I don't know what everyone else is talking about (building custom analog front ends is not something anyone suggested and is not written in any post!!)

[Marco]

Sometimes the terminology varies by manufacturer, sometimes it's a "development board". Sometimes it's just called a "high speed digitizer."

There's high speed and then there's high speed, for their 1 GSPS range Analog Devices has pretty open development boards ... for their 10 GSPS range ADC it's a non reprogrammable black box. TI similarly has non reprogrammable black boxes for appropriate ADCs, though they do give you a FMC+ to FMC adapter so you can use it on FPGA development boards. But that still means a whole lot of DIY and figuring out how to get everything to work together.

Teledyne SP and Keysight have digitizers at that speed which allow firmware development, but I doubt that's cheaper than the Acquiris.

The Gage Eon Express and Cobramax Express with their averaging firmware do seem an out of the box working alternative.

I don't know what everyone else is talking about (building custom analog front ends is not something anyone suggested and is not written in any post!!)

Well, the TI ADC EVM is AC coupled for instance ... if you want DC coupling you'll have to do some modding.

So if the Cobramax Express's 8 bits resolution is enough you can have a working system for 9000$ (website prices, so might be able to negotiate that down). Or you can get 4000$ worth of development boards with a whole lot of DIY hardware and software development. I don't think the saved money is enough to pay for the time spend.

[ezalys]

I'm not paid to develop something or cobble something together. It needs to work now.

I've seen these Gage cards and heard very good things. So far it's the gage card vs. the keysight/aquiris. Could you explain a bit more about how they compare?

[Someone]

the OP is asking for an integrated product rather than building something. Scopes have the front end required and triggering performance

Hi Someone,

I'm sure we are just talking about the same thing. It is extremely common in industry to pair what is essentially an FPGA development board with a high speed digitizer front end.

You can keep saying there is some other non-specific way to achieve the solution... but everyone else is providing examples. If you think there are other more appropriate examples, do provide them rather than just saying you know a better way to do it and we don't.

[moore]

I'm not paid to develop something or cobble something together. It needs to work now.

I've seen these Gage cards and heard very good things. So far it's the gage card vs. the keysight/aquiris. Could you explain a bit more about how they compare?

Aquiris hasn't been part of Keysight for a couple of years now.  My impression is it was just not a good match.  They are a Swiss company.  Like a lot of smaller firms, they are looking for the big customer who will develop something around their products and wants to buy 100-1000 eventually.  Not to say they won't work with a one-off customer, just be aware they may be focused on a different problem than you are.

There's a third option I'd forgotten about - Teledyne SP Devices.  I saw them at a trade show.  They have cards that seem pretty powerful but are much more than even the Aquiris.  I forget if they were PXI only.  If a giant company with top level customer support is important though, and money is no object, they may be the only choice of these three.

My suggestion is to define your must-need specs, and go through the data sheets of the most appropriate choices at the two/three companies, and get fresh quotes.  If they have demo software, get that (GaGe used to).  Look at the API documentation.  If you have a big budget and lots of time and flexible spec needs, it could be a headache to decide, but then maybe you can get a demo card from each.  If you have a modest budget and less time to shop and need to start working, then I'd say go for the GaGe and forget the others.   

[rsjsouza]

the OP is asking for an integrated product rather than building something. Scopes have the front end required and triggering performance

Hi Someone,

I'm sure we are just talking about the same thing. It is extremely common in industry to pair what is essentially an FPGA development board with a high speed digitizer front end.

You can keep saying there is some other non-specific way to achieve the solution... but everyone else is providing examples. If you think there are other more appropriate examples, do provide them rather than just saying you know a better way to do it and we don't.

The only ones I found more or less "ready to run" from the Xilinx page are from Curtiss-Wright. (4GSPS, 12GSPS and 25GSPS). If they are an aerospace company, purchasing one may be a difficult and costly option.