Just got the first batch of test boards. Feels really good to hold them. Onto the testing.
July 21, 2016 at 6:20 am
When will be it available commercially ?
July 25, 2016 at 10:25 pm
We don’t have a fixed date, but expect at lest a few months. We have to do another prototype run and make sure everything is working as expected.
July 31, 2016 at 2:18 pm
hello do you have an price range for theses boards ???
July 31, 2016 at 3:38 pm
We don’t have prices for the boards yet, but we would love to hear what do you think the price should be.
In general we use prices of boards like BladeRF and HackRF as a reference point. There are many things which can make the board significantly cheaper or much more expensive, though, and we can’t announce the price until we fix the feature set and understand demand (and thus quantity).
September 16, 2016 at 4:36 pm
Are there any details on the FPGA make & model?
September 16, 2016 at 7:16 pm
We’re using Xilinx Artix 7 XC7A35T for prototypes. We haven’t decided yet whether we’ll use it, or 15T or 50T in production. What is your application? We’re looking for more user feedback and use cases to make a proper decision.
September 16, 2016 at 7:21 pm
A NSF-funded project we just finished up developed a NoC and DSP framework for a Zynq Zedboard + AD9361 device. This provides a framework for a streaming interface to radio devices and signal processing blocks ( https://bitbucket.org/cusdr/cunoc_sdr ) plus a framework for produce bitfiles from a JSON specification of the components using a Docker-based Vivado system ( https://github.com/dirkcgrunwald/wiserzynq-docker ).
We’d be interested in using this on the XTRA boards or using the similar RFNoC from Ettus if licenses permit.
September 16, 2016 at 7:36 pm
Oh, and duh, forgot to say – having the bigger FPGA (even at an elevated price) would allow e.g. correlators and FFT’s to be in the FPGA, which would seem to be important.
The 50T looks to be ~$30 more than the 15T, but has slices similar to the Zynq 7020 we’re using (although fewer DSPs). That’s a pretty good compromise although another option would be to have two models (small & then 100T/200T variant, but that may not make sense).
Compared to the LimeSDR, which uses the same RF chip, the 50T would also seem to match their Cyclone resources.
September 16, 2016 at 8:39 pm
We’re thinking about two versions – like Lite and Pro with different feature set. Our current idea is that Lite will have 15T and pro will have 50T. Unfortunately it’s impossible to put 100T/200T in miniPCIe form factor – chip packages for these chips are just too big.
September 16, 2016 at 7:39 pm
I’m not certain where my earlier reply went to, but it disappeared. We developed a DSP-NoC system ( https://bitbucket.org/cusdr/cunoc_sdr ) with a generic VHDL AXI-Stream switch, AXI DSP components, etc and a Docker-based Vivado toolflow that simplifies generating bitfiles ( https://github.com/dirkcgrunwald/wiserzynq-docker ).
If we had time, we would have integrated it more e.g. the Pothosware interface ( https://github.com/pothosware/pothos-zynq/wiki ) so it wasn’t so one-off.
September 16, 2016 at 8:47 pm
Looks like an awesome project!
It should be well possible to integrate this with XTRX – we’re also using AXI streams to stream data. And we have AXI4-Lite like control interface.
PS Your earlier reply went to a moderation queue for some reason. Not sure why. Anyway, I’ve approved it.
September 25, 2016 at 3:21 am
I would love to have a 1pps input as well as an input for external 10MHz if possible.
September 25, 2016 at 7:28 pm
There is a 1pps and external clock input for multi-unit synchronization. Normally it will accept 30.72MHz reference, not sure it will accept 10MHz – I’ll check with the team.
September 28, 2016 at 11:30 am
Thanks for the reply. 30.72MHz would be fine. Any update on an estimated time frame. I wouldn’t mind even being a tester.
November 4, 2016 at 9:11 am
Any plan to have LTE OpenAirAlliance (with Eurecom?) stack run on this HW?
Is Jan 2017 a realistic date for commercial samples?
November 4, 2016 at 10:25 pm
Yes, we definitely plan to have OAI running. It might not be the first priority for us. Patches will be welcome, as usual!
December 27, 2016 at 12:36 pm
Are the designs and souce available? I’m in a completely different field that might be a blessing to co-opt your awesome work for neural recordings.
December 27, 2016 at 4:57 pm
Our current plan is to release drivers and libraries into open-source. FPGA is under discussion with good chances to be released as well. Hardware itself is not planned for release at this moment.
January 1, 2017 at 5:03 pm
I habe some question. I looking for an Client he want for an Bus Entertainment System something who can “Decode” all Available DAB(+) Transponder. Is there any way to get the Bandwidth for?
It also look very interesting for some Multi Channel System. Hopefully it will be possible use it in GNU Radio?
Is there any Plan to make a Version only for reception the cheap as possible way to use Multiple SDR in one System for different Bands. (For example its not necessary to have an huge bandwith when there is anything I am not interested in….)
February 11, 2017 at 7:54 pm
Apologies for late response. Missed your message in the moderation queue.
I’m not very familiar with DAB+. How much bandwidth do you need for it?
GnuRadio is already supported – Sergey has added XTRX support to gr-osmosdr. The code for it will be published with the public release of XTRX.
January 3, 2017 at 10:41 pm
Found you via your 33C3 talk, very excited! Can you please share some details about the analog front end? Is there a Power Amplifier for TX? Is there a Low Noise Amplifier for RX? Are there switchable filters, and what frequency curves? Ok, maybe too many questions but those are the basics.
February 11, 2017 at 7:35 pm
Cord, apologies for the delay. Missed your comment.
We have PAs, LNAs and TDD switches on a carrier PCIe board:
PA is capable of producing 100mW (P1=20 dBm) with flat 20 dB gain from 50 MHz to 4 GHz i.e. over the complete range available for XTRX.
LNA has NF<1 dB up to 2.5GHz, gain 20 dB at 0.7 GHz, 10 dB at 3.5 GHz.
TDD switch has >50 dB isolation in FDD mode (closed state).
XTRX itself doesn’t have any PA or LNA due to size and (more importantly) power consumption limitations. Mini PCIe has a limit of 2.5W and we’re already at 2.3W when both channels are receiving and transmitting.
XTRX also don’t have switchable filters, because they are application specific and we believe it’s better to use external filters specific for your application. just LTE has tens of bands and we can’t put filters for all of them in.
February 17, 2017 at 3:13 am
Any way i can get one of these to play with? I am just starting to look into SDRs and really not verry well off as for money. I would love to try one out though
February 25, 2017 at 7:47 am
We expect mass production at H2 this year. There will be prototypes available for selected early adopters in in Q2 this year, but they are limited to our partners who already have significant experience in SDR space.
I recommend you to look at getting something like LimeSDR to start with and get back to XTRX when you need something really industrial/embedable 🙂
March 17, 2017 at 10:47 am
Hello, would you be interested in getting someone to write a SoapySDR driver for the XTRX? I would be happy to do that in exchange of being a participant in the early adopter phase.
March 22, 2017 at 4:16 am
Matthias, thank you for the offer! We’re still working on the details of the early access program, but this is definitely interesting for us.
March 17, 2017 at 4:58 pm
I just want to know how much can we buy one if it has released?
I am a CE(Communication Engineering) student. I’m so interested in SDR but I can’t afford a USPR board.
March 22, 2017 at 4:18 am
Hi XL, thank you for your interest! We haven’t fixed the price for XTRX yet. Follow updates here to know when announce it.
March 22, 2017 at 8:39 am
OK, I will keep following your bolg.Do you have an idea when it will be released? I am glad to wait for your production conference
March 22, 2017 at 6:25 pm
Not sure about full production yet. It depends on the Early Access Program results – whether we’ll need to make any changes before going into production and how many customers we can accumulate for the first production run. Price for 10,000 units is quite different from 1,000 units 🙂
Fill in your details below or click an icon to log in:
You are commenting using your WordPress.com account. ( Log Out / Change )
You are commenting using your Twitter account. ( Log Out / Change )
You are commenting using your Facebook account. ( Log Out / Change )
You are commenting using your Google+ account. ( Log Out / Change )
Connecting to %s
Notify me of new comments via email.
Blog at WordPress.com.