# USRP's not Communicating **URL:** \ **Category:** GNU Radio\ **Created:** [August 25, 2015, 11:14pm UTC](https://www.ruby-forum.com/t/usrps-not-communicating/243144 "2015-08-25T23:14:20Z")\ **Posts on this page:** 5\ **Page:** 1
**Author:** ![Rafa\_F](https://www.ruby-forum.com/letter_avatar_proxy/v4/letter/r/958977/32.png) [@Rafa\_F](https://www.ruby-forum.com/u/Rafa_F)\ **Post date:** [August 25, 2015, 11:14pm UTC](https://www.ruby-forum.com/t/usrps-not-communicating/243144/1 "2015-08-25T23:14:20Z")
Hello, I want to transmit the data between two USRP’s and make them communicate with each other. But I guess the packets are not being received properly. I am connected the two USRP’s to the same laptop and trying it. Is that applicable? I mean, will it work if I do that? Or should I connect to two computers and perform that? I have been receiving this error. Code: ./benchmark\_rx.py -f 2.435G ./benchmark\_tx.py -f 2.435G linux; GNU C++ version 4.8.4; Boost\_105500; UHD\_003.009.git-144-g407e3086 Using Volk machine: avx\_64\_mmx – Opening a USRP1 device… – Using FPGA clock rate of 64.000000MHz… No gain specified. Setting gain to 56.250000 (from [0.000000, 112.500000]) UHD Warning: The hardware does not support the requested RX sample rate: Target sample rate: 0.050000 MSps Actual sample rate: 0.250000 MSps Symbol Rate: 25000.000000 Requested sps: 2.000000 Given sample rate: 250000.000000 Actual sps for rate: 10.000000 Requested sample rate: 50000.000000 Actual sample rate: 250000.000000 ok = False pktno = 53034 n\_rcvd = 1 n\_right = 0 ok = False pktno = 24 n\_rcvd = 2 n\_right = 0 ok = False pktno = 35 n\_rcvd = 3 n\_right = 0 ok = False pktno = 44 n\_rcvd = 4 n\_right = 0 ok = False pktno = 46 n\_rcvd = 5 n\_right = 0 ok = False pktno = 46 n\_rcvd = 6 n\_right = 0 ok = False pktno = 3872 n\_rcvd = 7 n\_right = 0 ok = False pktno = 12304 n\_rcvd = 8 n\_right = 0 ok = False pktno = 49 n\_rcvd = 9 n\_right = 0 ok = False pktno = 50 n\_rcvd = 10 n\_right = 0 ok = False pktno = 54 n\_rcvd = 11 n\_right = 0 ok = False pktno = 200 n\_rcvd = 12 n\_right = 0 ok = False pktno = 63 n\_rcvd = 13 n\_right = 0 Please suggest. Thank you Regards, Ravi ---
**Author:** ![softf5ajivanet](https://www.ruby-forum.com/letter_avatar_proxy/v4/letter/s/9f8e36/32.png) [@softf5ajivanet](https://www.ruby-forum.com/u/softf5ajivanet)\ **Post date:** [August 26, 2015, 4:19am UTC](https://www.ruby-forum.com/t/usrps-not-communicating/243144/2 "2015-08-26T04:19:04Z")
You can have each USRP on the same computer or not, that shouldn’t matter. I’ve done this test with some USRP2’s and I remember what we did was transmit a CW tone from one to another and measure a frequency offset (maybe 10s of kHz?). Basically, transmit a tone at some known frequency (like 200 MHz) and measure the received frequency (maybe something like 200.02 MHz) and then add the 0.02 into you center frequency for this test. You can also just synchronize the USRPs with a 10MHz reference and 1PPS. Devin ---
**Author:** ![softf5ajivanet](https://www.ruby-forum.com/letter_avatar_proxy/v4/letter/s/9f8e36/32.png) [@softf5ajivanet](https://www.ruby-forum.com/u/softf5ajivanet)\ **Post date:** [August 26, 2015, 4:26am UTC](https://www.ruby-forum.com/t/usrps-not-communicating/243144/3 "2015-08-26T04:26:06Z")
On 08/25/2015 10:17 PM, devin kelly wrote: > You can also just synchronize the USRPs with a 10MHz reference and 1PPS. > > Devin > The USRP2 uses a 2.5PPM master on-board clock, so at 200Mhz tuned > frequency, the max it can be off is about 500Hz. For narrowband > transmission, > that may be too much, but I’d be very surprised to see 20kHz offset > with a 200MHz carrier. ---
**Author:** ![softf5ajivanet](https://www.ruby-forum.com/letter_avatar_proxy/v4/letter/s/9f8e36/32.png) [@softf5ajivanet](https://www.ruby-forum.com/u/softf5ajivanet)\ **Post date:** [August 26, 2015, 5:42am UTC](https://www.ruby-forum.com/t/usrps-not-communicating/243144/4 "2015-08-26T05:42:02Z")
Ah sorry I wasn’t specific. This was from years ago, the numbers I used were just meant to be an example, not accurate. ---
**Author:** ![softf5ajivanet](https://www.ruby-forum.com/letter_avatar_proxy/v4/letter/s/9f8e36/32.png) [@softf5ajivanet](https://www.ruby-forum.com/u/softf5ajivanet)\ **Post date:** [August 26, 2015, 12:43pm UTC](https://www.ruby-forum.com/t/usrps-not-communicating/243144/5 "2015-08-26T12:43:48Z")
Dear Ravi, this is no good. We’ve told you several times now that you need to fix things, and you just come back with your same broken configuration and ask the same question. It’s totally OK if you just say “hi, sorry, I did not understand your answer, could you explain this and that”, but not addressing what we wrote is really getting frustrating, for me at least. Let me repeat what Tom and others have tried to explain to you: There’s no single “switch” you can flip so that your transmission works. We have pointed out methods to help you optimize your transmission, for example looking at the RX spectrum and adjusting gains until it looks “good”. We can’t, just by looking at the problems the receiver script has, tell you what you need to do. Really. Please stop asking the same question over and over again, and ask different questions. A clever one would, for example, be: I’ve got this spectrum (screenshot), and it doesn’t look like \. Increasing the gain just made things worse." I hope this leads you to a path that promises more success than what you’re currently doing. So: You’ve gotten /a lot/ of responses. You might need to consider how much consultation costs you would have caused if we weren’t doing this out of interest but for money – so please explain what you did not understand in our answers, _we really want to help!_ Here’s my rather comprehensive answer from last time: > You should really specify the gain. That defines how much power comes > > > > > 250kS/s. The effect of you not correcting your flow graph is that the > doing something else on it? Or are you running in a virtual machine? > > > No gain specified. > > Actual sample rate: 0.250000 MSps > > effect kind of cancels itself out – still, fix that! > > to me as “John”, so I’d suggest you pick one and stick with that, > because it’s less confusing. > > [1] “Communication Problems between 2 USRP” email thred Best regards, Marcus