usrp-users@lists.ettus.com
Discussion and technical support related to USRP, UHD, RFNoC
View all threadsNew Product Announcements from Ettus Research
---====================
New Product Announcements from Ettus Research
---====================
We are very excited to announce the immediate availability of the
latest low cost Software Radio product from Ettus Research, the
USRP(tm) B100. The B100 has the following features:
- USB 2.0 interface
- Xilinx Spartan 3A-1400 FPGA
- Compatibility with our entire daughterboard family
- Fully supported by UHD drivers
- Dual 64 MS/s 12-bit ADCs
- Dual 128 MS/s 14-bit DACs
- Onboard TCXO for precise frequency control
- 10 MHz and 1 PPS inputs for external references
- Flexible clocking from 10 MHz to 64 MHz
- 8 MHz of RF bandwidth with 16 bit samples
- 16 MHz of RF bandwidth with 8 bit samples
The price is $650 each, and the B100 is in stock and ready to ship.
---====================
The USRP E110 has a larger FPGA (Spartan 3A-DSP 3400) than
the E100, but is otherwise the same. This is perfect for those who
wish to offload DSP operations from the main ARM processor.
Both the E100 and E110 are now fully compatible with the GPSDO.
The price of the USRP E110 is $1500 and the E100 is $1300.
Both the E100 and E110 are in stock and ready to ship.
---====================
As always, you can purchase all of our products through:
http://www.ettus.com/order
and you can contact sales@ettus.com if you have any further
questions.
Thanks,
Matt Ettus
President, Ettus Research LLC
Hi Matt, great news!
I wonder, what is a frequency step of the reference generator in B100.
Could I tune it with at 1Hz precision or better?
On Tue, Oct 4, 2011 at 01:08, Matt Ettus matt@ettus.com wrote:
---====================
New Product Announcements from Ettus Research
---====================
We are very excited to announce the immediate availability of the
latest low cost Software Radio product from Ettus Research, the
USRP(tm) B100. The B100 has the following features:
- USB 2.0 interface
- Xilinx Spartan 3A-1400 FPGA
- Compatibility with our entire daughterboard family
- Fully supported by UHD drivers
- Dual 64 MS/s 12-bit ADCs
- Dual 128 MS/s 14-bit DACs
- Onboard TCXO for precise frequency control
- 10 MHz and 1 PPS inputs for external references
- Flexible clocking from 10 MHz to 64 MHz
- 8 MHz of RF bandwidth with 16 bit samples
- 16 MHz of RF bandwidth with 8 bit samples
The price is $650 each, and the B100 is in stock and ready to ship.
---====================
The USRP E110 has a larger FPGA (Spartan 3A-DSP 3400) than
the E100, but is otherwise the same. This is perfect for those who
wish to offload DSP operations from the main ARM processor.
Both the E100 and E110 are now fully compatible with the GPSDO.
The price of the USRP E110 is $1500 and the E100 is $1300.
Both the E100 and E110 are in stock and ready to ship.
---====================
As always, you can purchase all of our products through:
http://www.ettus.com/order
and you can contact sales@ettus.com if you have any further
questions.
Thanks,
Matt Ettus
President, Ettus Research LLC
USRP-users mailing list
USRP-users@lists.ettus.com
http://lists.ettus.com/mailman/listinfo/usrp-users_lists.ettus.com
--
Regards,
Alexander Chemeris.
On Mon, Oct 3, 2011 at 2:30 PM, Alexander Chemeris <
alexander.chemeris@gmail.com> wrote:
Hi Matt, great news!
I wonder, what is a frequency step of the reference generator in B100.
Could I tune it with at 1Hz precision or better?
The clock generator is the same as the USRP E1x0 series. It will not
generate 1 Hz steps by itself. If you need steps that fine you will need to
use an external reference and tune the reference in fine steps.
Matt
On Mon, Oct 3, 2011 at 4:30 PM, Matt Ettus matt@ettus.com wrote:
On Mon, Oct 3, 2011 at 2:30 PM, Alexander Chemeris <
alexander.chemeris@gmail.com> wrote:
Hi Matt, great news!
I wonder, what is a frequency step of the reference generator in B100.
Could I tune it with at 1Hz precision or better?
BTW, why would you ever need that? That's about 16 parts per billion per
step.
Matt
On Tue, Oct 4, 2011 at 03:32, Matt Ettus matt@ettus.com wrote:
On Mon, Oct 3, 2011 at 4:30 PM, Matt Ettus matt@ettus.com wrote:
On Mon, Oct 3, 2011 at 2:30 PM, Alexander Chemeris
alexander.chemeris@gmail.com wrote:
Hi Matt, great news!
I wonder, what is a frequency step of the reference generator in B100.
Could I tune it with at 1Hz precision or better?
BTW, why would you ever need that? That's about 16 parts per billion per
step.
To run OpenBTS reliably. GSM specs says your reference clock should
have 100ppb stability and precision, which is only 5Hz@52MHz. Our
experience shows that many phones still work with 500ppb=0.5ppm clock
offset, but this is only 25Hz@52MHz. So you can't get reliably working
OpenBTS if a reference clock can't tune with precision of at least
25Hz (5Hz is better). This is the reason we made ClockTamer to be
tunable with 1Hz step.
So, what is the minimum tunning step for E100/B100 exactly? I don't
have my E100 near my hands to test by myself at this moment.
--
Regards,
Alexander Chemeris.
A software-defined radio shouldn't need to incrementally adjust its
reference clock. Instead, it can maintain a tuning offset. Any USRP
can tune to fractions of a millihertz in the DSP. So if you're using
kalibrate or another program to estimate tuning error, you can make up
that error by retuning instead of reclocking.
In E100 and N210, the simple way to get a base-station-quality
frequency reference is to use a GPSDO -- the GPSDO kit Ettus Research
offers gives 0.01ppb accuracy averaged over 24h, which amounts to
0.5mHz @ 52MHz (or 0.01Hz at 1GHz carrier).
That said, for most non-carrier-grade applications, the stock 10MHz
reference of E100, E110, N200, and N210 is good enough for OpenBTS use
right out of the box, using Tom Tsou's OpenBTS-UHD branch.
--n
On Tue, Oct 4, 2011 at 12:43 AM, Alexander Chemeris
alexander.chemeris@gmail.com wrote:
On Tue, Oct 4, 2011 at 03:32, Matt Ettus matt@ettus.com wrote:
On Mon, Oct 3, 2011 at 4:30 PM, Matt Ettus matt@ettus.com wrote:
On Mon, Oct 3, 2011 at 2:30 PM, Alexander Chemeris
alexander.chemeris@gmail.com wrote:
Hi Matt, great news!
I wonder, what is a frequency step of the reference generator in B100.
Could I tune it with at 1Hz precision or better?
BTW, why would you ever need that? That's about 16 parts per billion per
step.
To run OpenBTS reliably. GSM specs says your reference clock should
have 100ppb stability and precision, which is only 5Hz@52MHz. Our
experience shows that many phones still work with 500ppb=0.5ppm clock
offset, but this is only 25Hz@52MHz. So you can't get reliably working
OpenBTS if a reference clock can't tune with precision of at least
25Hz (5Hz is better). This is the reason we made ClockTamer to be
tunable with 1Hz step.
So, what is the minimum tunning step for E100/B100 exactly? I don't
have my E100 near my hands to test by myself at this moment.
--
Regards,
Alexander Chemeris.
USRP-users mailing list
USRP-users@lists.ettus.com
http://lists.ettus.com/mailman/listinfo/usrp-users_lists.ettus.com
Hi Nick,
Thank you for your thoughts. My comments are below:
On Tue, Oct 4, 2011 at 21:59, Nick Foster nick@ettus.com wrote:
A software-defined radio shouldn't need to incrementally adjust its
reference clock. Instead, it can maintain a tuning offset. Any USRP
can tune to fractions of a millihertz in the DSP. So if you're using
kalibrate or another program to estimate tuning error, you can make up
that error by retuning instead of reclocking.
ADC and DAC will still operate with the frequency offset and this will
give you a clock skew. Some mobiles are tolerant to that, but some are
more picky. And this may give you strange behavior during testing.
In E100 and N210, the simple way to get a base-station-quality
frequency reference is to use a GPSDO -- the GPSDO kit Ettus Research
offers gives 0.01ppb accuracy averaged over 24h, which amounts to
0.5mHz @ 52MHz (or 0.01Hz at 1GHz carrier).
Yes, but it costs just as a USRP itself and is excessive for OpenBTS
lab testing. It may be necessary somewhere, but for OpenBTS something
like ClockTamer's GPS sync is good enough. Many people complain that
OpenBTS kit with USRP is already too costly.
That said, for most non-carrier-grade applications, the stock 10MHz
reference of E100, E110, N200, and N210 is good enough for OpenBTS use
right out of the box, using Tom Tsou's OpenBTS-UHD branch.
2.5ppm is clearly not enough. You may be lucky to get one which will
work or you may be unlucky and fall into all those clocking troubles.
OpenBTS mailing list is bloated with mails that "OpenBTS doesn't work"
because of bad clocks already.
So, I still wonder what is the accuracy of E100/B100 clock tunning?
Our calculations show it's about few kHz, which is too big.
--
Regards,
Alexander Chemeris.