Re: Max Output Power from the LFTX

GNU Radio
1

Thanks Marcus for your reply.

Probably somewhere in the neighborhood of +7dBm, maybe a little more.

The LF/BASIC series have no gain in either path, so you’re just looked
at buffered ADC ouput.

So, ~ +7dBm is the max output power I supposed to get from the LFTX
daughterboard? How do I get that?

Since I am only getting -28.13 dBm, does that mean I have some issue
with
my LFTX daughterboard?

Khalid

Message: 2
Date: Tue, 6 Jan 2015 10:13:45 -0330
From: “khalid.el-darymli” [email protected]
To: “[email protected]” [email protected]
Subject: [Discuss-gnuradio] Max Output Power from the LFTX
daughterboard
Message-ID:
<CACdmG=zCHATiwrtv=wBVLwG0_
[email protected]>
Content-Type: text/plain; charset=“utf-8”

Hi,

What is the maximum output power from the LFTX daughterboard when used
with
the USRP N200?

According to this datasheet [1], the N200 with the WBX daughterbaord
provide an output power of 15 dBm. However, when using the LFTX
daughterboard, I am getting a much less output power.
[1]
http://www.ettus.com/content/files/07495_Ettus_N200-210_DS_Flyer_HR.pdf

In GNU Radio with the USRP N200, we use a sinusoid with a frequency of
150
kHz and an amplitude of 0.98, fed into the LFTX daughterboard for a
center
frequency of 5 MHz. When the output of LFTX is plugged into a scope
terminated with a 50-ohm terminator, the scope reads 24.8 mV
(peak-to-peak). This is around -28.13 dBm.

Is this the max power one can get out of the LFTX daughterboard?

Thanks.

Best regards,
Khalid
-------------- next part --------------
An HTML attachment was scrubbed…
URL: <
http://lists.gnu.org/archive/html/discuss-gnuradio/attachments/20150106/37846ff4/attachment.html

Message: 3
Date: Tue, 06 Jan 2015 08:48:08 -0500
From: “Marcus D. Leech” [email protected]
To: [email protected]
Subject: Re: [Discuss-gnuradio] Max Output Power from the LFTX
daughterboard
Message-ID: [email protected]
Content-Type: text/plain; charset=UTF-8; format=flowed

On 01/06/2015 08:43 AM, khalid.el-darymli wrote:

Best regards,
Khalid
Probably somewhere in the neighborhood of +7dBm, maybe a little more.

The LF/BASIC series have no gain in either path, so you’re just looked
at buffered ADC ouput.

–
Marcus L.
Principal Investigator
Shirleys Bay Radio Astronomy Consortium

2

Since I am only getting -28.13 dBm, does that mean I have some issue
with my LFTX daughterboard?

Khalid

Output power is determined largely by baseband signal magnitude in that
case.

3

Output power is determined largely by baseband signal magnitude in that
case.
Yes, I understand. And I noticed that for baseband signal with an
amplitude *>
1, * the pass-band signal gets clipped off. So I assume that an
amplitude
of 1 for the baseband signal should deliver the maximum output power
(for
the pass-band signal).

My question was, what is the maximum power one can get for the pass-band
signal output from LFTX? In your earlier reply, you said it is around
+7dBm, am I getting this right? In my case, for a baseband amplitude of
around 1, I am only getting -28.13dBm, much less than what you said, I
am
not sure,* why?*

Thanks,
Khalid

4

is around +7dBm, am I getting this right? In my case, for a baseband
amplitude of around 1, I am only getting -28.13dBm, much less than
what you said, I am not sure,why?

Thanks,
Khalid

For a magnitude of ~1, you should see maximum amplitude signals coming
out of it.

You could try setting gain, but on any platforms with VGAs in the DAC,
there’s only about 6dB of gain range settable. Trying setting a gain in
your
sink block of, let’s say, 20dB, with a magnitude of 0.9. What power
output do you get? How are you measuring power? What is your
flow-graph
actually doing? Have you tried both connectors on the LFTX?

5

share this info since what you’re seeing might be a hardware issue
rather than something you can resolve with settings.

-Mike
The only active component that can really fail on those boards is the
differential-to-single-ended buffer. But you have to work failry hard
to make
those fail. Like maybe driving a dead short for an extended period,
or a big ESD zap on the RF connector.

6

Hi Khalid,

I had a similar issue with an LFTX that I had been working with for
about
two years. I actually had two boards running on separate USRP2’s and
both
were working fine for quite a while, but I recently noticed that one had
a
drastically lower output range. I ended up replacing the LFTX and all is
working well now, so I’m assuming it was a hardware issue that came up
at
some point.

I didn’t take the time to look for the problem, but I figured I’d share
this info since what you’re seeing might be a hardware issue rather than
something you can resolve with settings.

-Mike

On Tue, Jan 6, 2015 at 12:59 PM, khalid.el-darymli
<[email protected]

7

Trying setting a gain in your sink block of, let’s say, 20dB, with a
magnitude of 0.9. What power output do you get? How are you measuring
power?

This gives me 23.2 mV (peak to peak). This is the reading I get from a
scope connected to a T. For the input ends of the T, one end is
connected
to a 50-ohm terminator and the other end is connected to the output from
LFTX. This gives a power of: 30+10 log_10[{23.2e-03/[2*sqrt(2)]}^2/50]=
-28.711 dBm.

What is your flow-graph actually doing? Have you tried both connectors
on the LFTX?

Simple, a signal source (Sine) block directly connected to a USRP Sink
block.
Yes, both connectors on the LFTX daughterboard give the same voltage
reading.

Thanks,
Khalid

8

Thanks Mike. I appreciate your comment.
I’ll look into buying a new LFTX.

Khalid

9

Is the input impedance of your scope 1Mohm or 50 ohms? You sure you have
the voltage multiplier (1x/10x) set correctly?

On Tue, Jan 6, 2015 at 11:01 AM, khalid.el-darymli
<[email protected]

10

Yes, it is 1 Mohm, and my probe is set to 1X.

Khalid

11

Is the input impedance of your scope 1Mohm or 50 ohms?
Yes, it is 1 Mohm,

Well, in that case, your power consideration is not really correct; the
LFTX is designed for 50 Ohm load impedance; so, for meaningful
measurements, you’d have to measure the voltage over 50 Ohm.
I’m not really sure how much current the opamp on the LFTX is designed
to source, but I think adding a 50 Ohm load would reduce the current
through the negative feedback resistors R4/R7 (and in turn reduce the
bias over R33-R3/R35-R6) , which might change your measurements
significantly.

Greetings,
Marcus

12

On 01/06/2015 02:17 PM, Marcus M. wrote:

significantly.

Greetings,
Marcus
The AD8047 is designed for a max of 50mA drive current. If you do the
math, that’s plenty.

I believe that Khalid was terminating in 50Ohms resistive, so with a
1Mohm impedance across that, the scope input impedance will hardly make
any
difference at all–it should “see” the voltage developed across that
50Ohms, without noticeably changing it.

13

On Tue, Jan 6, 2015 at 11:17 AM, Marcus Müller
[email protected]
wrote:

the negative feedback resistors R4/R7 (and in turn reduce the bias over
R33-R3/R35-R6) , which might change your measurements significantly.

With a 50ohm term hanging off the other end of a tee, the effective
impedance is 50ohm, so I’d expect to see the rated 7dBm or so coming out
of
his setup.

Just for reference, I’m currently getting 5mW (7dBm, 700mV P-P) out of
an
LFTX into a 50 ohm load for another project.

–n

14

Hi All,

Just to update you in case somebody else comes across this problem.
Following Mike’s advice, we replaced the LFTX daughterboard with a new
one.
We’re getting now 1 V p-p. This is around 4 dBm.

Thanks very much for your help.

Best regards,
Khalid

15

Ah ok, I was (wrongly) assuming Khalid just leaves the TX port virtually
open.

Greetings,
Marcus

16

On 01/16/2015 02:08 PM, khalid.el-darymli wrote:

Interesting. The driver chip that’s used, as a unity-gain
differential-to-single-ended driver, is, according to the datasheets,
pretty robust. So, it would
be interesting to know exactly what the failure mode is.

17

We were using only one Tx channel with the Subdev Spec in the USRP Sink
is
set to A:AB

Could the failure be due to that the other (idle) Tx channel is left
open
circuit (i.e., not terminated by 50 ohms)?

Khalid

On Fri, Jan 16, 2015 at 3:41 PM, Marcus D. Leech [email protected]
wrote:

Best regards,

at some point.

of around 1, I am only getting -28.13dBm, much less than what you

with my LFTX daughterboard?

    daughterboard

frequency of 5 MHz. When the output of LFTX is plugged into a scope
-------------- next part --------------
To: [email protected]

scope terminated with a 50-ohm terminator, the scope reads 24.8 mV

18

I see. If you would like to diagnose the problem, we can ship you the
malfunctioned daughterboard?

Thanks,
Khalid

19

On 01/16/2015 02:42 PM, khalid.el-darymli wrote:

We were using only one Tx channel with the Subdev Spec in the USRP
Sink is set to A:AB

Could the failure be due to that the other (idle) Tx channel is left
open circuit (i.e., not terminated by 50 ohms)?
That seems quite unlikely. These aren’t “precious princess” RF power
transistors, but reasonably-robust differential-to-unbalanced drivers.
I’d
be very surprised if they couldn’t handle an open-circuit condition.

20

On 01/16/2015 02:54 PM, khalid.el-darymli wrote:

I see. If you would like to diagnose the problem, we can ship you the
malfunctioned daughterboard?

Thanks,
Khalid

Thanks. I personally wouldn’t have time/equipment to really diagnose
it. So it would be up to Ettus R&D as to whether they want it back to
see
what the failure mode is.

I’ve been doing Ettus support for 6 years, and LF_TX are not generally
items that show up with RMA requests very often.