RFX900 Failure

GNU Radio
1

Hi all,

Power output has significantly dropped, initially by 8dB, now much more.

My first thoughts are that the PA has failed, so is it possible to break
the transmitter…

  • by prolonged transmitting at high power (i.e. setting it to/near
    maximum)?
  • using a mismatched antenna?
  • mismatching resulting in VSWR effects? (err, without a load)?

I’m obviously going to have to test now, where to start, any
suggestions, like what voltage swing before and after the 3315 should I
expect?

Many Thanks
David

2

On 02/24/2010 12:42 PM, David E. wrote:

Hi all,

Power output has significantly dropped, initially by 8dB, now much more.

My first thoughts are that the PA has failed, so is it possible to
break the transmitter…

  • by prolonged transmitting at high power (i.e. setting it to/near
    maximum)?
    Probably not
  • using a mismatched antenna?
  • mismatching resulting in VSWR effects? (err, without a load)?

RF amplifiers don’t like transmitting into badly-mismatched loads. They
can definitely get damaged that
way. Anything below 2:1 VSWR will probably be OK, but beyond that,
you flirt with amplifier damage.

Discuss-gnuradio mailing list
[email protected]
http://lists.gnu.org/mailman/listinfo/discuss-gnuradio

–
Marcus L.
Principal Investigator
Shirleys Bay Radio Astronomy Consortium

3

On 02/24/2010 09:42 AM, David E. wrote:

I’m obviously going to have to test now, where to start, any
suggestions, like what voltage swing before and after the 3315 should I
expect?

I have seen this once before with someone who was transmitting at max
power continuously. The problem may be in the SAW filter, which would
make it easy to fix. You can just put a cap of anywhere between 50 and
1000 pF, size 0603 in the empty capacitor location which is in parallel
with the filter.

In order to tell if that really is the problem, you would probably need
to probe with an RF probe for your spectrum analyzer or vary fast
oscilloscope. You could probe at the antenna port and immediately
before the SAW filter, and if there is a big loss in the filter you know
that is bad.

If you don’t have the equipment to test this, it may be easier to just
put the cap in there and try it.

Matt

4

Hi,

Thanks for the fast response. Yes the SAW filter is broken and putting a
capacitor in as you suggested has brought the output power back up to
normal. :slight_smile:

So, a couple of questions please,

* Can the SAW be damaged by too much power from the PA? (I could
  find no info on the maximum power limits for this device, just the
  bandwidth and attenuation).
* Can the SAW be damaged by a mismatched load (i.e no load!). I've
  asked around about this and get differing opinions
      o Yes, because the reflected power will be additively doubled,
        and enhanced due to the high Q of the filter
      o No, because the filter is a passive device, and the power
        will just pass through.
* Also, apparently, SAW filters can easily be damaged due to
  physical shock, damaging the piezo electric material, so maybe
  this was just a one-off. I'll have to replace the chip anyway

Thanks again guys,
David

5

On 02/26/2010 10:05 AM, David E. wrote:

Hi,

Thanks for the fast response. Yes the SAW filter is broken and putting a
capacitor in as you suggested has brought the output power back up to
normal. :slight_smile:

Great! If you can survive without the filtering, you’re actually better
off now because you’ll have 1.5 to 2 dB less loss.

So, a couple of questions please,

  • Can the SAW be damaged by too much power from the PA? (I could
    find no info on the maximum power limits for this device, just the
    bandwidth and attenuation).

It can be. I had been told by Sawtek that the part was supposed to be
good to half a watt. While the RFX900 won’t put out that much power,
some of them will come within 1-2 dB of that. The longer you operate
near any parts maximums, the shorter its life will be. This is only the
second time in many many boards and over 4 years that we’ve seen this
problem, so you just might be on the unlucky tail of the distribution.

  • Can the SAW be damaged by a mismatched load (i.e no load!). I’ve
    asked around about this and get differing opinions
    o Yes, because the reflected power will be additively doubled,
    and enhanced due to the high Q of the filter

Not enhanced by the Q, but a high SWR can result in higher peak voltage
which can stress a device.

o No, because the filter is a passive device, and the power
will just pass through.

Each time the power passes through it is subject to insertion loss which
is dissipated, so reflected power will add a little to the dissipation
if you have a really bad reflection.

  • Also, apparently, SAW filters can easily be damaged due to
    physical shock, damaging the piezo electric material, so maybe
    this was just a one-off. I’ll have to replace the chip anyway

My best guess is that this is due to sustained high power over a long
period heating and aging the device.

Matt