DSA875, damaged?

[grg183]

HI,

I have a DSA875 which got accidentally fed a +26dBm signal at 2.5GHz. The DSA was setup with 30dB internal attenuation but unfortunately no external attenuator and by the time I realized it had already been running on this signal for a while. I don't recall seeing any overload warnings from the DSA but the max hold plot was saturated on the +20dBm mark so I thought it was on the limit but still safe (I should have added an external attenuator, I know). After a while of this, I started reading a lower signal. The frequency and shape of the signal were correct but the amplitude went down to 0dBm (for the same signal as before).

I don't have a way to test with a separate signal generator or second analyzer at the moment so to test it out I connected the TG from the same DSA directly to its input expecting to see a flat plot. The attached screenshot is what I got with the TG set on -10dBm full span. Below around 3.2GHz, there is roughly a -25dB attenuation but above 3.2GHz it seems to be measuring correctly. I loaded the system preset before doing this measurement.

It is interesting that below 3.2GHz it seems to still be measuring but with an attenuated reading.

Am I missing something or is it damaged? Anyone had this happen?

[tv84]

Have you tried a settings reset?

[tautech]

Have you tried a settings reset?

I loaded the system preset before doing this measurement.

It seems so.

FWIW I've seen the same stepped sweep at 600 MHz in a Siglent unit that has also been damaged.
If the TG wasn't in use at the time it's likely okay and just the RF input damaged.

[RFDx]

It is interesting that below 3.2GHz it seems to still be measuring but with an attenuated reading.
Am I missing something or is it damaged? Anyone had this happen?

It looks like the SA is using two receiver frontends to cover the full 7.5GHz span. Something happened to the frontend responsible for the 3.2GHz range.  The strange thing is that the TG plot is flat up to 3.2GHz. It doesn't seem like something got damaged just permanently attenuated by about 26dB.

[grg183]

It looks like the SA is using two receiver frontends to cover the full 7.5GHz span. Something happened to the frontend responsible for the 3.2GHz range.  The strange thing is that the TG plot is flat up to 3.2GHz. It doesn't seem like something got damaged just permanently attenuated by about 26dB.

Yes that's what I thought too, they must have 2 receivers, one for <3.2GHz, and another for >3.2Ghz, which probably makes sense considering there is another model in the DSA800 series limited to 3.2GHz.

FWIW I've seen the same stepped sweep at 600 MHz in a Siglent unit that has also been damaged.
If the TG wasn't in use at the time it's likely okay and just the RF input damaged.

The TG itself should be fine, I've barely ever used it and it was not in use at the time. At the time of the damage I had it set on 2.508GHz with 10MHz span, 30dB internal attenuator, 20dBm reference level and the signal was displayed top flattened on the 20dBm mark, later I found the connected device was outputting around 26dBm.

It seem like something in the measurement path got degraded resulting in an attenuation but the receiver itself seem to still be good. I've tried a few known signals and they are displayed perfectly just attenuated.

[RFDx]

At the time of the damage I had it set on 2.508GHz with 10MHz span, 30dB internal attenuator, 20dBm reference level and the signal was displayed top flattened on the 20dBm mark, later I found the connected device was outputting around 26dBm.

The damage level with 30dB attenuation is above 30dBm. I don't think the frontend is damaged. Does the attenuator still work? Does the preamp increase the level by ~17dB? Your 26dBm input signal and the additional attenuation of exactly 26dB doesn't seem to be a coincidence. Did you somehow change the "reference level offset" by 26dB? A level setting somewhere in the menu could be off.

[tautech]

At the time of the damage I had it set on 2.508GHz with 10MHz span, 30dB internal attenuator, 20dBm reference level and the signal was displayed top flattened on the 20dBm mark, later I found the connected device was outputting around 26dBm.

The damage level with 30dB attenuation is above 30dBm. I don't think the frontend is damaged. Does the attenuator still work? Does the preamp increase the level by ~17dB? Your 26dBm input signal and the additional attenuation of exactly 26dB doesn't seem to be a coincidence. Did you somehow change the "reference level offset" by 26dB? A level setting somewhere in the menu could be off.

Running Preset will fix that.
It's damaged for sure.

[grg183]

The damage level with 30dB attenuation is above 30dBm. I don't think the frontend is damaged. Does the attenuator still work? Does the preamp increase the level by ~17dB? Your 26dBm input signal and the additional attenuation of exactly 26dB doesn't seem to be a coincidence. Did you somehow change the "reference level offset" by 26dB? A level setting somewhere in the menu could be off.

That's an interesting observation, it does seem weird for it to be a coincidence! However as tautech says the Preset should fix that. Its still really weird that it's still measuring but just attenuated.
I confirm that the pre-amp and internal attenuator are still working correctly and they are applying the expected amount of gain and attenuation also in the < 3.2Ghz range.

I will try to double check that it is not the TG but that was not in use at the time and I was seeing the same attenuation on my signal as well, so that is unlikely.

 Contacted Rigol service center and they say it is a damaged receiver, their only option is a ~€4k full motherboard swap
I will probably just get a SSA3075X-R to replace it and keep the DSA875 as a spare knowing that the reading is off, or even just as a generator. I don't see it making sense spending that much money to repair it.

[RFDx]

I confirm that the pre-amp and internal attenuator are still working correctly and they are applying the expected amount of gain and attenuation also in the < 3.2Ghz range.

This means the damage must be somewhere in the 1. IF section. A fault in the fixed frequency IF would manifest as a flat but attenuated amplitude plot. The upper 3.2...7.5GHz range probably downconverts directly to the 2nd IF and still works as expected.

[grg183]

I confirm that the pre-amp and internal attenuator are still working correctly and they are applying the expected amount of gain and attenuation also in the < 3.2Ghz range.

This means the damage must be somewhere in the 1. IF section. A fault in the fixed frequency IF would manifest as a flat but attenuated amplitude plot. The upper 3.2...7.5GHz range probably downconverts directly to the 2nd IF and still works as expected.

I think that is an excellent observation! It's probably 'just' the IF amplifier damaged. I will definitely have a look inside and see if I can identify the IF section and troubleshoot it somehow.
I'm definitely not spending 4k+ to repair this.