EEVblog Electronics Community Forum
Products => Test Equipment => Topic started by: KlausKragelund on September 22, 2021, 06:46:10 pm
-
Hi
I am the happy owner of a SVA1015X. Well, happy until it seems it has a burned front-end. I do not know if it was this way when I got it, but most likely I have had an accident in the lab, with a transient it could not survive
When I apply 0dBm, it reads -17dBm
I have a scope hooked up in parallel, since it does not act as 50ohms any more. When I measure the resistance with a Bode100 it has 7.7ohm above 100kHz
When the scope is hooked up alone to the generator it reads 445mVrms (6dBm). The 6dB extra comes from unloaded generator since the scope is 1Mohm
When I then add the spectrum analyzer in parallel, the scope reading falls to 60.7mVrms (-22.6dBm). This actually corresponds well to the 7.7 ohms load resistive divider
A photo of the input stage with the 49R9 ohm resistor
https://www.electronicsdesign.dk/tmp/siglent-sva1015x-spectrum--network-vector-analyser-teardown_28380822677_o.jpg (https://www.electronicsdesign.dk/tmp/siglent-sva1015x-spectrum--network-vector-analyser-teardown_28380822677_o.jpg)
I suspect I can get it running if I change the resistor
Any one else that has experience with SVA1015x, so other stuff I should look out for while I have it open?
Regards
Klaus
-
Since the protection diodes in your VNA's front end aren't populated, my best bet would be that (at least) the first RF switch, the HMC1118, is shot. These aren't too expensive so I'ld jut give it a try and replace it and then (if the VNA still isn't fixed), work up the signal path...
-
A close look at the picture makes me suspect the solder joint on the connector too, it looks cracked to me.
-
Since the protection diodes in your VNA's front end aren't populated, my best bet would be that (at least) the first RF switch, the HMC1118, is shot. These aren't too expensive so I'ld jut give it a try and replace it and then (if the VNA still isn't fixed), work up the signal path...
The picture is actually from the teardown by Dave Jones.
About the switch, what is it actually for? (I thought the 50R resistor would be connected always)
-
There might be some help here:
https://www.eevblog.com/forum/repair/fixing-the-rf-front-end-of-a-spectrum-analyzer/ (https://www.eevblog.com/forum/repair/fixing-the-rf-front-end-of-a-spectrum-analyzer/)
While that is for the SSA3021X the front end should be very similar.
-
There might be some help here:
https://www.eevblog.com/forum/repair/fixing-the-rf-front-end-of-a-spectrum-analyzer/ (https://www.eevblog.com/forum/repair/fixing-the-rf-front-end-of-a-spectrum-analyzer/)
While that is for the SSA3021X the front end should be very similar.
Very nice indeed, thanks :-)
-
The video shows the repair of an SSA3000X spectrum analyzer, the input circuitry of your SVA is slightly different.
The purpose of the mentioned switch is to route the input signal to the 49.9 ohms dump in case the input level is too high. In order to evaluate the input level, a portion of the signal is routed through the passives R5 / R6 / R10 / C19 to a voltage doubling detector consisting of D9 and D10. This DC signal is further on processed on the other side of the PCB probably by a comparator (digital or analog) and used to generate the control voltage for the input switch U3.
As mentioned before, U3 either dumps the input signal into R4 or passes it on to the second input selector switch U5 which forwards it to the programmable input attenuator U4. The other input to U5 and thus to the SVA's receiver is once again split by a switch (U103) whcih permits to select between an internal test signal or the output of the directional coupler at the output of the tracking generator in order to measure the reflected power (VNA mode).
Since the input level protection circuitry reacts with a short delay, sufficiently high signals may actually damage all the semiconductors at the input path, which would be in sequence the input protection switch U3, input selector switch U5 and also the attenuator U4. I wouldn't expect much damage to happen after the attenuator. In most such cases it's just the input protection switch that gets damaged.
Since you didn't publish a photo of your damaged SVA's PCB, we cannot tell form visible discoloration / charring what additional components may be damaged. You may try to measure the reflected power in VNA mode to check if this also reads too low. In this case, it's likely that the selector switch or the attenuator got damaged as well. If this level reads correctly, the fault should be limited to the input protection switch U3 and possibly surrounding passives.
Cheers and good luck,
Thomas
-
The video shows the repair of an SSA3000X spectrum analyzer, the input circuitry of your SVA is slightly different.
The purpose of the mentioned switch is to route the input signal to the 49.9 ohms dump in case the input level is too high. In order to evaluate the input level, a portion of the signal is routed through the passives R5 / R6 / R10 / C19 to a voltage doubling detector consisting of D9 and D10. This DC signal is further on processed on the other side of the PCB probably by a comparator (digital or analog) and used to generate the control voltage for the input switch U3.
As mentioned before, U3 either dumps the input signal into R4 or passes it on to the second input selector switch U5 which forwards it to the programmable input attenuator U4. The other input to U5 and thus to the SVA's receiver is once again split by a switch (U103) whcih permits to select between an internal test signal or the output of the directional coupler at the output of the tracking generator in order to measure the reflected power (VNA mode).
Since the input level protection circuitry reacts with a short delay, sufficiently high signals may actually damage all the semiconductors at the input path, which would be in sequence the input protection switch U3, input selector switch U5 and also the attenuator U4. I wouldn't expect much damage to happen after the attenuator. In most such cases it's just the input protection switch that gets damaged.
Since you didn't publish a photo of your damaged SVA's PCB, we cannot tell form visible discoloration / charring what additional components may be damaged. You may try to measure the reflected power in VNA mode to check if this also reads too low. In this case, it's likely that the selector switch or the attenuator got damaged as well. If this level reads correctly, the fault should be limited to the input protection switch U3 and possibly surrounding passives.
Cheers and good luck,
Thomas
Thanks Thomas for a very good explanation. I have disassembled the unit, and no discoloration is seen. It seems the added attenuation due to the fault is almost excactly the same as the video, so I hope it is just the switch. I have ordered a switch to change it. In the meantime I will have to live without the spectrum analyzer, although I use it a lot. (for EMC testing)
In the video comment section there is a suggestion to only connect the signal to the spectrum analyzer when the DUT has been turned on, and disconnecting the analyzer before turning the DUT off to avoid high voltage low frequency transients.