Siglent SSG3000X and SSG5000X series RF Signal generators

[Martin72]

My dear Rob, I'm just a little fool.
Not a big one.
But I can't see very well, the matter of whether or not there is an IQE module is clarified very quickly without measuring weight.

[Nokia6681]

Hello from Germany,

I have a question for you, I have the Siglent SSG3021X which no longer has an output as of today, I had not previously set up a device with it and then suddenly the output was DEAD, at the absolute output maximum there was still a low signal level but no signal full power more.

Now the question is what could be defective here, with a circuit diagram I would certainly be able to determine it straight away, but where can I get something like that from?

Do any of you have a circuit diagram of the HF output?

[tautech]

Hello from Germany,

I have a question for you, I have the Siglent SSG3021X which no longer has an output as of today, I had not previously set up a device with it and then suddenly the output was DEAD, at the absolute output maximum there was still a low signal level but no signal full power more.

Now the question is what could be defective here, with a circuit diagram I would certainly be able to determine it straight away, but where can I get something like that from?

Do any of you have a circuit diagram of the HF output?

First Preset then adjust Level and Frequency to values you can easily test.
Have you checked the BNC LF output is working correctly ?
Then work through the SM checks:
https://int.siglent.com/u_file/download/19_11_22/b8cd0afabb.pdf

[Nokia6681]

I never use the LF output.
 a check using the service instructions does not reveal any errors. I think that the device that I previously compared may have sent and destroyed the output Amp.
Unfortunately, Siglent only specifies a maximum DC input voltage at the output but does not indicate the maximum amount of HF that can enter the output. I haven't disassembled it yet to get a closer look at the components used.

[tautech]

I never use the LF output.
 a check using the service instructions does not reveal any errors. I think that the device that I previously compared may have sent and destroyed the output Amp.
Unfortunately, Siglent only specifies a maximum DC input voltage at the output but does not indicate the maximum amount of HF that can enter the output. I haven't disassembled it yet to get a closer look at the components used.

Yet it is a check the rest of the generator is working correctly and any damage is confined to HF RF output.

Not heard of damage/failures in these before so do keep us posted with your findings.

[Nokia6681]

I'm going to carry out more tests today, especially regarding the LF output, but I think I destroyed it with the radio, it's quite possible that I was transmitting with it and forgot to put it back on the dummy load and the generator was damaged as a result .

My dealer said 4 years ago that nothing is normally repaired on this device but that it is either repaired or replaced at a flat rate, but spending over 1000 euros on it makes me reluctant to look for a circuit diagram, but I will do it this afternoon as soon as I can I have time to disassemble and see what generates the output here, the driver stage must still be working because with the full output power setting, around 10-20µV still comes out of the output.

[tautech]

I'm going to carry out more tests today, especially regarding the LF output, but I think I destroyed it with the radio, it's quite possible that I was transmitting with it and forgot to put it back on the dummy load and the generator was damaged as a result .

My dealer said 4 years ago that nothing is normally repaired on this device but that it is either repaired or replaced at a flat rate, but spending over 1000 euros on it makes me reluctant to look for a circuit diagram, but I will do it this afternoon as soon as I can I have time to disassemble and see what generates the output here, the driver stage must still be working because with the full output power setting, around 10-20µV still comes out of the output.

I'm 100% sure many could benefit from your findings and any imagery of PCB and devices.
Best of luck.

[Nokia6681]

YES i will do my best, Pictures i will Record some of this , i hope sharp !

[TurboTom]

Reverse power events are the usual culprits of such failures in RF signal generators.  Even though virtually all these instruments have some reverse power protection at their outputs, if you happen to supply "enough" power in reverse, these measures either aren't fast enough or just won't be able to cope with the amount of power (some single-digit wattage will most probably be tolerated for a short time, but if it's in the 100W ballpark, it's "Hasta la vista, Baby").

These defects can be repaired in most cases since they are usually very localized, but it's questionable if the instrument is still in spec after the repair and even more so if you will be able to get a schematic. Higher reverse power events also may charr some components and even damage the PCB. I was lucky that two repairs of R&S SM300s with similar problems turned out okay for me (also no schematics, but at least a decent block diagram available). I think the Siglent generator will be easier to repair simply due to the more "organized" mechanical construction. But you will definitely need some experience and RF technology knowledge, and identifying components may be somewhat awkward as well.

Depending on your own experience, it may ba a good idea to disassemble the instrument and post some good, detailed photos of the relevant section here for supportive opinions before starting to actually "rework" the instrument.

Good luck with the repair and all the best,
Thomas

[Nokia6681]

Today I removed the mainboard and checked, R620 (49.9R) burned out, but the U70 is still supplied with 3.3V voltage, U46 should get voltage via C755, as does U45 via C754, at least I did with U45 There is voltage on both pins of C754, and all other components that look like this are apparently suppression chokes + C
U70 is the same chip as the Siglent SSA 3021 digital attenuator.
The LF output works perfectly, I have now ordered the H1118.
I was also able to measure the HF at full power up to U39, but after that the HF continues to decrease, but it could be that the U70 is blocking the data due to the 30W that went into it and now other ICs can't switch properly here.

I have to put all the pictures on my server because I can only post 2 pictures in full resolution here

[Nokia6681]

Here is the Link to my Server Folder with more Pictures uncompressed !

https://www.asz-muenchen.de/reserve/Siglent/

[Nokia6681]

In Germany if you need spare parts there is absolutely nothing left, you can't even get the 49.9 ohm resistor anywhere, these days you have to order everything from somewhere abroad, it's a shame and you then wait for these parts for weeks, It's no fun anymore to repair any device. Just buy and use everything and throw it in the electronic waste.

Is that also the case in your countries? Or can you get parts better here, Mauser Elektronik has pretty much no idea where they get these parts from, but you don't really want to order everything for 20 euros or so much that you don't need in order to save on shipping do it.

Have you looked at the pictures, a super clean layout, but beware of something broken like mine is currently, then good advice is very expensive.

[TurboTom]

My approach for initial fault finding would be to remove U70 (Hittite / Analog H1118) and to jumper pads 14 and 3 on the pcb with a wire as short as possible. If the output signal is re-established at the proper level, replace U70 and you should be ready to go.

The tricky part will be removal of U70 since it's got an exposed pad, the board appears to be at least a hybrid composition, maybe Rogers, and the "white" (RF) side of it isn't typically known for its rigidity... Traces are easily pulled off if the solder isn't completely liquefied. Lead-free solder that Siglent used for assembly doesn't simplify things either. You need a decent SMT rework setup and some experience to get it right. At least, I wouldn't approach the job with single-sided heating of the PCB, due to the multiple layers of copper inside.

I hope that the "dump switch" U70 has saved the rest of the circuitry!

All the best,
Thomas

[Nokia6681]

Yes, I'll do that as soon as the 1118 is delivered. Mauser has withdrawn the order because only certain companies can order it but not on my Company  , I have now ordered it via Farnell. It should supposedly arrive tomorrow, if it is true, I will be surprised.

Soldering and unsoldering is no problem for me, hot air and very good flux, I do these and BGA almost every day, just don't drink coffee beforehand!

I'll do the connecting of the input and output after it's outside, I measured it again yesterday with the Speci, which comes out at the back with the 2.2V set (10-12µV), so to speak, nothing

As soon as I have the new one, I measured it without the IC and then with the new one back in, I'll report back.

[markus_jlrb]

Dears,

I would not like to open a new thread, as my question is partially related to the SDG6kX device
in combination with the SSG3kX.

I would like to ask if someone of you had already tried to pass a IQ signal generated by the SDG6kX
to the internal ports available in the SSA3kX (without IQ HW option)?

See the internal SSG3kX pic's listed few postings above.

Markus

[TurboTom]

Won't work...unfortunately 

If you follow the traces around J1 and J2, you will find two two-way switches U18 and U22 that either link the RF signal internally (via C122, R40 Jumper, C124) or externally to J1 (output to the I/Q modulator module) and from J2 (input from the external I/Q modulator module) towards the AM modulator section. Without this module, the SSG3kX won't do I/Q modulation.

Cheers,
Thomas

[Martin72]

Hi...

After I was contacted by a forum member, I have now understood it and am of the opinion that this is indeed still a bug:
Screensaver function...
You set it to say 1min, then it actually runs - but not after a restart.
Even if you set the power-on state to "last", it does nothing.
According to the menu, the system then remembers that the screensaver was set from "off" to 1 minute, but it is no longer activated.
Only if you change it back to "off" and then again - and then it only works for the current "session", after a restart the screensaver is no longer executed as described above.

Martin

[Nokia6681]

The H1118 was delivered just one day after ordering.

Apparently I burned the first IC while soldering it in, but I didn't use the second IC again with hot air but used it conventionally and so far everything is working as before.

The problem with soldering in and out is the circuit board, so much heat is drawn off via the circuit board that you can't actually solder it in here without a heating plate on the underside and I don't have a heating plate, all the other ICs do So far in my life I have only replaced it with hot air, but this... and at a maximum of 280 degrees, which this IC can handle according to the specifications, you don't really get very far when soldering in, even if the tin is included Its melting point is 185 degrees, the board eats it like nothing.

The main thing is that the generator is now running again inside, what more could you want.

But I don't want to remove or solder this IC again without bottom heat!

[markus_jlrb]

@TurboTom,

according to the UI, an external IQ signal could be enabled in the SSG3kX.

That's  what confusing me. But on the other hand no external IQ Input BNC
are available, but the case is already prepared for this purpose.

Btw the UM claim that external IQ modulation via a SDG6kX is possible, but
Siglent support decline this statement written in the manual.

Markus

[Martin72]

The whole IQE story stands or falls on whether you have an IQE model.
This requires the board inside and the two BNC sockets on the rear panel.
If you don't have such a model, you probably don't need to think about it any further.

[2N3055]

@TurboTom,

according to the UI, an external IQ signal could be enabled in the SSG3kX.

That's  what confusing me. But on the other hand no external IQ Input BNC
are available, but the case is already prepared for this purpose.

Btw the UM claim that external IQ modulation via a SDG6kX is possible, but
Siglent support decline this statement written in the manual.

Markus

Please find attached the very User Manual.

Page 74:

2.8 I/Q Modulation
I/Q modulation, i.e. two orthogonal signals (carriers with the same frequency, phase
difference of 90°, generally expressed by Sin and Cos) and I (In-Phase, in-phase
component) and Q (Quadrature Phase) signals are respectively modulated by carrier and
transmitted together, thereby improving spectrum utilization.

Note: Only some models are equipped with IQ modulation function. For specific models,
please refer to the SSG3000X data sheet.


User Manual is for both I/Q model and the basic model without I/Q module.
Data sheet also plainly says there are two models..

[Nokia6681]

If this option was not ordered when you purchased it, the device does not have it and cannot be selected on the front of the display. My device does not have this function, if I had wanted it I would have had to order it with an additional payment when purchasing it.

I don't think Siglent offers to buy these separately because the firmware software also needs to be expanded.

[TurboTom]

The generator probably also needs to undergo an extended calibration since the I/Q modulator has to operate over the full bandwidth of the instrument. Thats one of the disadvantages of the "direct frequency generation" approach.

Other instruments like the R&S SM300 feature a frequency mixing approach to generate the desired output. There, the I/Q modulation is much easier since it can be applied to the "fixed frequency" mixer input. Moreover, wide frequency sweeps are possible. The advantage of the "direct" approach is that it's way less difficult to generate very low phase noise signals. So, both methods have their pros and cons.

[markus_jlrb]

@TurboTom,

thanks for explanation - I was not aware of this concept Siglent is using inside the SSG3kX device.

Other instruments like the R&S SM300 feature a frequency mixing approach to generate the desired output. There, the I/Q modulation is much easier since it can be applied to the "fixed frequency" mixer input. Moreover, wide frequency sweeps are possible. The advantage of the "direct" approach is that it's way less difficult to generate very low phase noise signals. So, both methods have their pros and cons.

Due to the fact that internally we have two ports on the main PCB for I/Q I was convinced that there must by a
mixer present to modulate the carrier with the fed in I/Q signal.

But when the mixing is done inside the FPGA you are right that my request could not be fulfilled.

Nevertheless I'm keen to check what will happen if I provide a small I/Q signal to the both internal ports.
Have to wait until my grantee for my device is expired and will explore in this time the opportunity to
get an I/Q module from Siglent repair support team in Germany.

Will report my findings as soon I get a proper hint on this topic.

Markus

[Nokia6681]

I don't think that will help you at all because you can't select anything in this menu. Software nor ready for this.