Rohde SM300 several issues

[wd5jfr]

Just got an SM300 which passes all 10 self tests, has nice clean but low output to 3GHz and will run for several hours and then no RF output.  Running self test shows all 10 self test as Error Timeout.  Shows 481 hrs run time with 456 starts, Running right now all night about 10 hrs, see what happens. Got $600.00 in it and don't have the capability for a component fix, need to replace circuit board if they become available???  So I wonder if there's reasonable repair service out there or I should give up and sell it?
Thanks
Hank

[wd5jfr]

After running non-stop over 13 hrs no errors.  Changed frequency setting  and did a few sweeps, still low on output.
The question: Buy, Sell, or Hold?
Any ideas appreciated.
It's kind of a nice small well designed unit easy to set-up. Now I'm wondering if there's a Rigol, Siglent, etc with similar features to look for if I can't get it repaired.
Thanks
Hank

[TurboTom]

I also yesterday received an SM300 that I bought with the well-known defect of not being able to output (almost) any signal above -26dBm setting, paid approximately the same price.

Already fixed the PA problem but I've got to say, this piece of equipment is a complex design. One can really see the quality difference compared to a "real" budget instrument. Every critical part is double- and triple shielded with RF absorbing foam layers in between. Structural shielding parts are machined from aluminium and afterwards silver plated. Everything "smells" quality... It's no surprise that these instruments sold new twice or three times as expensive as Rigol&Siglent (R&S???  ) instruments nowadays.

So if the SM300 is (half-way) operational, I'ld say it's a keeper, especially considering what has to be spent on something with even basic complex modulation abilities. I'ld consider the chance to source a working spare RF assembly to be virtually non-existant, at least reasonably priced, since this is the module that usually fails on these generators.

@wd5jfr: I didn't completely understand the exact failure mode of your instrument - you say it's low on output, does that mean the level is low over the whole range, or is it (more or less) accurate below -26dBm (no modulation enabled) and drops when you select output levels above that figure? That would be a clear indication for a faulty output MMIC (read below). If it's low throughout the whole level range and shows some discontinuity at the -26 -> -25dBm threshold, it's still a strong indication for a faulty output MMIC but in conjunction with a possibly broken output RF switch (AW002R2-12). If the level is adjustable continuously over the whole range but just too low, I'ld look at the output circuitry behind the switch. Only if there are several, repeating discuntinuities, I'ld search the fault at the attenuators. The weak spot of these generators appears to be the tolerance of the output circuitry against reversely applied RF power (i.e. testing a transciever's receiving performance and accidentially hitting p.t.t...  ).

Reaching the critical components in the RF section isn't a job for the faint of heart, there are about 100 screws of well mixed sizes to be removed in order to get access to the relevant area, several layers of conductive foam to be peeled off (possibly without damaging them) and finally the defective components have to be identified and replaced.

Since the board cannot easily be powered up when exposed and separated from the digital / ch2 AF board, real fault testing is difficult. I've been able to gradually apply the PA bias voltage to find that the constant current source for the output stage MMIC (AGB3303) maxed out already at 2.5V, so obiously this chip was faulty. I also noticed that some previous repair work had been executed on this board, probably after some catastrophic damage had been done to the final microwave switch with some pyrotechnical effects going off... At that time, the AGB3303 must have been replaced as well (as the flux residues witnessed). Since I didn't find any more defects, I went ahead and replaced the defective MMIC. But I didn't have a spare AGB3303 so I compared datasheets and took the risk to install an SBB5089Z which works surprisingly well in that application despite it's designed for constant voltage operation and not with a resistive dropper (like the original) or even a CC source as supply.

Now the generator works in all ranges almost corretly and reaches 13dBm or more but obiously, it's a db or slightly more off level. So here comes the million dollar question: Does anyone have or is aware of a calibration instruction for the SM300? I'm pretty sure with my SA and an HP power meter, I'ld be able to get the calibration done well enough to meet its spec.

I searched the web and didn't find anything...any help would be appreciated!

[chick0n]

Someone on Youtube Reversed the Generator Board a Bit...



But its in Spanish...

[tv84]

But its in Spanish...

A language that has more native speakers than English... 

[tv84]

Someone on Youtube Reversed the Generator Board a Bit...

My spanish is not very good and the signal generation area is worse  but... I had a quick look at some parts of the video to see what the author was showing. Besides doing a general description of the various board blocks, the author focuses on the IC AD7417 that supervises the temperature of several areas of the board. He had a problem of early shutoff of the board and decided to place a server fan (as he shows) in order to cool down the board. I think his stability problems stopped with that improvement.

A knowledgeable Spaniard could help you more. 

[COSMOS2K]

I also yesterday received an SM300 that I bought with the well-known defect of not being able to output (almost) any signal above -26dBm setting, paid approximately the same price.

Already fixed the PA problem but I've got to say, this piece of equipment is a complex design. One can really see the quality difference compared to a "real" budget instrument. Every critical part is double- and triple shielded with RF absorbing foam layers in between. Structural shielding parts are machined from aluminium and afterwards silver plated. Everything "smells" quality... It's no surprise that these instruments sold new twice or three times as expensive as Rigol&Siglent (R&S???  ) instruments nowadays.

So if the SM300 is (half-way) operational, I'ld say it's a keeper, especially considering what has to be spent on something with even basic complex modulation abilities. I'ld consider the chance to source a working spare RF assembly to be virtually non-existant, at least reasonably priced, since this is the module that usually fails on these generators.

@wd5jfr: I didn't completely understand the exact failure mode of your instrument - you say it's low on output, does that mean the level is low over the whole range, or is it (more or less) accurate below -26dBm (no modulation enabled) and drops when you select output levels above that figure? That would be a clear indication for a faulty output MMIC (read below). If it's low throughout the whole level range and shows some discontinuity at the -26 -> -25dBm threshold, it's still a strong indication for a faulty output MMIC but in conjunction with a possibly broken output RF switch (AW002R2-12). If the level is adjustable continuously over the whole range but just too low, I'ld look at the output circuitry behind the switch. Only if there are several, repeating discuntinuities, I'ld search the fault at the attenuators. The weak spot of these generators appears to be the tolerance of the output circuitry against reversely applied RF power (i.e. testing a transciever's receiving performance and accidentially hitting p.t.t...  ).

Reaching the critical components in the RF section isn't a job for the faint of heart, there are about 100 screws of well mixed sizes to be removed in order to get access to the relevant area, several layers of conductive foam to be peeled off (possibly without damaging them) and finally the defective components have to be identified and replaced.

Since the board cannot easily be powered up when exposed and separated from the digital / ch2 AF board, real fault testing is difficult. I've been able to gradually apply the PA bias voltage to find that the constant current source for the output stage MMIC (AGB3303) maxed out already at 2.5V, so obiously this chip was faulty. I also noticed that some previous repair work had been executed on this board, probably after some catastrophic damage had been done to the final microwave switch with some pyrotechnical effects going off... At that time, the AGB3303 must have been replaced as well (as the flux residues witnessed). Since I didn't find any more defects, I went ahead and replaced the defective MMIC. But I didn't have a spare AGB3303 so I compared datasheets and took the risk to install an SBB5089Z which works surprisingly well in that application despite it's designed for constant voltage operation and not with a resistive dropper (like the original) or even a CC source as supply.

Now the generator works in all ranges almost corretly and reaches 13dBm or more but obiously, it's a db or slightly more off level. So here comes the million dollar question: Does anyone have or is aware of a calibration instruction for the SM300? I'm pretty sure with my SA and an HP power meter, I'ld be able to get the calibration done well enough to meet its spec.

I searched the web and didn't find anything...any help would be appreciated!

Hello:

What you say is correct, the video I did it, and in the I explain that when it is done, they usually jump some alarms by excess temperature, so the device entered protection mode and stopped giving RF output.

I have already repaired some units of this generator, and I have fallen enough to find the spare materials, I have come to the point of knowing that these RF cards were manufactured in India and there are currently 3 in the warehouse and its value is approximately $ 1500

The MMIC that you have put is correct, since any AGB3303 unit that you can buy in China will be falsified, I bought more than 40 AGB3303 units being all false and remarked, and last few months of operation.

This equipment has no software calibration for the output power, if you set the output signal between -25 and + 13db is not leveled, but between -26 and -127db are controlled by a step prior to the HMC307 in which There are 2 MMIC (HMC313 and a Gali5 +) and an operating variable attenuator controlled by an operation that takes a sample of the Gali5 + output by means of a Schottky diode detector and necessary decoupling components, then there are 4 attenuation stages formed by HMC307, Where one of them is from 0 to 30 dB in 2 dB steps, for the smallest attenuation steps, a DAC is used, which controls the HMC346-C8 operational.

To adjust the output levels between -25 and + 13db, the PI -4DB attenuator is to be altered, which is just before entry to the AGB3303. Curiously the SBB589Z when mounted gives quite similar values ​​to AGB3303.

I use this generator daily and really the high power of the signals, although it does not have a certain precision does not matter much as long as they are not several dB, mine is about 0.2db below 0db ya + 13db gives 0.3db more . If I am concerned about the signal from the -26db until -133db since it is here where your calibration is much more complicated and is usually used for recipient repair and adjustment and as we know that from a certain signal it is complicated to measure the power of by Example -100dB unless we have a very complete laboratory or a professional test receiver. Therefore, if I give importance to low-value signals.

This vector generator is designed so that users can not repair it. If it is a mini computer that manages the BF and AF board. It is built with cost connectors and some complicated components of finding equivalences.

Greetings