doing in house EMC pre-compliance testing - where do I start?

[Simon]

Well rather unimpressed with this spectrum analyser, the manual is a mess, the machines UI is a mess, once you get into modes like EMI menu's change and duplicate others, it's not clear what menu's apply in what mode and I can't get it to read my USB drive to do the firmware update. Why did I take the advice of those assholes at telonic?

No explanation on the USB stick requirements, it's like a list of the buttons names and that is it!

[tautech]

Well rather unimpressed with this spectrum analyser, the manual is a mess, the machines UI is a mess, once you get into modes like EMI menu's change and duplicate others, it's not clear what menu's apply in what mode and I can't get it to read my USB drive to do the firmware update. Why did I take the advice of those assholes at telonic?

No explanation on the USB stick requirements, it's like a list of the buttons names and that is it!

8GB is preferred and FAT32 is a must !

The current manual covers 3 ranges all with the same basic SA capabilities and same EMI mode.
Have you not followed the links I posted ?

But hey if I need hold your hand I will as my SVA1032X has the exact same EMI mode as your SSA3015X Plus.
First practice taking screenshots with it to USB to post here but be sure to have appropriate menus captured in them.
You preference but the Normal image rather than Inverted ink saver shows up best here on the forum. Enter the System menu to find where to change it.

[Simon]

Yes I have 2 versions of the manual, the larger has stuff that the machine appears not to.

The PC software is missing DLL's, plugging it in on USB does what? computer just blips, only driver I can find fails to install, the Ethernet requires a driver despite it being a server...... it just goes on, a shed full of promises and a broken window pane of actual delivered product.

First I would like to update the firmware, that seems to have special requirements. 8GB stick? I will see if I can find one.

[tautech]

Yes I have 2 versions of the manual, the larger has stuff that the machine appears not to.

The PC software is missing DLL's, plugging it in on USB does what? computer just blips, only driver I can find fails to install, the Ethernet requires a driver despite it being a server...... it just goes on, a shed full of promises and a broken window pane of actual delivered product.

First I would like to update the firmware, that seems to have special requirements. 8GB stick? I will see if I can find one.

IIRC I have used 16GB sticks in these but I don't normally.

Don't use the PC SW, just don't unless you really really need to and instead just use the webserver.
Connect to your LAN and enter the I/O interface and give the SA a unique IP but on your LAN subnet then Save it.
Enter this unique IP into the address bar in your PC browser and hey presto we have a remote connection that you can do everything you could sitting in front of the SA.

[Simon]

well luckily it takes a 16GB stick, firmware now updated. At some point you may answer the question actually asked.

[Simon]

Yes I have 2 versions of the manual, the larger has stuff that the machine appears not to.

The PC software is missing DLL's, plugging it in on USB does what? computer just blips, only driver I can find fails to install, the Ethernet requires a driver despite it being a server...... it just goes on, a shed full of promises and a broken window pane of actual delivered product.

First I would like to update the firmware, that seems to have special requirements. 8GB stick? I will see if I can find one.

IIRC I have used 16GB sticks in these but I don't normally.

Don't use the PC SW, just don't unless you really really need to and instead just use the webserver.
Connect to your LAN and enter the I/O interface and give the SA a unique IP but on your LAN subnet then Save it.
Enter this unique IP into the address bar in your PC browser and hey presto we have a remote connection that you can do everything you could sitting in front of the SA.

You mean a bit like my oscilloscope that silent no longer admits to the existence of, yea..... We do PC software but it's vapourware. Struggling with ethernet but I probably will on a corporate network run by security nuts.

[tautech]

We do PC software but it's vapourware.

No.
TBH I haven't used EasySpectrum since the early Siglent SA's, those before the Plus series however since the addition of the webserver in Plus, X-R and SVA models it's so much simpler to use as it mimics the instruments UI on a PC plus within the Instrument control panel you have a couple of additional features too.

With any of Siglents Easy**** SW care need be taken to how everything is initially set up and particularly the requirement for the bloated NIVISA SW of which a full install is north of 500 MB !
We can get away with just the Runtime component for the required connectivity drivers but it's still north of 100 MB FFS !

Have my SVA1032X alongside me to grab a screenshot to help with getting your connected to your LAN should it be required.

[tautech]

LAN reconfigured to home network after a recent show.

System>Interface>LAN Config> IP Config>DHCP = ON allows your network to negotiate an IP with the SA after which change DHCP to Static and selecting the last digits of the IP with the touch screen use Back to delete them and enter a higher # so to prevent future IP clashes with later added devices. Then Apply.
Enter the saved IP into a PC browser and you are greeted with the Welcome page.

A couple of captures for guidance attached.

[tautech]

BTW you can change the screenshot to Normal from the default Inverse from within the Display menu.
Far right 4th button down is Save that's used for screenshots which you can then add a filename or just punch Enter to push the screenshot to USB.

While you're there have a fiddle with some of the other bits and bobs in the Display menu.....Annotation and Display Line are both features that can be useful.

[tautech]

With some poking around to get a feeling for how everything works venturing into the EMI mode shouldn't be so daunting or should it ?
As you've discovered even when within a Mode UI you still need pop out of it occasionally to make/add adjustments and then re-enter the EMI mode to set up to capture the EMI measurements proper.

Here things get more complex where you need know which tests to use and what detection strategies to use including those in the Amplitude menu where your inputs are to be controlled and NF probes will likely need the Preamp ON.

These 2 screenshots are just an introduction as Amplitude is out of the EMI mode then we need use the Mode button to return to EMI mode.
Although a BS  display with open input and full sweep soon after a Factory default Preset.....another little tool you need to learn about so to be able to start each project with a perfectly clean state and zero remaining settings.

Anyways you need select the appropriate CISPR filter within the Scan Config menu and a Span to match and while in the EMI mode you can back out of any menu with the blue circular back arrow.
Use the Stop/Start button to perform the EMI sweep then examine the Signal list results.

Consulting the manual and videos linked earlier should help further once you are this far.

[Simon]

yes when you are in EMI mode the trace menu is bolloks but it is accessible, the detector menu replaces it, one big mess in the UI as you don't know what is active and what is not. I should have learnt my lesson with that scope and realized that the droids at telonic who were supposed to be technical were just looking at £250 more for the siglent instead of the rigol I was looking at. Siglent and telonic, never again either!

[fourfathom]

Have you tried just using the SA as a SA, without all the EMI software and other crap?  You don't even need the fancy CISPR filter if you are just trying to do preliminary comparison and evaluation measurements.  You might already be done by now. When you spend more time on the tools than the job, you're doing it wrong.

[Simon]

I've not been using the EMI mode, I fail to see how it works or what it is good for. but I guess if I use the EMI filter (in normal spectrum mode) then the 25dB I want to loose is more real that guessing with the gaussian filter.

i fail to see how to activate the line of the standard but that is pointless as I am comparing only anyway. Turns out it's the crappy meanwell power supply, reading between the lines on the datasheet and looking at the construction they may have just screeched under the line - or the line. When they tell you that the test was done on a 360x360x1mm metal sheet you start to think WTF meanwell, don't you know that ferrous and non ferrous are two different things but you did not specify, they fail to say if it was grounded. When they say that the unit is meant to be part of a piece of equipment it's a bit of an admission that it actually fails radiated unless you enclose it with the load on a short lead.

[tautech]

yes when you are in EMI mode the trace menu is bolloks but it is accessible, the detector menu replaces it, one big mess in the UI as you don't know what is active and what is not.

Bollocks !

Your unfamiliarity with this instrument and unwillingness to learn how it works is screaming at us you think this is too hard to negotiate yet most of what you need is right in front of you on the display calling 'I'm here' !

Take a step back and understand what every symbol and number around the display actually means when after some time using these instruments for general less complex SA work one becomes very familiar to identifying settings at a glance.

Rather than dive balls deep into a real project just use a coax RF loop and start sniffing all manner of appliances, phones, tablets, laptops, SMPS walwarts to get a feel for what EMI looks like and the range of frequencies you'll likely encounter.
Such exercises can be very revealing and will properly drag you into a decent learning state of mind and generate more familiarity with the tools you are working with for you to better follow the videos and App notes already linked previously.

[Simon]

I came here asking about the setup, you are the one that demanded to know the brand, as I said at the start, the brand is irrelevant to what I wanted to know, but as you wanted to know what I was using I've told you and I have told you what I think of it!

My very first experiment was to use an RF signal generator to do as you suggest, I learned from that that what working in dB means is that if my transmission goes up or down by 10dB so does the output and what I measure, this means that if I profile the failed unit I can see in a relative measurement where I need to be using the report because everything I am doing is relative.

I'm not overly sure that trying to mimic a test lab is the best use of the instrument which is what the EMI mode is for.

[fourfathom]

I'm not overly sure that trying to mimic a test lab is the best use of the instrument which is what the EMI mode is for.

Yeah, unless you have a calibrated antenna / test-site then the EMI pass/fail templates aren't much use.  EMI response filters might be good to use, but as long as you can identify the original failing signal then any relative measurement should tell you what you need to know (unless your fixes actually change the nature of the signal, such as a quasi-random spreading might do.)  I will admit I've never used a modern analyzer for EMI work, so there may be other advantages to those modes.

[Simon]

Well having the EMI filter (quasi peak) is useful as it means that the measurements are processed in the same way as the lab, so 20dB off for my measurement should be 20dB off in the lab. Unfortunately I can't modify the failing product as it is a meanwell power supply, the specs of the EMC on the datasheet are rather dubious. I have tried filters but we just force the same energy out somewhere else. As the PSU is in an aluminium box and I think it is radiated off the supply the only way to stop it is use a larger compartmented box with firewalls so that once filtered the signal stays filtered. The alternative non meanwell PSU we have ordered has a lot of baffling on the PSU so it looks like they actually tried to pass EMC rather than fudge around it with dubious statements.

[fourfathom]

Unfortunately I can't modify the failing product as it is a meanwell power supply, the specs of the EMC on the datasheet are rather dubious. I have tried filters but we just force the same energy out somewhere else.

I can't see the system, so this might not apply, but can you isolate all supply inputs and outputs using common-mode ferrite chokes?  These should be as close to the supply as possible, to eliminate any "antenna effect" of the cabling.  And I mean *effective* chokes, using multiple turns on the core, appropriate ferrite material, and perhaps several cores in series.  This obviously isn't a production solution, but it will tell you if an external solution is even possible.  Don't just slap a random ferrite clamp on the cables and think you're done -- that might only make a dB or two difference.  You need to effectively disconnect the cables at RF frequencies.

If this doesn't work and you can't modify the supply then you are left with external shielding and keeping the inside-the-shield noise from being coupled to any cables that leave your shielded enclosure.  This is usually a much harder and more costly task.

[Simon]

Well it's a machine that the power supply plugs into. The power supply is laptop power brick style. We put the meanwell PSU into an aluminium die cast box as it is open framed. The supply is a UHP-1000-48. With a nearfield probe plenty of just is coming out of the end of the supply as there is no barrier, it's like the final 2 transformers/inductors are just radiating. With that in an aluminium box we have no chance, the PSU needs mechanical measures that separate it from any filter we use so that as fast as we filter it does not pick up again on the cable after the filter.

I am confused as there are sets of capacitors mounted to daughter PCB's when they could just be on the main PCB. There is a folded sheet metal can over one inductor / transformer with a sil pad stopping it touching the EMI grill that covers the top of the supply, it's like it failed EMC and make shifting the inductor to be a shielded one worked, they just did not need to shield the other two in their cachanded testing. I suppose that could be something I look at doing but not much space in there.

[fourfathom]

The nearfield probe will help locate sources, but it may not tell you much about how they radiate.  You can have a strong current-loop that couples into the probe, but if the loop dimensions are small it won't actually put out a significant signal at a distance.  Isolate your cables / wires / connections, if only to understand the problem.

[Simon]

basically if you orient the near field probe around the right way whilst holding it in front of the output terminals you see plenty of signal, turn it 90 degrees and it's gone. so stuff is radiating out of the open end and I suspect the two transformers there are the culprits.

[fourfathom]

basically if you orient the near field probe around the right way whilst holding it in front of the output terminals you see plenty of signal, turn it 90 degrees and it's gone

You will see this polarity sensitivity with all sorts of radiators, including cables -- few sources are truly isotropic.  I can't see your setup, but do take to heart my suggestions about choking off *all* the cable currents.

[Simon]

Well I tried a pi filter but this did nothing, I tried an old RS mains common mode filter, nothing, I have a common mode choke from Wurth just arrived I will try but I think we are just barking up the wrong tree with this power supply. Tellingly we get higher peaks with no load than with a light load, there is "excess energy" in the supply that has nowhere to go.

Aluminium reflects more than it absorbs so any RF energy is getting plenty of opportunity to get onto our wires after a filter.

[fourfathom]

Well I tried a pi filter but this did nothing, I tried an old RS mains common mode filter, nothing, I have a common mode choke from Wurth just arrived I will try but I think we are just barking up the wrong tree with this power supply. Tellingly we get higher peaks with no load than with a light load, there is "excess energy" in the supply that has nowhere to go.

What common mode choke from Hurst?  I'm talking about getting some big honking ferrite toroids or clamp-cores and running the wires through those, perhaps three turns, and perhaps several cores in series.

Again, I can't see what you've got and perhaps you've done everything you can, but a PI filter has capacitors to ground (or at least to a common point), and where you connect that common point can make a big difference.  Just *try* the simple common-mode choke, an effective one, and see what happens.  And back up when doing the radiated measurements.  As I've mentioned, stuff you can pick up with a near-field probe may not be the same stuff that is actually radiating.  It may be the same frequency, may be the same actual signal as what is radiating, but until you know *how* it's radiating you don't know much.

Aluminium reflects more than it absorbs so any RF energy is getting plenty of opportunity to get onto our wires after a filter.

"After a filter" has me worried.  Aluminum is an excellent shield, and you're not going to fix this with a magic absorptive enclosure.  If you put a filter at the supply, then run wires from that to a hole in the Al box you are going to be very disappointed.

Stop looking for the production solution.  First find out what you have to do, then figure out how to do it practically.

I'll back off now and wish you good luck.  I'm always happy to give advice.

[Simon]

Well given that we are 20dB out I guessed that just a toroid core with the wires wrapped round it may not be enough, so was looking to use a full blown common mode choke. Yes all common points go back to the case but that involves separate wires. Yes aluminium is a shield but does not absorb like steel, so anything that get's out of the supply will be bouncing around in the box to be picked up. The cable is actually sort of shielded in that it's armoured and that is grounded to the case.