Rigol 815-TG not working correctly, To RMA, or not RMA?

[KD2FVM]

Greetings!
I have a Rigol 815-TG that doesn't work correctly. I noticed it wouldn't correctly display a properly tuned set of UHF duplexer cans (checked okay on the IFR-1500). I called Rigol and they started an email exchange with me. All the testing suggested the TG was sending some RF, but the levels seemed off. I was able to read RF inputted into the spec analyzer (signals from IFR, or other sig gens I tried) until I applied the firmware upgrade (sent by the tech at Rigol). Now the spec is deaf. It now also fails to display 0 mHz signal that all spec analyzers show, so the receiver isn't working right. I did a factory reset,no help, and now they want it to be RMA. They tell me if they replace the main board its gonna be about $700. I bought it used, and the void if removed sticker is intact so I'm not sure if I should crack it open first or just bite the bullet and hope it can be fixed for less than the $700.

I cant seem to find a whole bunch of repair information on this unit, so not sure about digging into it.

Does anyone have experience with Rigol repairs?
Will they care about the sticker (if I open it) since its long out of warrantee?
If they replace the board, will it be effectively a "new" 815? that is will they install the latest board?
thoughts?

[KD2FVM]

I've been recovering from the flu so today was the first opportunity to get to the SA and gather the info you requested.

So the firmware question is complicated. Tech support sent me a firmware update to try and things got worse, then I've been playing with different versions and find the unit must have some hardware issue.

When I started the firmware was as follows:
Main Board:00.03
RF FPGA: 00.05
Digital board:00.04
Firmware:00.01.08
Boot: 00.01.02

The update from Tech support changed it to:

Main Board:00.03
RF FPGA: 00.05
Digital board:00.04
Firmware:00.01.19
Boot: 00.01.02

I ran the latest update from Rigol web site:
Main Board:00.03
RF FPGA: 00.05
Digital board:00.05
Firmware:00.01.19
Boot: 00.01.02
Not sure why the dig board changed.

Now the SA receives nothing, and shows a noise floor of -50dBm. I have injected a -20dBm signal on various frequencies with no response at all from the SA.

There are no error messages. I think its time to consider cracking it open and seeing if there is something obvious wrong.
Maybe i will send it in, though no-one has responded with any experiences with rigol RMA service.

[YetAnotherTechie]

Wild guess: apply the upgrade from tech support again...

[KD2FVM]

Maintenance mode achieved, no signals detected even with the Int Ref on.

[KD2FVM]

Screen shot included

[ted572]

Hello KD2FVM:

Seeing your 'Screen shot - test.pdf (55.04 kB)' I think that there is a good chance that your DSP IF circuitry is very likely OK, and that your issue is probably between the 1st RF Amp and the last analog IF Amp.  Don't let this scare you, because if this is the case it can be easy to troubleshoot (and repair) seeing you have a IFR that you can use to find where the issue is.  By the way even a Boonton RF Millivolt Meter, etc. could be used to troubleshoot the RF Front-end through the last analog IF.

We need to see if your  TG is working:

So please turn the DSA815-TG's Tracking Generator ON with the SA Span set at 'Zero Span'.  Any frequency is Ok to use, but to keep it simple use 100 MHz. The default TG level is -20 dBm and this will be fine. Now please connect your IFR to the DSA815 TG Output connector and look for the DSA815s 100 MHz TG signal, and it should be there at ~ -20dBm. If so, then all is good so far in my opinion.  Edit: Meaning by this that yes there still is a problem, but it should be manageable to locate a fix the issue.

If the issue is between the SA Input and the DSP I will be able to walk you through the RF/IF section to find the fault, that is if you are comfortable working with and on SMT PCBs/equipment.

Please let us know what you see for the TG Output?       Ted

[nctnico]

Is there some kind of self calibration procedure? If yes, then I'd try this first.

[KD2FVM]

Okay, so the plot thickens...
I used the IFR tuned to 100MHz and set zero span on the TG. Nothing, so I pulled out the Avcom SA I have and searched for the TG signal, it was around 126MHz. I tried changing to a couple of frequencies and see the TG will generate a signal thats pretty far off of the target, then slowly drift towards the actual set frequency. This was verified by the RD 1992 counter. The 10MHz output (on the back) is 10.00025 by the way, I checked it to make sure the internal ref wasn't spastic.

Examples
TG set to:      frequency starts at:       frequency ends at:
100MHz          133MHz                    127.874 (after 5 mins)
500MHz           265.9MHz                 271MHz (after 5 mins)

[KD2FVM]

It has a cal feature, but that doesn't really help either.

[nctnico]

OK. Seems like time for a repair to me. Did it ever work right?

[ted572]

Hello KD2FVM:

It is disappointing to hear that your DSA815's TG isn't working.  If you feel qualified you can trace through the 3 VCOs and see if they, or any of them are working as expected by selecting SA test frequencies and using Zero Span for basic CW 'LO' frequencies to probe around for.  If you have a basic understanding of PLLs then you should be able to get closer to where the issue is, that is if this is something you want to pursue further.  Otherwise you may want to look into an alternative Spectrum Analyzer, or going with Rigol for a RMA and having them install a replacement Main PCB assembly (an easy alternative, although expensive).

If you are interested in additional analysis of your failure you can use the following as a resource: https://www.eevblog.com/forum/testgear/spectrum-analyzer-rigol-dsa815/msg1059060/#msg1059060  My document here will help you find the VCOs as I have identified the corresponding LED indicators for each of the 3 VCOs. They are located next to their corresponding VCO. I have their operating frequency ranges identified, and I provided all of the IF frequencies, and the various LO frequencies. This should provide all the key info to understand this area of the DSA815. Others have posted photos that I have used for various PCB sections that can assist with documenting signal levels and flow. I have not made up a block diagram, as this would additionally simplify appreciation of the circuitry, but this shouldn't be necessary for technically competent people with the necessary skill and understanding of the circuitry. If you have serious technical questions and feel that you have the ability to dig into this I will be happy to assist you further.

       Regards, Ted

[ted572]

Hello KD2FVM:

The thing you could do first if you decide to look into this issue further, is to verify that the LOs are present and at the correct frequency/approximate level (~ +6 dBm, +/- 6 dB is good enough for now). You can probe each of the 4 mixers (3 each for the SA section, and 1 for the TG section). I identified them all by Ref. Designation and Vendor P/N, so they are easy to find on the PCB. Of course you are going to have to remove the Sectioned Aluminum Shield from the Main PCB RF/IF section. Then return most of the hold-down screws using spacer stand-offs ~ equivalent to the height of the Aluminum Shield's height, so that the PCB makes good contact for it's ground connections to the chassis/bottom shield. Locate all of the mixers and their LO inputs so that you can easily probe them with a 500 ohm input impedance RF Probe (Note: approx. 20 dB loss in the probe).  For the probe use 50 ohm coax cable between your IFR SA input (or whatever you intern to use for making these measurements with a 50 ohm input). On the free end of the coax add ~ 450 ohm resistor in series with very shot leads (this is your probe (input) Tip), along with a very short ground between the coax shield and the Mixer's ground return for it's LO input.  Please let me know the results.  I sent you a PM with my phone number, so you can call me if you need any additional info or clarification. If you don't feel comfortable doing this fine, I understand.     Ted

[KD2FVM]

I'll see if I can get some free time this weekend to crack her open. I'd like to at least understand what went wrong before sending it in for a refurb.

[ted572]

KD2FVM:

If you want to see ahead of time what you will find inside your DSA815-TG you can view it here from a safe distance (for you and it):
If it's construction looks familiar and basic to you - go for it.  Although if it looks somewhat complicated, or if you don't have experience working with RF/microwave circuits, please consider staying away from it. To me it is very straight forward and it's easy to follow the signal flow, but then I'm involved in RF equipment and system design.

Initially there is no reason you would have to remove the Main Board, just the Aluminum Shield is enough to gain sufficient access to determine where the primary issue would be.  Cheers and good luck, Ted

[N8AUM]

What does it look like with the TG connected with short piece of coax to the input ?

[ted572]

KD2FVM:
If you want to hone your skills on removing Warranty Void Stickers, now is your opportunity.    When you 'crack her open'
See the Links I provided below, but add a little heat from a Hot Air Gun or Hair Dryer.  And you will see how easy it is to do.
     Cheers, Ted

[KD2FVM]

So, I sent this in to Rigol finally. They tell me the UI was completely screwed up, and they dont know why. They said they have seen this before on 1 or 2 units, and the remedy is a fresh main board. The problem is parts aren't coming over right now so.....they said they'd swap it with a different unit and send it back. It should be here soon so hopefully it will be okay.

It wont have the licenses (must've been cracked licenses as I bought it used I have no idea) might look to see if the new one is hackable, though I'm not sure I'll need it.