CMU200 failed loop tests.

[eurgenca]

Hello.
After lots of repair my cmu now stuck on I suppose one last but hardest error.
Cmu fails internal loop tests on some frequencies for some dB and all band on rf loop path. Same on 1>4 and 3>2 path.

SG looks like work, according to my power meter I have loss of cca -1dB. But SA is cca -5dB off on whole band. On pic you can see -35dB signal connected to RF4, 2 and 1. It is best on RF1 input. I have read that problem could be in frontend, rx/tx or dig module. What is your opinion or it could be calibration?

[eurgenca]

Am i think correct?
External loop path (both 1-4 and 3-2) use (half RX) red path. Internal loop path (1-1 and 2-2) could use blue and green path. This will be logical, BUT why then RF1 like input is o.k.? It use green path too on my opinion.

[RF_fanatic]

This is a problem in your RX/TX module

You can see this based on the gap of the Band 4 respons

If you would sweep the spectrum you would see a big gap when the band changes to band 4. So your picture with a signal at 1000 Mhz does not give a real indication of your rf path.

Set your CMU to the following:

RF out > RF1
RF IN > RF1
Spectrum : Min 10 Mhz, max 2700 Mhz
Spectrum setting: max hold
Generator on: max power
Step manually in steps of 0.1 Mhz by spinning the rotary dial and you will see a response which will not be straigt at all but with huges dips

[Michael-VK2BEA]

The RX has two first IFs (842 MHz and 1817 MHz). 1817 MHz is used for RF signals up to 1200 MHz and above 2200 MHz (from memory). 842 MHz is used between 1200 MHz and 2200 MHz. (actually above 2200 there is an additional down converter)

Check the three ranges to see if you have a loss on all or just some of these ranges.

I have a unit here that had bad IF filters. In the RX part there are three Murata monoblock ceramic filters. In my unit the copper surface of two of these were corroded and the filters had bad loss. I had to replace these filters. (It appears that there is some reaction between the RF absorber and the copper surface (the top surface that contacted the absorber is the only surface that showed corrosion .. and the 2nd identical 1817 filter was clean)

[eurgenca]

Hello, Michael.
Today I have time to test cmu and I prove that cmu have level dip between 1.2GHz and 2.2GHz on RX. So I am open RX/TX module and like on your picture filters are very corroded.
Maybe is the same error like on your module. Can you please tell me markings on big filter, because I can't read them. Where are you bought this filters I will replace them anyway. Thanks for answer.

[Michael-VK2BEA]

The 842MHz IF filter (used between 1200MHz and 2200MHz) is a 3 pole ceramic Murata DFC3R836P025HHD. The marking on the filter is "836 CD". The filter is used also in on the TX (look on the other side of the board and you will see this. (It is not in a cavity with RF absorber and is thus OK). Note that this is an 836MHz filter but is 25MHz wide and so covers the actual IF frequency of 842MHz.
Its shown in this datasheet: https://pdf.datasheet.live/datasheets-1/murata_manufacturing/DFC4R947P025BHA.pdf
I have not found any of these available (after dozens of queries). I did find a version with a slightly different footprint (BHD instead of HHD) that should fit by installing the filter down one pin.
This alternate footprint version is available here for a reasonable price https://www.ectransistors.com/10-PCS-INDUCTOR-TRANSFORMER-MMCL-DFC3R836P025BHD-3R836P025-DFC3R836P025BH.html
(this website looks suspicious... I don't know if its legit)

The other 1842 MHz IF filter (used up to 1200MHz and between 2200MHz and 2700MHz) - Murata DFC31R84P075HHA is available here: https://www.rf-microwave.com/en/murata/dfc31r84p075hha/1842-5-mhz-ceramic-band-pass-filter-gigafil-3-poles/fg-1842-a/

[_Wim_]

I have posted about a similar issue here: https://www.eevblog.com/forum/repair/cmu200-rx-level-issue-failed-loop-test/new/#new

As corrosion on the filters is way less severe on my unit, and the level jumps are very flat across their frequency band,  I think that at least in my case the issue is not related to these filters. Did any of you manage to replace them and did it fix your unit?

[Michael-VK2BEA]

Hi Wim,
   Yes, I saw your post and the situation look very similar. The filters that are corroded are at the IF and so the loss is consistant accross the band that uses the IF filter. (see previous messages regarding what RF frequncies use what IFs). Since finding this issue I've heard from three other people who have had this issue filt the corroded filters.
Yes, I replaced the filters and it fixed the loss I was seeing.

Michael

[_Wim_]

Since finding this issue I've heard from three other people who have had this issue filt the corroded filters.
Yes, I replaced the filters and it fixed the loss I was seeing.

You can probably update that number of people to 4   

[_Wim_]

You can probably update that number of people to 4   

Yep, 4 it is 

[eurgenca]

Hello.
Thanks to _Wim_ and Michael-VK2BEA I was able to replace both coroded filters. I can comfirm that RX/TX module now work like a new.
Thanks to all for help.
Sebastijan.

[Ringmodulator]

Hi,

my unit seems to have the same poblems.

Can someone point me to a source for these filters, which delivers to Germany?

Chris