Thanks so much Andrew -
73 de Marty, KN0CK
I think you have the TIA set too high ā¦
And i wonder why your noise floor is so low when PGA is maxed ā¦ with mine (no case) noise floor is around 70db ā¦ not 100db
I also tested both RX1 input (L/W), reworked soldering and verified connections on RF transformer from RX1L, all is OK
Less sensitivity to reception on that input. I use it different softwareās, a very good antenna and HF filter, compared to Airspy, HackRF and RX1W input reception, the RX1L is very insensitive.
Many Thankās for your help Marty !
1 Like
Hi ricsi,
Iāve also tested my v1.4 unit (comparing with RTL-SDR) and the results are not good, so, Iāve repeated the quick test (WCDMA source with provided file and loopback) and after some trials the result is (channels A and B):
Apparently, both channels are not working with the same levels (apart from other observed differences), so, Iāve several doubts about how to discard or to confirm if something is wrong with this board (or it depends on other reasons Iām not considering, like proper grounding, powering, shielding, etc).
Also, I donāt like so much it is not possible to remove the unit in a safe way (before unplug it) when Myriad USB driver is used (i.e. in comparison with Zadig drivers).
Best Regards
BelmY
Yes, I have similar behaviour. Channel 1 is 15dB below Channel 1 on the WCDMA loopback test. I though they should be identical.
Thanks gerryk,
So, I think somebody from LMS should confirm if this behaviour (different levels between A and B channels, different values in WCDMA spectral density and constellation, wrt v1.2, possible phase noise degradation, receiver with lower sensitivity and other differences) is a symptom of a possible issue or problem with our v1.4 units or, at least, to explain in a deeper way how to discard it (assuming if so, v1.4 has worse performances than v1.2).
I was also reading some posts about some proposed Calibrations (TX and Rx), affecting once are completed (i.e. DB populated) also to working OpenBTS applications. Anyway, they are still not part of the Quick test (or a more extended one, with an step by step guide), and they should be the first steps to be completed by arriving enthusiasts (amateur, academic or professional) supporting the project, to verify in a rigorous way a full functionality of the unit, before to start a disappointing investigation and learning on the matters really we are interested in.
On the other hand, I fully support the preparation of an specific and more detailed User Manual, with descriptions, recommendations (to avoid thermal issues, ESD, with assembly guidelines, installation procedures, warnings, troubleshooting, calibrations, etc) apart from the Wiki.
I hope these issues can be clarified and solved in short time. May be I can contribute to it.
Sorry, but Iād like my expectations on this nice piece of HW were fully covered and, at this point (with a lot of work to do and stuff to learn with this board also from my side) Iām a little bit frustrated.
Best Regards
BelmY
2 Likes
I albo also noticed signal level degradation in LNAW_B channel. Around 10db smaller signal with respect to LNAW_A.
Hi @Zack
Would you be in a position to comment on this?
Are RX_1 and RX_2 supposed to have the same levels when looped back in the WCDMA self-test?
/ Gerry
Quick update: weāre working on a diagnostic procedure. More details soon.
5 Likes
An initial diagnostic back from the engineering team has been added to the wiki:
1 Like
Sorry ā¦ I had a very busy time, and could not test.
The good news is that I have borrowed a second LimeSDR and a HackRF.
When I have a little more time to breath I will report back here more.
I have a an SRH789 antenna connected to LNAH.
Noise floor is very low below -100, no case, just the bare board.
I do not want to install it in the casing while I do not know that it works as expected.
Same here.
Self Test looks like this here ā¦ the two channels are different.
Also when I go to read Temp I get -215.8 degrees celsius.
It is definitely MUUUUUCH hotter, I can barely touch the chips or the ground plane on the side.
Hi @andrewback
Thanx for the procedure.
But I fear it has to go to much more detail, as currently I do not understand what is requested.
1 Are RF matching networks identical?
2 Are all A and B channel settings identical?
3 Has any RF matching changes been carried out?
4 Was proper ESD protection used?
1 I have not changed anything ā¦ how can I check?
2 I have not changed anything ā¦ how can I check?
3 no (if you mean modifications of the board like desoldering coils etc.) I wanted to wait until I understand the RF performance of my LimeSDR.
4 yes
āClearly state the exact difference in gain in dB and the frequency it was measured atā
Can you get this from the screenshots posted here?
If not tell me which screenshots you need.
" Provide details, including versions of:
FPGA gateware and FX3 firmware"
Library version: v17.02.2-myriadrf1~xenial
Build timestamp: 2017-04-30
Interface version: v2017.2.0
Binary interface: 17.02-1
" Lime Suite
Application software"
Installed Ubuntu packages as recommended by LimeMicro.
" Wherever possible supply a Lime Suite .ini file"
No changes made after installing it from your ubuntu packages.
Please tell me what additional information you need.
Also please keep in mind that I am not an electrical engineer.
Also I still do not understand the different receive paths.
If I connect a SRH789 to LNAH1 I can listen to strong FM radio stations.
On LNAL1 and LNAW1 I barely receive anything.
I do not understand this, as the WIDE (100kHz-3.8GHz) and LOW (100kHz-1.5GHz) should be better suited as HIGH (1.5GHz-3.8GHz), but nevertheless LNAH1 works much better for FM reception than the other two ā¦ WHY??
Any ideas @andrewback
I have the opposite (but expected) behaviourā¦ L is much stronger then H and W.
Iād say the input network on L is possibly faulty on your device.
I hope that somebody from LimeMicro will comment on it.
The first thing I did after unpacking was selftest and my result was as follows:
But I thought it was normal.
It looks as though you may have a defective board. Just to confirm again that you have not made any modification? Assuming not, will see if @zack has any other suggestions.