RM520N-GL R03 firmware A0.301/A0.303: SA NR CA improvements?

unpicked0110 · September 9, 2026, 2:44am

Hi all,

I’m running an RM520N-GL (R03 branch) on T-Mobile US, currently on firmware RM520NGLAAR03A04M4G_01.204.01.204. Using it as a portable 5G gateway in SA mode (AT+QNWPREFCFG=“nr5g_disable_mode”,0).

Performance has been solid: getting 555–620 Mbps DL / 60–65 Mbps UL on SA n41 with dynamic NR CA (n41+n71, n41+n25 under load). However, I’ve never been able to achieve dual n41 CA (e.g. n41 ARFCN 520110 + 502110), even though QSCAN shows both cells available with good signal.

I noticed two newer R03 builds:

- **RM520NGLAAR03A04M4G_A0.301.A0.301** (2025-06-12)

- **RM520NGLAAR03A03M4G_A0.303.A0.303** (2026-01-29)

I emailed support@quectel.com asking for release notes but was redirected here, so I’m hoping someone from Quectel or the community can help with:

1. Release notes for A0.301 and/or A0.303: any changelog or summary of changes?

2. NR CA improvements: specifically, has SA TDD+TDD CA (n41+n41) been improved or enabled in newer firmware?

3. SINR reporting: older R03 builds had a known issue with intermittent invalid SINR readings (−32768). Is this fixed?

4. T-Mobile MBN updates: any changes to the T-Mobile commercial MBN profile?

5. Official download / checksum: is there a way to verify firmware package integrity?

I also saw that another T-Mobile user reported [SA n41+n25 CA dropouts with packet corruption]( RM520N-GL NR5G SA Carrier Aggregation Dropouts ) on R03. Would be curious if newer firmware addresses that.

Current setup details:

- Mode: SA (nr5g_disable_mode=0, mode_pref AUTO)

- Serving tower: gNB 1103064, PCI 402, n41 ARFCN 520110

- Signal: RSRP −82 to −84, SINR 23–27

- Antenna: stock (2×2 MIMO, router has 2 SMA ports)

Thanks in advance for any info!

silvia · September 14, 2026, 7:32am

Dear @unpicked0110
It supports n41B.

I have sent newer version to your email, please check.

TommyD · September 16, 2026, 12:43pm

Hello, I have an RM520N-GL running RM520NGLAAR03A03M4G_A0.305.A0.305 on T-Mobile US.

I am also using 5G SA n41, but my download speed is around 218 Mbps despite excellent signal conditions (approximately -62 dBm RSRP and 27–28 dB SINR). AT+QNWCFG="nr5g_mimo_info" shows only 2 MIMO layers.

Could you please provide the latest R03A03 firmware available for my module, particularly the firmware discussed in this thread with the SA n41/NR CA and performance improvements?

Thank you.

unpicked0110 · September 17, 2026, 6:07am

Hi Silvia,

Thanks for confirming n41B support and for sending A0.304 — I received it and reviewed the package.

I checked the A0.304 changelog against my scenario (T-Mobile US, SA, Commercial-TMO MBN): the delta vs 01.204 is Vodafone L2NR PS HO (EU), the QGDNRCNT CLAT return-value fix, SUCI calculation, and TS.25 — no n41 CA combo changes, no TMO MBN updates, and the -32768 SINR issue in AT+QENG is still unaddressed. So I’ve stayed on 01.204.01.204 for now.

The core question remains: since the UE supports n41B, why does n41+n41 never aggregate in practice? My two n41 carriers (520110 + 502110, same gNB 1103064 / PCI 402, RSRP -81/-83, SINR 27-30) are both strong, yet across dozens of AT+QCAINFO samples (idle and under load) the SCC is always n25 — an n41 SCC has never appeared.

Is SA n41+n41 purely gNB-scheduler-controlled, or is there a UE-side CA combo table / MBN setting / AT flag gating it?
Is there an AT command on R03 to query the UE’s supported NR CA combos, so I can confirm n41B is actually in the table?
If gNB-controlled: any UE-side trigger (measurement config, AT+QNWCFG, etc.) that makes the second n41 carrier eligible as SCell?

Thanks,
Elton

silvia · September 17, 2026, 7:10am

Dear @TommyD
RM520NGLAAR03A03M4G_A0.305.A0.305 is the latest one, no newer version.
You can follow the combination of antannas.

silvia · September 17, 2026, 7:33am

Dear @unpicked0110
Do you confirm your operator has n41B combination?

No AT command to query UE’s supported CA combos.
Do you try to local N41 and see the result?

unpicked0110 · September 21, 2026, 3:11am

Hi all,

RM520N-GL on T-Mobile US (SA), firmware RM520NGLAAR03A04M4G_01.204.01.204, MBN Commercial-TMO (0x0A01050F, 202402061). Using it as a 5G gateway in SA mode. I stayed on 01.204 because the A0.304 changelog (2026-07-17) shows nothing about n41 CA or TMO MBN changes.

I’ve been trying to figure out why n41+n41 intra-band CA never happens on this modem and ran a few experiments. Sharing the setup and results in case anyone has seen something similar, any pointers appreciated.

Experiment 1, baseline: AT+QSCAN=3,1 shows two strong n41 carriers from the same gNB/PCI:

+QSCAN: "NR5G",310,260,520110,402,-83,-11,27,1,10D4D812D,2C6300,273,41,32,6,-
+QSCAN: "NR5G",310,260,502110,402,-81,-11,30,1,10D4D8137,2C6300,245,41,32,0,-

Under DL load, AT+QCAINFO consistently shows PCC n41 520110 + SCC n25 396970. Across dozens of samples (idle and loaded, over several weeks), the SCC is always n25 (occasionally n71), an n41 SCC has never appeared. So no general CA failure, just this one combo.

Experiment 2, same-site cross-check with a Pixel 10: At the same location on the same cell (PCI 402), a Pixel 10 under load showed 4CC at 100+20+20+90 MHz (from an NSG log, excerpt attached), and later NetMonster showed it camped on 502110 (90 MHz) with aggregated BW 90+20+20+100 MHz (screenshot attached). So the network here does hand out n41+n41, the modem just never gets it.

Experiment 3, MBN swap: I switched the active MBN from Commercial-TMO to the generic ROW_Commercial (0x0A010809), reset the modem, re-ran the load test with verified traffic (about 84 MB over 3.5 min), and sampled AT+QCAINFO 8 times: still PCC n41 + SCC n25 every time, zero n41 SCC. Then switched back to Commercial-TMO and confirmed everything is back to normal. Behavior looks identical under the generic MBN, so the TMO carrier policy doesn’t seem to be the deciding factor.

Given all of the above, I’m wondering whether the R03 A04M4G baseband CA combo table might simply not include n41 intra-band 2CA, i.e. the UE never advertises it so the gNB never configures it. A couple of questions:

Is there an AT command on R03 to query the UE’s supported NR CA combos, so I could check whether n41+n41 is in the table at all?
If it’s not in the table, is there a firmware or MBN release on the R03/A04 branch that adds n41 intra-band CA support?
Minor and separate: I also see intermittent invalid SINR -32768 in AT+QENG on 01.204, is that a known issue?

Full AT logs (QGMR / QMBNCFG / QENG / QSCAN / QCAINFO idle+load) available on request.

Attachments:

NetMonster screenshot (Pixel 10, camped on n41 502110 at 90 MHz, aggregated BW 90+20+20+100 MHz)
NSG log excerpt (serving-cell band labels for both n41 ARFCNs + ServiceState bandwidths before/after 4CC; location records stripped)

Thanks,
Elton

quectel-forum-nsg-excerpt.txt (1.5 KB)

silvia · September 24, 2026, 7:45am

Dear @unpicked0110
As mentioned, it only supports n41B, the total bandwidth doesn’t exceed 100MHz.
So did you confirm operator supports this combo? lock n41 and try?

unpicked0110 · September 24, 2026, 11:50pm

Hi Silvia,

To close the loop: my two n41 carriers are 100 MHz (520110) + 90 MHz (502110) = 190 MHz total, which exceeds the n41B 100 MHz cap, so n41 intra-band CA is impossible on this build. I ran the n41-only lock test you suggested (6x QCAINFO under load): PCC-only every time, no SCC, the second n41 carrier never added. My Pixel 10 on the same cell aggregates 100+90+20+20 — the 100+90 within n41 alone already exceeds class B — so the network side demonstrably supports wider-than-B n41 aggregation.

One direct question: is n41B the hardware limit of the RM520N-GL (SDX62), i.e. does it not support n41C or any wider n41 intra-band class? If so, no firmware or MBN can change that. Please confirm.

Thanks,
Elton