Just don't take it to North America, you'd be real disappointed:Damn ... that looks like a pretty sweet road warrior device !!
LTE FDD bands 1 (2100MHz) / 3 (1800MHz) / 5 (850MHz) / 7 (2600MHz) / 8 (900 MHz) / 20 (800MHz) / 28 (700MHz)
If you ask me, it's actually USB data that's missing on the hAP ax lite... Since it be slick if the hAP appear as ethernet device to a laptop (and draw power from laptop at same time) – that be one cable then to internet, an ethernet switch, and "basic" Wi-Fi. AFAIK that's not possible. Maybe in the North American versionSo POE definitely missing here for the use case.
I am totally with you there but ...If you ask me, it's actually USB data that's missing on the hAP ax lite... Since it be slick if the hAP appear as ethernet device to a laptop
Drawing 2.4A from your laptop ? Then you need PD capability on your laptop, yes ?(and draw power from laptop at same time) – that be one cable then to internet, an ethernet switch, and "basic" Wi-Fi.
Got it !Ordered one, should have it by end of the week, latest early next week.
Just in time before we go to France for 2 weeks so can test it there as well
I'm not jealous at all...Got it !
Lots of work to do so probably will not be able to play with it today.
NO! It looks like it's not reading SINR – that seems like a bug. Now it could be either Android app not reading/charting, OR RouterOS isn't reading it from the modem. Might want to check with winbox/cli to see if it's reading SINR from modem there... But it should be showing the SINR.Ps that line on the graph is when I switched screens to take the screenshot
@holvoetn thanks for your test, very interesting...Ax lite lte powered by 10.000mAh power bank
Well, if your in hotel room etc...it generally obvious the window be better placement than say the bathroom that presumably have lower signal... Does the knowing amount of dB loss matter there? And not like you can install new antennas etc... So irrelevancy is kinda relative.Checked on PC/Winbox, no SINR either.
[...]
So probably a bug for this type of modem (FG621-EA, same as in Chateau AX LTE). Or this reading is irrelevant ?
My opinion is that all signal values tell some story. I view it as RSRP is the beginning and the CQI is the end of story (& that's actually the order Mikrotik uses in winbox). SINR include noise and interference, so that part of the story it tells. But the CQI incorporates SINR data into a number from 0-15, with 15 being best – so that one is pretty easy to get some quick/general sense of the current speeds.I have found that looking at SINR usually gives the best indication of overall situation. 4 is quite low, but workable. Try to get it higher
Since you're already installing extra-packages, you can also add the dude package. And the dude can monitor the router it's running on, so with few added probes you'd have some nice charts for stuff like cellular and voltage*. The dude is quite handy for data recording IMO... And Dude is so old, I'd imagine the modern ARM in the hAPaxLite won't notice much & your'd be monitoring one device (itself, locally).Also installed zerotier in the mean time.
Changed the script to output LTE data every 10 minutes that text is being send (RSRP / RSRQ / RSSI / SINR (which is not there yet)).
7h33 uptime, powerbank still shows 2 leds.
Good plan. AFAIK, there is a lot of timing involved in. The CQI and SINR have to do a lot more math on the modem – so are typically reported less frequently & I think RouterOS gives up waiting for data, so remove the value in the CLI/app/etc. e.g. it's probably there, but RouterOS gives up waiting...Anyhow, I launched a sup-ticket, will see what they answer.
You can go to fcc.id/TV7L41GXD and look at Internal Photos.You can find one disasembly of Hap Ax lite on YT and there you will see, that there is a space for soldered modem.
Not bad results at all... Let's assume that you have 25% battery left, that's about 7500 mAh consumed for over half a day, okay there was no much load on it but still not so bad.13h30 and still 1 led on powerbank.
Noticed using SNMPWalk the oid for SINR is present. It's simply not populated.Good plan. AFAIK, there is a lot of timing involved in. The CQI and SINR have to do a lot more math on the modem – so are typically reported less frequently & I think RouterOS gives up waiting for data, so remove the value in the CLI/app/etc. e.g. it's probably there, but RouterOS gives up waiting...Anyhow, I launched a sup-ticket, will see what they answer.
And hAPaxLiteLTE uses a Fibocom modem, not Quectels in other models, so even more likely timing/something is different...
But couldn't find oid for interface uptime ?
That one is far from obvious... I think the auto-scan stuff will add a "network", but it does a lousy job, so disabling that is also a good idea & add the device(s) "by hand" to map ...Already figured that one out as well using a "dummy device" to link to
But in case of power bank that wouldn't be useful anyway because power bank outputs stable 5V all the time and i presume that hap ax lite accept only 5V input so maybe that's the reason there is no voltage monitor ?It's a shame there isn't voltage in /system/health – with a battery pack, that's kinda useful...
Confirmed by support it is missing for now for this type of modem.So probably a bug for this type of modem (FG621-EA, same as in Chateau AX LTE). Or this reading is irrelevant ?
I'll ask support already about it (SUP-124867).
Yup, fair enough... Sure your right...the battery pack or router would disconnect USB before you'd see any voltage drop.But in case of power bank that wouldn't be useful anyway because power bank outputs stable 5V all the time [...]It's a shame there isn't voltage in /system/health – with a battery pack, that's kinda useful...
Yeah I had enough faith in Android that it does just add extra lines in a screenshotConfirmed by support it is missing for now for this type of modem.So probably a bug for this type of modem (FG621-EA, same as in Chateau AX LTE). Or this reading is irrelevant ?
I'll ask support already about it (SUP-124867).
Will be added later.
[admin@MikroTik] > /interface lte at-chat lte1 input="AT*PROD\?"
output: +CME ERROR: Operation not supported
That only works on the original R11e-LTE AFAIK. The hAPaxLite has Fibocom modem, so even if allowed... likely be a different command.Also it looks like you cant change IMEI on this one.Code: Select all[admin@MikroTik] > /interface lte at-chat lte1 input="AT*PROD\?" output: +CME ERROR: Operation not supported
Basically it's like cloning the MAC-address on a cable modem.Why would you ?
No, it's another reason: since 3G there isn't a central device certification authority and MNOs have to perform their own interoperability tests (and some device vendors even require that before enabling their devices to work with particular networks and/or features, such as VoLTE, Appke was pretty notorious about that with LTE networks roll-out). The positive thinking says it's for the better for users because if device is allowed to work in network, operator knows that it'll be relatively trouble-free. The negative thinking is that operators use it to force users to use whatever crap they're selling.... but some carrier look if the device's modem's serial# (the IMEI) matches the SIM card# (the ICCID) to allow network access.
Sure, why not ?@holvoetn, so your wife lets you test routers at vacation ?
My just said only wires I wanna see in our suitcase are phone and laptop chargers
... Orange fiber.
I'd imagine the modem likely doesn't turn off, when the kernel exits from the shutdown.EDIT: Not bad at all expect power draw when turned off...
The modem shutdown too would actually make more sense . I just know it doesn't do that with modem on other platforms. Seems like a reasonable request (or bug depending on POV).But this is good to know... If for some reason you leave router connected to power bank it is possible that router would drain powerbank...
Just connected my wAP ac LTE with R11e-LTE (CAT4) to my usb meter, it is connected to a usb-c power bank via PD adapter set to 16V.IDLE (LTE connection up, wireless enabled but no clients, no wired devices)
Input voltage: 5.08V
Current draw: 0.478A
Power draw: 2,428W
That's good news, it means the power can be cut to modem on it – that be step one to be able to do it via a shutdown . Maybe you should file a feature requestI tested command you gave and what happen is LTE disconnects and lte1 interface disappear for about 40-45 seconds from interface menu.
100% agree. I'm hoping (absolutely not expecting) the US/NA* version have PoE IN myself.Only thing I still do not like: It has no POE capabilities.
Feature request, Mikrotik please make that shutdown command actually shuts down routerThat's good news, it means the power can be cut to modem on it – that be step one to be able to do it via a shutdown . Maybe you should file a feature requestI tested command you gave and what happen is LTE disconnects and lte1 interface disappear for about 40-45 seconds from interface menu.
Mode button nothing yet, and user led to show LTE interface activity.I'm curious, everyone, what did you set mode button and user led to?
Time to steal your ideas.
https://git.eworm.de/cgit/routeros-scri ... -button.mdI'm curious, everyone, what did you set mode button
https://git.eworm.de/cgit/routeros-scri ... to-leds.mdand user led to?
Please do!Time to steal your ideas.
LOL, yeah at least 50 pages... Problem is there isn't perfect solution to queues – because LTE is pretty variable (often related to time-of-day) as you note. The hAPaxLiteLte6 uses 2.4Ghz Wi-Fi... so that acts as a bit of "queue" already assuming the 2.4Ghz Wi-Fi is using a 20Mhz channel however... A larger 40 or 80Mhz channel is likely to cause more problem since you'll likely get Wi-Fi re-transmission, and 80Mhz (and likely 40Mhz) Wi-Fi is actually faster than LTE speeds.Queues have been useless to me cause I dont know lte bandwidth.
Maybe one day I'll read that 50 page thread on queues without falling asleep.
I just did speedtests (with a PC client to speedtest.net) at different times of the day.Curious: How to you set those limits ?
I thought it was not possible on interface ?
Can you show the relevant part of the config please ?
Always interested to learn ...
If it bothers you that much, feel free to ask support.
And do let us know the response, please
Modern PCs with ACPI support offer reading DMI table which contains lots of hardware information. Such as information about "Memory Device", a.k.a. RAM modules. So reading this information it's easy to get size of physical memory. Then there's "Memory Array Mapped [Address]" which contains amount of memory, remapped by BIOS. And then there's amount reported by kernel as total memory ... which accounts all the un-remapped RAM regions and memory used by kernel (and drivers).We will see if we can somehow make it show total and reserved, so it's clear, but even linux kernel does not show this reserved memory, so it will not be easy.
The "32MB per wifiwave2" makes sense in theory, but the empirical evidence tells otherwise. When disabling or deleting the wifiwave2 package from the AX Lite/LTE6, total-memory still shows at 224.0MiB, no changes from when wifiwave2 is installed and active except for the wireless interface disappearing from the system.All I know is must be a decent enough device if the complaints are in the calculation of memory.
But on a hAPaxLite, the 32MB per wifiwave2 would make sense. One 2.4Ghz interface, so 224+32=256, which the memory spec for hAPaxLite.
Actually that explains it may not just storage space why wifiwave2 doesn't fit on cAPac/hAPac[2|3]/wAPac — if wifiwave2 driver is mapping 32MB x 2 of physical RAM, that half of memory on 128MB RAM devices...
I was hoping that the 04 version would fix the "sim is locked" after modem or device reboot when SIM PIN is enabled. Unfortunately this isn't the case. Too bad the firmware update changelogs for the modem aren't made public.There is seldom an announcement about new version.
I already asked about changelog in the past to support, they don't give it.
From 16121.1034.00.01.01.01 to 16121.1034.00.01.01.03:
https://upgrade.mikrotik.com/firmware/FG621-EA/16121.1034.00.01.01.01/latest
https://upgrade.mikrotik.com/firmware/FG621-EA/16121.1034.00.01.01.01/image
From 16121.1034.00.01.01.03 to 16121.1034.00.01.01.04:
https://upgrade.mikrotik.com/firmware/FG621-EA/16121.1034.00.01.01.03/latest
https://upgrade.mikrotik.com/firmware/FG621-EA/16121.1034.00.01.01.03/image
Fair question, IDK. But perhaps physical memory reservation happens in the main package in expectation of the wifiwave2 package. At the end of the day, the kernel will allocate at least some physical memory that wouldn't be available to a usermode process's heap... so this is never going to equal the total memory*.The "32MB per wifiwave2" makes sense in theory, but the empirical evidence tells otherwise. When disabling or deleting the wifiwave2 package from the AX Lite/LTE6, total-memory still shows at 224.0MiB, no changes from when wifiwave2 is installed and active except for the wireless interface disappearing from the system.
At this point, all that we can do is speculate. However, that's still 32MiB unaccounted for, wifiwave2 or not. I used to own a hAP ac2 and whether it was running ROSv6 ot ROSv7, you still see the full 128MiB total-memory.Fair question, IDK. But perhaps physical memory reservation happens in the main package in expectation of the wifiwave2 package. At the end of the day, the kernel will allocate at least some physical memory that wouldn't be available to a usermode process's heap... so this is never going to equal the total memory*.The "32MB per wifiwave2" makes sense in theory, but the empirical evidence tells otherwise. When disabling or deleting the wifiwave2 package from the AX Lite/LTE6, total-memory still shows at 224.0MiB, no changes from when wifiwave2 is installed and active except for the wireless interface disappearing from the system.
*unless one changes the definitions of what's shown
Back in the day of plain BIOS and stupid MMUs, what happened was this: physical memory was contiguous address space and started from 0 to the RAM size (e.g. 33554431 for 32MB size). If there was a peripherial device with its own memory and with fixed address range (e.g. from 16777216 to 20971519 for 4MB memory region in range between 16MB and 20MB), BIOS only excluded RAM from being used in that region but didn't remap it to different address region. This was quite often as peripherial devices tended to be pretty simple, but so was BIOS. So the example ended up with 28MB of usable RAM (and 4MB of video / frame buffer) ... and kernel reported 28M as total RAM. All this happened before kernel even started, kernel only received RAM table so it was aware that region between 16MB and 20MB was not available as RAM. So this didn't depend on some driver being loaded or not. What driver did do (after eventually being loaded) is to figure out that certain memory address space was mapped to certain peripherial functionality and then driver used that address space accordingly.But perhaps physical memory reservation happens in the main package in expectation of the wifiwave2 package.
Would it be possible to update the modem and apply multiple incremental patches at the same time (for example, by appending all files into one)? I would also like to know if /interface/lte/firmware-upgrade command is able to fully upgrade the modem in one execution (if more than one patch is available) or if multiple commands are required.For those interested, I tried getting the full firmware for the FG621-EA via the update URL. I was disappointed to find out that the updates were actually firmware patches implemented as binary diffs from the last version, which the patch was based on. This means that the update files are only smaller in size and not the full firmware release for the modem. But it is good to know that the update mechanism is implemented securely using signed firmware files.
I was trying to study its filesystem and get some list of supported AT commands along the way. In contrast, the R11e-LTE6 modem and its variants are provided full firmware files via the update URL.
Here are the update URLs for this modem:You can open "image" using any capable file archiver and see the files contained therein. You can also study how the patching process works. And yes, you can also use firmware=image when upgrading the modem whenever OTA upgrading fails for some reason, after downloading it to the local filesystem and verifying the required version before updating.Code: Select allFrom 16121.1034.00.01.01.01 to 16121.1034.00.01.01.03: https://upgrade.mikrotik.com/firmware/FG621-EA/16121.1034.00.01.01.01/latest https://upgrade.mikrotik.com/firmware/FG621-EA/16121.1034.00.01.01.01/image From 16121.1034.00.01.01.03 to 16121.1034.00.01.01.04: https://upgrade.mikrotik.com/firmware/FG621-EA/16121.1034.00.01.01.03/latest https://upgrade.mikrotik.com/firmware/FG621-EA/16121.1034.00.01.01.03/image
/system leds settings set all-leds-off=never
Late suggestion:Oh, it is blue? Pretty sure I have not seen it on then.
From here (ROS6 example but it still works for ROS7, I tested it on AX Lite)I'm a new to mikrotik world and I have just bought this new LTE router. So far so good. I would like to program the Mode button to switch on/off the wifi. Could someone post a script please with instructions where to put it? I have tried some scripts from other models but did not work.
I have noticed under the routerboard menu there is NO "Mode button" selection.
Thanks in advance.
/system script add name=test-script source={:log info message=("1234567890");}
/system routerboard mode-button set on-event=test-script enabled=yes
Not necessarily different cell tower but definitely different band (2600MHz now vs. 1800MHz before).Other cell tower and other band so may be normal.