I observed on my international version, EU plug hAP ac² that it reports more RAM than described in product specification, product brochure and block diagram. In all documentation 128 MB is mentioned, however “/system resource print” shows 256 MB:
Are there multiple HW revisions on the market? Not that I would need more than 128 MB memory for my SOHO routing requirements, but I am still interested. Maybe it would worth to update documentation.
As I undestand it, sometimes you may get lucky and receive a little more than what’s in specification. Not that double RAM helps much, even 128MB is probably more than what this device will ever need. I wouldn’t dare to say the same about flash. But hey, it’s free, no need to complain.
Yes, and also about her teeth. But. But.
Ah, of course, i will live with this. I already bought hAP ac². I’m wait it from EuroDK. May be i will more lucky and my hAP ac² will have a SD-card slot or something else???
How did you checked that? Did you disassembled your unit? I could only check with command “/system resource irq print”, and there I see 2 entries with “ipq4019_ahb” which indicates that it is IPQ-4019 as you said, however the specification on Qualcomm site says 1 GB RAM for 4019, and 256 GB RAM for 4018. So hard to says without disassembling the unit, but one guy disassembled it, and on his picture the SoC is clearly 4018. And one of his screenshot in the article also shows 233 MB RAM.
Edit: Ok, I see in Wireless/WiFi Interfaces, under the Type field that it is IPQ4019. However I still believe just the driver is named like 4019, the SoC itself is IPQ-4018.
It’s strange, but I see only two antennas, and the block diagram shows 4 antennas. For the first time I see that Mikrotik would use dual-band combined antennas. How do the transmitters not jam the receivers of the other range, while maintaining a decent sensitivity?
Labels on the circuit board says “2G CH0 + 5G CH1” … the other is obscured by antenna itself, but I guess print would be similar (but with swapped chains).
The frequency of two WiFi bands are (radio wise) far away from each other with plenty of other occupied bands in between so that filters (both in TX and RX branches) can do very decent job. TX needs decent band-pass filter not to emit energy outside of ISM band (that would be illegal). RX needs band-pass filter as well to reject some adjacent-band interference which might lower usable sensitivity. Further more, RX needs channel-pass filter and this one is usually implemented in DSP software (or else channel switching would be a hard thing to do). Decent band-pass filter would have out-of-band attenuation of at least 40dB. Given that there are two band-pass filters in play (e.g. TX filter on 2G and RX filter on 5G), overall out-of-band attenuation inside single device should be at least 80dB … if TX power is +20dB, that would mean -60dB at input of RX pre-amplifier. Which is not good, but not useless either … given the nature of OFDM being used for recent WiFi standards where sensitivity is less of an issue than with other (legacy) technologies.
