Wireless TX Power, why not Max?

[Rednroll]

The Asus router is the 1st router that I have owned where it allows you to adjust the wireless power transmission level.

The default transmission level is set at 80mW.

Setting description:
Set the capability for transmission power. The maximum value is 200mW and the real transmission power will be dynamically adjusted to meet regional regulations.

So, I have to ask. Why wouldn't you want to set this to the max 200mW? There has to be some trade-offs huh? Why else would 80mW be the default if a higher setting could boost wireless performance? Or is there something I'm misunderstanding?

 

[L&LD]

As mentioned; the Tx power is dynamically adjusted.

In my testing I have found 100mW to be the best for range and throughput (smaller home).

Increasing the Tx power also increases noise the amps generate; therefore higher power may actually give you less usable signal (same as what I have found in actual use).

The 80mW is the default because for most it doesn't need to be higher.

If you want the best the router is capable of; try 100mW instead.


I still have to try slightly higher Tx power levels (105-120 mW) as that works well for some people, but my defaults are still 100mW right now and what I would personally recommend.

 

[DrTeeth]

Increasing the Tx power also increases noise the amps generate; therefore higher power may actually give you less usable signal (same as what I have found in actual use).

Sri to be picky, but higher power does not mean more noise* as such. Loss of throughput or problems with higher power are likely to be related to the quality of the receiver and closeness to the transmitter.

The doubling of the power output will not give as much of a boost as you would think - I would not expect it to turn a marginal system into a max throughput one.

Running max power if not needed is just antisocial, as it will impact on others nearby; just like talking louder than necessary will bother others. The general rule is use the lowest possible power to achieve the desired result, up to the legal maximum in your area.

*noise is something quite specific, more like https://en.wikipedia.org/wiki/Electromagnetic_interference, though there are other meanings.

 

[L&LD]

No, be picky. But you're still mistaken, I think. Increasing the power output would not be attempted for a 'too close' receiver (or at least it shouldn't). We are talking, I assume, about medium to far coverage.

I agree that a marginal signal would not become maxed either - but that doesn't mean it won't become usable though.

The low quality amps (when used at max output) are what is to blame for the noise I'm talking about.

Using max (legal) power is not antisocial: it is how the systems have been designed to be used. Using less than maximum power impacts performance. As my testing from 80mW to 100mW Tx power shows for me. When it shouldn't if we believe the theory (that 1/5 Tx power increase for a 10dBm increase in signal at medium/far distances should not be possible).

Antisocial is using more bands (40MHz width) than the immediate wireless environment allows, imo.

Also, as has been stated many times here; simply having the signal strong does not degrade neighbors network performance - it is the % utilization of that signal that would affect them.

To use your analogy; talking loud is not noise; it is about the router enunciating clearer. An that is what it can't do that at maximum power level (at least not on any of the consumer routers I've used).

If we had access to the highest quality amps this wouldn't be an issue - but we don't and it is.

When I tested my Tx levels at 120mW or higher, the throughput decreased. Even as the signal showed (slightly) stronger.

Theory only works if the equipment can perform at that theoretical level.

I haven't found any yet that do. At least; not at a price I would personally pay for.

 

[DrTeeth]

The low quality amps (when used at max output) are what is to blame for the noise I'm talking about.

Using max (legal) power is not antisocial: it is how the systems have been designed to be used. Using less than maximum power impacts performance.

"Low quality amps" does not mean anything. Do you mean the quality of the electrons themselves ? Seriously, what do you mean by that term?

Max power IS antisocial if a lesser power would do, just like if I were to use my maximum allowed transmitting power to talk to another radio ham across town. Less than maximum power does not always impact performance - believe me, I hear it all the time. Turning down the wick will not make my signals less readable across town, quite often the opposite applys due to a strong signal overloading the other receiver (not noise).

The Wi-Fi system used on routers is no different to any other transmitting system and the same 'rules' apply.

 

[stevech]

Sri to be picky, but higher power does not mean more noise* as such. Loss of throughput or problems with higher power are likely to be related to the quality of the receiver and closeness to the transmitter.

The doubling of the power output will not give as much of a boost as you would think - I would not expect it to turn a marginal system into a max throughput one.

Running max power if not needed is just antisocial, as it will impact on others nearby; just like talking louder than necessary will bother others. The general rule is use the lowest possible power to achieve the desired result, up to the legal maximum in your area.

*noise is something quite specific, more like https://en.wikipedia.org/wiki/Electromagnetic_interference, though there are other meanings.

Correct in general... an add-on amp will usually not have adequate linearity (because it costs a lot to do) so the higher data rates in WiFi OFDM tend to get distorted in the outboard ramp. Technical: The ratio of peak to average/RMS power in the higher bit rates requires a reduction in amp output power to "make room" for the peak/average ratio. This is typically a 4-6dB reduction (backoff) in power and should be done by the low level firmware, irrespective of the user's desired power via the config screens.

The infamous OFDM backoff is needed because getting a power amp that has the headroom without backoff quadruples the cost of the amp and that's a large portion of the total cost.

The user screen interface power setting in WiFi, most especially DD-WRT, is largely ignored at about about 30mW in OFDM modes.

 

[DrTeeth]

Correct in general... an add-on amp will usually not have adequate linearity (because it costs a lot to do) so the higher data rates in WiFi OFDM tend to get distorted in the outboard ramp. Technical: The ratio of peak to average/RMS power in the higher bit rates requires a reduction in amp output power to "make room" for the peak/average ratio. This is typically a 4-6dB reduction (backoff) in power and should be done by the low level firmware, irrespective of the user's desired power via the config screens.

The infamous OFDM backoff is needed because getting a power amp that has the headroom without backoff quadruples the cost of the amp and that's a large portion of the total cost.

The user screen interface power setting in WiFi, most especially DD-WRT, is largely ignored at about about 30mW in OFDM modes.

Many thanks for that extra information. Much appreciated and thanks for taking the time to share.

DrT

 

[L&LD]

"Low quality amps" does not mean anything. Do you mean the quality of the electrons themselves ? Seriously, what do you mean by that term?

Max power IS antisocial if a lesser power would do, just like if I were to use my maximum allowed transmitting power to talk to another radio ham across town. Less than maximum power does not always impact performance - believe me, I hear it all the time. Turning down the wick will not make my signals less readable across town, quite often the opposite applys due to a strong signal overloading the other receiver (not noise).

The Wi-Fi system used on routers is no different to any other transmitting system and the same 'rules' apply.

If you think that antenna amps (or anything else for that matter) are manufactured to the same quality levels, then I won't be trying to educate you on that.

Comparing the ham radio's data (audio) vs. the WiFi radio's GB's of data that it needs to send is stretching the comparison a little thin.

Because WiFi is different than ham radio. Read the WiFi spec's closer. Ham simply broadcasts irregardless of whether another transceiver is listening or broadcasting itself, whereas WiFi is paying very particular attention to it's immediate environment (and will be doing so much more going forward with AC and beyond).

This is why it is not antisocial to be using 100% Tx power: the routers are behaving whether we want them to or not (again; a strong signal without a high utilization of the unwanted network does not impact it's neighbors throughput as much as you might think).


Lol... quality of electrons... it is not the electrons themselves that make the quality; it is the very precise ordering of those electrons that make a difference at the levels we can notice (and in my case; notice easily).

How high that precision is defines the quality that I'm talking about.

 

[TonyH]

Hi,
Rule number 1, when power is increased it generates more heat>more shot noise>dirty signal>distorted signal>spurious signal, also when power is increased beyond optimal level for good overall stable operation, it wastes energy, causes adjacent interference, band swamping, desentisize the RX front shortens equipment life......

 

[DrTeeth]

If you think that antenna amps (or anything else for that matter) are manufactured to the same quality levels, then I won't be trying to educate you on that.

<slaps head> Sorry for a senior moment, when I saw 'amps' I thought you were talking about current and not amplifiers. My bad .

 

[L&LD]

Lol... no problem, I seem to get those moments more and more.

 

[lancer73]

And in this discussion we all forgot about the client side. If you can "hear" the ap, that doesn't mean the ap can "hear" you.


Sent from my iPhone using Tapatalk

 

[L&LD]

lancer73,

the client was not forgotten: I was specifically talking about medium to long distances and not at the extreme edge of the range.

The client could already hear the AP - the higher (allowed) Tx power allowed the client to connect at a faster connection rate.

 

[TonyH]

Hi,
When client RX is barely on the knee threshold of full quieting. Once RX reaches that point increased power does not make any difference.

 

[jim769]

I dont think it really matters were you set the tx power i have ran all my Asus routers at 150mw and never seen much difference one way or the other. I can say it has had no negative impact on the cpu temp or radio temps if its at 80mw or 150mw. I think the router runs what tx power it wants to reguardless of were the tx power is set.

 

[RogerSC]

I dont think it really matters were you set the tx power i have ran all my Asus routers at 150mw and never seen much difference one way or the other. I can say it has had no negative impact on the cpu temp or radio temps if its at 80mw or 150mw. I think the router runs what tx power it wants to reguardless of were the tx power is set.

I think that you're right about that. While increasing the tx power via the web interface can be fun, the results are really somewhat unpredictable. If the router follows your directions unconditionally, you can easily end up with more noise and distortion and your effective throughput goes down even though the signal gets a little stronger. I've had this happen *smile*...the signal is stronger, but throughput goes down. Or there can be no apparent effect at all, no change in either signal strength or throughput, which most likely means that the router is already at an optimal point relative to signal strength and throughput.

On a side note, I have a Netgear R7000, that has it's power set to "Auto" by the firmware. This seems to be a mode that uses maximum useful power, as far as I can tell. Before the router was set to "Auto" by default, it was set to 71mw, which produced a lower signal strength and throughput. When I tried adjusting the tx power on it manually, I would see the throughput go down after a certain a point. I find the "Auto" mode is much better, I don't have to think about it, and it appears to stay at the max strength, max throughput point that provides coverage for my whole house. Must be built into the hardware, like the power amp has feedback of some sort. I'm guessing from what I see, anyways.

Interesting stuff.

 

[ankhazam]

Hmm.. so this thread in TLR means
1) Asus detects the region and if I bought the router from a legitimate EU (Polish) retailer... I get all the EU channels with their full proper channel?
2) Asus slacks and has cut the TxPower and channer for one set to rule them all globally?

I believe it's the second implementation as clearly from what I can choose does not map 1:1 to https://en.wikipedia.org/wiki/List_of_WLAN_channels#5_GHz_(802.11a/h/j/n/ac/ax)
and mine reports only: wl_channel_list_5g = ["36", "40", "44", "48", "52", "56", "60", "64", "100", "104", "108", "112"];

@RMerlin , could you comment on that... have you seen any bits of code that could hint the first implementation?

BR,
Ank

 

[ColinTaylor]

and mine reports only: wl_channel_list_5g = ["36", "40", "44", "48", "52", "56", "60", "64", "100", "104", "108", "112"]

Those are the correct options for the EU if you have set the channel bandwidth to 80MHz.

 

[ankhazam]

So you say by #36/80 it actually makes a 42?
http://www.revolutionwifi.net/revolutionwifi/2013/03/safely-using-80-mhz-channels-with.html
And by this one https://metis.fi/en/2018/02/5ghz-channels/ it is most reasonable to set #100 for max power?

 

[ColinTaylor]

So you say by #36/80 it actually makes a 42?
http://www.revolutionwifi.net/revolutionwifi/2013/03/safely-using-80-mhz-channels-with.html

Correct.

And by this one https://metis.fi/en/2018/02/5ghz-channels/ it is most reasonable to set #100 for max power?

I don't know you'd have to try it for yourself. You might get a little bit more. Check the signal strength on your client device.