Taranis X9E

[Martyn K]

Thanks Gonzo, I agree that your maths are better than mine..

However, I am still getting considerably better than 300m range with my Taranis.

However, my XJT firmware is the 150122 version supporting both EU X8/X6/X4 receivers but also existing V type (D8 mode) receivers.. The latest version is 150413 which does not support D8 mode - only D16

It is possible that the latest XJT firmware runs at reduced power

Martyn

[Martyn K]

I have just posted the following question on the FrSky forum..

**LINK**

Hello

We are investigating reports of low range on EU Firmware XJT modules.

Can you confirm if the XJT module with firmware v1.8.1 (XJT_EU_150413) transmits at the same power level as an XJT module with EU firmware 1.7.1 please? (XJT_EU_150122)

Thanks

/snip

Lets see if I get a response

[GONZO]

Martyn K, as you've got the earlier EU firmware are you getting similar range when used in D8 and D16 mode? Is it possible that you've only used it on only one or other of the modes(D8)?

Edited By GONZO on 22/09/2015 18:44:57

[Martyn K]

Hi Gonzo (the great)

i am using both D8 and D16 with various models. Not noticed any range problems with either and quite happy to push the theoretical limit as well. 300m isn't really that far and often exceed that on long approaches at our strip because the way the trees are organised.

I still don't think this a protocol problem but may be a power problem with the latest Tx module firmware

Martyn

Edited By Martyn K on 22/09/2015 19:34:27

[MartynColes]

Hi All,

I'm a different Martyn (Coles) , and if I can comment :-

I upgraded my Taranis at the same time as Martyn K to the early EU firmware (Feb I think). In fact at his house with his kit. (he was aware of me being there! )

I have been using X4R, X6R, X8R and V8R7 receivers since with no noticable difference in range. One of my models has been a slope soarer and was flown at a long distance from the slope. I had RF signal strength indicated by telemetry, although I can't remember the actual figures, I remember thinking it was more than plenty.

I have also flown 60 size power models and not had any concerns about RF strength. I'm curious enough to have set up to read out loud the RF and battery levels on application of a switch, so do it now and again out of routine.

So for me with my version of firmware all appears to be fine. Does this add weight to suspect later versions?

Martyn Coles

[Bob Cotsford]

Posted by Martyn K on 22/09/2015 19:17:40:

300m isn't really that far and often exceed that on long approaches at our strip because the way the trees are organised.

Not doubting you Martyn....

When I fitted an altimeter I was amazed just how far away 400' is, let alone 300m - nigh on a thousand feet! At 1000' the Cularis is starting to look pretty small, all 100 odd inches of it! I'd be hard pushed to keep orientation on a 60" model at 1000'.

But them myopia is my middle name

[Peter Christy]

BEB is right - it is an inverse square law, so you cannot use a simple linear progression. Also, you need to know what the total path budget is to enable any meaningful response to be made. For most 2.4 GHz equipment, the path budget is in excess of 100dB, so even following the linear model, a 14 or 15 dB drop in signal would only have a 15% effect. Factor in the inverse square law, and its even less. Simple logic does not always apply to RF - its more like black magic - especially when spread spectrum is involved!

I updated my early Taranis to the ETSI 1.8.1 spec when it came out, but did not subsequently update it any further. As a result, mine works with the new X series receivers just fine, but also is still compatible with the older D series too. Just as well, as I have a few of them, and the reason I stopped updating! There is nothing illegal in this, as my system pre-dates the rule change ("Grandfather rights"!). I have not experienced any range issues with new spec X series receivers. I have checked the output on my spectrum analyzer and it is identical to the output of my DSX-9, which has given many years of faithful service.

I have heard a claim that the little plastic aerials fitted to the new receivers can be problematic, and that this can be fixed by replacing them with the simple wire aerials as previously fitted. I haven't experienced this, though.

If people are having issues, then the cause must lie elsewhere, and until someone does an in depth case analysis on a suspect system, we are in the realms of speculation. To date, my Taranis - and my DHT "hack" module - have proven just as reliable and effective as any other brand, and it has my full confidence.

I have to say that the EU has seriously blundered over this issue. The original stated intention was simply to clarify the existing regulations, which were open to misinterpretation (the reason the Dutch and others initially banned model control on 2.4 GHz!). Instead, they've changed the basic rules, caused confusion in the market, and made an agreed international standard no longer standard - and for no good reason!

Time to give the Eurocrats a good kicking, methinks!

--

Pete

 

Edited By Peter Christy on 22/09/2015 22:46:29

[GONZO]

So, from your first paragraph, are you saying that with a Tx o/p of 100mW(20dBw) and a Rx with a sensitivity that gives a range of 1.5Km(1500mtr) if the Tx o/p signal power is droped by 15dB to approx 3mW then you would only loose 15% of range or 225mtr from 1500mtr and still give a range of 1275mtr.

Or, conversely, a drop in Rx sensitivity from say 100dB to 85dB would result in a similarly small loss of range.

[Peter Christy]

It depends not only on the power output and receiver sensitivity, but also on the environment. Someone I know flies - or used to fly - gliders to extreme range on 459 MHz, using only 10mW power. He never had a problem, and that was just using basic PPM! Similarly, when I was developing my 459 MHz system, I modified one of Roy Lever's Merlin PCM systems to 459 MHz and flew a small trainer to a considerable distance with NO Tx AERIAL attached! It was just flying on what was leaking from the BNC socket on the transmitter, and never gave a hint of going in to failsafe.

What 100mW does buy is added security. Just because we CAN fly at 10 mW doesn't mean it will always be secure. It depends on what else is around! 459 MHz is generally pretty quiet - or was, until the drone boys discovered it!

2.4GHz, on the other hand, is quite a widely used band. There is no easy way of saying what effect a given power reduction (or loss of receiver sensitivity) will have as there are simply too many variables. You can predict what would happen under ideal conditions, but this has little application in the real world. In ideal conditions, even a 10mW transmitter ought to be more than adequate - but 100mW will give you added security.

The point I was trying to make - and I know I was grossly over-simplifying a difficult concept - was that you cannot simply string numbers together the way Martyn did and get a meaningful answer. No disrespect to Martyn, by the way. What he did is logical, but our 2.4 GHz systems operate on a different level of logic! And I haven't even started to consider "coding gain"...........

To use an analogy: Say your car does 30 mpg under ideal conditions. You have a 10 gallon tank. You might expect a range of 300 miles. But that does not allow for hold-ups due to road works, accidents, hills, etc. So you always make sure you have more fuel than you actually need before setting off on a journey.

100mW is more than we actually need - MOST OF THE TIME! But it does give us added security in less than ideal conditions.

But to answer your question directly (although I'll use simpler figures to keep the arithmetic simple!): If the power level dropped by 10dB (a ten-fold power loss), then you might reasonably expect the range to drop by 1/sq.rt.10=0.316 So if the original range was 1500 meters, it would *probably* reduce to around 1025 meters.

Again, this is an over-simplification, but it does give an indication of how it all works! Hope this helps....

--

Pete

[Paris Skoutaridis]

Actually,loosing 10dB in transmit power means your link can tolerate 10dB less path loss before your signal quality becomes too poor to maintain lock, all else being equal.

At 2.4GGHz 1500m range translates to about 103dB of path loss. If you lower that By 10dB your max sustainable path loss is 93dB and the corresponding range at 2.4GHz is roughly 460m.

Hope this helps

[Phil Green]

Posted by Martyn K on 22/09/2015 12:04:06:

...resulting in a loss of sensitivity - down by about 14dBm

 

Did Frsky themselves actually say its down by 14dBm (eg -104 to -90) , or 14dB Martin?

 

Edited By Phil Green on 23/09/2015 12:58:57

[Geoff S]

Posted by Phil Green on 23/09/2015 12:43:07:

Posted by Martyn K on 22/09/2015 12:04:06:

...resulting in a loss of sensitivity - down by about 14dBm

dB, not dBm

Phil, db is about the most abused unit of measurement in existence. I've just given up trying to explain that it's a ratio, not an absolute

But I guess you knew that already.

Geoff

[john stones 1 - Moderator]

I thought that was the Pint Geoff...I've only had the one officer

John

[Phil Green]

Theres a fourth option I've not seen mentioned - keep the old datarate but increase the latency?

[GONZO]

Peter,

From your penultimate para using your figures and method:

Power loss 10dB; sq.rt.10 = 3.16; 1/3.16 = 0.316; range of 1500mtr x 0.316 = 474; new range 1500 - 474 = 1026

For other levels of power loss:

Power loss 20dB; sq.rt. 20 = 4.47; 1/4.47 = 0.224; range of 1500mtr x 0.224 = 336; new range 1500 - 336 = 1164

Power loss 5dB; sq.rt. 5 = 2.23; 1/2.23 = 0.448; range of 1500mtr x 0.448 = 672; new range 1500 -672 = 828

Must be something wrong here as the power loss goes up the loss of range goes down and as the power loss goes down the loss of range goes up.

Although, it could be me as I've taken my strong pain killers for the pain I suffer in my left foot (nerve damge caused by diabetes) and am some what spaced out.

Edited By GONZO on 23/09/2015 13:13:24

Edited By GONZO on 23/09/2015 13:17:42

[GONZO]

Phil,

You overlook the importance the add men place on claiming the lowest latency figures.

At one point FrSky did actually quote these figures on there forum but quickly withdrew them.

Edited By GONZO on 23/09/2015 13:16:51

[Paris Skoutaridis]

I stumbled across this thread and since I have been thinking of getting a Taranis it grabbed my attention.

Being a communications engineer I thought I'd lend a hand to help with the power and loss calculations.

The formula for path loss is 20*log(4*pi*d/lamda) were d is the distance in metres and lamda is the signal wavelength calculated as the fraction of the speed of light (3*10^8) over the signal freequency (2.4 * 10^9 Hz in this case)

At a max "useful" range of 1500m you get a path loss of 103.5dB.

If you then reduce your transmit power by 10dB, the result is that you can maintain the link (i.e. maintain receive lock) at a distance that corresponds to 10dB less path loss than the original range of 1500m. I.e. you can maintain the link with a maximum path loss of 93.5dB.

At a frequency of 2.4GHz, the range at which the total path loss is 93.5dB is about 460 metres.

Keep in mind the above is quite simplistic as it only takes into account free space loss. The receiver will also have to contend with multipath fading (i.e. several delayed versions of the same signal arriving from different angles at the receiver), block and absorption from surroundings and the model's structure and so on.

Hope this helps somewhat

[Peter Christy]

Hi Gonzo,

Actually, we're both wrong! That's the problem of doing complicated stuff in a rush (that's my excuse, anyway!)

That figure for the reduced range that I gave should have been 475 meters, not 1025! (1/10 of the power, roughly 1/3 the range - all other things being equal!).

But remember that dBs are a logarithmic scale: dB=10log(P1/P2) So a loss of 20dB represents 1/100th of the power, 30dB is 1/1000, etc, etc.

Thus your second calculation should read 20dB loss: sq.rt.100=10 ; 1/10=0.1 ; 1500 metersx0.1 = 150 meters.

Sorry about the silly mistake! (Wanders off kicking himself hard.....!)

--

Pete

[Martyn K]

Posted by Phil Green on 23/09/2015 12:43:07:

Posted by Martyn K on 22/09/2015 12:04:06:

...resulting in a loss of sensitivity - down by about 14dBm

Did Frsky themselves actually say its down by 14dBm (eg -104 to -90) , or 14dB Martin?

Edited By Phil Green on 23/09/2015 12:58:57

They stated receiver sensitivity on EU was -90dBm as against -104dBm on International..

Regards

Martyn

(P.S. - I understand that dB is a ratio between 2 values and dBm is a logarithmic reference to 1mw (0dBm))

[Martyn K]

Oh..and back to the O/P, I assume that no-one has bought a X9E yet (and the way this thread is going we'll probably frighten them away for life)

Martyn

[GONZO]

Peter,

Told you I was spaced out, its taking me ages just to type!

But, lets look at FrSky's original statement that they had lost 14dBm sensitivity with a version of the EU firmware. Does this not equate very closely to the claimed low range of up to 300mtr in real life with some recent X9E, Taranis and X series Rx combinations?

I hope this makes sense,its taken me almost 15min to type this.

[Phil Green]

The difference between -104dBm and -90dBm is 14dB, not 14dBm.

[Peter Christy]

Gonzo: Yes, that looks pretty close - back of the envelope calculation and lets assume 13dB, not 14 as its easier to do without a calculator and 1dB won't make any significant difference!

10 dB loss = 1/10; 3dB loss = 1/2, therefore 13 dB loss = 1/20 (If you add/subtract logarithms, you multiply / divide the actual numbers)

Square root of 20, approx 4.5, so (1/4.5) x 1500 = 333.333

What I cannot understand is why a minor change to the protocol (and it is minor) would have such a dramatic effect. And why haven't I experienced any range issues with new X series receivers? Not even a hint of an RSSI warning - which I do get very occasionally on older D series receivers? And the D series give warning LOOOONG before they run out of range. In flight its usually a momentary aerial orientation thing - not uncommon at UHF frequencies - and has no effect on the model.

There has to be something else at work here.

--

Pete

[GONZO]

Phil, Yes typo. Just come round from partially sleeping off the tablets unfortunately I'm due some more within the hour.

Pete, I have no personal experience of this range issue(all my stuff is on pre 1/1/2015), just trying to find an explanation. FrSky quoted the figures for receiver sensitivity; EU Rx -90dBm, International Rx -104dBm. As you've confirmed this looks very suspect when taken with the real life reports of lack of range. Your lack of experience of this issue could just be down to when you purchased your Rx's. Could be something else or perhaps just a short period of time where the firmware used had this affect on products produced. FrSky then rectified matters on all subsequent production. Who knows, will we ever know???

[Peter Christy]

One thing that does puzzle me is that receiver sensitivity is usually quoted with reference to field strength. Field strength is usually quoted in terms of Volts per Metre. For receiver sensitivity, this is more often uV/M - often abbreviated to uV (micro-volts). In the days of 27 MHz, everyone was struggling to make receivers capable of detecting 1 uV!

So quite what they mean when they speak of dBm for receiver sensitivity, I'm not sure. At what distance? Even if its dB - not dBm - then the question still remains "compared to what?" As it stands, the phrase is meaningless.

The added complication is "coding gain" resulting from the spreading and de-spreading of the signal. It could be that they are talking about a reduction in the coding gain, or even a reduction of the total path budget. I suspect that the answer lost something in translation, because as it stands, it doesn't make a lot of sense.

The nearest equivalent to the "path budget" we refer to for spread spectrum was the "dynamic range" of 27 / 35 MHz systems. Most AM receivers had an AGC range of around 60 dB, and could probably go another 10dB once the AGC had run out of range. So a total "path budget" of around 70 dB seems a reasonable assumption. Compare that to even the reduced figure quoted by FrSky (90 dB), and you should still be 20dB (100 times) better off than you were on 27 / 35 MHz!

I have a sneaky suspicion (and it is only a suspicion!) that the problem is more likely related to those plastic encapsulated aerials that come as standard on the current receivers. I suspect they are more directional than the basic aerials used previously. If I were a gambling man, I wouldn't mind betting that replacing those with the standard wire ended aerials would make a lot of the issues go away....

--

Pete