Kiwi's XPS test results are up.

[gareth.ky]

	Quote: Originally Posted by BZFrank
	Thats an artefact and not real - remember that the sweep time of the WiSpy-2.4 is 165ms while FASST jumps must faster. Frank

Gotcha

[Daemon]

	Quote: Originally Posted by Simpleton
	Call it placebo if you will, but I can tell a difference between a Futaba PPM RX and a DSP filter RX such as a Berg. Both are FM PPM, both use the same TX module. So I suppose it's conceivable that if one went from a PPM FM DSP Berg to an XPS system, one could indeed perceive a difference?

That's because Bergs are *slugs*. I can tell the difference between regular analog FM Rx
and Bergs. I continue to not understand why people like Bergs so much, because
I've had nothing but problems with them. Slower response time, they move the servos
slower, and they literally lock up or glitch and fall out of the sky anywhere inside
about a 10 square mile area or so around where we fly. There are faster DSP Rxs
like the Sombra gear.

My XPS gear feels pretty responsive, but so do some of my 72Mhz Rxs, so I
won't make any claims either way.

BTW, someone can correct me if I'm wrong here, but it seems that with a packet based
system fed from PPM input it should be possible that the response could *feel* different,
like this. Normally the PPM 20ms frame (roughly 13ms of signal, and 7ms sync pause)
is fed serially to the Rf Tx and then to Rf Rx on the other end of the link in the receiver, and as
each pulse for each servo comes along, the Rx filters it out and sends it to the appropriate
channel. If the input contains movement on 8 servos, then channel 1, gets updated
first, then 2, then 3, then 4.. and so on for 13ms. Then there's a 7ms sync pause and the
next PPM frame comes along. Now let's say the Tx rolls up all the stick inputs
in the Tx PPM frame after the first 13ms, sends a packet that takes 2ms (yes that's 2m
latency beyond the original PPM signal pulses, had it been transmitted serially). Rx
receives the packet and instead of feeding all the servos serially one after another, it sends
every servo their pulse starting at the same moment. This might *feel* different.
It's not faster, it's not significantly slower either, but it might be perceived
differently. BTW, I don't know if XPS does it this way or not, but it could.
Any 2.4Ghz packet based system could. Maybe I'll get out the silly scope.

ian

[Simpleton]

	Quote: Originally Posted by Daemon
	That's because Bergs are *slugs*. I can tell the difference between regular analog FM Rx and Bergs. I continue to not understand why people like Bergs so much, because I've had nothing but problems with them. Slower response time, they move the servos slower, and they literally lock up or glitch and fall out of the sky anywhere inside about a 10 square mile area or so around where we fly. There are faster DSP Rxs like the Sombra gear.

Maybe all of that is true, but the premise was that somebody couldn't tell the difference between XPS and PPM FM. Bergs, like them or not, are PPM FM RXs. I contend that it would be possible to tell the difference between this particular PPM RX/TX combo and the XPS combo.

FWIW, I've never had so much as a glitch with Bergs. They do slow down if heavy interference is encountered, possibly why they are so "slug"gish to you.

[XJet]

	Quote: Originally Posted by Simpleton
	Maybe all of that is true, but the premise was that somebody couldn't tell the difference between XPS and PPM FM. Bergs, like them or not, are PPM FM RXs. I contend that it would be possible to tell the difference between this particular PPM RX/TX combo and the XPS combo.

I wasn't thinking of DSP-PPM but good old "dumb" PPM where each pulse is fed out to the servo as soon as it's received.

The Bergs strong DSP requires a delay of at least one whole frame (20mS) in order to validate the integrity of the data it's received and you're right -- it *is* noticeable.

I remember going back to a "dumb" PPM system after flying the old Berg 5/II receivers and I was astonished at how much more immediate the model's responses were (mind you this was a 3D profile with HS5925 (0.08Secs) servos).

The newer "mild DSP" receivers such as the Coronas seem to introduce far less delay -- I suspect they're checking each pulse as it's received and passing it on without waiting for a full frame to be constructed/buffered.

	Quote:
	FWIW, I've never had so much as a glitch with Bergs. They do slow down if heavy interference is encountered, possibly why they are so "slug"gish to you.

The Castle-made Bergs are rock-solid but the last batch of the old Bergs had some *real* issues and caused many crashes when they simply went to sleep in mid-air for no apparent reason.

[Simpleton]

	Quote: Originally Posted by XJet
	The Castle-made Bergs are rock-solid but the last batch of the old Bergs had some *real* issues and caused many crashes when they simply went to sleep in mid-air for no apparent reason.

All of mine are pre-Castle, pre-napping Bergs.

Funny that some of the indoor precision/freestyle guys run those sluggish Bergs. You'd never guess by watching them that their receivers sucked.

[XJet]

	Quote: Originally Posted by Simpleton
	All of mine are pre-Castle, pre-napping Bergs. Funny that some of the indoor precision/freestyle guys run those sluggish Bergs. You'd never guess by watching them that their receivers sucked.

They don't suck - they just have a higher latency than most other receivers. You soon become used to this and simply learn to input your movements a little sooner. It's not even a conscious thing, the brain does it automatically.

You *do* notice the difference however if you're switching back and forth.

You haven't noticed those people who've switched from PCM to 2.4GHz suddenly flying better just because they've lost a little latency do you?

[Daemon]

Around here, they suck. All of them. Pre-CC ,post-CC, every model, Berg 4, 5, 6, 7.
They lag, they glitch, they fly out of range at 100 yards and fall from the sky.
I *literally* don't trust them any further than I can throw them, and several times, visiting
pilots have come to our slopes or fields and had planes glitch, lock-out badly in the air, or crash
outright, and we've discovered that they're using a Berg, again, both old and new models.
I've even discovered it by recognizing the symptoms and asking "Do you have a Berg
in there?" and getting an affirmative response, having not known before.
I owned 3 of them at one point, and gave them all away to out of staters.
By contrast, old "dumb" Rxs like the Hitec 555, Electron 6, Superslim 8,
and all manner of FMAs work flawlessly, and we regularly fly them out over half a mile
(as we also do with various 2.4Ghz systems including XPS, Spektrum, Airtronics FHSS).

I wish I knew why, but knowing that they work so badly here, makes me not trust them
anywhere else, because we don't know what causes the problem so don't know
where else we're going to encounter it. And it's not just one field or one slope. It's roughly
10-15 square miles of area near the mountains west of Denver. I'm not trying to drag this
thread off topic but it's sort of an interesting case study, because it demonstrates how huge
the difference in experience with a particular product can be. We're busy talking about how
badly XPS sucks, while some people have had no problems at all with them, and
it mirrors my experience with Bergs which some people swear by, but I've lost
expensive planes to.

ian

[cptsnoopy]

I could tell the difference from my electron6 rx's compared to the DX-6 2.4ghz. The difference was not so much in speed but in stability. I was flying an Ark400 heli when I first tried the DX-6 and was actually amazed at how stable the heli became just by installing the AR-6000. All the little twitches (I would guess baby glitches) went away and the heli felt sooo much more connected. I immediately dumped all of my 72mhz stuff and started buying more AR-6000's. The XPS/evo12 feels about the same as my DX-7 so I would only assume that if I started with XPS it would have been the same difference from 72mhz.

[USN_POPS]

	Quote: Originally Posted by cptsnoopy
	I could tell the difference from my electron6 rx's compared to the DX-6 2.4ghz. The difference was not so much in speed but in stability. I was flying an Ark400 heli when I first tried the DX-6 and was actually amazed at how stable the heli became just by installing the AR-6000. All the little twitches (I would guess baby glitches) went away and the heli felt sooo much more connected. I immediately dumped all of my 72mhz stuff and started buying more AR-6000's. The XPS/evo12 feels about the same as my DX-7 so I would only assume that if I started with XPS it would have been the same difference from 72mhz.

I as well noticed the difference when I went to XPS. I'm not a world class flyer, so use a lot of Expo to keep me from overcontrolling. With the Futaba 9c and Fm RX, I had the Expo set to 60 %, after using XPS, I was overcontrolling again, so on one particular plane that I fly most, I went to 80% to get back to where I didn't have so much problem with it.

[XJet]

	Quote: Originally Posted by Daemon
	Around here, they suck. All of them. Pre-CC ,post-CC, every model, Berg 4, 5, 6, 7. They lag, they glitch, they fly out of range at 100 yards and fall from the sky. I *literally* don't trust them any further than I can throw them, and several times, visiting pilots have come to our slopes or fields and had planes glitch, lock-out badly in the air, or crash outright, and we've discovered that they're using a Berg, again, both old and new models. ian

It sounds as if you've got a lot of noise on the image frequency around where you fly -- which is something that *all* SC receivers have trouble dealing with.

I'm not actually that impressed with the way that Peter Berg tried to downplay the importance of dual conversion in RC receivers. The Berg receivers are designed on the premise that the DSP will take care of any signal that gets through on the image frequency -- but this only has a very small chance of success.

[XJet]

	Quote: Originally Posted by RichardCorby
	I as well noticed the difference when I went to XPS. I'm not a world class flyer, so use a lot of Expo to keep me from overcontrolling. With the Futaba 9c and Fm RX, I had the Expo set to 60 %, after using XPS, I was overcontrolling again, so on one particular plane that I fly most, I went to 80% to get back to where I didn't have so much problem with it.

A couple of things...

80% is a *LOT* of expo! Are you sure your throws didn't increase when you switched to XPS?

Did you actually check to see if the throws were different when you switched systems?

The plane won't know what RC gear is being used and if a control surface is deflected by x degrees then the resulting response will be the same. If the control input acts a little more quickly due to reduced latency then it'll also respond more quickly when you remove the control input so overcontrolling shouldn't be an issue.

Remember also that we're talking about 10ms or so of reduction in latency over a PCM setup. With most servos being around the 0.15 second mark for 45 degrees deflection (0.3 deg/mS), which means even if you peg the stick, the control surfaces on an XPS system are only going to be 3 degrees or so ahead of the standard system in terms of servo movement.

Enough to make a real difference? I don't know.

If it's just reduced latency (which I doubt) then all you've got to do is fly a while until your brain adapts to the more immediate response.

If the latency is reduced then your plane should be *easier* to fly (after retraining your brain) because the control surface deflections will more accurately follow your finger movements.

I always found it a *lot* harder going from low-latency to hi-latency setups than the other way.

Adding expo *won't* change the speed at which your servos react to stick movement and 80% is a wicked-high amount of expo on any plane.

On my 3D profiles I have 55% expo and that's with huge surfaces with over 50 degrees of throw either side of neutral and I have no issues with it being too touchy or oversensitive in straight and level flight.

Even my 30% Extra 330L gasser has just 55% expo on elevators and ailerons and it's set up for 3D throws as well. I don't use dual rates (just another switch to flick (or not flick) at the wrong moment) -- I just learn to move the sticks a little less aggressively when flying on the wing.

This is probably the subject for another thread but how much expo are others using?

[bob_nj]

Subscribing to the volumes of information hoping to eventually digest

[uninverted]

Not the place I know.
If that much expo is being used, maybe a look at the link positions on the servo/horns should be looked at.
In on servos, out on horns, all about mechanical advantage. Works for me. Reduces throws, but if overcontrolling who needs wild throws.

[Daemon]

	Quote: Originally Posted by XJet
	It sounds as if you've got a lot of noise on the image frequency around where you fly -- which is something that *all* SC receivers have trouble dealing with. I'm not actually that impressed with the way that Peter Berg tried to downplay the importance of dual conversion in RC receivers. The Berg receivers are designed on the premise that the DSP will take care of any signal that gets through on the image frequency -- but this only has a very small chance of success.

Agreed 100% on both points. Any time the issue comes up, everyone quotes Peter's article
on the subject, and I think it's a fundamentally flawed premise. That said, some SC PCM
Rxs work just fine here.

ian

[tadawson]

	Quote: Originally Posted by Daemon
	Agreed 100% on both points. Any time the issue comes up, everyone quotes Peter's article on the subject, and I think it's a fundamentally flawed premise. That said, some SC PCM Rxs work just fine here. ian

Hmmm . . .

I run older Bergs in all my electrics - planes and helis, and they are the **only** thing I have found that will run clean in the helis, short of PCM . . . . but never a glitch in a one of them, and response feels fine to me . . .

I guess we all see things differently . . . .