Thread Tools
Aug 01, 2010, 05:45 PM
multirotor/hotliner pilot
saturnine's Avatar
Thread OP
Discussion

Spektrum satellite input


I know this has been discussed a bit in previous threads but I believe these are concerns that haven't been touched on.

I have a DX7se which is supposed to send 11mS frames, which is great, but it appears that the satellite alone only receives at 22mS. I timed it and it was about 22.2mS per 16 bytes. My questions are as follows:

1) Is the satellite incapable of utilizing the 11mS/11-bit frames from a hardware perspective? If it is capable, does one need to bind it to a receiver that is designed to receive 11mS/11-bit frames to get it to work? If it isn't capable at all, this raises all sorts of interesting questions about utilizing 3-axis gyros that use satellites alone as receivers, since you're technically missing some speed as a result of those "upgrades".

2) The data comes in as 10bit (1024 step) frames every 22mS from the satellite, but what I found interesting was that my analysis showed me that the DX7se is limiting throws to a range of ~ 680 steps (extreme to extreme) per channel. The center points are all around 511 as they should be, but they all do not go anywhere near the 0 or 1023 thresholds, but are bound within ~ 170 steps of each edge. Does anyone know why this is the case? It seems pretty wasteful and possibly even deceiving to suggest there is 1024 steps when your radio can only use about 66% of them. Thoughts?

For example (in hex):

Aileron center 5 FE, left 7 52, right 4 AA, center 510, 680 step range
Elevator center 9 FF, down 8 AF, up B 52, center 511, 674 step range
Rudder center D FC, left F 4E, right C A9, center 508, 677 step range
Throttle bottom 0 AA, up 3 59, 687 step range

This gives me a glimpse at how "accurate" these pots are on these radios; not very. Anyway, anyone know why they are so limited? I suppose we could assume that 11-bit would give about double the range of ~ 1360 steps, far less than promised as far as I can tell, unless I'm missing something?

Thanks,

Damon
Sign up now
to remove ads between posts
Aug 01, 2010, 05:56 PM
Registered User
A couple of questions:

Which receiver are you using?

Are you setting your end points to 150 %. If at the default of 100%, of course you are not going to get all the resolution available.

From your comments it sounds like you are using a non-2048/ non high speed receiver. The Tx knows this from binding and is defaulting to 1024/22 msec to suite the receiver.
Aug 01, 2010, 06:08 PM
multirotor/hotliner pilot
saturnine's Avatar
Thread OP
Thanks for the response.

I am working directly with a satellite receiver with no "mother".

The radio is set stock, so to whateve that is, 100% as you say? Excuse my obvious ignorance, but why is 100% not 100% but 150% is? I'm sure there are useful reasons, but I'll need to be enlightened.

Yeah, it seems the satellite alone is not a high speed, high res receiver. An interesting question though is whether it becomes one if bound through a high speed, high res receiver? Any ideas?

Thanks,

Damon

Quote:
Originally Posted by teamdavey
A couple of questions:

Which receiver are you using?

Are you setting your end points to 150 %. If at the default of 100%, of course you are not going to get all the resolution available.

From your comments it sounds like you are using a non-2048/ non high speed receiver. The Tx knows this from binding and is defaulting to 1024/22 msec to suite the receiver.
Aug 01, 2010, 07:14 PM
Registered User
In order for 150 % to be available at all, the overall system resolution must accomodate the stretch so seeing only ~70% of the range used when the end points are set to 100% is exactly what you would expect.

Now - all satelittes are the same, but all receivers are not. The Tx determines the capability of the receiver at binding and acts accordingly. I am not party to the communucation between the Rx and satelite, but I know from playing that there is a difference between the data sent from the Tx when there is a 2048 receiver out there versus a 1024. Since the core receivers (excluding 6200) aslo contain a satellite or satelites it is reaonable to assume the data from the satelite is also differnet in the two cases. (flat out has to be).

So how are you binding the satelite? If you are using a non 2048 receiver, the transmitter is smart enough to only send 1024 data. If you are using a 2048 receiver, it will send 2048 data. I would guess the update rate issue is in the same class.

I should emphasize that what the Tx thinks it is communicating with is set when you bind the Tx to whatever.

Note that the DX7SE only claims high speed "with a capable receiver".
Last edited by teamdavey; Aug 01, 2010 at 07:16 PM. Reason: Update
Aug 01, 2010, 09:20 PM
multirotor/hotliner pilot
saturnine's Avatar
Thread OP
See below in red, thanks.

Quote:
Originally Posted by teamdavey
In order for 150 % to be available at all, the overall system resolution must accomodate the stretch so seeing only ~70% of the range used when the end points are set to 100% is exactly what you would expect.

I understand that, what I don't understand is the purpose for anything beyond 100%? In other words, why not just have the channels always using the full range, or at least much more so than 66%? I can imagine some value in buffer room, but 33%? Any explanation what advantages are afforded by the extra throws would be helpful.

Now - all satelittes are the same, but all receivers are not. The Tx determines the capability of the receiver at binding and acts accordingly. I am not party to the communucation between the Rx and satelite, but I know from playing that there is a difference between the data sent from the Tx when there is a 2048 receiver out there versus a 1024. Since the core receivers (excluding 6200) aslo contain a satellite or satelites it is reaonable to assume the data from the satelite is also differnet in the two cases. (flat out has to be).

Yes, but it is also possible that the satellites have the ability to switch between 2048/11-bit and 1024/10-bit. My testing shows 1024/10-bit (no surprise), but I am aiming to determine whether 2048/11-bit is possible with or without the use of other receivers.

So how are you binding the satelite? If you are using a non 2048 receiver, the transmitter is smart enough to only send 1024 data. If you are using a 2048 receiver, it will send 2048 data. I would guess the update rate issue is in the same class.

I've binded directly to the satellite using my own software. There came the question whether or not binding through a high res/low latency receiver would put the satellite in a similar mode that can be used on its own? Of course this would be easy enough to test with such a receiver, but unfortunately I don't have one at my disposal at the moment. I was hoping others might know the answer.

I should emphasize that what the Tx thinks it is communicating with is set when you bind the Tx to whatever.

Note that the DX7SE only claims high speed "with a capable receiver".

Yes, I read a review about how it can transmit either in 11mS or in 22mS depending on the receiver. My questions are more specific to the satellites themselves and their physical capabilities in situations that I currently am unable to test.
Aug 01, 2010, 09:49 PM
Registered User
OK - here goes:

"I understand that, what I don't understand is the purpose for anything beyond 100%? In other words, why not just have the channels always using the full range, or at least much more so than 66%? I can imagine some value in buffer room, but 33%? Any explanation what advantages are afforded by the extra throws would be helpful."

The quick answer is because thats the way it is. The long one is because the system has to account for additive mixes, use of subtrim etc and this "margin" allows for all that.

Personally I frequently use over 100% base setting and in CCPM use on a heli the full 150% is often hit during combination stick movements.

"Yes, but it is also possible that the satellites have the ability to switch between 2048/11-bit and 1024/10-bit. My testing shows 1024/10-bit (no surprise), but I am aiming to determine whether 2048/11-bit is possible with or without the use of other receivers."

Of course, but the Tx has to think that is is talking to an Rx that can handle it - this is determined by the transaction that occur during binding and that will depend on the host Rx software - in this case your software............................

"Yes, I read a review about how it can transmit either in 11mS or in 22mS depending on the receiver. My questions are more specific to the satellites themselves and their physical capabilities in situations that I currently am unable to test."

Satallites are a common design to all types of receiver. You can swap them at will.

I hope this helps.
Aug 01, 2010, 10:08 PM
multirotor/hotliner pilot
saturnine's Avatar
Thread OP
Interspersed below:

Quote:
Originally Posted by teamdavey
OK - here goes:

The quick answer is because thats the way it is. The long one is because the system has to account for additive mixes, use of subtrim etc and this "margin" allows for all that.

I can see use in mixes, etc, that helps me understand, but wow, very interesting to me that there is so much room "left over" so to speak.

Personally I frequently use over 100% base setting and in CCPM use on a heli the full 150% is often hit during combination stick movements.

I suppose this is where the "eletronic E-Ring" features of some newer radios comes into effect. Of course it seems amazing to me that this feature is relatively new at all.

Of course, but the Tx has to think that is is talking to an Rx that can handle it - this is determined by the transaction that occur during binding and that will depend on the host Rx software - in this case your software............................

Well, I wish I had a 11-bit/11mS capable receiver to bind it with to see if that put the satellite in that mode. More interesting of course would be to know what I need to send the satellite to make it and the radio beliece it's high res/low latency capable.

I hope this helps.

Thanks, it does
Aug 01, 2010, 10:38 PM
Registered User
"Well, I wish I had a 11-bit/11mS capable receiver to bind it with to see if that put the satellite in that mode. More interesting of course would be to know what I need to send the satellite to make it and the radio beliece it's high res/low latency capable."

To be sure what we are saying here - I do not believe the satelite has modes. It is a conduit from the the Tx to the baseband element of the receiver. Once the Tx "knows" what is out there evertyhing is in the hads of the Tx and the satellite simply passses that to the baseband stuff.

For example, if you bind a 9303 (2048 capable) Tx to a 1024 capable reciver and then bid it to a 2048 capable receiver (on the same mode address of course) it will still talk to the 1024 capable receiver, but the results at the servos are chaos.
Aug 01, 2010, 11:08 PM
multirotor/hotliner pilot
saturnine's Avatar
Thread OP
Yeah, maybe we are misunderstanding eachother a bit? I'm with you with regard to what you say below about the satellite being a conduit, I think. It's clear to me that the satellite alone only intends to do 1024/22mS but assuming it is bound with a 2048/11mS receiver and then used alone will it be capable of receiving, by itself, 2048/11mS or is it that even when used with a 2048/11mS receiver it's passing through only at 1024/22mS to plug in gaps in data with the faster mother receiver?

The bottom line questions is: Under any circumstances, can the satellite receiver, when use on its own as my receiver, deliver 2048/11mS to my model?

Damon

Quote:
Originally Posted by teamdavey
"Well, I wish I had a 11-bit/11mS capable receiver to bind it with to see if that put the satellite in that mode. More interesting of course would be to know what I need to send the satellite to make it and the radio beliece it's high res/low latency capable."

To be sure what we are saying here - I do not believe the satelite has modes. It is a conduit from the the Tx to the baseband element of the receiver. Once the Tx "knows" what is out there evertyhing is in the hads of the Tx and the satellite simply passses that to the baseband stuff.

For example, if you bind a 9303 (2048 capable) Tx to a 1024 capable reciver and then bid it to a 2048 capable receiver (on the same mode address of course) it will still talk to the 1024 capable receiver, but the results at the servos are chaos.
Aug 02, 2010, 03:10 AM
Registered User
"I've binded directly to the satellite using my own software"
The satellite does not respond during binding so perhaps you are instructing the tx to use 1024.

Is the AR9000 2048 capable and does it use the same satellite as you have? If yes then I'd expect it to be able to take 2048. Is the sat sold with any of the rx's capable of 11ms? Again you will need to know how to instruct the tx to use that frame length.

The 100% relates to the pulse width that most radios use by default, ie: around 0.9 to 2.1ms. Calling 'normal' throws 100% is a reasonable approach, 120% would be more than normal and 80% less than normal.
Regards, David.
Aug 02, 2010, 03:44 AM
We want... Information!
Bruce Abbott's Avatar
Quote:
Originally Posted by saturnine
This gives me a glimpse at how "accurate" these pots are on these radios; not very.
"not very" compared to what? The sticks in your radio are centered to within 0.3% of each other, and the endpoints are within 1%. Compare this to the rated tolerances of a high quality potentiometer (Bourns 3852); Linearity and Effective Electrical Angle: 5%, Total Resistance 10%.

Quote:
my analysis showed me that the DX7se is limiting throws to a range of ~ 680 steps
680 steps is more than enough to provide smooth control response. I did a little experiment to see how many distinct mechanical positions I could produce by moving the throttle stick with thumb and forefinger - the best I could do was 5 steps per detent, which corresponds to a total of 250 steps. That makes the radio about 3 times better than my fingers!

Here's what Spektrum have to say about resolution:-

"there comes a point where the resolution of a radio surpasses the resolution of the servos or ESC, making incredibly high resolution claims irrelevant."
Aug 02, 2010, 09:17 AM
multirotor/hotliner pilot
saturnine's Avatar
Thread OP
They all use the same satellites. Now, does anyone know how to get it into that mode? It seems everyone is convinced they can do it, now I'm interested, specifically in how.

Thanks.

Quote:
Originally Posted by David T
"I've binded directly to the satellite using my own software"
The satellite does not respond during binding so perhaps you are instructing the tx to use 1024.

Is the AR9000 2048 capable and does it use the same satellite as you have? If yes then I'd expect it to be able to take 2048. Is the sat sold with any of the rx's capable of 11ms? Again you will need to know how to instruct the tx to use that frame length.

The 100% relates to the pulse width that most radios use by default, ie: around 0.9 to 2.1ms. Calling 'normal' throws 100% is a reasonable approach, 120% would be more than normal and 80% less than normal.
Regards, David.
Aug 02, 2010, 09:23 AM
Registered User
Quote:
Originally Posted by saturnine
They all use the same satellites. Now, does anyone know how to get it into that mode? It seems everyone is convinced they can do it, now I'm interested, specifically in how.

Thanks.
What do you init the sattelite rx with on bind, ie the pulse train at power up?
Aug 02, 2010, 09:25 AM
multirotor/hotliner pilot
saturnine's Avatar
Thread OP
I'm not an electrical or mechanical engineer, so my reference is my own (apparently unreasonable) expectations. If you say it's normal, fine it's normal. My personal (uninformed) opinion is that it's not very accurate. It seems to me this could all be calibrated in the factory, but I suppose that adds time and ultimately cost they are not willing to spend either resource on.

If ~ 680 steps is more than enough, they why do the vendors bother selling 2048 at all? I'm not going to debate about the merits of more or less control steps, that's way outside the scope of this thread. My question most closely related to your comments here is about what's available (1024, or 2048, depending) versus what's used (~ 680 and presumably ~ 1360, depending). If Spektrum or other radio manufacturers themselves sell 2048, I'd say your comment and quote below demonstrate a bit of deception in marketing, don't you? After all, by yours and their logic, 2048 is pointless, right? Anyway, this isn't my aim.

Thanks,

Damon

Quote:
Originally Posted by Bruce Abbott
"not very" compared to what? The sticks in your radio are centered to within 0.3% of each other, and the endpoints are within 1%. Compare this to the rated tolerances of a high quality potentiometer (Bourns 3852); Linearity and Effective Electrical Angle: 5%, Total Resistance 10%.

680 steps is more than enough to provide smooth control response. I did a little experiment to see how many distinct mechanical positions I could produce by moving the throttle stick with thumb and forefinger - the best I could do was 5 steps per detent, which corresponds to a total of 250 steps. That makes the radio about 3 times better than my fingers!

Here's what Spektrum have to say about resolution:-

"there comes a point where the resolution of a radio surpasses the resolution of the servos or ESC, making incredibly high resolution claims irrelevant."
Last edited by saturnine; Aug 02, 2010 at 09:33 AM.
Aug 02, 2010, 09:31 AM
multirotor/hotliner pilot
saturnine's Avatar
Thread OP
From memory, I believe it's 4 pulses of 120uS (01010101) to get it into bind mode. I've seen light talk about other pulse sequences, but I've not really spent a lot of time messing with them.

Quote:
Originally Posted by Turbotronic
What do you init the sattelite rx with on bind, ie the pulse train at power up?


Quick Reply
Message:

Thread Tools

Similar Threads
Category Thread Thread Starter Forum Replies Last Post
Sold new, spektrum ar6200 complete with satellite, 6ch full range justplainold Aircraft - General - Radio Equipment (FS/W) 0 May 27, 2010 11:33 AM
Discussion Spektrum Satellite Receiver extension? Tram Radios 5 Mar 29, 2010 09:42 PM
Wanted Spektrum 6 (or higher) channel Rx w/ satellite sting35 Aircraft - General - Radio Equipment (FS/W) 3 Jan 02, 2010 12:18 PM
Sold Spektrum Satellite receiver BRAND NEW! Quad_Addict Aircraft - General - Miscellaneous (FS/W) 1 Dec 24, 2009 02:05 AM
For Sale Spektrum 6100E Receiver - Sale or Trade for Spektrum Satellite SpeedVision Aircraft - Electric - Helis (FS/W) 2 Aug 06, 2009 08:25 PM