Spektrum Rx Min Voltage Study

[Pinecone]

I have stated in several threads, that IMO, some of the problems with "lockouts'" , "shutdowns", and "reboots" are due to Rxes that are more sensitive to voltage than others.

So I spent some time today checking most of my Spektrum Rxes for minimum operating voltage. I used my Mastech 3030 Variable power supply. This is capable of up to 30 volts and up to 30 amps. I made an adapter to plug into Rxes with a DVM to measure the voltage. I had at least one servo (uninstalled Rxes) hooked up and as I tested the voltage, I operated 1 channel with the servo.

I slowly dropped the voltage as I operated the servo and kept an eye on the LED. I would lower the voltage until the LED went out and/or the servo stopped working. I then raised the voltage to see how the relink went, then lowered again to confirm.

I have a total of 18 Spektrum Rxes (yes, that many), and I tested 16 of them. The two I did not test are installed and hard to get to the plugs to test.

I tested 8 AR6100s, 5 AR6100E, 2 AR6200, and one AR7000.

Spektrum specifies that the Rxes will work with at least a minimum of 3.5 volts. So I started my testing at 3.7 volts.

In my testing the worst performance (highest drop out voltage) was 3.08 volts. WELL below spec. The best Rx worked down to an amazing 2.90 volts. The AR6100s tend to work at slightly lower voltages than the AR6200s and the AR6100Es. But we are talking about a total voltage range of 0.18 volts from best to worst.

Most Rxes would stop working and the LED would go out at the same time. Interestingly, several of the Rxes would go into an intermediate mode, where they would stop working (servo not responding to the stick movement) and the LED would dim, but not go out. As soon as the voltage was raised, they started working with a good link. One AR6200 had the sat Rx dropping out (LED going out) maybe 0.1 volts higher than the main Rx.

Most of my Rxes would reconnect immediately upon raising the voltage above the threshold by a few hundreds of a volt. A few would take several seconds to reconnect. So it does seem that the some of the newer Rxes do reconnect faster than others. The DX6i manual states that Rxes made after July 2007 have quick connect where they remember the last used channels and check them for a valid GUID before scanning. BTW the Tx stayed on the entire testing time, so it was on the same two channels.

IMO, if anyone has a real lockout or reboot inflight either their Rx is resetting at a higher than normal voltage, or they have SERIOUS power problems, as their system has to be dropping to almost 3 volts.

Also, I would check your reconnect performance and if you have some the take longer, use them in planks, not your helis.

If someone has an Rx or two that they feel has had inflight problems, please test it. Or if you want to send it to me, I will test it and let you know. I would really like to see some data from suspect Rxes.

Yes, I will return it. :)

Unless you don't want it back, I will take donations. :)

[brgsstm]

Interesting, how did the 7000's fare? any difference seen as though they are powereing the satellite receiver aswel?

Tom

[Danal Estes]

Pine, thanks for some hard data! Keep it coming.

[zagiman]

I'd be interested in the 7000 data as well. Thanks for all the time and effort :noteworthy you put into doing this test.

[Squirrlybird]

Pinecone,

PM sent.:thumbup:

[Pinecone]

OK, totally revised as I actualy opened the Excel spreadsheet with my data.

AR7000 - Sat dropped at 3.02 votls, main at 3.01 volt.

AR6200 #1 - Sat dropped at 3.07 volts, main at 3.02 volts

AR6200 #2 - Both dropped at 3.02

AR6100Es ranged from 2.96 to 3.05 (one data point was recorded wrong, I will have to retest)

AR6100s ranged from 2.90 to 3.00, except for the one Rx that dropped at 3.08.

Still, all a very tight range. But The Es seems to need a bit more voltage than the top pin versions.

[zagiman]

AR7000 ( I only tested one), seemed to have both Rxes drop at the same time, the same with the second AR6200.

Do you remember what voltage it dropped at?

[Aberdeen]

The fix would be to have Isolated power source for the reciever and the Servos.
or some type of current limiter installed on the servo power buss. this can be done by using a seperate BEC to the BUSS, ( along with the ESC's BEC feeding the RECIEVER) but internal Modification of the reciever would be needed. (cut the buss tracing on the pc board, and install a lead for external BEC power)

There might be a flaw in the desing of the PC boards of the recievers that is causing high resistance, causing the Voltage drop..
one day I will look in to that..

In the meantime, my JR 10X, STAYS IN USE.


Anthony

[Mercuriell]

Great work Terry - I suspect may so called brownouts are due to signal shielding - the 2.4GHz are very prone to shadow and I think the Rx placement is often secondary to signal considerations

[Pinecone]

Great work Terry - I suspect may so called brownouts are due to signal shielding - the 2.4GHz are very prone to shadow and I think the Rx placement is often secondary to signal considerations

I agree. It is easy to point at the Rx and say it failed.

I have had one offer to send me an Rx that has had in flight problems for testing. That will be interesting to see how it fares.

[Pinecone]

The fix would be to have Isolated power source for the reciever and the Servos.
or some type of current limiter installed on the servo power buss. this can be done by using a seperate BEC to the BUSS, ( along with the ESC's BEC feeding the RECIEVER) but internal Modification of the reciever would be needed. (cut the buss tracing on the pc board, and install a lead for external BEC power)

There might be a flaw in the desing of the PC boards of the recievers that is causing high resistance, causing the Voltage drop..
one day I will look in to that..

I can see your proposed fix for the higher end setups. But easier is 1) Make sure your power supply is robust enough to handle the loads without serious voltage drop, and 2) make sure your Rxes are working to spec.

Funny thing is, there are a couple of theads on RCG about problems with the ParkZone T28. Seems many are being delivered with two digital servos instead of one. Serious glitching, lockouts, and some burned up ESCs. Adding an external BEC or replacing the digitals with analog both seem to help. And this is on 72 MHz. :)

[Steel Butcher]

Great work Terry - I suspect may so called brownouts are due to signal shielding - the 2.4GHz are very prone to shadow and I think the Rx placement is often secondary to signal considerations

Why is the "standard" installation procedure on the AR7000's, to slap the extra receiver against the CF frame? I would think that this would defeat the purpose of the redundant rx.

Several months ago someone posted an install with it on the tail servo, which is what I have been running. I have the main unit far back in the frame (450) so that the antennas are very exposed.

[Aberdeen]

Terry,
I would be intresting to know the associated BEC people used when they experienced these lockouts. I do agree lockout are due to faulty power systems causing voltage drops. And 99% of the culpruit is the high gauge wire and sub quality connectors from these ESC's / BEC manufactures use..

Align uses some crappy wire..... Kontronic and Jeti, are one of the few that uses HQ wire and connectors.. also, the connection piont from ESC/BEC to Reciever should be very clean and treated with Deoxit, or similar types of contact enhancing chemicals.

Then we have the problem of using a BEC that is normaly overated for the application..

I wonder how many people experienced these lockouts using a 6 v Nicad???
because from the quick benchtesting i have done... i would say none....:wink:

[Mercuriell]

Why is the "standard" installation procedure on the AR7000's, to slap the extra receiver against the CF frame? I would think that this would defeat the purpose of the redundant rx.


I couldn't agree more - maybe its habit coming from planks - even GF has some signal attentuation for 2.4. Personally I mount my extra receiver on the landing skid with it's antennas sticking out in the air - I think sometimes cosmetic considerations outweigh functinal ones ! Extra length extension cords are available and sometimes useful.

[Pinecone]

1) I mounted mine on the side of the lower plate so that antenna is below the frame.

2) I have some stuff coming to test various BECs for voltage drop versus amp draw. SO check back in a couple of weeks. :) And then I will be asking to borrow ESCs/BECs for testing. :)

[Danal Estes]

But...

The difference between and LV reboot and shadowing should be crystal clear by looking at the symptoms of the event:

LV reboot results in total loss of control for many (3 to 7) seconds. No controls change position, and throttle should not change either (some ESCs may shut down when getting no pulses from the rebooting RX).

Shadowing literally physically cannot be sustained in a moving aircraft for more than a fraction of a second, and regain of control should be instantaneous once good frames start coming in again. Should look/feel very different than a LV, should be just a momentary freeze, if that.

Also... It is easy to determine if a given receiver on an AR9000 is being shadowed. There are two RXs in the base unit, and a satellite (and optionally, a second satellite). The little data logger gadget shows dropped frames on each RX (called "Antennas" on the logger), frames dropped (meaning that every antenna missed that frame, on both frequencies) and lockouts (meaning that many frames in a row were lost and the RX started outputting failsafe positions).

I've flown the 9000 with logger on Carbon framed heli with the RXs in various positions. Most flights results in zero drops of any sort. Occasionally I'll see 1 or 2 drops on one antenna, meaning it was "nulled" or "shadowed". By definition, if I only see 1 drop, that antenna could only have been shadowed for 1 frame time, with is about 20 millisecond (or 1/50th of a second, if you prefer). Of course, with two drops, it is impossible to say if that was two events of 1 frame, or 2 frames back/back... so it could have been 40ms (about 1/25th of a second). By the way, I have never seen a frame dropped (all antennas missed it) and of course, never a lockout. This is a Logo 10 extreme, all carbon and aluminum frame, carrying a big honking 2:1 gearbox (big metal mass) in addition to the motor. Main RX is on upper rear shelf, satellite is out on tailboom. The main RX (two of the "antennas") is sitting on a carbon shelf, with the antennas literally resting on all the excess servo wire coiled up. Not the best looking install, if you are used to the old ways... the data logger proves this is actually fine for 2.4G stuff.

But anyway, that logger verifies that a given single antenna cannot stay shadowed, even attached to a CF heli, with large metal masses and a 6s battery, for more than a tiny fraction of a second. Even though I hover this guy nose in a lot, putting all that metal, battery, and carbon DIRECTLY between the TX and RX. I also fly it out far away in all orientations (the 2:1 gearbox gives this guy massive power and it is a wonderful 3d toy!).

So, if you think shadowing is a big issue, it is easy to experiment with an AR9000/logger and see if it really is. Given the two RXs in the base, you could even put the base in an OK place and intentionally put the satellite in various "abusive" spots and see what happens.

Based on my experiences with the 9000/logger (and just flying other spektrum setups), shadowing is not much of an issue. Even in spots where it looks like it might be... and it should be really easy to tell a shadowing event from a LV reboot event.

[Pinecone]

BTW I just tested all my AR6100s, Ar6200s, and AR7000s (AR6100Es will all be after July 2007) for relink time, and all but one of them, relink after Rx power is removed in about 1 second or less. This tracks with what the DX6i manual says that Rxes made after July 2007 have a mod to reduce relink time. Now, in testing at times, the relink would take longer, so not sure what is going on, but in most cases (probably 70 - 80% of the time) the relink was almost instantly.

I talked to Horizon (I can tell you all this because I only have one Rx to send back myself :) ) that they WILL mod any older Rxes that take an excessive time to relink. If they are under warranty, just send them back with a letter of explanation. It is a PHYSICAL mod to the Rx, but can be done to your Rx (not replacing them). And since I have one that was not a v1.2 that I sent back to have the firmware update done, I think they may also do the other mod at the same time, as it relinks quickly.

I also noticed, that the older AR6100 has a matt finish to the plastic "case" while all the ones that came as v1.2 have a shiny plastic.

And the one that I have that showed poor relink speed, also died during testing and will not link at all. No LED even after 20 seconds. So I tired to rebind and with the bind plug in, the LED comes on very weak and solid. So back for repair or replacement. So it may not be the relink situation, but just a dying Rx.

So I have had 2 out of 18 problems Rxes. 1 AR6200 that the main Rx would not bind (Horizon replaced with a new one), and 1 AR6100 that died without ever having been flown.

[Steel Butcher]

But...
Main RX is on upper rear shelf, satellite is out on tailboom. The main RX (two of the "antennas") is sitting on a carbon shelf, with the antennas literally resting on all the excess servo wire coiled up. Not the best looking install, if you are used to the old ways... the data logger proves this is actually fine for 2.4G stuff.


So, if you think shadowing is a big issue, it is easy to experiment with an AR9000/logger and see if it really is. Given the two RXs in the base, you could even put the base in an OK place and intentionally put the satellite in various "abusive" spots and see what happens.
.

Had you done any experiments with the satellite in other spots? I'd like to move it off the tail if it doesn't matter, but I don't have any logging capability. Thanks.

[JesusFreak]

If your using the Spektrum logging equipment, my understanding is that is there is full power loss, the log data is lost, so you would not see any errors. If your using some other equipment that may not apply. This is per some notes on their web site regarding a glitch with one of Quique's large planes.

But...

The difference between and LV reboot and shadowing should be crystal clear by looking at the symptoms of the event:

LV reboot results in total loss of control for many (3 to 7) seconds. No controls change position, and throttle should not change either (some ESCs may shut down when getting no pulses from the rebooting RX).

Shadowing literally physically cannot be sustained in a moving aircraft for more than a fraction of a second, and regain of control should be instantaneous once good frames start coming in again. Should look/feel very different than a LV, should be just a momentary freeze, if that.

Also... It is easy to determine if a given receiver on an AR9000 is being shadowed. There are two RXs in the base unit, and a satellite (and optionally, a second satellite). The little data logger gadget shows dropped frames on each RX (called "Antennas" on the logger), frames dropped (meaning that every antenna missed that frame, on both frequencies) and lockouts (meaning that many frames in a row were lost and the RX started outputting failsafe positions).

I've flown the 9000 with logger on Carbon framed heli with the RXs in various positions. Most flights results in zero drops of any sort. Occasionally I'll see 1 or 2 drops on one antenna, meaning it was "nulled" or "shadowed". By definition, if I only see 1 drop, that antenna could only have been shadowed for 1 frame time, with is about 20 millisecond (or 1/50th of a second, if you prefer). Of course, with two drops, it is impossible to say if that was two events of 1 frame, or 2 frames back/back... so it could have been 40ms (about 1/25th of a second). By the way, I have never seen a frame dropped (all antennas missed it) and of course, never a lockout. This is a Logo 10 extreme, all carbon and aluminum frame, carrying a big honking 2:1 gearbox (big metal mass) in addition to the motor. Main RX is on upper rear shelf, satellite is out on tailboom. The main RX (two of the "antennas") is sitting on a carbon shelf, with the antennas literally resting on all the excess servo wire coiled up. Not the best looking install, if you are used to the old ways... the data logger proves this is actually fine for 2.4G stuff.

But anyway, that logger verifies that a given single antenna cannot stay shadowed, even attached to a CF heli, with large metal masses and a 6s battery, for more than a tiny fraction of a second. Even though I hover this guy nose in a lot, putting all that metal, battery, and carbon DIRECTLY between the TX and RX. I also fly it out far away in all orientations (the 2:1 gearbox gives this guy massive power and it is a wonderful 3d toy!).

So, if you think shadowing is a big issue, it is easy to experiment with an AR9000/logger and see if it really is. Given the two RXs in the base, you could even put the base in an OK place and intentionally put the satellite in various "abusive" spots and see what happens.

Based on my experiences with the 9000/logger (and just flying other spektrum setups), shadowing is not much of an issue. Even in spots where it looks like it might be... and it should be really easy to tell a shadowing event from a LV reboot event.

[Danal Estes]

Had you done any experiments with the satellite in other spots? I'd like to move it off the tail if it doesn't matter, but I don't have any logging capability. Thanks.

I haven't moved the satellite itself, but based on the logs from the two antennas in the base RX, and based on looking at a lot of other folks installations that fly with no perceptible issues, I'd say you can put the satellite alongside the frames or whatever.

Just curious, why do you want it off the tail?

[Danal Estes]

If your using the Spektrum logging equipment, my understanding is that is there is full power loss, the log data is lost, so you would not see any errors. If your using some other equipment that may not apply. This is per some notes on their web site regarding a glitch with one of Quique's large planes.

Of course I'd loose log data in a LV reboot. I'd also know that I had no control of the chopper for 3 to 7 seconds! With a helicopter, non-control is extremely obvious; they are inherently unstable. When you see one in a nice stable hover, the pilot is making continuous minute corrections, even on a calm day. Since this is helifreak, my apologies if I'm telling you things about helis you already know... I just include that for non-heli readers.

My point with all the logging examples was that dropped frame events last about 1/25th to 1/50th of a second, and an LV reboot lasts many seconds. You aren't going to mistake one for the other, contrary to suggestions earlier in the thread that some LV reboot incidents might actually have been "shadowing" incidents.

My more subtle point is that shadowing is much less of a concern than you think, even on a single antenna setup, given that the logger shows that loosing even 1 frame on a given antenna is a very rare event, even with antennas laying on coiled up wire that's laying on carbon that's "behind" a motor/gearbox/battery when hovering nose in. So you're just not going to see shadowing events that are seconds long. Therefore, no confusion with LV reboot.

So, LV reboot and shadowing are very different, and very easy to tell apart, logger or no logger.

[JesusFreak]

Danal:

This is true in general. However, some of the newer receivers can reboot fast, per Pinecones post, so it may not be a 100% safe assumption. It seems as though sufficient Power supply is the key to success though. ;-)

Cheers!
Lee


Of course I'd loose log data in a LV reboot. I'd also know that I had no control of the chopper for 3 to 7 seconds! With a helicopter, non-control is extremely obvious; they are inherently unstable. When you see one in a nice stable hover, the pilot is making continuous minute corrections, even on a calm day. Since this is helifreak, my apologies if I'm telling you things about helis you already know... I just include that for non-heli readers.

My point with all the logging examples was that dropped frame events last about 1/25th to 1/50th of a second, and an LV reboot lasts many seconds. You aren't going to mistake one for the other, contrary to suggestions earlier in the thread that some LV reboot incidents might actually have been "shadowing" incidents.

My more subtle point is that shadowing is much less of a concern than you think, even on a single antenna setup, given that the logger shows that loosing even 1 frame on a given antenna is a very rare event, even with antennas laying on coiled up wire that's laying on carbon that's "behind" a motor/gearbox/battery when hovering nose in. So you're just not going to see shadowing events that are seconds long. Therefore, no confusion with LV reboot.

So, LV reboot and shadowing are very different, and very easy to tell apart, logger or no logger.

[th3tick]

trextuning appears to be doing something similar here (http://www.trextuning.com/blog/?p=197#more-197).

Their results are a little different - higer-voltage shutdown observed, and no apparent knowledge of the reset time change ;)

It is interesting that their conclusion appears to be that the ESC's BEC must be providing no power for ~1 second on occasion...

[Azlum]

This stuff is scary :-\

[CatchmyCorsair]

Does every material mask/shadow 2.4Ghz signal? Does it matter if the frame is plastic, CF, GF?