Hi, a few times - but not always, I've seen my spirit shake the servos in a really weird way on startup.

I'd like to know if this is normal/expected behaviour? What should I do to avoid this? The system does have an R2 buffer - but this as you can see in the video is NOT attached to the power system at the time of startup.

Jive Pro 120+HV
Spirit
BK-7001HV servos

[youtube]https://youtu.be/MKIxiD82nYo[/youtube]
https://youtu.be/MKIxiD82nYo

_________________
--
John Clayton

Hi,

no, this is not correct.

Unfortunately I know many cases where these servos (especially BK and RJX that are identical) are behaving in a weird way and causing various problems.
Manufacturer is aware of this and there is a firmware available for these servos, that should solve at least some of these issues.
Changing BEC voltage to a lower (under 6.6V) often helps, but sometimes there are shorts between enclosures of these servos and they are picking a noise like an atenna. Then one servo can affect other - yes, it is really weird.
In this case even motor beeping or touching servos can cause servo jittering (even that it is not connected to a unit but just receiver).

You can verify this with a multimeter - e.g. with Resistance measurement - one probe on the bolt of a servo, second probe on the GND from the cable.

_________________
Spirit System developer

What kind of value should I expect to be measuring with the resistance - the fact that I have some at all is presumably an indication of connection and thus bad?

Wouldn't I expect the jitter to be constant if it was a short? It all stops once the spirit initialized though.

_________________
--
John Clayton

Virtually there shouldn't be any measurable value.
If so, then in range of Mega Ohms. Mostly all classic multimeters will show just Infinite (1000) or something like this.

_________________
Spirit System developer

Ok, so I put one end on the bolt holding the servo together and another on the neg wire of the same servo right?

_________________
--
John Clayton

Yes, exactly. There should be nothing - no difference between attached probes.
Mostly the problem is that their servo enclosures are too sharp and can cut partially the cables (even that there is a rubber grommet). When the cable is replaced, the problem can be completely solved (if it was really this case).

From point of engineering, it takes some time so that everything is stabilized (especially voltage). So there might be some initial noise. But servos are too sensitive for yet unknown reason for such small differences and reacting to it.
The unit itself is outputting servo signal immediately, so you shouldn't see anything like this.
I guess that with a different servo(s) you will not observe this.

You can also try to connect these servos directly to the receiver. Also, the issue could disappear when you disconnect some servo, if this one had some issue.

_________________
Spirit System developer

I checked gnd/neg of the servo bolts against the neg bus of the heli and there is no conductivity. I wobbled the servo wires close to their rubber grommets and there is no change there.

Ill just keep an eye on the thing, as im not sure what else i can do to trouble shoot at this point.

_________________
--
John Clayton

Great,
maybe update of the servo firmware might solve the issue. But there are many users with these servos without a problem. Generally these servos are performing very well, but there are often pieces that are naughty.

Can you describe how you are powering the unit/servos exactly (how are connected cables)? Or take a photo.
Do you have some spare ferrite rings?

Lowering the BEC voltage can also fix it or at least you will know if it will help.

_________________
Spirit System developer

Servo firmware for BK servos? I didn't think they we're programmable...

On the protos v2 I run power as follows:
- Jive Pro 120+HV running at 7.4 BEC output
- Master and Slave both have ferrite rings
- master goes to the Spirit AUX port (via powerbus, see below)
- slave goes into the receiver (via powerbus, port 3, 4 o r 5 - can't recall, doesnt matter)
- MSH power bus; ESC port for master, RPM port for slave - at the back of the heli ESC port goes into spirit AUX and the RPM port leads to the rx. This power bus is apparently rated for 40A.

that's about it - anything I missed?

I haven't seen this behaviour in the last power-ups - which doesn't make me any more comfortable as that just indicates an intermittent error.

I do know the Jive 120+HV in general is sensitive to power draw within the first few milliseconds of power on (several posts related to R2 buffer packs or Bavarian Demon FBL's on helifreak will confirm this), the JIVE does not immediately init to 7.4v until these first few ms are over. From what I've seen if there is any kind of power draw on the ESC during this startup time, the ESC outputs about 3.3v (IIRC), and doesn't change until reset. I have no way of confirming this because I'm not an electronics expert nor do I have the oscilloscope to prove these theories myself (but hey, that's why we have public forums - helifreak to the rescue).

I know that doesn't bring us any closer to a solution.

Next week I'm out in Crotia with Hugo Markes and his crazy group of friends/customers for a week of flying the protos and goblin 570 - we'll see if it holds up

_________________
--
John Clayton