Add an extra cell to a nimh pack

[Andy J]

3 hours ago, Nigel R said:

 

By the way - is the 350mA current at idle - or when the sticks were moved?

 

 

The 0.35amp is indeed the system idle current with no stick commands for the 4.8v pack which reduced to 0.18amp for the 6.0v pack.

Edited November 25, 2022 by Andy Joyce

[Phil Green]

Are the two elevator servos individual to separate elevators, or mechanically linked (ie paralleled)?

Linked servos need some means of sharing the load and continued control should one lock solid.

More than once I have seen two servos hard-linked together, where the slightest difference in throw/gain/linearity

will result in the two servos fighting each other with almost stall currents  🙂

This is one mechanically sympathetic way to do it, note the white clevises dont touch the middle servo arm,

if yours are linked, do you have something similar Andy?

 

 

[Andy J]

Afraid the two elevator servos are connected by separate pushrods Phil, so that is not the issue.

After a quick search on Bing found one post on Flying Giants of a person with a similar problem who cured the issue of telemetry alarms by increasing his BEC supply from 5.2v to 6.0v. So the issue is related to the RB20 not the source impedance.

The page also linked to a dedicated RB20 page by our American friends. This is 40 odd pages in length which I am halfway through. Found one post of a similar issue but unfortunately the post never got answered.

The thead did recommend updating the RB20 software to cure telemetry issues so will do that first as well as binding both Rx8 Pros with telemetry On as rhis could be an issue given the RB20 should auto switch between the two feeds.

Will post on the American site to see if the supply voltage is citical and report back unless anyone has direct experience of the RB20 and can advise further.

[Andy J]

Tried a SW update in the RB-20 this afternoon and whilst I got the impression it reduced the frequency of overcurrent alarms the system is still generating them with no Tx commands being given say one very 3 to 5 mins. Our friends in the US have yet to add much to the debate other than to advise 4.8V packs are way too small for an 8 servo set up. Given I am only driving 4 servos at the moment will ignore this advice but changing the second pack also to 6V may be the way to go if only to close down the issue of the overcurrent alarms when using the 4.8V pack.

 

[Andy J]

Still unable to identify the cause of intermittent over current alarms.

With a supply voltage of around 4.9v to 4.8v and no control demands an alarm is still being occasionally generated. Inspection of the log files shows nothing as the frequency of the recorded telemetry is far too slow to observe any spikes in the current drawn. Did wonder if there is a method of speeding up the telemetry logs but thought the simplest solution was to increase the supply voltage and upgrade the 4.8v pack to a 6v pack.

[Martin Harris - Moderator]

Would it be worth contacting the manufacturer for comment? 

[Denis Watkins]

Have been through the RB20 instructions at least 3 times now and Power Supply instructions are very Specific

With very many Do's and Dont's, less specific on Voltage, but more akin to Over Current Supply and Cut off

With many instructions of not connecting so many named sockets.

It would appear to be very easy to get an RB20 into Over Current cutoff

[Andy J]

To be honest Denis I tend to ignore the printed instructions given with all FRSky products as find most totally confusing and far from helpful. Also the printed text could do with increasing in size.

[Denis Watkins]

There are just so many Dont's, look here

 

[RottenRow]

RB-20- manual-1.pdf

 

Page 3 of the RB20 manual suggests that the overcurrent / undervoltage indication is given if the supply voltage drops below 4V. I would suspect with 8 servos connected, even in an idle state (no stick stirring) a sufficient number of the current spikes caused by the servos jittering would occasionally coincide to cause a drop in voltage sufficient to cause the alarm to be raised. The very transient nature of this would be too quick for the actual overcurrent trip to come into effect (which cuts one channel output off).

 

Any slight high resistance in the supply path would exacerbate this (such as connectors, a switch, etc).

 

Increasing the cell count by one will give you 1.3V more 'headroom' before the 4V low voltage limit is reached.

 

Do you have access to an oscilloscope Andy? Perhaps somebody in your club has one? If you are able to, monitoring the battery voltage at the RB20 might well reveal the presence of short term dips in the supply voltage, with very occasional deeper ones.

 

Brian.

[Frank Skilbeck]

Andy have you eliminated the switches between the batteries and RB20, I had one set up where the telemetry alarm went off setting the model up on the bench, I thought it was the battery but it turned out to be the new heavy duty mechanical switch, it had a high internal resistance, waggles the sticks and the voltage alarm went off.

 

BTW even when you aren't waggling the sticks the servos are still supporting the weight of the control surfaces so there will be a small load on the servo.

[Nigel R]

10 hours ago, RottenRow said:

a sufficient number of the current spikes caused by the servos jittering would occasionally coincide to cause a drop in voltage sufficient to cause the alarm to be raised

 

Mmm. There's a thing. Digital servos have very short but sharp current spikes caused by the PWM drive. Stack a few of those up and that could cause this, if the RB-20 isn't doing much filtering.

 

On 26/11/2022 at 15:59, Andy Joyce said:

Given I am only driving 4 servos at the moment

 

If you put the servos in

1) one at a time

2) gradually add more servos

 

do you get the alarm in both scenarios?

 

I'm wondering if it can be nailed to one particular servo.

 

The other thing that strikes me is that the RB-20 isn't doing quite what it should do here.

 

Sometimes, with open source kit, the customer ends up being the beta tester. C'est la vie.

[Andy J]

Given up trying to fault find the problem as now have added an extra cell to the 4.8V pack which should have cleared the issue as others have found. Also because of the MPX power connector style used on the RB-20, don't have another battery which is compatible with the RB20 to continue further testing using 4.8v by possibly removing servos one at a time.

 

One thing I did note is I thought the RB20 had a soft circuit break in each channel and if an over current was detected it would isolate that channel, but on all occasions of an alarm occurrence all 4 servos currently connected worked as normal.

 

[Andy Gates]

I did notice that when I read the instructions just out of interest, that the power requirement did seem extreme.

 

"make sure both supply the maximum continuous current of 40A and maximum peak of 120A"

 

All the 3700mAH 4 cell packs I can find are only rated to 30A peak.

[Martin Harris - Moderator]

There's similar advice on the Jeti Centralbox of which the FRSky unit appears to be a close copy.  It is to ensure that sufficient current can flow in order to trip the servo port overload in the event of several simultaneous short circuited servos.  Assuming the circuitry is as similar as the packaging, it only needs to detect in excess of 15A per port to trip and this will be easily supplied by most receiver packs for short durations - especially as there will normally be two working in parallel and multiple short circuits are unlikely.

 

Having experienced a short circuited aileron servo which tripped the overload detection in flight on my Jeti version, powered by two 1800mAh 30C 2S LiFe packs [remember that these are not aimed at being powered by ESC BECs] I can confirm both the usefulness and effectiveness of such units. There's a strong possibility that it saved the model from heat damage or even loss of control.

[RottenRow]

It's also interesting that the RB20 manual states, on page 1, "Recommended input voltage range: DC 4~8.4V (1~2s Lipo or 4~6s NiMH)" yet it has a low voltage alarm that gets triggered at 4V.

 

A 1s LiPo will only give 4.2V fresh off charge, that will drop to near 4.0V during 'normal' use, which must surely produce repeated over current / low voltage alarms (which are the same thing). I can't really see how it could be used on a 1s LiPo, even ignoring the fact that the specifications for most servos will be 4.8V plus.

 

1 hour ago, Andy Joyce said:

One thing I did note is I thought the RB20 had a soft circuit break in each channel and if an over current was detected it would isolate that channel, but on all occasions of an alarm occurrence all 4 servos currently connected worked as normal.

 

That is most likely because the alarms were being triggered by low voltages rather than high currents (and very short term ones at that).

 

1 hour ago, Andy Joyce said:

Given up trying to fault find the problem as now have added an extra cell to the 4.8V pack which should have cleared the issue as others have found.

 

Seems like a reasonable outcome; you will still have the overcurrent protection as pointed out by Martin above.

 

Brian.