Gute Neuigkeiten!
Gute Neuigkeiten!
Hi!
Jeti hat mir eben die Erlaubnis erteilt, den Schriftverkehr zu veröffentlichen.
Also dann hier die erste eMail vom 09.01.:
##############
Dear Julian,
We measured Your servo and receiver you send us. As you can see, this servo Doesn´t have a filtration of DC motor commutator and electric discharge, which are on commutator and it generates extensive range of disturbing frequency. This signal is on servo cable (impuls) and passes directly to processor and influences function of its osccilator.
Receivers R6 and R8 have a greater input capacity and filtrate this disturbing signal.
DC motor of servo probably doesn´t have filtrations capacitors which are necessary for filtrattion a electric discharge of commutator. According to this fact, this servo hasn´t got CE certification. For CE is controled intensity and spectrum of disturbing signals.
Follows informations from our measurement.
Picture 1 . Scren021.jpg
Measurement: spectrum analyzer (Agilent N9320A, 9kHz – 3GHz)
You servo is not connected.
No. 1 and 2: GSM
No. 3: Duplex 2,4 GHz
------------------------
Picture 2 . Scren022.jpg
Measurement: spectrum analyzer (Agilent N9320A, 9kHz – 3GHz)
Your servo is connected.
No. 1 and 2: GSM
No. 5: Duplex 2,4 GHz
Another peaks: noise from servo commutator
-------------------------
Picture 3 . scope_1.bmp
Measurement: oscilloscope (Agilent DSO5014A, 100 MHz, 2GSa/s)
Servo is connected.
Green: power supply ( 4 x NiCd accu)
Yellow: servo impuls
------------------------
Picture 3 . scope_2.bmp
Measurement: oscilloscope (Agilent DSO5014A, 100 MHz, 2GSa/s)
Servo is connected.
Green: peak zoom - power supply ( 4 x NiCd accu)
Yellow: peak zoom - servo impuls
The measurement is not correct, because the maximum frequncy of oscilloscope is 100 MHz.
#######################
Bilder siehe Anhang.
Hier die eMail, die ich soeben erhalten habe:
###################
Dear Julian,
Regarding your questions:
>Does the processor have some kind of watchdog implemented, which could
restart it at the occurrence of a crashed processor?
In the processor is used watchdog with next safety component (brown-out
detector, etc.). Disturbing of this servo in range of hundreds MHz
probably influenced the osccilator function so that processor worked on
high frequency than is allowable. This may be reason for failure a
correct receiver function .
>Would it be possible to built future generations of those recievers with
lowpass filters between the processor contacts, and the servo plugs?
Actual production has protection for this event. We would like to
emphasize, that this trouble was only with using servos which don´t
follow EMC.
>Would it be possible, that noise is injected into long servo cables by
strong RF noise, thus leading to the same problem we discuss?
Not suitable conduction a long wires to these servos (which generates
strong disturbing) can occur to malfunction also other equipment.
Firstly, if wires make loop (coils) around these equipment.
We agree with publication these information, but with respect to trade
concern third parties, wouldn´t be presented name of servos
manufacturer. We would like to be informed about important things in
this matter.
We thank You for your work, which you helped to improve our new
products. According your finding, we have made decision to modify output
filter by R4 and R5 receivers so that it isn´t possibly to influence
function these receivers with using thus "wrong" servos.
Thank you for your understanding. We send your servo, modified R4 and
small (DUPLEX) present back.
With regards
######################
Na das freut mich doch zu hören. Die in der aktuellen Produktion befindlichen R4 und R5 Empfänger sind also mit einer Art Filter ausgestattet.
Ein Watchdog gegen Brownouts ist vorhanden. Die Störimpulse allerdings haben wohl zu einer Art unzulässig hoher Übertaktung geführt.
Wenn also Störimpulse durch neue Filter ferngehalten werden, und gegen normale Ausfälle Watchdogs vorhanden sind, sollten alle Sorgen ein Ende haben.
Den Servohersteller habe ich ja nun schon genannt, aber ich denke, dies ist auch richtig so, um maximale Informationstransparenz und Aufklärung sicherzustellen.
Ciao,
Julian
PS: Hat Jeti da gerade einen zeitlichen Rekord von "Problemmeldung" zu "Problembehebung" aufgestellt?
