2015-07-21 11:55 AM
I need some help on a drive controller that I designed that uses two L6482's. The problem
is sometimes while running, one motor will shut off. Reading the status register shows an over
current condition. The testing current is operating at 3 amps but will go to 5 amps RMS in the end
motor application.
I am sure it is a set up problem with the gate charge and rate. I am operating the system in !/4 step
and the maximum motor shaft speed is 7 rev/second. 800 steps for one revolution. The over current
condition is while at 3 rev/second.
The output FET's are ST's STP40NF10L with a typical Qg = 46nC, max = 64nC.
Setup is as follows:
System voltage is 72 volts from a switching power supply.
GATE DRIVING:
IGATE = 32mA
TCC = 1750nS
TDT = 500nS
TBLANK = 500nS
BOOST = DISABLED
VCC = 15 V
UVLO = 11 V
PHASE CURRENT CONTROL:
TON MIN = 4uS
TOFF MIN = 21uS
TOFF FAST = 10uS
FAST STEP = 16uS
TSW = 48uS
Predictive current = ENABLED
Motor Specs:
R = 0.65 ohms
L = 5.4mH
Rated amps = 5.6A
Torque = 1260 oz-in
Bi-polar parallel winding connection.
Two Observations:
1) When I increase TCC by one increment, the motor will not run and is in a short circuit condition. I have changed
IGATE to 64mA and TCC to 875nS and the motor will not run. It is in a over current state. Not sure what to do here.
2) The 220uF 100V filter caps used for each drive system runs hot. ST did not provide the ESR for the filter caps on the
design information. I installed a United Chemi-Con filter cap with an ESR of 0.023 ohms, 390uF 100V. The filter cap ran
cool but I noticed the over current fault happened more frequent. Any ideas?
Please help me as I am not clear what the next step to try.
------------------------------------- Hello, Layout for your review. The power comes in on the top right of the board. M1 is on the right side and M2 is on the left side. The output for each motor controller are on the bottom layer and are on the edge of the board. This is a 2 oz copper fab. I have connected extra power and ground wires from the header to the filter cap of M2. No change in behavior. #l6482h2015-07-23 06:15 AM
The configuration for the driving of the MOSFETs seems ok.
For the current control algorithm I suggest you to increase the TON_MIN at 5 us (it must be > than 2 x tCC + tDT + tBLANK). The layout shows some critical point, in particular for the long nets connecting the IC with some MOSFETs. I suggest you the following: - connect a 220 nF ceramic capacitor between each HS MOSFET drain and the ground in order to filter the current pulses form body diode recovery - Reduce the parasitic inductance between LS MOSFET sources and shunt resistors (avoid vias) What is the OCD threshold value? Do you have some screenshot showing the output voltage during the commutations?2015-07-23 09:19 PM
Hello,
Some additional values are: OCD_TH = 17 hex or 37 amps TVAL_RUN = 13 hex or 3.45 amps I have included some screen shots and pictures for you. I suspect that I exceeded the formula for TON_MIN as you have provided below. I will try all in the next few days and report back. Regards, Robert ... ________________ Attachments : 72V_BATT_CUR.jpg : https://st--c.eu10.content.force.com/sfc/dist/version/download/?oid=00Db0000000YtG6&ids=0680X000006HzHN&d=%2Fa%2F0X0000000bJA%2FUmoJFcNWR3h6aBQneGVD1jtU1RPbHtc6zpq12ZvLKoA&asPdf=false72V_BATT_VOL.jpg : https://st--c.eu10.content.force.com/sfc/dist/version/download/?oid=00Db0000000YtG6&ids=0680X000006HzHD&d=%2Fa%2F0X0000000bJ9%2F0_KXa9ZcGPuCxdU5vlz7yjFjkZwaUNBUBC8RD_aWY5U&asPdf=false72V_PHASE_CUR.jpg : https://st--c.eu10.content.force.com/sfc/dist/version/download/?oid=00Db0000000YtG6&ids=0680X000006Hz7s&d=%2Fa%2F0X0000000bJ8%2FDbiiythq9TzFm3CnvvjbJN1_V.2pBRd75U6InR2whl0&asPdf=false72V_PHASE_VOLTAGE.jpg : https://st--c.eu10.content.force.com/sfc/dist/version/download/?oid=00Db0000000YtG6&ids=0680X000006Hz3J&d=%2Fa%2F0X0000000bJ7%2FqFL_AcSKNh_GncSQuU7PYgZZA.nQWHMXY5ceUyvsKto&asPdf=falseDSCF2001.jpg : https://st--c.eu10.content.force.com/sfc/dist/version/download/?oid=00Db0000000YtG6&ids=0680X000006HzGG&d=%2Fa%2F0X0000000bJ6%2Fk3FsAmNXE23yfa.8Xrdo2RcbNHmwNBhXvIR1gdunmGk&asPdf=falseDSCF2002.jpg : https://st--c.eu10.content.force.com/sfc/dist/version/download/?oid=00Db0000000YtG6&ids=0680X000006HzEj&d=%2Fa%2F0X0000000bJ5%2FnstiOi16CQd9ELX0UK5YB5MCqgUHIS6G4uTD13X0qk4&asPdf=falseDSCF2003.jpg : https://st--c.eu10.content.force.com/sfc/dist/version/download/?oid=00Db0000000YtG6&ids=0680X000006HzDr&d=%2Fa%2F0X0000000bJ4%2F.zWOjokW.Xu3ZMorSVbfZjQoBbf._QKhaScKqDk5xSE&asPdf=false2015-08-02 09:52 PM
Hello Enrico,
The TON_MIN was the key to fixing this problem. I increased to the value from your equation and the situation was better. It would still over current but infrequent. I increased TON_MIN one more increment to 5.5uS and no problems. I think this is solved. - Thanks. I added at you request 220nF caps on the high side drivers. I have included some shot for you. I can't see any large difference with the caps but it may be helping. 4 images: VS filter cap surge current with high ESR cap and new low ESR cap, Phase voltage without 220pF cap and phase voltage with new 220pF cap. Regards, Robert ... ________________ Attachments : HIGH_ESR_FILTER_CAP_CURRENT.JPG : https://st--c.eu10.content.force.com/sfc/dist/version/download/?oid=00Db0000000YtG6&ids=0680X000006HzGo&d=%2Fa%2F0X0000000bJ2%2FOAe2iLuBTCgKxds8xC0TJlaUKWnXallJgFIzXlNiiGs&asPdf=falseLOW_ESR_FILTER_CAP_CURRENT.JPG : https://st--c.eu10.content.force.com/sfc/dist/version/download/?oid=00Db0000000YtG6&ids=0680X000006HzGe&d=%2Fa%2F0X0000000bJ3%2FD4BzD5zAfq18BfW4SsQTLZAYIlf6njjRKrA72gwMY.A&asPdf=falsePHASE_VOLT_CAPS_ADD.JPG : https://st--c.eu10.content.force.com/sfc/dist/version/download/?oid=00Db0000000YtG6&ids=0680X000006HzGU&d=%2Fa%2F0X0000000bJ1%2FOgi9KKWAGl8kdySrUZ9LmcEEgNROCT8E.HO_0uRfSgA&asPdf=falsePHASE_VOLT_NO_CAPS.JPG : https://st--c.eu10.content.force.com/sfc/dist/version/download/?oid=00Db0000000YtG6&ids=0680X000006HyvJ&d=%2Fa%2F0X0000000bJ0%2Fdjl5NUtI6BEtRuAVMed7DV65DJzrJnwsASreRk1SzCE&asPdf=false2015-08-03 03:09 AM
Hello Robert,
Glad to know that the application is now operating as expected. In this case the current probably was out of control and the OCD was triggered. The 220 nF cap should help filtering the noise caused by the turn-off of body diodes. In order to evaluate the effectiveness you should look closely the rising and falling edges of the outputs. Kind Regards Enrico2015-08-26 09:45 PM