Other Parts Discussed in Thread: MSP430F6736A, MSP430F47183,
Regards,
Dear TI team. This is regarding enquiry no: CS1713050 for migration to new chipset of MSP43067771a/ msp430f6736a for 1 PH energy meter.
We are planning to migrate to new chipset of TI, MSP430F679991A from the older one(MSP430F47183) which we are using currently.
We have performed a comprehensive comparison of the chipset by integrating the MCU in our 1 PH energy meter. Hardware type test as per iec.62052-11.2003 and equivalent Indian Standard IS15885/IS16444 are performed on the two chip set. Our 1 PH energy meter essentially have two PCBs one is SMPS board and other is MCU board.
For testing purposes same SMPS board is used and the MCU cards are tested one by one.
Electrical sub assemblies of shunt and CT current sensor is same for both the variants. We are using Shunt as current sensor in Phase current path and CT as current sensor in neutral current path.
Our observations are:
01) MSP430F47183 series MCU card passes all the tests. All EMI EMS test and 35KV ESD tamper test on the energy meter's Phase and Neutral input. Meter with this MCU card is immune in EFT Test (IEC61000-4-4). And is also not perturbed in 35 KV ESD test.
35KV ESD test is performed by Indian utilities on the Terminal inputs(P_IN, N_IN, N_OUT, P_OUT) pins for the duration of 10 mins. And meter should not get hang during this test or if it gets hang must come out from hang state by its own. Our energy meter integrated with MSP430F47183 series is immune to this test with the existing shielding arrangement.
02) 1 Ph energy meter using new chipset MSP430F67771A was found to fail in EFT 61000-4-4 test. During this test there should not be any reset or hang and meter accuracy should lie between +/-4%. This chipset gets reset during the test. We than replaced the shunt sensor by CT in the phase current path and than the meter passed in EFT test. Hence we request the TI team to give us the solution where we can use shunt current sensor in energy meter. This is sometimes mandatory requirement from Indian Utilities. As CT current sensor provides the galvanic isolation the conductive EFT noise was not effecting the anti aliasing filter and hence accuracy was also not perturbed. We highly doubt on the common mode conductive noises disturbing the MCU from Shunt sensor to anti aliasing and than to ADC pins of controller. New chipset is very sensitive to noises as compare to older one, MSP430F47 series.
03) In 1PH energy meter integrated with new chipset MSP430F6771A series , 35 KV ESD test was performed at the meter terminals. The meter gets hang immediately even though the discharge gun is 1 foot away from the meter terminals. The RF noise is effecting the MCU instantly. To resolve it we have used 10nF decoupling capacitor at NMI pin of MCU instead of 2.2 nF decoupling capacitor. This problem of hanging immediately due to RF noise was atleast resolved.
However when 35 KV conductive ESD is applied at the input pins(Phase and Neutral) of energy meter. It is observed that after 2 to 3 min the MCU again gets hang. The effect of noise is even worse if SHUNT is used as current sensor in phase current path. 35 KV ESD test is to be performed for the total duration of 10 min.
New microcontroller is effected (hanged and reset both) during this test. We have configured NMI pin to reset functionality.
We request TI team to give us suggestions to integrate new MSP430F67771a series effectively such that both EFT and 35KV ESD tests are passed even with SHUNT as current sensor.
I have read TI docs of isolated shunt sensor but we don't want to go for that. As existing MSP430F47183 microcontroller is completely immune to both the tests, we aspire same performance with new chipset of MSP430F67771a also.
