Hello MPS support team,
I am designing an industrial pool controller using the MIE1W0505BGLVH to isolate the analog measurement front-end (pH and ORP electrodes, 24-bit ADC ADS1220, LMP7721 femto-amp op-amps) from the digital/power side of the board. The isolated rail also feeds an LM27762 charge pump that generates +/-2.5 V for the analog signal chain.
I noticed there is some inconsistency across the forum regarding the Y-capacitor value between GND1 and GND2:
- Topic 3494 (May 2024) - https://forum.monolithicpower.com/t/mie1w0505bglvh-y-cap-value/3494: MPSNow_Yu recommended 6.8 pF for a CAN bus application
- Topic 5108 (October 2025) - https://forum.monolithicpower.com/t/mie1w0505bglvh-emi-suppression/5108/1: Krishan.FAE recommended anywhere between 4.7 nF and 100 nF for EMI suppression
- Topic 5573 (January 2026) - (link not alowed): Krishan.FAE again confirmed Y-cap addition as the solution for radiated emissions in the 30 MHz to 1 GHz range, and added that the voltage rating should match the insulation voltage of the internal transformer (2.5 kVRMS) with headroom, and an SMD package with low ESL
The datasheet itself does not mention a specific value, only shows the footprint position.
My application context:
- The product will go through CISPR 32 Class B for residential/commercial market compliance
- The product is installed outdoors near pool equipment, with potential exposure to surges from lightning, pump motor switching, and AC line transients
- There is no critical low-leakage-current requirement (the controller is not patient-connected and is mains-powered through an external supply)
- The isolated side feeds extremely sensitive analog circuitry (3 fA bias current op-amps reading high-impedance pH glass electrodes around 100 megohms to 1 gigohm), so the cleanliness of the isolated GND2 is critical for measurement accuracy
- The MIE1W is on a 4-layer PCB with separate GND planes on primary and secondary sides, following the datasheet layout guidelines, with 3.5 mm minimum creepage
Based on the forum recommendations and my application requirements, I am planning to use:
- Capacitance: 10 nF (middle of the recommended range, balances EMI filtering with leakage current)
- Class: Y1 (8 kV pulse rating, fail-safe open behavior, provides margin above the 2.5 kVRMS isolation of the MIE1W considering possible lightning-induced surges in outdoor pool installations)
- Voltage: 250 VAC rated
- Package: SMD 1206 or similar with low ESL
- Example part: Murata DE1E3KY103MA4BL01 or equivalent
My questions:
-
Is 10 nF Y1 250 VAC an appropriate choice for this kind of industrial outdoor application with CISPR 32 Class B target, or would you suggest a different value given the analog precision requirements on the isolated side?
-
Could the higher capacitance (compared to the 6.8 pF previously suggested for CAN bus) inject any switching noise from the primary side into the sensitive analog ground on the secondary side, given that the MIE1W internal switching is around 1 MHz? In other words, is there a risk that the Y-cap becomes a noise injection path rather than a noise drain at the analog measurement frequencies?
-
Would you recommend adding a second Y-cap position (DNI by default) to allow tuning during EMC pre-compliance testing, or is a single well-chosen value generally sufficient?
Thank you in advance for your support. The MIE1W0505BGLVH has been a very nice solution for our isolation needs and we are committed to using it correctly.
Best regards,
Uilter