Electronics > Projects, Designs, and Technical Stuff

Approach to design a multi-stage RLC low-pass filter?

<< < (12/12)

T3sl4co1l:

--- Quote from: kreyszig on April 04, 2023, 12:49:58 am ---Thanks, I will look into those. I have looked at p-channel MOSFETs that could handle a short circuit, and it seems difficult to find one with a ~0.1 Ohm resistance and a low VGS voltage. I would most likely need to power the amp with 10V, but then it would slow down the current limiting circuit. Since this will be connected to the MP4576 that has itself over-current, hiccup short-circuit protections and a soft-start feature, how important do you think it is for the pre-filter to have short-circuit protection?

--- End quote ---

This thread has been a refreshing exercise, but alas, now we come to the X-Y problem: what are you really doing here?  Would you like to discuss your system (inputs and outputs, blocks within)?

As I said before... if you don't need current limiting, don't bother!  Is this powered by the MP4576?  Or is that powered by it?  Which way isn't clear.  (Maybe it's both..!)  Do any other features make sense e.g. voltage limiting, reverse protection?  If you're not sure, then what is the environment -- battery, DC adapter, automotive, USB, etc.?  Most likely, you don't need it at all, and just the filter is fine.

As for EMI filtering, what's it connected to?  List all ports/connections.  Any particular levels to comply with?  EMC is a holistic subject, and can be deeply affected by what's happening completely across the system -- you literally cannot provide too much information here.  A single board (requires at least schematic and layout, and enclosure and wiring harness if applicable) can be evaluated for EMC, but that's no guarantee it will behave when placed inside a larger system.  We can't encompass everything of course; sooner or later we'll need to use a model to represent a vehicle, or mains wiring, which introduces uncertainty (but within the standards with prescribe such methods, that's simply how it is).

Tim

kreyszig:

--- Quote from: T3sl4co1l on April 04, 2023, 06:14:48 am ---This thread has been a refreshing exercise, but alas, now we come to the X-Y problem: what are you really doing here?  Would you like to discuss your system (inputs and outputs, blocks within)?

As I said before... if you don't need current limiting, don't bother!  Is this powered by the MP4576?  Or is that powered by it?  Which way isn't clear.  (Maybe it's both..!)  Do any other features make sense e.g. voltage limiting, reverse protection?  If you're not sure, then what is the environment -- battery, DC adapter, automotive, USB, etc.?  Most likely, you don't need it at all, and just the filter is fine.

--- End quote ---

The filter is to be powered by 12V DC through a ~1m cable that is getting power from one of the control ports of a HVAC air handler's PCB (the port has +12V, ground, low power +5V, RX, TX). The filter powers a SMPS that uses MP4576 (c.f. https://www.eevblog.com/forum/projects/surge-protection-grounding-and-fail-safe-biasing-for-isolated-rs-485). Since the filter itself can produce a current surge at startup, and due to the ability of the active circuit to achieve better damping than a filter without an inline resistor, I thought it was a good idea to use the proposed design


--- Quote from: T3sl4co1l on April 04, 2023, 06:14:48 am ---As for EMI filtering, what's it connected to?  List all ports/connections.  Any particular levels to comply with?  EMC is a holistic subject, and can be deeply affected by what's happening completely across the system -- you literally cannot provide too much information here.  A single board (requires at least schematic and layout, and enclosure and wiring harness if applicable) can be evaluated for EMC, but that's no guarantee it will behave when placed inside a larger system.  We can't encompass everything of course; sooner or later we'll need to use a model to represent a vehicle, or mains wiring, which introduces uncertainty (but within the standards with prescribe such methods, that's simply how it is).

--- End quote ---

The SMPS is powering a non-isolated MCU, and also an isolated full duplex RS-485 transceiver. The MCU acts as a translator between the air handler's serial protocol and the RS-485 RTU protocol. The RS-485 transceiver is to be connected (A, B, Y, Z and ground) to other transceivers, some of which have only floating grounds through their power supply. Unfortunately, I have little information on the air handler's PCB, other that I know it can provide a bit over 3W of power based on some scarce documentation I could find, and other than what I can measure and reverse engineer. I don't have a particular EMI standard to comply with. I mostly want to protect the air handler's PCB and not cause interference issues that could degrade the RS485 transmission.

T3sl4co1l:
Ah yeah, that thing.  Well, it only has to start up once, and if it works at all (source doesn't lock you out from momentary overcurrent?) it'll keep working, so I wouldn't worry about it, at least without trying it first, or with more information on the PCB.

An LC at the inlet is probably enough, so, what you've done here, and the isolation will do a good job with EMI, give or take maybe a CMC on the power or '485, or some R+C (on the order of 100R + 100pF maybe?) across the isolation barrier to dampen ringing.  Or maybe even just ferrite bead(s) on the power or 485 cables, instead of a CMC.  But most of those won't really mean anything without testing and it's probably fine without, at least until you notice problems.

Tim

Navigation

[0] Message Index

[*] Previous page