EEVblog Electronics Community Forum
Electronics => Projects, Designs, and Technical Stuff => Topic started by: omglol on December 25, 2015, 10:15:54 pm
-
I am making synchronous rectifier for an automotive alternator (12V). I have some basic questions that (I believe) many of users here know how to answer! :)
1) I will be using CASR 50-NP (http://www.farnell.com/datasheets/1639867.pdf (http://www.farnell.com/datasheets/1639867.pdf)) current transducers to measure phase currents and DC current. The transducer has output voltage between 0.325V and 4.625V, with 0A at 2.5V. The MCU I will be using has ADC with 3.3V inputs, so I need to scale the current transducer output. I've read "3v tips'n'tricks" by Microchip (http://ww1.microchip.com/downloads/en/DeviceDoc/chapter%208.pdf (http://ww1.microchip.com/downloads/en/DeviceDoc/chapter%208.pdf) >> TIP #16) and they are proposing design with voltage divider + op-amp as voltage follower to do the job. Is this good practice? Can someone verify my design (attachment)?
2) I would also like to have digital signal for frequency of the alternator (shaft). I will be calculating this from voltage frequency of one of the three phases. I was thinking of creating 1.65V source with an op-amp, and then somehow compare phase voltage with this 1.65V. Whenever the phase voltage is greater than 0V the output should be 3.3V, otherwise it should be 0V. I also need to add some hysteresis to that. What solution do you suggest?
Thank you very much for your answers!
-
You're going to use a microcontroller to do this..??!
My crystal ball is telling me a cloudy future for you... no, literally, clouds of smoke...
The preferred approach is something like http://www.linear.com/product/LT4320 (http://www.linear.com/product/LT4320) but you have some other issues to consider:
- Maximum operating frequency is quite high (~kHz), making this part unsuitable
- Three phases needs two of these (and I doubt you'll find an IC that'll do three phases / 6 FETs in one)
- Robustness is needed. Automotive conditions are harsh: high temperature, high current, high voltage (in a load dump event), and transients from switches, relays, ignition, etc.
As heat and efficiency matter very little in the application, I would be more than happy to use bog standard general purpose junction diodes for rectification. They're usually press-fit into the alternator case itself.
(Unless you're talking about a non-automotive application of an automotive alternator.)
Tim
-
Would be easier to replace each diode with something like this (http://www.edn.com/design/analog/4368525/Use-a-self-powered-op-amp-to-create-a-low-leakage-rectifier). With faster and higher voltage opamp obviously, need to do some math to keep clear of opamp offset voltage.
-
Its a research project and no, I am not going to use anything else. Also, it will be only bench tested and not mounted in an actual car. Can anybody answer to my questions?
-
Maybe you can reuse sync rectifier that Liv used in his power supplies (https://www.eevblog.com/forum/projects/diy-bench-power-supply-psl-3604/) ...
(https://i.imgur.com/Mp1IHoK.jpg)
LTspice model is in the attachment ...
-
1) I will be using CASR 50-NP (http://www.farnell.com/datasheets/1639867.pdf (http://www.farnell.com/datasheets/1639867.pdf)) current transducers to measure phase currents and DC current. The transducer has output voltage between 0.325V and 4.625V, with 0A at 2.5V. The MCU I will be using has ADC with 3.3V inputs, so I need to scale the current transducer output. I've read "3v tips'n'tricks" by Microchip (http://ww1.microchip.com/downloads/en/DeviceDoc/chapter%208.pdf (http://ww1.microchip.com/downloads/en/DeviceDoc/chapter%208.pdf) >> TIP #16) and they are proposing design with voltage divider + op-amp as voltage follower to do the job. Is this good practice? Can someone verify my design (attachment)?
What micro are you using? Do you really care about the higher measurement range? (It's not relevant to rectification.)
-
Micro type does not matter. If I want to measure 150A I need to scale 5v to 3.3v, as simple as that ...
-
Micro type does not matter.
If the micro has an integrated input buffer there is no need for any additional buffering, the transducer already has an integrated output buffer.
-
This is a research project? What... what is your thesis? How are you hard set on this methodology? This sounds all kinds of horrible.
Tim
-
You people are aware that you can do a self-controlled active rectifier with just a few MOSFETs and resistors (and some Z-diodes, for higher voltages)?
At least for 50 Hz it's a fine solution. Gate capacitance will ruin it for kHzes, though.
-
Horrible? Why?
-
You people are aware that you can do a self-controlled active rectifier with just a few MOSFETs and resistors (and some Z-diodes, for higher voltages)?
Only for a resistive load, not going to work with a smoothing capacitor.
-
Horrible? Why?
Smoke?