Dear Ti
I use LSF0204pwr for dsp output gpio buffer(uni-directional) and communication out/in signal buffer(bi-directional).
I understand LSF0204 is bi-diretional multi voltage level translator in web information.
Please check if there is a problem with the circuit I designed. ( I used TMS320F28335 )
Best Regards.
Hello,
A couple notes here:
Top Application (LED Driving)
1. This will probably work, but its not an ideal setup. Note that the LSF is just a switch internally, so it does not have any sink or source current. All sink current is from the driver previous to the LSF, and all source current is provided by the pullup on the output.
2. Because this is not ideal, I would recommend going with another device that will work much better in such an application
The SN74LVC07A device is a 6 channel open drain buffer. It can be powered from 3.3V and tolerate 5V at the output. The device will be able to sink current to turn the LED on when the input is low, and it will go high-impedance when the input is high, thereby turning off the LED. This device is a much better solution to the problem you are seeing, so I would highly recommend it over the LSF0204.
Bottom Application (3.3V to 5V translation)
Doing 3.3V to 5V translation is a perfectly acceptable application for this device, but please take the following into account.
I would recommend looking at this application note to understand the LSF device operation: http://www.ti.com/lit/an/slva675b/slva675b.pdf
1. Pulldowns on the input: If you are not actively driving this line low with some other driver, the output will not ever go to 0V. If you refer to the app note linked above, you will see that the internal structure is a switch. If you have a pulldown on the input of 2.2k and a pullup on the output of 1k, your output low voltage will be about 3.44V because it is just a voltage divider. You will need an active driver on the input of the LSF to ensure that the output low voltage is truly low (it needs to overcome the 1k pullup). See the structure below.
2. Line Capacitance: If the traces are long from the LSF to the other two devices, then the rise time of the up translating signal will be dependent upon your pullup and the line capacitance. If your pullup is 1k, and your line capacitance is large, then it may rise more slowly than expected. Please keep in mind that you may need to decrease your pullup size to get more drive strength and reduce the RC value of your rising edge.
Might I also suggest looking into using the LSF-EVM to help evaluate this device for your application:
http://www.ti.com/tool/LSF-EVM
Best,
Michael
