FTDI FT232RL USB-Serial port converter giving junk data continuously

Hi,

I know this is not related to MATLAB. But I hope it will be posted here. I have a UI which accepts data through serial port. Since I do not have a serial port I use FT232RL chip to create a virtual COM port. I did not program the chip, but am using the default settings. When I connect the UI to the virtual COM port I continuously get junk data (like A5 B5 B5 A5 B5......etc). This comes a s soon as I connect to the port. Beacuse of this I cannot give commands through the port and get the data I require. Is this a problem with the chip or because I did not program it? The FT232RL is USB powered and acts ad USB to RS232 converter.

Answer 1

[Walter Roberson]

Is this data coming through as soon as you connect the USB connector, before you turn on the device? If so then you probably have some kind of short in your USB connections or possibly in how you wired the FT232RL (did you use a socket or solder it directly?)

You have not said anything about what the UART portion of the chip is connected to. If that device is producing bogus signals perhaps because of a free-floating line with no pull-up, then that signal would get transmitted.

As we don't know anything about what is connected to the UART, we don't know whether the data is being transmitted with parity. Serial ports usually assume 8 bits no parity these days, but the transmitting device might have different ideas.

If the stream of data consisted of just A5 A5 A5 then that is a common pattern which consists of alterating binary 0's and 1's. For such patterns the baud rate setting is usually correct: if the baud rate setting was wrong then you get different relatively recognizable patterns that can often give you a hint about what the correct speed is; those patterns are, however, difficult to describe but obvious at a glance with sufficient relevant experience. (It has been too many years for me since I last needed to recognize these.)

[Sreedu]

My system is like this. I read data from a sensor through a

microcontroller. The UART interface to the MC has max3221 for voltage level shifting. The output from the max3221 IC is connected to the FTD232RL. The junk output comes when I switch on the microcontroller(even without the sensor connected to it). Is this because of some voltage difference at the pins? The MAX3221 IC gives out +9V and -9V outputs for logic level 1 and 0. But since the FTD232RL is USB powered, is the transmit and receive pins pulled back to +5V and 0V? Since the max3221 output is connected to these pins, does this affect the data transmission? I am not so familiar with the USB to UART conversion.

[Walter Roberson]

If I read the FTD232R data sheet properly, the maximum input voltage for the UART is +5V. See the datasheet, http://www.ftdichip.com/Support/Documents/DataSheets/ICs/DS_FT232R.pdf and look at the bottom of PDF page 14, which says,

"An integrated level converter within the UART interface allows the FT232R to interface to UART logic running at +1.8V, 2.5V,

+3.3V or +5V."

See also Table 15, PDF page 17, which lists the absolute maximum DCC Input Voltage ("All other inputs") as -0.5V to (VCC + 0.5V).

An input voltage down to -9V would blow the FT232R almost certainly, and an input voltage of +9V would probably be enough beyond +5.8V to blow the FT232R.

I agree that these specifications are no-where near what one would expect for RS232, http://en.wikipedia.org/wiki/RS-232#Voltage_levels, which has to be able to stand bursts up to +/- 25V and routinely operates with negative voltages. I do not understand how a device can claim to be a UART and not allow negative voltages. None-the-less The -0.5V slop margin is the _only_ place in the entire FTD232R data sheet that I find a negative voltage allowed.

I think you may need to contact FTDI's support on this. You need to confirm that either the device cannot do what you need, or that the manual is (possibly) wrong.

[Sreedu]

Thank you. I will try that.

[Bob]

My experience with the FTDI chips are that they are very robust. My guess is you are looking at "reverse polarity" where the input to the chip is "low marking", but the chip expects to see a high voltage as logic "1". If you have a scope, check this, otherwise just throw an inverter on the signal and then feed it to the FTDI. Buad rate and polarity are the typical issues. Hope that helps ...