View Full Version : ftdi ft232r clock

03-17-2007, 03:04 AM
does anyone know how or if it's even posible to keep the ft232r outputing a clock signal out of one of the cbus pins even if there is no usb conectin to chip? I have a design where the propellar chip is getting it's clock signal from the ft232r cbus0 pin but the clock signal is gone soon as the usb cable is un-pluged. is there anyway to keep the chipn active? would sinking a pwren# cbus pin to ground keep the chip and clock active?


T Chap
03-17-2007, 03:06 AM
power the chip from the board, not the cable

easy fix

note in the data sheet there are examples of both cases, plus switching modes

03-17-2007, 03:24 AM
it is powered from the board. my design is basicly like the dlp prop in self powered configuration. I have a dlp prop and whenever I remove the usb cable even in selfpowered mode it will lose its clock signal and stop running. http://www.dlpdesign.com/dlp-prop-dsv11.pdf

Mike Green
03-17-2007, 03:25 AM
Someone else was working with FTDI's version of the USB Propstick which uses the FT232R's clock to provide a clock for the Propeller. He had problems producing video (TV) output. It turns out that the FT232R's clock is not stable enough to use to generate a decent quality video signal. If you don't need that kind of stability, it's probably fine.

03-17-2007, 03:59 AM
I'm not planing on video but very good thing to know. this is the email I recieved from dlp design about the clock signal

"The USB chip will only drive the oscillator when it is both powered and when
it sees the 1mS USB frames from the host. When the 1mS frames stop, the IC
assumes the host has entered Standby mode and shuts down to meet the USB
spec of less than 500uA current draw in Standby more."

my question is if there is any way around this to prevent the ft232r from going into standby, or some how keep its clock output going. Has anyone had any experiences with this?

Mike Green
03-17-2007, 04:23 AM
The FT232R is required to go into standby. If you're running the Propeller off the FT232R's USB power, that will likely go away as well as the clock. Is there some requirement that you can't or don't have room for a crystal or resonator? Another option might be to use the PWREN# pin to enable the Propeller to switch to its internal clock when the FT232R goes into standby and back to the FT232R's clock when the USB interface is reinitialized. You'd need some source of power other than the USB bus since that would go away too. If you use the Propeller's low speed internal clock, the power consumption could drop way down, suitable for use with a backup battery.

03-17-2007, 04:34 AM
I was trying to save space by using the FT232R's clock and I can't use the prop's internal resonator because I need serial comunication. Looks like i'll have to adda crystal to my design. Too bad the board I designed comes tomorrow and I already need to change the design, oh well I'm new to this and it's all part of learning. atleast I whent through batch pcb and didn't spend a fortune on the pcb (just had to wait over a month for it...) thanks for the help now I know what i'm up against.


T Chap
03-17-2007, 04:55 AM
Why can't the USB frames be simulated on the Prop? You would always have to kick start the thing with a cable though, but let the Prop provide the frames after disconnect.

Just jumper a wire(use current limiting resistor maybe 4.7k) from an unused Prop pin to the USB frame in, have the Prop first sync to the frames, then AND to the frame input via the resistor. Then pull the cable

1. Monitor the frames on Pin
2. Sync to frames
3. Switch to outputting sync frames on PIN
4. pull cable

Post Edited (originator) : 3/16/2007 10:02:00 PM GMT

03-17-2007, 05:27 AM
good idea but I won't be able to kick start it through usb when in'm using it in the field. looks like i'm going to have to jumper a crystal to the prop untill i redesign for V2 of my pcb board.


T Chap
03-17-2007, 05:37 AM

I could be mistaken on this theory.

Could the Prop generate the exact clock/frames at boot up to make the IC think it is a USB cable? I don't know the speeds required for the clock, someone else may know. If it is "just a clock"", it could be produced by a number of methods, even an SX MAY be able to emulate it. If this is correct, it is a cheap solution.

If a USB device were plugged in, you would have to switch back to the cables clock/frames.

Mike Green
03-17-2007, 05:40 AM
If you don't need high speed serial communications, you could still use the internal oscillator and use the incoming data to recalibrate. You're always going to have a start pulse and a stop pulse of known logic level. If you know the first character or first several characters that you'll get, you can locate a known bit or sequence of bits. Once you have an approximate clock speed, you can recalibrate on every received character since you'll know when to expect the bit transitions.