Driving control lines with capacitance, do I need a buffer/driver?

nmz787

I'm driving a Toshiba TCD1304AP CCD with my propeller, it seems to work OK, but I wonder if some glitchy behavior might be related to the capacitance of the control lines... I've read about this before with regards to switching losses in MOSFETs, but I imagine it could cause glitching if the slew rate was too low or something else maybe. The datasheet shows drivers/buffers, but I don't know how to figure out if the GPIO are just fine for driving these CCD control lines.The lines are:
MCLK typ 10pF
SH typ 600pF
ICG typ 250pF

How can I use this information? It must be given in the datasheet for some reason. right?

P.S. the datasheet doesn't mention slew rate or rise time, but the MCLK freq is between 0.8 and 2.4 MHz, and the SH is a 50% duty cycle pulse with a max freq of 100KHz... ICG is just a once-in-a-while pulse, with edges about 5 or 10 microseconds between them.Thanks!

Phil Pilgrim (PhiPi)

The Prop pins' internal resistance is about 25-30 ohms. You can use this, along with the external device's input capacitance, to determine what the rise and fall characteristics will be and if they're fast enough to meet specs. If not, you can always interpose a MOSFET driver, which is designed to drive high-capacitance loads.

-Phil

Duane C. Johnson

Hi nmz787;

nmz787 wrote: »

The lines are:MCLK typ 10pFSH typ 600pFICG typ 250pFHow can I use this information? It must be given in the datasheet for some reason. right?P.S. the datasheet doesn't mention slew rate or rise time, but the MCLK freq is between 0.8 and 2.4 MHz, and the SH is a 50% duty cycle pulse with a max freq of 100KHz... ICG is just a once-in-a-while pulse, with edges about 5 or 10 microseconds between them.Thanks!

The MCLK typ 10pF would be of no concern.
The SH typ 600pF could be a problem.
The ICG typ 250pF maybe.

The best way to evaluate this would be use an O-Scope and measure the transition times.

A rough estimation can be made using the Prop pin source and sink currents.
Micro_P8X32A_Pin-Source.png Micro_P8X32A_Pin-Sink.png
Looks to be at least 20mA.
So Time = C*V/A = S
Lets say you need s change of 3V:
10pF * 3V / 20mA = 1.5nS
600pF * 3V / 20mA = 90nS
250pF * 3V / 20mA = 38nS
This is quite a crude calculation but should be in the ball park.

Duane J

nmz787

Thanks Phil and Duane! I should be able to check things with my oscilloscope, I'll post my results compared to the rough calculation Duane gave.