6,4VDC to TTL bidirectional bus shiftting

[killabyte]

hi, i need to achieve fast bidirectional communication between two data level devices, the 1. is an open collector I/O device powered from a battery and unregulated. the batt swing between 6,4 and 5,0 VDC.

the 2. is a bidirectional port in a MCU unit. the maximum allowed input for this device is 5,3 VDC.

a 5VDC zener would be, but i guess its rise and fall time is not fast enough and will bring problems.

a diode from input to 5VDC and from GND to input would limit voltage levels but just to 5+0,7 = 5,7 V so the port will be damaged.

regards

killa

 

[itsmoked]

Sure be useful if you specified the CPU type...
Certain ones I know of I can supply solid answers to others will only be general suggestions.

Also what is the swing of your open collector device? You say it could be as high as 6.4 or 5.0V? What is the low?

To damage a cmos part or even latch it (undesirably) requires more than just voltage, it require current. If you use a schottky diode to VCC and Ground they are faster than the processor and have a lower Vd then 0.7V. Add some resistance and you won't be hurting the processor.

Also you can put reversed diodes in parallel and cut out some voltage between the processor and the other device.

 

[nbucska]

Let's mark the battery fed device "B" and the micro etc. --
which I assume is fed from AC supply -- "A".

How many lines are A--->B , B--->A and A<--->B ?
What is their speed? What is Vmin and Vmax of the
battery?

Do you need one or for mass production? Any other limitation or requirement?







<nbucska@pc33peripherals.com> omit 33 Use subj: ENG-TIPS
Plesae read FAQ240-1032

 

[killabyte]

the cpu is a motorola 68hc908gp32 working @ 5VDC , the opencollector swings between 6,0 and 0 VDC (digital signals).

the data rate is between 10KBPS and 100KBPS sometimes slow sometimes fast.

the lines are A<--->B and is just one line but later will go up to three lines: A<-->B, C<-->D, and E<-->F

it is not intended foer mass production but we will build some units.

regards

killa

 

[nbucska]

Now just one more question: What are the devices on the
"B" side with inputs coming from "A" ? Who makes them?
Are this lines really open collector or 3-state ?

It is unusual to use battery for half of the circuit. Couldn't you use a common supply?


<nbucska@pc33peripherals.com> omit 33 Use subj: ENG-TIPS
Plesae read FAQ240-1032

 

[itsmoked]

Use one of these(*) on the processor pin going from the pin to VCC. In the case of the BAS40 of course pin 3 would go to VCC.

Then put a 2k Ohm resistor between the processor pin and your I/O. This limits the current to your processor which allows the diode we just added to clamp to the 5.3V limit.

*

http://rocky.digikey.com/WebLib/Infineon/Web Data/BAS40_BAS40-04_BAS40-05_BAS40-06.pdf

 

[benta]

What am I missing?
It is stated that it's an open-collector output. So why not just connect the pull-up resistor to the positive supply of the microcontroller instead of the battery voltage?

 

[killabyte]

hi, the b side is unknown, just certain that it is an output collector and is a bidirectional one.

a common supply is not allowed, the b side device is a black box and cann not be opened, it just have two pins: gnd and dataI/O. the pullup resistance is already connected to its own supply inside the b device.

it is true open collector.

the diodes and resistance works but increase rise time, it works up to 88 kbps because parasitic capacitances, so im looking for another choice.

regards.

killa