m@x232 based RS232 & PICs

[Piccy]

Can someone please explain the purpose
of R4 on this RS232 interface example from
Microch*p. Its not for current limiting during shorts is it. I'll need a fat high wattage res for that won't I?

Its from the "caring and feeding of
the 1674 pdf" - its not explained there.

Also what effect does higher value caps
have - I want to be able to use 1uf or 10uf caps.

Thanks.

Uploaded file: **broken link removed**

 

[neuralc]

Hi Piccy,

The R4 are for current limiting, protecting in some way the PC RS232. Because of the high impedance of both input the current carried by the GND is little, so you don't need high power resistor.

Instead of 0.1 uF you can use 1uF in m@x232, I have already tried and everything works fine.

Hope helps,

NeuralC

 

[Piccy]

Thanks neuralc,

I'm not an electronics pro to figure out for my self which is why I asked.

What kind of protection? If the protection is for accidental connections / wrong wiring then I'm thinking why use a 10 ohm for current limiting (such a low value). Say you got +9 volts to that pin - thats 900mA through the res? No?

Looking at it another way if the current through the res is always going to be low because of the high input impedenaces you talked about then why use a 10 ohm res?
To limit against what value assuming you might be able to get more than 1v across
the res in an unlucky event?

Thanks for the 1uf confirmation (I'm happy) but what is the trade off, advantage for using higher values? Sorry to be a pain!
I still want to be able to use it at speeds of 115.2kbps. I haven't the hardware to test yet.

Thanks.

 

[neuralc]

Hi,

You are right, you can use higher values (such 47 ohm or more) but they put only 10 ohm, because they don’t know what cable the user connect and what is the cable length, so to get some protection (in wiring problems) and don't limit the length of the cable, they put only 10 ohm.

I normally use 1uF (on the app note from m@xim/st they use 1 uF) because it's very easy ( and much more cheaper) to find electrolytic caps with this value.

Hope helps

NeuralC

 

[james]

Piccy,

The PIC16C74 DEMO SCHEMATIC you've posted make use of m@x232A. This device has an higher slew rate, compared to m@x232, supporting up to 200kbps transfer rates.

In case of m@x232A the dual charge pump converter capacitors may be limited to 0.1uF (not polarized ones) still supporting m@x declared speed.

You've to increase their values in the case only your circuit require voltages for ausiliaries circuit (i.e. analog) of +/-10V supplied from m@x232A (from pins 2/6).
In all other cases 0.1uF will suffice. Small multilayer capasitors will be enough and cheaper.

About 10 ohms resistors, your argument about 9V and 900mA would works only if the generator (in our case the tranreceivers in PC) has unlimited current capabilities.

Luckely that's not absolutely true. The tranreceiver in PCs are limited to 20mA and virtually protected from any shorts to ground (short current of 22mA).

So, your question about the use of 10 ohms resistor come back again.

As above said the m@x 232A has an high slew rate. Its value is 30V/us m@x. So, the dv/dt is much higher of simple m@x232. In case of longer wires (let say 30m or longer) you should to consider the wire inductance to be significant. At high slew rate, that is high dv/dt, the wire inductance may have detritement effects on the signal and high ringing effects may be seen especially if the resistance of the cable is low.

To reduce this effects, a dumping resistor to ground, or couples of resistors on TX and RX are normally used. This resistors have the effect to increase the equivalent series resistance of parasitance inductance of the cable.

Obviously this is applied when a sustained trasnfer rate have to be achieved. If you need 9600 bauds, you can forget it.

Hope this help you

james