Help me improve a simple IR modem

[Kekkaishi]

Modem

I need to construct simple infra red modem. the schematic is like below.

After constructing it on breadboard, I think I got few new ideas to enhance it but I need some advice from the expert.

**broken link removed**

For case
1: I think I would lik t replace all of the component with just a normal voltage regulator like LM7805 but after a googling I found out that LM7805 with two capacitor(to regulate to 5v voltage) only produce current <1mA. So may I know is there any other component that I can use to replace them?

2: Is there any dual 2input NOR gate THT component? I only fount SMT component for this NOR gate to replace 4001B.

3. I think I want to increase current for the collector pin of 2N4401 so that I can increase the intensity of the IR LED. In order to achieve it, I found out to ways, one of them is to use bjt that have higher Beta and the other one is to increase the base current of the 2N4401. To increase the base current I need to redue the 10k resistor vaue but I need help on the calculation. May I know what other bjt that can be used which has higher beta value than 2N4401? and anybody can help me with the calculation of base current?

Thank you for all your helps.

 

[banjo]

Modem

1. This design works by stealing power from the PC's serial port. To do that it pulls power from 2 pins of the serial port. You cannot do that with the LM7805. You could replace the 100 ohm resistor and the zener with an LM7805.
2. Thru-hole versions are available, but are probably obsolete. You will find them in some of the surplus stores.
3. This is a digital circuit, so the base current is easy to calculate. It is (VCC-0.7)/10K, at least in the first approximation which is usually good enough. When Q2 is ON, Q1 is OFF which makes the calculation easy. However, there are some hard limits to the collector current of Q2. Remember that you are stealing current from the serial port, therefore you power source is limited. If you make Q2 conduct too hard, then VCC will drop.

 

[Kekkaishi]

Re: Modem

1. This design works by stealing power from the PC's serial port. To do that it pulls power from 2 pins of the serial port. You cannot do that with the LM7805. You could replace the 100 ohm resistor and the zener with an LM7805.
2. Thru-hole versions are available, but are probably obsolete. You will find them in some of the surplus stores.
3. This is a digital circuit, so the base current is easy to calculate. It is (VCC-0.7)/10K, at least in the first approximation which is usually good enough. When Q2 is ON, Q1 is OFF which makes the calculation easy. However, there are some hard limits to the collector current of Q2. Remember that you are stealing current from the serial port, therefore you power source is limited. If you make Q2 conduct too hard, then VCC will drop.

Thanks for your reply banjo. for case 3, vcc after been regulated is 5v. that's mean base current = (5-0.7)/10k = 0.43mA. Before that may I know how to find the beta value for this 2N4401 since in its datasheet beta value vary with Ic and Vce. May I know what is the resistor value that I can replace that 10k resistor so that I can have max current for collector pin and in the same time willl not burn the IR led.

 

[banjo]

Modem

The max current for the IR LEDs requires the datasheet for that part. Usually, the pulse current on an LED is quite high. The only way to kill an LED is to drive the current really high constantly.
If you continue to run from the serial port's power, I think this power source will limit you long before the LED's spec becomes an issue.
From the 2N4401 datasheet, you can use either the Beta min of 40 or look at the curves for a typical collector current and choose beta from there. From the Fairchild datasheet, 150 is a typical value. I realize that these two numbers differ by almost 4X. The reason is that in switching circuits, you drive the transistor into saturation and allow the collector load to determine the current rather than the base drive. With the circuit as it is, beta = 150 means the transistor will attempt to sink up to 64.5mA. With the minimum beta of 40, 17.2mA of collector current.
Use an oscilliscope and look at the waveform at the collector of Q2. If the collector drops within a few hundred milli-volts of gnd, then the collector drive is fine and increasing the base current will not get you any additional range.
Overdriving the transistor has its own problems. First, you waste more power in the base junction and you do not have power to spare. Secondly, the deeper into saturation you drive the transistor, the longer and slower it is turning back off.

If I was trying to tweak this circuit, I would look at eliminating the MAX232. On the transmit side, this part inverts the signal and limits it to 0 to 5V. However, you have spare gates in the 4001 to do the inversion and the limits can be done with two diodes and a resistor. On the receive side, the MAX232 again inverts the signal and then shifts it to about +-10V per the RS232 spec. However, it is widely known that on cables under six feet, RS232 works fine at 0 to 5V. Therefore, you can either use a BJT to perform the inversion, or incorporate the inversion back in the PIC. (The MAX232 consumes power running a charge pump and its internal circuitry that you could be using for the IR LEDS.)