First post, by Darmok
- Rank
- Newbie
Hi all!
In order to play older games that don't support soundcards and only use a PC speaker, I connected the output of the speaker on the motherboard to the PC_SPK AWE64 input. This led to a sharp increase in noise at the output of the sound card, as a result of which I had to remove the gain at the speaker input with a mixer and increase it again if the game required a speaker. I got fed up with it and decided to do something.
A brief study of AWE64 showed that noise and interference at the PC_SPK input are inevitable due to the very simplified Creative solution. In effect, the input, with little filtering, is connected directly to the +5V rail of the motherboard, with all its powerful noise. Even worse, this input does not have a ground pin, which means that it is simply impossible to get a low noise level at this input. The good news is that the TAD and PC_SPK inputs are essentially the same input, which means that the TAD input can be used to connect a speaker while still allowing the mixer to adjust the signal level. In the absence of the TAD input, you can use the CD_IN input, and if the CD_IN input is occupied, then it is desirable to keep the ability to use the PC_SPK input, trying to reduce the noise level.
In my opinion, the only cardinal way to get rid of interference is the use of optical isolation. In order not to mess with a galvanically isolated power supply, which in itself is a source of interference, I decided to use the photovoltaic effect in a pair of LED - photodiode. The experiment showed that a photodiode loaded with a resistance of 1 kOhm develops a voltage of 0.45 V on it and has sufficient speed for our purpose.
I did not want to lose the opportunity to hear the caressing ear of BEEP at the start of the computer 😀, and at the same time I needed to ensure silence during nightly games with headphones. Therefore, I implemented a timer that turns on the speaker for 30 seconds after the power is turned on or when the reset button is pressed.
The use of the timer entailed the use of its own speaker driver, and since I added a driver, why not give it additional features.
Thus the PC Speaker Sound Device was born.
The following application modes are possible:
1. We connect only the speaker and set J1 to the Enable position. You can connect a useless switch Keylock or Turbo (if it is not used) to J1, in which case it becomes possible to turn the speaker on and off as needed. This mode corresponds to the standard use of a speaker with a peak power of 120 mW (8 ohm speaker). You can also jumper pins 1-2 of the CN5 connector and get a maximum peak output power of 315 mW. Or connect a potentiometer and be able to adjust the volume from 0 to maximum. This mode also allows the speaker to be used on motherboards that are designed to use a piezo buzzer and have a high internal resistance of the output driver, as a result of which a standard 8 ohm speaker sounds too quiet.
2. Everything is the same as in the previous mode, but we connect the output of the sound card amplifier to the AMP_IN input. This allows you to use sound cards that do not have any sound inputs at all (for example, Adlib). In this case, it is better to install a speaker with a larger diameter for better sound quality.
3. Connect the speaker, connect the sound card to the LINE_OUT or PC_SPK output and set J1 to Auto. In this mode, the speaker will be connected for about 30 seconds after the power is turned on or the reset button is pressed, allowing you to hear the start-up BEEP or alarms if something happens. In games, we listen to speaker sounds through stereo speakers connected to the sound card.
4. Everything is the same as in the previous mode, but we connect the output of the sound card amplifier to the AMP_IN input and use only the built-in speaker in mono mode.
Before making a prototype of the device, I would like to ask your opinion about it. What other useful functionality could be added? Maybe all this is superfluous and only the basic possibility of a low-noise output is needed? Criticism and suggestions on the electrical circuit are also welcome.
Thanks for support.
Update: I have made changes to the circuit based on the results of testing the device.