Greetings I have two questions First, I have 3 small ISD voice chip recorder circuits, they play back 120 seconds of pre-recorded sound/voice to an 8ohm set of speakers, (Cheap desktop computer speakers) individually each ISD circuit is clear and crisp on the speakers, as I gang the circuits into the speakers, (Radio shack multi input to single output jack) the sound quality degrades rapidly. Even a single unit playng while the others are attached to the speaker becomed horrible. Any suggestions to prevent degradation in sound quality? (besides using separate speakers) how would I isolate each set of inputs to a common set of speakers? Second, Where might I find a low voltage (3 vdc) audio amp circuit? Cheers Steve
It's a PNP transistor, so everything is the 'wrong way round'. I've a feeling you'll have to either run this from a 9V battery, with its positive terminal to ground, or take the power from the negative rail & ground of a split-rail supply used by the main amplifier. Caps are in uF but C1 & C3 could be a lot smaller AFAICS. To answer the original query, one way is to use an opamp summer as the mixer, should work with a small audio IC such as LM386.
AFAICS? The Opamp Summer looks like it would work to combine "channels" like a mixer as suggested. Just need to figure out what values for the resistors. All channels are identicle in output. so would the r values be the same? or increase as mentioned on the site. Cheers Steve
here ya go: edit: well, it looks like cpemma beat me to it. that is exactly what my circut is edit2: As Far As I Can See (i think)
Thanks alot folks, seems to work for what I need it to do, three questions for you; 1. in the diagram above, should I place any isolation between each of the inputs (between the input jacks and first trio of 10K) to prevent audio or voltage leakage between inputs? 2. Momentary DPDT switch, is ganged to 3 separate voice playback modules, all 3 are trigger via a momentary contact. Should I add anything to each trigger line to isolate each module? Diode? (to prevent any leakage up the lines between modules) 3. I have a 9Vdc 500ma power supply running this project, one circuit ( LED color organ) "strobes" when that power is applied, yet functions properly when a 9Vdc battery is applied, the battery is around 45ma, might the excessive Ma source cause the interferrance? If so, how would I go about reducing the Ma? Cheers Steve
In reverse order, (3) I think is maybe due to the psu ground not being at the same potential as the other(?) ground. A battery has an isolated ground so works OK. Explain more about what power supplies are involved. If it's not that, the adaptor supply may be either too rough and need smoothing caps adding, or it's actually well over 9V at the low load. (1 & 2) It may help to put a 200nF-1uF capacitor in series with each input to get rid of any DC bias, which would be amplified. The op-amp end of each summing resistor is held at a fixed voltage (nominally ground) so interference between inputs is minimal. The link explains: One snag I can see with the 2-battery method is that the battery running the LM386 & 741 will go flat long before the other one supplying 741 alone. Using a single-rail supply for both isn't much harder. edit: now I re-read your original post, you already have the active speaker side, so the opamp mixer alone would do, you don't need the LM386. Connect the opamp output through a 220n-1uF cap to your speaker input, try that.
Using this diagram/circuit, I get rid of the distortion caused by joining all the channels. But the 10K pot has no effect on the volume, did I do something wrong? Cheers Steve
I'd like to give the single-rail supply a try to see if it changes the results I am getting. I've breadboarded this circuit, but I cant seem to get a rail-supply to work, would you mind posting a quick and dirty diagram to see where I am going wrong? Cheers Steve
First of all, check your wiring. (It seems like you reversed the + and - inputs of the opamp.) What's the impedence of your speaker? Your circuit will only work if the speaker is high impedence (500 ohms or higher).
With a single-rail amp, you provide a virtual earth at 1/2 supply using R1 & R2 and block DC at input and output with capacitors for normal opamps. The LM386 has the input bits internally and only needs an output cap. If it's an active (powered) speaker I can't see why your volume control doesn't work. With a plain non-active 8R speaker it will be either on or off and could damage an opamp. Well spotted, and I copied it...
Thanks. Changed polarity, replaced the 741, marked improvement. its working! Using the diagram I posted (with the power in the right direction), is there any way to increase the output more without using the LM386? I'm at full volume on the speakers, it's just above a normal talk level in output. Not looking for a earth-shaker, just enough to grab your attention. Cheers Steve
Increase R5. 20k will double the voltage gain, 47k around 5x, 100k 10x. Experiment, or fit a 100k log pot there and forget the other pot.
but there is a limit on what the 741 can do - driving headphones is pushing it. you could try a mosfet amp... hook the gate up to the output, throw a 22 ohm resistor on drain to +v and the 8 ohm speaker on source to ground. i would reccomend a bipolor, but it would sound crummy AFAIK
I have an LM386 guitar headphone amp using the floating ground technique, works a charm but has nasty distortion at high gain, we're talking completely oversaturated and nasty clipping at that. However at normal gain levels it sounds nice, and doesn't create too much distortion
