Stereo to mono summing question.

I've been building these little amplifiers that only have 1 speaker. In the past i've always just used an 1/8" input jack with the left and the right terminals connected together. This has always worked fine for me and has never created any issues. However, i'm understanding that this is not the correct way to do it. Searching the web I see that people use 2 resistors, one for each channel, before connecting them together. This makes a sort of sense to me as I imagine the electricity takes the path of least resistance and therefor won't try to feed backward through the opposite channel. Again, i've never experienced any issues with just connected the left and right channels together without using any resistors... is this actually something that is necessary?

Assuming it is, my 2nd question is about resistor values. Schematics i've found on the web use values as low as 475 ohm up to 10k. When doing my own testing, adding resistors reduces the volume. Even using 1k resistors reduces the output volume a noticeable (and therefor significant?) amount. So, what are the lowest value resistors that i can use and be "safe"?

Thanks in advance!

There is a proper way to do this and a simple way to do it.

Firstly, directly linking the channels isn't a good idea. It makes little difference to your amplifier but consider that whatever is poducing your signal could be overloaded, possibly pushing power from one channel to the other instead of into a load.

The simple method is just to use two resistors, it has nothing to do with 'path of least resistance', it is simply to isolate the direct connection between inputs while not reducing the level too much for your amplifier to use. The values are not critical as long as they are at least as high as the source is designed to drive. Increasing the value reduces the load on the input source but also reduces the signal available to your amplifier, reducing them will highten the risk of overloading your source but drop the volume less. What you need is a compromise, I would suggest a resistance of 10% of the amplifier input impedance. Without a specification or schematic I can't tell you what that value is.

The 'proper' method is to use a summing amplifier. It completely eliminates signal passing between the inputs by connecting them to a 'virtual ground'. That means each input sees only a load, there is no signal at the other end of the load so as far as it is concerned that's the end of the line. The ground isn't real though, you create it using feedback from an amplifier. With the correct values, the amount of feedback can exactly compensate for the incoming signal and cancel it out at the summing point (the virtual ground). The amplifier output is the sum of the two signals it is trying to cancel and is therefore the 'mono' signal you need for the amplifier you build. The gain should be zero but with some adjustment of values it is possible to increase the signal if necessary.

Brian.

Simply shorting L and R works well for signal sources with medium output impedance, e.g. 150 to 600 ohm. Many line level sources have it, but not all of them. So if you are not sure about the kind of output, use a resistive combiner.

Here is more on summing amplifier circuits

https://www.allaboutcircuits.com/vol_3/chpt_8/8.html

https://www.electronics-tutorials.ws/opamp/opamp_4.html

https://www.circuitstoday.com/summing-amplifier

betwixt said:

What you need is a compromise, I would suggest a resistance of 10% of the amplifier input impedance. Without a specification or schematic I can't tell you what that value is.

Click to expand...

Brian, thanks for the information! I did not build the amp circuit this time, and am instead using **broken link removed** that i purchased on Ebay. It uses the TDA7297 chip, but I don't see any schematic information available.

The input source is either a phone, or **broken link removed** bluetooth receiver when going wireless.

betwixt said:

The 'proper' method is to use a summing amplifier. The gain should be zero but with some adjustment of values it is possible to increase the signal if necessary.

Click to expand...

I would not want any gain, as i already leave the input device's volume at about 80-85% to avoid distortion. Would this method result in 0 loss, unlike the straight resistor method? Would the summing amplifier require input power from the battery?

Thanks for the advice everyone!

- - - Updated - - -

FvM said:

Simply shorting L and R works well for signal sources with medium output impedance, e.g. 150 to 600 ohm. Many line level sources have it, but not all of them. So if you are not sure about the kind of output, use a resistive combiner.

Click to expand...

I expect the input devices to be phones. However, I am giving myself a wireless option this time, so that would change the input source to **broken link removed** thing. The description gives me the output KHZ, but i don't see anything about output impedance.

Based on the photograph of the PCB, the input connections go to the top of the volume control and it's other end is ground. It would be safe to assume the input impedance is approximately the same as the value of the volume control. Look at it closely and it will have the value marked on it, probably as a letter and three digits, something like 103A which indicates 10K logarithmic track. The first two digits are the value, the third is the number of zeroes following it, for example 103 means 10,000 = 10K.

A summing amplifier can have zero gain or a higher gain, it is set by the scaling of the feedback resistor values. It will draw some current but it should be very small, probably less than 1mA so it shouldn't make much difference to battery life when compared to the 3,000mA the main IC can take.

The bluetooth adapter should work well with the summing amplifier although to be fair, it should work with a resistive isolator as well. The frequencies the advert quotes are sampling rates, and misleading too in the context of Bluetooth communicaton. They have litle relevance to the audio you hear through the amplifier.

Brian

The E-bay amplifier is rated at 15W + 15W. Then the total output power is 30W (30 Whats, not 30 Watts).
But the supply is only 12V at 2A which is 24W. The amplifier heats with 12W so the maximum output power is only 12W which is 6W per channel with HORRIBLE clipping distortion. The datasheet for the amplifier IC shows it beginning to clip when its output is a little more than 3 Watts into 8 ohms per channel.

Audioguru said:

The E-bay amplifier is rated at 15W + 15W. Then the total output power is 30W (30 Whats, not 30 Watts).
But the supply is only 12V at 2A which is 24W. The amplifier heats with 12W so the maximum output power is only 12W which is 6W per channel with HORRIBLE clipping distortion. The datasheet for the amplifier IC shows it beginning to clip when its output is a little more than 3 Watts into 8 ohms per channel.

Click to expand...

Ouch, interesting. Well, i'm actually only using 1 of the channels to drive the single 8 ohm speaker, hence me needing to convert sterio to mono before going into the amp. Would you be able to suggest a better chip / amp for my (novelty) project? My specs:

The speaker/driver: **broken link removed**

The speaker "box" is a 5 liter mini-keg. I gather that 5 liters is something like 305 cubic inches, or roughly 5000 cubic centimeters.

Part of that internal speaker area will be displaced by a 12v UB1250 battery. I am open to other battery suggestions so long as I can make it fit. The battery will power the amp, but there will also be a cigarette lighter outlet where the keg spout used to be. Most of the time this will be used in conjunction with a 5v USB adapter for charging the occasional phone, but could also be used to provide quick power to 12v accessories while camping. Whatever kind of battery would last the longest given that type of usage is what i'm after.

Sorry for going off-topic here, but Audio-Guru is making me question my choices! Thanks for any advice you can offer!

You have a nice modern little speaker, but it is not sensitive because many speakers produce 2 or 3 times more sound for the same power input. Do you also have a nice tweeter for the highs?

The TDA7297 bridged amplifier operating from a 12V supply has an output from each bridged amplifier of about 7.5 peak to peak so the total is about 15V peak to peak. The RMS voltage is 5.3V then the real power into 8 ohms is 3.5W.

A better amplifier will have less loss so it can drive a 4 ohm speaker in a car and produce about 6.25W into 8 ohms or about 10W into 4 ohms.
I had a car that came with 2 ohm speakers and its little amplifier produced about 29W into each speaker when the battery was 13.8V.

Your speaker can handle 40W so use an LM3886 amplifier IC operating from a plus and minus 28V supply for 38W at very low distortion.

Thanks Guru. No, there is no tweeter for the highs. I only have room for 1 speaker, so I tried for something that was listed as "full range" to try to get the best sound i could from a single 5-1/4" speaker. The LM3886 requires a minimum of 20 volts, and i think i should stick to 12 volts because of space limitations. I know i can't fit 2x 12v batteries in there, and i'm not finding any 24v batteries that are small enough for my application. What amp chip would you suggest for a 12 volt system?

Years ago, Philips and ST Micro made many car radio amplifier ICs. Today most are not made anymore since the Chinese have taken over.
A few of the amplifier ICs also had a built-in voltage doubler so the output was 55W into 4 ohms since when the supply voltage is doubled then the output power is about 3.5 times more.

There are a few amplifier ICs available today that produce about 30W into a 2 ohm speaker when the supply is about 13.8V.

Recently, a guy at the park had a full size cooler with a motorcycle battery, two 8" woofers and two tweeters and it sounded very good even far away.

About 30 years ago I made a small stereo sound system for the beach that sounded much better than the "boom boxes" that everybody else had. I used a Ni-Cad battery and National Semi amplifier ICs.