Stereo to mono, is it so difficult?

[ms660]

Cheapo mono amp, cheapo PSU if needed to power it, sum existing stereo speaker outputs with resistors that have a value of at least 10x the nominal output impedance and feed into the cheapo mono amp, no internal wiring for matey to do, just plug it all in and go, and matey can adjust the mono volume to match the existing system volume or to suit what's needed.

Basically as said in Post number 3.

Lawrence.

[mhennessy]

Originally Posted by GMB

I searched the net for some example circuits and what I am seeing is various designs of bridge output stages that are not so symmetric, the two parts interact in various ways.

Can you link to some examples? I'd be very interested to see a design that takes that sort of risk in a car radio

[Sound Man]

Would this suit your needs - KEF-Audio-Stereo-to-Mono-Transformer

David

[GrimJosef]

If the transformer looks like it would be useful then it might be worth moving quickly. KEF have discontinued it and relatively few retailers have any stocks left.

Cheers,

GJ

[Boater Sam]

This is more like what I was expecting to find. I will suggest it and that he gets a move on.
Is anything along the same lines made by any other firm?

[Herald1360]

Since the TX effectively operates at low impedance, it wouldn't be too onerous to handwind your own- the turns required will be <<the number for a normal valve OP TX primary, more like its secondary. The "hyrid combiner" version with its single centretapped winding looks easiest. It reminded me of the ones in the whiffletree arrangement inside the 8 module RF combiner in Racal's TA1860.

[GrimJosef]

Since this is wound on a single core I'm still struggling to see how, unless the inter-winding coupling is very loose and inefficient, the transformer is not tending to force the signal in all three windings (apart from DC isolation) to be the same. In which case we're back to having either L or R at the input to the amp's channels but L+R at the output. Or am I missing something ?

Cheers,

GJ

[kalee20]

I agree with GJ - unless it is as I put in post #13. But that is dependent on having a single common return.

If both outputs are bridge, then a pair of windings on the same core (with the mono speaker hung between the centre-taps) will work, as long as the bridge outputs are truly equal and opposite. Otherwise, (the general-purpose case) two transformers will be necessary, each with a centre-tapped single winding - unless I have missed a configuration!

[Trigon.]

If the car radio/cd was built in the last 15/20? years then it will almost certainly have bridged outputs and KO's solution in post#2 will work just fine.

However a solution that will work for any output configuration is a simple 1:1 transformer on one channel, with it's output in series with the other channel. I'm sure any slight difference in response will be insignificant through a waterproof speaker outdoors on a boat!

A more symmetrical but frankly overkill version could have a 1:1 transformer on each channel with their outputs connected in series.

And no need for spendy audio transformers - I've just tried it using the secondaries of a 0-25 + 0-25 30VA mains toroid as a 1:1 and it sounds remarkably good!

Cheers

[Boater Sam]

A Kef transformer has been ordered, I await the fitting and outcome. It seemed the best option for the guy as he has no knowledge.
Would some knowledgeable member who has referred to bridged outputs on current car radios/cd players please explain to me what it is all about.
Are the 2 stereo outlets not from 2 separate amplifiers any more?
Sam.

[GMB]

It is nothing to do with stereo, so each channel has a its own amplifier.

In the old days the amplifier would have fed its output to one of the LS terminals with the other terminal connect to ground.

To boost the output, the so-called bridged output is that each channel has two amplifiers, so one feeds one LS terminal and the other inverts the signal and feeds the result to the other terminal. This doubles the voltage that the LS sees.

This gives the simplistic view of it:
https://en.wikipedia.org/wiki/Bridge...led_amplifiers

The snag is that real circuits have feedback connections that do not treat the two halves the same.

[Trigon.]

A typical car audio four channel ("4x40W"..!) output stage in a single IC:-

[mhennessy]

Originally Posted by GMB

The snag is that real circuits have feedback connections that do not treat the two halves the same.

Sorry to repeat myself, but could you provide links to some of the examples you mentioned in your earlier post?

Thanks in advance,

Mark

[mhennessy]

Originally Posted by Boater Sam

A Kef transformer has been ordered, I await the fitting and outcome. It seemed the best option for the guy as he has no knowledge.
Would some knowledgeable member who has referred to bridged outputs on current car radios/cd players please explain to me what it is all about.
Are the 2 stereo outlets not from 2 separate amplifiers any more?
Sam.

Hi Sam,

If the radio in question did use bridging - as they mostly do - then it meant that converting to mono was just a matter of making the correct connection - no extra parts needed at all.

I don't have a car stereo sat around at home, but I do have a hi-fi amplifier that I made a few years back that supports bridged working. If you need convincing of the principle that Karen put forward, I'd be only too happy to put it on the bench and hook it up.

When you're stuck with a fixed power supply rail, increasing power delivery to the loudspeaker is a matter of lowering the impedance (hence car audio is 4 ohms), and then bridging to double the available voltage swing (theoretically quadrupling the power output).

Another advantage of bridging is there is no need for an output cap. These are large - typically 1000uF or (hopefully) more - and so occupy significant board area. Whereas the chips are tiny.

Another option would be to use a step-up audio transformer at the output to boost the power, but I haven't seen that done for a long time! The ones I saw were not very good because manufacturers cut corners on the transformers.

Today, the big car amps use a switched-mode power supply to turn the incoming 12V into much larger supply rails - meaning no need to bridge to get the desired power output - and they are usually symmetrical rails, so no need for an output cap.

Now, I'm not completely up to date with modern "head units", as I believe they call them, but it's certainly not impossible that they've started to squeeze switched mode DC-DC converters into the radio/CD unit. If so, then they might not use bridging. If this is the case, Karen's suggestion won't work - that is why I asked for the make/model of the unit, as it might have been apparent from the instruction manual or sales guff...

Hope that helps in addition to the other explanations,

Mark

[Boater Sam]

I know only that the new set is a Kenwood car radio and cd player, I thought they were all food mixers.
I will update when the guy has fitted the transformer, hopefully he will be sorted.

This new fangle stuff with SMC and chips is beyond my patience, I prefer the stuff I know and can fix.

[stuarth]

If the KEF transformer is just a 1:1:1 transformer with 3 separate windings, then won’t it do the same as connecting the two outputs together, with the current limited only by the coupling in the transformer? I can see a risk of damage to the power amps.

The single 1:1 transformer described in posts 9and 29 sort of does the same, except that the speaker impedance is added to the path between the amplifiers, so unless the speaker impedance is too low (it should be twice that of each speaker in a normal stereo setup), the current in each amplifier is only that seen in a normal setup, and no damage should occur.

Stuart

[Boater Sam]

Oh I do hope not Stuart, it will be my failure if it does.
At least using this transformer will guarantee DC isolation.

[boxdoctor]

This gets complicated if each output is a true bridge circuit, as really, in practice, nothing can be connected to anything from the other channel as they are carrying dissimilar signals. (Remember it's stereo) There may even be some standing D.C. on either or both outputs due to the basic design and/or mid-point offset inaccuracies in each amp.
I would think that to be sure of non-roasting of the amps all output connections will have to go through D.C. blocking capacitors before they go to anything else.
Then a simple resistive mixing network can be used to generate a mono signal and feed it into an amplifier of suitable output for the appliction.
Most solid-state amps do not require a low impedance load to operate properly, and the circuit can be designed to work in the low Kohms range, thus enabling film type capacitors in the 1 or 2 uf range to be used. If a low load impedance is required, an 8 ohm resistor can be fitted across each amp's output directly. Bear in mind that the external amplifier will have the input earth side at D.C. earth, too. Tony.

I also prefer the KO solution, elegant and simple. There well may be a bit of DC offset much the same as using the bridge output conventionally. No worries there.

[mhennessy]

Originally Posted by boxdoctor

This gets complicated if each output is a true bridge circuit, as really, in practice, nothing can be connected to anything from the other channel as they are carrying dissimilar signals. (Remember it's stereo)

Well, the dissimilar signals is the whole point of Karen's suggestion.

Let's take just the front speakers for simplicity. Take the front left speaker. You have +L and -L. For the right speaker, you have +R and -R.

If you connected a speaker between +L and +R - which is all you can so with a non-bridged amplifier - you end up with the difference between L and R - in other words, the stereo content (the "S" signal). This is not what we need here, but used to be commonplace in receivers back in the '70s that had pseudo-quadrophonic working - it involved little more than wiring an extra pair of speaker sockets that were fed from +L and +R (perhaps via a low-value power resistor as the rear speakers were generally nearer the listener). That worked - it still works today (see "Haffler"). I remember playing around with it when I was a student, which involved a good friend bringing his Wharfedale Diamonds into my room for the evening. He graduated as a Tonmeister - between the two of us, we knew some audio basics

Of course, we don't want L - R; we want L + R. As we have -R, we can connect the speaker between +L and -R, and that gives L - (-R), which is L + R.

In terms of doing this in practice, all power amps will share a common bias point, so there will be no DC offset issues (and if anyone is worried about this, it takes 5 seconds to check with a DMM, and not much longer to add a DC blocking cap). The amps won't mind in the least about doing this - remember that we're connecting to the rear outputs, that aren't currently connected to anything. Even if rear speakers were present, you'd still get away with it providing you don't play heavily compressed material at full volume (the IC might enter non-destructive thermal-shutdown).

In short, it will work if the radio in question is like the vast majority of car radios. It's a shame we don't have the model number, because if we did, we'd probably be able to give a 100% definitive answer.

It's strange that so many people have spoken out against this as a solution, given how obvious and simple it is. Perhaps that's precisely why - maybe it just "feels" too easy, somehow. Or perhaps there is some misunderstanding of how bridged output stages actually work?

But the same basic principle is used in countless places where stereo to mono conversion is done. Or indeed, stereo to M+S, and M+S back to stereo (M&S working is sometimes used in professional audio). More here: https://www.soundonsound.com/sound-a...-actually-work

It works; no additional components needed. Of course, there might be "outlier" examples of power amplifier circuits where this isn't practical, but no-one has brought one of these to the table yet...