Hey,

I am trying to make a volume control system for a 70V PA system. Something that is able to split one amplified audio channel (100W-200W) into 16 audio channels (0-40W each)

I have been looking around and frankly, I can't seem to find an ideal way to do this. I did ask this question some time ago but all that was suggested was to use an Autotransformer. This method would work very well if I could manually turn the knob but I need something that can be adjusted digitally.

I have been experimenting with some high power resistors and relays to mimic a linear db reduction but that requires a lot of expensive parts (unless you don't need the 40W resistors to deal with the power transmission on 70V audio?) Plus, no one keeps the kind of stock I would need for these and manufacturers have huge wait times for them.

Does anyone have an idea that can help me out with this? Are PCB mountable (and small) autotransformers with multiple tap levels available? Do I really need 40W resistors to handle the Audio signals? Is there some other off the shelf product that can do this for me?

 

Your title says volume control but then you introduce multiple channels with digital control. In short, I can't really tell what you want. More details will get you better feedback.

Volume is best controlled at line level, not speaker level.

 

16 x 40W= 640W and cannot be handled by a 100W to 200W amplfier.
100W/16= 6.25W and 200W/16= 12.5W.

Audio/video stores sell 70V auto-transformers on an electrical plate with a knob for restaurants and homes with about 10 volume settings for each speaker. A transformer for 6W to 12W is too big for a pcb. They are not digital.

 

So think of it like this.
1. There is one 70V audio line coming into a pcb.
2. The PCB has 16 audio lines coming out.
3. Each output audio line (channel) has a volume level that can be independently set, controlled by the PCB.
4. The PCB controls are connected to a microcontroller which is connected to a webpage.
5. Values set on the webpage control the volume on each of the output audio channels.

This is basically like one of those Y splitters for headsets but with volume control for each output controlled digitally by a microC.

 

16 x 40W= 640W and cannot be handled by a 100W to 200W amplfier.
100W/16= 6.25W and 200W/16= 12.5W.

Audio/video stores sell 70V auto-transformers on an electrical plate with a knob for restaurants and homes with about 10 volume settings for each speaker. A transformer for 6W to 12W is too big for a pcb. They are not digital.

Yes I realize the amp is not big enough for 16 40w speakers. There will be multiple 4w speakers hanging on each line with a total power that the amp can handle.
Yes I know that there are autotransformers in existence and that they are too big for PCBs. I think an autotransformer by definition is not digital as well. An autotransformer with tap levels can have its output controlled with relays though. If such things exist.

 

Nobody makes a digitally controlled 70V attenuator.

Think about using a 10W amplifier for each speaker with a digital volume control at the input of each amplifier.

 

Much better, thanks.

So, the audio power is all coming from the incoming signal with no downstream amplifiers, and you need to distribute that power and control the level on each leg?

Just thinking out loud here, could a bank of MOSFETs, one per channel, be configured to act as variable resistors in line with each leg, thereby increasing the impedance and dropping volume? (Albeit in an inefficient manner.) It would be fairly easy to control such a thing. I have an article on using a MOSFET as a resistor.

 

As a hypothetical question, what would you use to power the relays?
Also how would you send a digital signal to the relays?

Since you will need some kind of power, the solution is to build a satellite amplifier/speaker that is powered from a wall adapter. Then you only need a few mV signal from the 70V PA line.

You can choose any means of digitally controlling each unit, e.g. IR remote controller.

Edit: AG beat me to it. We're both thinking the same thing.

 

As a hypothetical question, what would you use to power the relays?
Also how would you send a digital signal to the relays?

Since you will need some kind of power, the solution is to build a satellite amplifier/speaker that is powered from a wall adapter. Then you only need a few mV signal from the 70V PA line.

You can choose any means of digitally controlling each unit, e.g. IR remote controller.

Edit: AG beat me to it. We're both thinking the same thing.

I am currently using GPIO expanders on a SPI bus plus Darlington Array transistors to drive the relays off of a 12V or 24V power supply. SPI drives IO, IO drives transistors, Transistors drive relays.

Okay so I don't want to control each speaker. I want to be able to control the line level with my website. There will be multiple speakers on each of the 16 outputs and some of them will have individual volume controls.

For example: I have a multi story office building with each story on an individual audio output from this device. With this device I can adjust the volume of the speakers on the entire floor, changing their maximum volume. But there are also individual auto transformers in the offices so they can turn the speakers down below the maximum volume level set by my device if they want to. Like they are in a meeting and don't want to listen to announcements. However, the hallway speakers do not have their own auto transformer so they are set to the maximum that my device chooses.

Hopefully that made sense.

The website is actually just some webpages that appear when I connect directly to my micro but eventually I will have it setup to receive UDP or TCP packets over the network.

 

AG beat me to it. We're both thinking the same thing.

I'm not! The OP would need 16 power supplies, 16 amplifiers, and routing of 16 control lines to implement the satellite strategy, when he apparently already has the main power source and doesn't need more amplification. The satellite strategy would clearly work (as opposed to my proposal), but I just don't see it as practical.

Oops, the previous post wasn't available when I wrote the paragraph above.

 

Much better, thanks.

So, the audio power is all coming from the incoming signal with no downstream amplifiers, and you need to distribute that power and control the level on each leg?

Yes.

Just thinking out loud here, could a bank of MOSFETs, one per channel, be configured to act as variable resistors in line with each leg, thereby increasing the impedance and dropping volume? (Albeit in an inefficient manner.) It would be fairly easy to control such a thing. I have an article on using a MOSFET as a resistor.

I was thinking something like that could work but would the MOSFETs degrade the signal quality any? Can they support that amount of power and the voltage? Can you link that article so I can take a look at it? I might get some inspiration from it. Were you thinking of using some D to A converter to control the MOSFET resistance? Could I just setup two of them to work as voltage dividers?

 

I'm not! The OP would need 16 power supplies, 16 amplifiers, and routing of 16 control lines to implement the satellite strategy, when he apparently already has the main power source and doesn't need more amplification. The satellite strategy would clearly work (as opposed to my proposal), but I just don't see it as practical.

Oops, the previous post wasn't available when I wrote the paragraph above.

So yeah, you have the right idea. This is all being driven from a server rack so that many amps and that many control devices would be insane and impractical for what I need. I was hoping for something relatively small that can fit on a pcb smaller than 16.5" x 6" because it is going into something else as part of an overall device.

 

Here's the article.

View attachment vishay_fet_cvr_an FET as Voltage controlled resistor.pdf

I've been thinking about building one of these, so please post back if you do.

 

Remember that 70V line audio has a maximum peak to peak voltage level of about 200 Volts, so it's not really a practical option to use FETs as VCRs in that application.

Also, the idea of using power resistors is not really viable. Wastes far too much power.

If you really don't want to use the suggestions relating to lots of smaller power amps -and I see your reasoning why not - then remote control of tapped auto transformers, using relays to select tappings, is almost certainly the only practical way out for high voltage audio distribution systems.
The units won't be won't be small, and they won't be that cheap!

If you do find a 'sideways thinking' alternative, there's a whole public address industry waiting to buy your answer, because they haven't found another way, as far as I know... and it's been talked about for at least 40 years in my own experience... and probably for a lot longer than that....

 

...that 70V line audio has a maximum peak to peak voltage level of about 200 Volts, so it's not really a practical option to use FETs as VCRs in that application.

I had to re-read that paper to see it wasn't very helpful, but there are ways to configure a MOSFET to behave like a resistor - you just have to control the gate in between the threshold voltage and the saturation voltage. Why do you say it's not practical?

Also, the idea of using power resistors is not really viable. Wastes far too much power.

It wastes less than playing the audio at full blast and letting the speaker dissipate the heat. Power drawn from the amp is reduced when the series impedance goes up. I don't think the OP cares too much about efficiency as long as he gets the control he wants.

 

I actually am concerned about the power consumption of the design as any designer would be but it isn't a huge breaking point.

As to the suggestion of just setting up a bunch of autotransformers with steps that I can just control with relays, I am thinking of doing that. I would just have to make a separate module or enclosure for it because it is going to be big.

On that note, I was wondering if anyone knows where I can find some either chassis mount or PCB mount Tapped Autotransformers or if normal Step Down transformers would work. And if you know where any tapped versions of those can be found.

 

On that note, I was wondering if anyone knows where I can find some either chassis mount or PCB mount Tapped Autotransformers or if normal Step Down transformers would work. And if you know where any tapped versions of those can be found.

Maybe you can buy the tapped transformers from a manufacturer of the "speaker volume controls" that are made for homes, restaurants and PA systems.

 

but there are ways to configure a MOSFET to behave like a resistor - you just have to control the gate in between the threshold voltage and the saturation voltage. Why do you say it's not practical?

As you say, there are ways of configuring MOSFETs to act like resistors --although in a linear fashion, over a wide range of AC voltage inputs, I have always found to be a little tricky, without the introduction of significant distortion into the signal chain.
Never tried dealing with a 200V AC input range though. I would think it be a little difficult to guarantee the linearity of the attenuation, by maintaining the required G/S 'on' voltage over a 200 volt peak to peak input range? But I may have missed something -- if I have I'm sure someone will advise on how to do it.


With regard to the use of power resistors, rather than tapped transformers, to 'lose' the excess power, I've always found the transformers to be a much more efficient method.

The point is, you need to attenuate the signal to the loudspeaker via it's local transformer, without affecting the actual 'line' voltage.
In this case 70 volts...so you need to either tap the primary, so that it presents less of a load to the line directly, or tap the secondary, so that you are presenting less of a load to the line indirectly, via the transformer.

If you simply attempt to attenuate the 'line' voltage resistively, then your added resistor simply dissapates the power no longer 'required' by the loudspeaker, as heat in the resistor.

But as I mentioned, maybe a new 'sideways' view of the best way of controlling 70 volt line volume levels is what will come out of this thread.

After all, the 'old fashioned' way has been going for a very long time.....

 

With regard to the use of power resistors, rather than tapped transformers, to 'lose' the excess power, I've always found the transformers to be a much more efficient method.

I'm sure that's true. Resistors really don't help solve the OPs problem anyway, unless you had maybe 3 or 4 of them to represent different volume levels. I only brought it up as a conceptual stand-in for the MOSFET-as-resistor idea.

If you simply attempt to attenuate the 'line' voltage resistively, then your added resistor simply dissapates the power no longer 'required' by the loudspeaker, as heat in the resistor.

It's not really that bad, since adding resistance reduces the total power consumed by the combined speaker+resistor. At the highest point, the resistor would only be wasting just as much as power as the speaker is using and the combined power would be half that the un-attenuated speaker.

That all said, it might sound like crap for the reasons already cited. But a PA doesn't have a real broad frequency range anyway, so there's a chance? I'm almost tempted to try it with my home intercom.

 

But a PA doesn't have a real broad frequency range anyway.

A CRAPPY-sounding PA sounds as bad as a telephone or AM radio with very high distortion, no bass and no treble. Half the time you do not know what is being said because the important high frequency consonant sounds are missing.
A good PA is high fidelity so it sounds real with very low distortion and produces all the consonant sounds of a person speaking up to 14kHz.

I watch a TV show each day that is on stations in Toronto, Buffalo, Calgary and Vancouver at different times due to the time zones. I can barely hear what is said on the station in Buffalo because its sound is hollow and echoey and high frequencies are missing. Also its compressor works too hard. Speech sounds like HONK, HONKEY, HONK.
The other stations sound PERFECT.

I complained to my local TV station that the anchor people reading the news sound very muffled but outside people and commmercials sound normal.
They don't care because they are deaf and speech is 300Hz to only 3kHz isn't it? (Yes, for vowels but people do not talk in vowels. Vowels are grunts and groans).