Well, you could use the p-channel MOSFET to disable the relay (thats shorts the speaker) after power on. That means the relay is turned on with power on and the off when RESET is gone. This would require 200mW only during RESET. The question is: what is faster? The plop or the relay?

Another way to avoid the turn-on pop is to use an amp that doesn't generate one. The TDA7052 amplifier uses a bridge output (Both speaker leads return to the IC, neither to ground). At turn-on, the DC level on both output pins rise together and so there is no turn-on pop in the speaker. It also has a greater output than the LM386, is designed for battery-powered equipment and will function down to a 3V supply and a single supply decoupling capacitor is the only external component needed. It comes in an 8-pin package. I have used this amplifier many times in 5V systems and it is an excellent performer and very easy to use.

Be warned - the TDA7052 has a higher gain than the LM386 so you may need to increase your resistor values to avoid overdriving it.

_________________
Shift to the left,
Shift to the right,
Mask in, Mask Out,
BYTE! BYTE! BYTE!

Sorry, late to the party on this one. I haven't read every post so I'm not sure if this has been ruled out yet. But why not consider using the Propeller from Parallax?

AY-3-8910 sound chips are still easy to find but you mentioned the 3V3 issue. So, if you are OK with using a micro-controller, the Propeller would be hard to beat as it has almost perfect emulation of the AY-3-8910, SN76489 and the SID chip. All using ONE cog. You could devote another cog or two to handling I/O and get 4-6 SID chips, for example, quite easily.

Plus, the Propeller is a 3V3 device.

_________________
Cat; the other white meat.

All the suppliers I've looked at list the TDA7052 as either end of life or obsolete.

_________________
Want to design a PCB for your project? I strongly recommend KiCad. Its free, its multiplatform, and its easy to learn!
Also, I maintain KiCad libraries of Retro Computing and Arduino components you might find useful.

One we're using at work is the TDA8541 which has a soft-start, and you can control the ramp-up time with an external capacitor. It has a bridged output, so you can't ground either side or make it common as with a 3-conductor stereo jack.

The LM386 can put out a half a watt or more but definitely not at 5V. It needs a much higher Vcc to do it. One of the common loads we drive in our products is 300Ω and one side must be grounded, so running it at 5V is not an option in that case. We have other circuits as well that won't work at 5V, partly because of standards we're left with in aircraft communications from the 1950's or maybe even earlier (which we don't have any control over), so we run those at 12V. If it's battery-powered (like from a 9V battery which has to work down to about 5V), we use a switching regulator to kick the voltage up to 12V.

_________________
http://WilsonMinesCo.com/ lots of 6502 resources
The "second front page" is http://wilsonminesco.com/links.html .
What's an additional VIA among friends, anyhow?

I found something that could be used in a bridge configuration. Usually bridge configuration is used to boost the power (doubling of voltage and current). But you don't need to do that. And as the speaker is wired between two amps outputs there is no need for a coupling cap. No cap no plop.

Whilst looking through some old magazines (1968 to be exact), I found a cure for the turn-on thump created by IC amplifiers.

The thump is caused by the speaker coupling capacitor charging through the speaker.

To cure the problem, connect two capacitors in series across the power supply. Each capacitor should have a value half that of the normal speaker coupling capacitor. Then connect the speaker between the IC output and the junction of the two capacitors.

At turn-on, the capacitors charge in series from the power supply and no charging current passes through the speaker if the two capacitors have the same value. In practice this greatly reduces the turn on thump.

_________________
Shift to the left,
Shift to the right,
Mask in, Mask Out,
BYTE! BYTE! BYTE!

Remember that there's unbalanced preceding circuitry feeding the power amplifiers' inputs which will also cause a pop. A bridged amplifier shouldn't cause its own pop, but it will faithfully amplify the pop fed to it at its inputs, unless there is some kind of power-up mute feature.

_________________
http://WilsonMinesCo.com/ lots of 6502 resources
The "second front page" is http://wilsonminesco.com/links.html .
What's an additional VIA among friends, anyhow?

You only had to look upthread, Paul!

You've described it nicely, though.

_________________
In 1988 my 65C02 got six new registers and 44 new full-speed instructions!
https://laughtonelectronics.com/Arcana/ ... mmary.html

A problem with that is that the power supply may not be very quiet. Even if it's a regulated linear supply, the regulation may not be as quiet and tight as you'd like, and you may be referencing your speaker output to signals or noise from other circuits. For the same or less board space though, you can use two 386's per channel in a bridged configuration (eliminating the output capacitors) and connect the - input of one to the + input of the other and vice-versa. You'll get four times the maximum output power. It also takes care of the problem of the power amp itself producing a pop, but not the circuitry that precedes it and feeds its inputs.

_________________
http://WilsonMinesCo.com/ lots of 6502 resources
The "second front page" is http://wilsonminesco.com/links.html .
What's an additional VIA among friends, anyhow?

Oh yeah! Doh! (Feels somewhat stupid now!)

I saw the idea and thought it would do for this thread. Quite forgot that it had been posted already. Silly me.

_________________
Shift to the left,
Shift to the right,
Mask in, Mask Out,
BYTE! BYTE! BYTE!

I've ultimately decided not to worry about the plop. Its not ideal by any means, but board space is limited (as is my wallet!) so they'll just have to cope with it. Perhaps an improvement for a Version 2.

---

However, I've been having trouble picking a speaker for the circuit... I've found This One, What do you think of it? I know i'm not going to get much out of it in terms of Bass, but that's unavoidable for a tiny speaker of its ilk. I could add a bass-boost circuit (one is shown in the LM386 datasheet) easily enough, but I'm concerned that then I might get clipping. I think the 1M mixing resistors we're using put it near clipping already?

_________________
Want to design a PCB for your project? I strongly recommend KiCad. Its free, its multiplatform, and its easy to learn!
Also, I maintain KiCad libraries of Retro Computing and Arduino components you might find useful.

Check out the data sheet. The response falls off below 2kHz. Any sound played through this is going to sound very tinny and lacking not just in bass but mid-range as well. Unless you are really strapped for space, I suggest you go for a conventional speaker of 2 to 3 inches diameter - it will sound much better.

_________________
Shift to the left,
Shift to the right,
Mask in, Mask Out,
BYTE! BYTE! BYTE!

I was aware of the frequency response issues, hence my comments about applying a bass boost circuit to the amplifier. I agree that a 3-inch speaker would be much better, but since its going on a handheld computer's single board, that's already so small I'm putting components on both sides..... I'm not really sure where I would even attach a 3 inch speaker, let alone find the space for it. I would expect that under most circumstances the user would use the headphone jack anyway.

_________________
Want to design a PCB for your project? I strongly recommend KiCad. Its free, its multiplatform, and its easy to learn!
Also, I maintain KiCad libraries of Retro Computing and Arduino components you might find useful.

_________________
http://WilsonMinesCo.com/ lots of 6502 resources
The "second front page" is http://wilsonminesco.com/links.html .
What's an additional VIA among friends, anyhow?