TDA2030 DC at output

[anotherjim]

Look out for brick psu's for  older medium sized printers and the like. AC out or unregulated DC. Some were 18-24v range & at least 2.5A. Shame is, there must be thousands of those in landfill now!

[MrStab]

i'd love to come across a few of those for personal use, they seem to be popular for fully-fledged amps. wouldn't the nearest fixed regulator down from 24 be 18V? would this be a job for one of the variable regulator ICs, for something like 22V? i've never used that kind before. i could probably justify a regulator inside, so long as Xerox have taken care of the rest.

the more i push this, the more i might as well switch to bipolar, but i'd rather not. at least not just yet. when i get some time i'll find a good maximum gain and shorten the tube (it's way too long), and see how the best-case scenario with the 2030 @ 18V goes.

[MrStab]

i know this is a whole topic unto itself, but is an SMPS totally unacceptable here, or is it just a case of the hassle involved? i'm struggling to think of an analogue audio device which would even have an external switched supply at that voltage, but maybe there's some product?

i'm guessing the generic supplies from component retailers wouldn't be friendly on audio. i'd imagine a series resistor in the input, to form a LPF with the 2,200uF cap, could only help so much. a few of the supplies i'm looking at just now have fairly-comprehensive datasheets, but i can't pick out any parameters which might help.

[Rob Strand]

analogue audio device which would even have an external switched supply at that voltage, but maybe there's some product?

There's nothing wrong with it as such but it does lend itself to additional noise sources.   Most of the higher powered card amplifiers use a switch mode to boost from 13.8V to +/-35V or so for the power amplifiers.   There's a lot of baggage in term of inductors and caps for filtering as well as very careful attention to PCB layout and physical placement of the parts.

There's some degree of hassle and usually lower reliability.

As far as efficiency goes many mains-fed switch-modes aren't very efficient and a typical class AB audio amplifier is quite inefficient.

I have seen switchmodes in those cheap-arsed surround sound systems, partcularly those with a low profile case.   They tend to use class-D amplifier chips for the audio power amps.

I'm all for grabbing those bricks anotherjim mentioned.  They are typically unregulated and single output.  The old HP printers used a transformer which was unregulated.

[bool]

...  The old HP printers used a transformer which was unregulated.

But a bit underpowered for a speaker-driving amp.

[MrStab]

thanks for the post, Rob. car amps crossed my mind at one point, but i'd need 230VAC input, then suffer from the efficiency problems you mentioned.

do people typically regulate the output of the printer blocks, or are they often used as-is without issue? are there huge caps in there? LM338 apparently goes up to 22V, but maybe there's a better IC.

[j0shua]

Hi again guys , i be out of internet since 2005, but now im back, i proud to say i survive from Cancer, was a hard battle but now im 100% free of sickness.

About TDA2030 is a pentawatt 18w is a sustitute of tda2006, is not good amplifier because need to much components and have to much distortionx anyway, if you build a practice amplifier are best options to get the full power of tda2030 need to use 18v. I hope this information help you, to avoid distortion add a ceramic cap ( 102 ) to the pin 4 ( audio out ) and another one in the V+ better if you can put a poliester cap

[Rob Strand]

@bool

ut a bit underpowered for a speaker-driving amp.

The ones jim and I are thinking of are quite beefy.    Adaptors from later models are rated somewhat lower but you do still see 2A units around.

do people typically regulate the output of the printer blocks, or are they often used as-is without issue?

At the end of the day there's no rules, you just use these things however you like.
I'm not a fan of using regulators for power amps.  If the power amp can handle the voltage you are always better off using the higher voltage.  This lets the amplifier produce more output,  especially on the initial attack part of a guitar signal, which reduces audible clipping/distortion.   However if you have got a regulated switchmode which does the job then you should use it!

are there huge caps in there?

You can't tell as there are many types:  AC,  DC unfiltered unregulated, DC filtered unregulated, DC regulated (which are largely switchmodes these days).

Many off the shelf DC filtered unregulated "plug packs" have a 1000uF cap.   Custom units like stuff from HP and the like are all over the place.  The beefy HP ones I was thinking about were unregulated but I can't remember if they were filtered or not.   

A small detail is when you filter an unfiltered plugpack the amount of current you can draw from the filtered rail is about 0.62 times the current rating on the plug-pack label!

LM338 apparently goes up to 22V, but maybe there's a better IC.

The best way to handle preamps and power amps is to simple create another power rail for the preamp.  So the power amp rail voltage does no impose any limitations on the preamp components.  There's many solutions, all have been used in consumer products.
1) Derive the preamp power rail from the power amp power using an RC network.  The series resistor is selected to provide enough voltage drop to keep the preamp voltage in a good range.  The same idea is used in tube amps.
2) Use a regulator for the preamp only.  This can be an IC or a zener diode.  Usually this is done if the preamp pulls too much current or you want to be 100% sure there's no hum on the preamp power.

Let's step back a bit from all the details.
From what I can see you are getting distortion on the loud notes/strums - yes?  This is pretty normal for a small guitar or hifi amp. 
The solution is to get the most power for the amplifier chip.
That means use a power supply with the highest voltage, assuming it has the current rating.

How to do this depends on your constraints   There's  a big difference in designing a commercial product where you have to carefully select parts according to some cost constraints, using off the shelf parts (eg. you might be able to find something cheap), and using up stuff you have around in your junk box.

I'm not sure what your constraints are, what parts you are looking to buy, what parts you have on hand.

[Rob Strand]

i proud to say i survive from Cancer, was a hard battle but now im 100% free of sickness.

Welcome back and congratulations.

[anotherjim]

I've never seen an instrument amp with a regulated supply for the power section. The problem with regulation is is usually brings some kind of power limit.
SMPS type can be very sensitive to a momentary overload causing rapid effective disconnection of supply, possibly for very short, but audible periods.
Look up some amp schematics, it's always transformer to rectifier to filter caps to power amp and also from the filter caps thru dropper resistors and filter cap to the pre-amp.

We are wary of building an AC supply for safety reasons, but also financial because you can often get one complete for less than you can buy the transformer. That is always the case with small regulated supplies, but for a power amp, especially if you want it all in one case, we may have to put on the rubber gloves and build the AC supply part.

If you dig out old DIY electronics project articles, you'll find building the complete PSU was often part of the project.

[MrStab]

sorry for the delay and thanks again for the responses, guys. i wasn't around here when you left the internet, Joshua, but welcome back and a huge congrats! i think i've read some of your posts throughout the years.

i've never, ever seen a regulator on an amp's power rails in real life, Rob. i have seen it online though, and my paranoia with noisy pedal supplies lately has rubbed onto this. keeping the preamp stage regulated seems like a good idea. i think i saw that in a schem the other day and was scratching my head because they used one way of a way-lower output voltage than the supply voltage. now i realise it wasn't necessarily the voltage itself that was important, but the regulation.

so from Rob's explanation of the symptoms and PSU, and the unexpected SMPS behaviour Jim's mentioned, i think the path ahead is clear. the hunt for a 24V linear PSU is on.

fun story: i tried making an SS amp back when i knew a bit less, and signed up to a relevant forum for tips. they shot me down harder than RG and PRR would if i said "i can categorically state that this op-amp sounds better". lol
i'm glad they did, it was too risky for me to be messing with mains AC at that point. i've since realised 75% of the problem was just a crap speaker, and the rest because of the problems outlined in this thread.

thanks a lot guys! i'll let you know how it goes. it's a bit of a back-burner project.

[Rob Strand]

i've never, ever seen a regulator on an amp's power rails in real life, Rob.

Regulators for preamps are very common.
Linear regulators for (large) power amps are pretty much non-existent; well except for crazy DIY audio guys.
Regulated switch-modes for power amps are showing up on medium power units.

so from Rob's explanation of the symptoms and PSU, and the unexpected SMPS behaviour Jim's mentioned, i think the path ahead is clear. the hunt for a 24V linear PSU is on.

You can avoid the issue Jim mentioned (which is real) simply by using a PSU with higher current rating; this also improves reliability.  If the power supply is rated to handle the *peak* current with the lowest impedance load on the amp then the power supply limiter will not kick-in.

As Jim mentioned, you can get cheap external Switch Mode units, often with beefy current ratings.

Also as Jim mentioned, if you don't want a plug pack then you have to use a built-in transformer.

[Rob Strand]

Something worth noting:

If you use a 24Vrms transformer (say 25VA with 10% regulation) it will give you a loaded voltage of 28VDC, similar to the design tabulated in the datasheet.  You will get about 10W into 8 ohm.  *However*  with a 24V rms transformer, if you consider the transformer regulation and worst-case main supply variation the worst case supply voltage will exceed the chips maximum voltage of 36V!    If you use a TDA2030A instead of the TDA2030 then you are safe (and maybe you can get more power).

To prevent this issue you need to decrease the transformer voltage.   The next easy to get transformer is 18Vrms.  In So as to no lose too much power you need to make sure the ripple is low and the supply does droop.  To do this you use a large transformer say 50VA and large caps say 2x4700uF.   After taking all these measures you will get maybe 7W into 8ohm.   The loaded supply voltage will be about 22V to 23V.

However, if you used a regulated switch-mode you can push the voltage pretty close to the 36V mark.  Since it is regulated you don't have to worry about supply regulation with load and mains voltage fluctuations.

Anyway something to consider...

[anotherjim]

The easy way to figure out the transformer is to take the secondary volts - which should be the RMS rating - and multiply by 1.5. This gives the peak output which the DC supply cap will charge to if there's no load. Actually I think it's x1.404 less rectifier diode drops, but 1.5 is quick and safe side reckoning.

So 24v gives max 36v, which by no coincidence is what Rob said.

Don't break your neck trying to eke out a few more watts. 10W is not significantly louder than 7w.

[PRR]

> i've never, ever seen a regulator on an amp's power rails

Dynaco Stereo 120.
http://www.updatemydynaco.com/Stereo120Repair.htm

And.... that's the only one comes to mind (outside obsessed one-offs).

[MrStab]

i was wondering about the loading issue. procedure tips noted, thanks guys.

audiophiles... *shudder*. what's the absolute worst power supply out there, and how many times should i hit it with a hammer? lol

so i'm less deterred by the SMPS now. if i'm trying to take the microphone back into the pedal and blend it with the clean signal, thus having that PSU's ground connected to the signal ground going into the "real" amp (albeit a pedalboard using an iso-regulated 8-output supply), is this still advisable with the right amount of filtering?

the guy was amping the mic through a second guitar amp instead of through the PA and liked it anyway, so i figure it's do-able!

[MrStab]

i used to have this f****** massive Cambridge Audio supply years ago when 5.1 surround was all the rage, which eventually failed and would trip the circuit breaker in my house whenever you turned it on. lol

[PRR]

OH! And Taylor of this forum offers PCBs with an audio power amplifier and a regulator.

But this is expedient. There's lots of 10V-16V car-sound power chips. There's LOTS of Dell 19V laptop power-warts on eBay, real cheap. Taylor saw that you could get nominal 12V out of 19V with a regulator, and the combo was cheaper than any other plan.

That old Dynaco was "special" many ways. It was pioneer hi-fi transistor. The circuit is today unfathomable. The parts were worked very close to their ratings. Closer than you can expect wall-outlets to be. The 2x40W version ran unregulated, but to reach 2x60W without popping costly transistors they had to nail the supply to 70V on the dot and rig it to shut-down on excess current. Improved devices have made that now silly.

[MrStab]

i was hoping to keep everything on one board as the speaker (and forthcoming jacks/pots) take up so much room, but now i know where to look if i change my approach there, thanks for that. i know it's always power amp kits and PCAs you hear about being counterfeit and incendiary, but i'd still trust a blank PCB from the community more than some of these "every circuit" sites. the benefits of a dedicated layout vs. my stripboard might convince me more.

19V Gateway laptop supply at the moment - it's a lot quieter with a 2,200uF cap, but still has some noise seeping in. would this one be fine for a voltage increase?
https://www.rapidonline.com/ideal-power-25hk-ab-240a125-d5-class-i-desktop-psu-24v-dc-1-25a-30w-18-1498
datasheet: https://www.rapidonline.com/pdf/18-1498.pdf

i am in fact using the TDA2030A. total dissipation is just over 6.2W (maybe?), rest of the circuit would be 400mW or less depending on op-amps, so i'd assume it's fine.

that Dynaco amp sounds a bit like what they'd call factory overclocked nowadays. for people discontent with the standards of the time.

[Rob Strand]

Here's my calculations:
Supply Voltage, Vcc: 24V
Load Resistance, RL:   8 ohms
Peak output current, Ipeak: 1.38 Amps
Output power, Pout:   7.6W
Average power from DC rail: 10.5W   (sine wave)
Average power from DC rail: 16.5W   (square wave)
Device Power dissipation, Pdiss = 4.5W (sine output, worst case conditions)
Required heatsink:  21 deg C/W  (or less)

You can see the peak current is a bit higher than the rated output current of 1.25A.
ie. the *peaks* are 1.1 times more than the rating.   In reality I can't imagine the PSU overload
cutting in at 1.1 times the rating.   If you are extremely unlucky and the power supply is
ultra-sensitive then the current limit might cut in.     As a back-up plan you could add a
couple of rectifier diodes in series with the power rail to reduce the peak current (and the
output power) just enough put the peak current under the radar.

The uncertainty in my calculations is probably in the order of 5% to 10% anyway.

One assumption is the speaker is 8 ohms!!!!  A 6 ohm speaker would increase the
peak current.