Can I use a 2A transformer to power 4 TDA2050 amplifier chips?

[Tesladude]

So I want to make a tda2050 amlifier with 4 of these chips, and after posting a lot of questions about different audio chips here, the chips always turn out to draw a few amps EACH as said in the datasheet.
On this chip it says a max of 5 amps and people always tell me to supply them with a transformer what can supply that much for each chip (the max). but I don't see why because I have been powering the tda2050 with a 400ma transformer reading that the chip only draws about 200 MAX when playing audio with a 4 ohm load. So why can't I just use like a 2A transformer to power 4 of these chips?

 

[Tesladude]

CORRECTION I am only reading about 50ma!

 

[Jony130]

What is your power supply voltage? And it lil we symmetrical or single supply?

 

[Tesladude]

supply voltage is a regulated 18v supply from a center tabbed transformer, the tab is ground.

simple circuit from the datasheet
http://html.alldatasheet.com/html-pdf/25046/STMICROELECTRONICS/TDA2050/6499/4/TDA2050.html

 

[jim hardy]

CORRECTION I am only reading about 50ma!

Wait'll it hits a booming bass note...

 

[sophiecentaur]

The point about class B amplifiers is that they only demand a lot of current when they are actually delivering a lot of signal power. 12W of audio at 12V will need 1A at any frequency - it's just that most music doesn't involve maximum amplitude sinusoids. Amplifier designs (even down to the size of Heat Sinks used) rely on this.

 

[Tesladude]

Wait'll it hits a booming bass note...

I made sure I was using a song with a ton of bass, and evan when I cranked the Ipod on full to the point where I was hearing a ton of distortion out of the amplifier I still had like 100ma going through.

ok, so would it be okay to run four of these amplifiers as a 2 channel amp (bridge mode,) with like a 3 or 4 amp transformer?

 

[jim hardy]

supply voltage is a regulated 18v supply from a center tabbed transformer, the tab is ground.

simple circuit from the datasheet
http://html.alldatasheet.com/html-pdf/25046/STMICROELECTRONICS/TDA2050/6499/4/TDA2050.html

The circuit you linked is for split supply,
that is three wires: +18, -18, and common ('ground') .

You say your supply is "a regulated 18v..." - is it just two wires +18 and common, or is it three wires +18 -18 and common?

Regulators are typically current limiting. How much current can your regulated supply deliver?

next question in next post.

 

[meBigGuy]

The peak to average ratio of music is, what, 5 to 1? maybe more. Your meter is only reading the average. You WILL saturate to the supply voltage at times, so calculate accordingly. You may be OK with the voltage dropping when you saturate, but be sure you understand it.

So assume you drive an +-9 volt squarewave into a 4 ohm speaker? (is it actually +- 9?) The peak current is 9/4 - 2.25 amps from 1 side then 2.25 amps from the other side. If you limit the maximum drive somehow, you can adjust accordingly. Power is 9*2.25 = 20.25 watts.

 

[jim hardy]

I made sure I was using a song with a ton of bass, and evan when I cranked the Ipod on full to the point where I was hearing a ton of distortion out of the amplifier I still had like 100ma going through.

If you have only 1/10 amp at 18 volts you cannot get more than 1.8 watts , but that is enough to make quite a bit of noise.

Given this amplifier's impressive capability I do not believe you are getting anywhere near full power from it.

What is measured power supply voltage at point you begin to notice distortion?

What kind of speakers are you using? What is their DC resistance by ohm-meter, and what impedance is written on their nameplate? And what power rating?
This amp should have enough power to smoke little computer speakers. Has that happened?

ok, so would it be okay to run four of these amplifiers as a 2 channel amp (bridge mode,) with like a 3 or 4 amp transformer?

Let's do simple arithmetic:
To get the 30 watts this amplifier can easily deliver into 4 ohms requires how much current?

watts = amps^2 X ohms
so amps = sqrt(watts/ohms)

amps = sqrt(30/4) = ~2.74
and that's for each amplifier.
If you plan to use four amplifiers get a ten or twelve amp [STRIKE]transformer[/STRIKE] power supply.

And don't neglect to use a LARGE heatsink. The amp will shut itself down if it gets hot.

That looks to be a very nice amp by the way.



Here's a thread where a beginner built an audio amp, albeit one more modest than yours:

http://www.discovercircuits.com/ubbthreads/ubbthreads.php?ubb=showflat&Number=7783&page=1

follow it far enough to see the nice speakers he built for just a ten watt amp !

good luck, and good listening !
Put your effort into the speakers,
and don't scrimp on the power supplies. I like toroid linears not switchers.

 

[Baluncore]

What confusion.
We do not know if a single or split power supply is being used.
We do not know the voltage(s) of the supply/supplies.
We do not know where or how the current is being measured; nor if it is average, RMS or peak.

Consider a pair of amplifier modules, being used together in a bridge configuration to drive a single 4 ohm speaker. The power supply to both amplifiers can be a single 0V(ground) and +18VDC rail.
Current flows from one rail, through one amplifier, on through the speaker then through the other amplifier to the other supply rail.

A full amplitude square wave will result in a supply current of 18 / 4 = 4.5A
Now two channels will need 9A total from the single 18VDC supply.

A bridge configuration is used when AC coupling is not possible or there is insufficient supply voltage available. So why use two amplifiers in a bridge configuration on 18VDC when you can use one amplifier on 36VDC (= split +/-18V) for the same power?

If you do not require full power you should be able to increase efficiency and reduce heat generation by lowering the power supply voltage or increasing speaker impedance.

 

[Tesladude]

It IS a split supply, from an 18-0-18 transformerI am running it on the split supply circuit I showed up above. I heard distortion as I cranked the input audio up to max(which is to be expected) not while changing the 18v supply. So I put everything into this amp and the MAX I could read with a multimeter that I was drawing even without a regulator was as I preveously mentioned 50ma. And when I was reading 50ma the speaker was VERY loud as it should be.

(In bridge I am only going to use 8ohms. but let's just focus on the single amp for now)

So it just really is weird that you guys are saying it's supposed to drive a couple amps for each amplifier but I am running this one off of a wimpy 400ma transformer and it works great!

I am hooking up the amperage meter on the positive line. am I supposed to do something different with a split supply?

 

[Baluncore]

With a split supply, all current to the speaker passes through the ground connection. But that current is AC. Half the time that current comes from the positive rail, the other half of the time it goes to the negative rail. You should be measuring the current in both the +Vs and -Vs. It is possible that there is a DC offset on the amplifier so the currents may not be the same. Is that the case?

An “18 – 0 – 18” volt transformer secondary would be rectified with a 4 diode bridge and the charge stored in two capacitors, one for each rail.
18VAC has a peak voltage of 18VAC*1.4142 = 25.45VDC
When rectified it will fall by one Vpn, 25.45 – 1.0 = 24.45VDC peak.
So your rails should have peak voltages of –24V and +24V.
Have you measured –Vs and +Vs ?
What are those measured voltages ?
What is the capacitance in uF of each of the two storage capacitors?

 

[sophiecentaur]

It IS a split supply, from an 18-0-18 transformerI am running it on the split supply circuit I showed up above. I heard distortion as I cranked the input audio up to max(which is to be expected) not while changing the 18v supply. So I put everything into this amp and the MAX I could read with a multimeter that I was drawing even without a regulator was as I preveously mentioned 50ma. And when I was reading 50ma the speaker was VERY loud as it should be.

(In bridge I am only going to use 8ohms. but let's just focus on the single amp for now)

So it just really is weird that you guys are saying it's supposed to drive a couple amps for each amplifier but I am running this one off of a wimpy 400ma transformer and it works great!

I am hooking up the amperage meter on the positive line. am I supposed to do something different with a split supply?

The fact is that a few hundred mW of music into a cheap speaker will produce a lot of sound. High quality monitor speakers tend to be much less efficient and may need driving quite hard to produce a 'loud' sound. But you can't get away from the fact that you won't get any more audio power out than the real DC power you are putting in. 'Intuition' can lead you to the wrong conclusions! Loud may not mean high power - even if your parents tell you to turn it down.

 

[meBigGuy]

Let's get this straight. If you put 18V across a 4 ohm speaker you will draw 4.5 amps. period. End of story. Assuming the supply will handle it, the current will peak at 4.5 amps as it clips.

If you use a 400ma supply the supply will go out of regulation and the voltage will reduce. You are distorting.
If you drive a sine wave you will hear the distortion easily. Also, the peak to average ratio for a sine wave is fixed (I'll make you work that out).

So, drive a 200Hz or 1KHz sine wave into the amp (make a wav file with cool edit or something) and listen carfully to determine when the distortion starts.

The RMS current you measure with a sine wave is meaningful. Current measurements made during music are not.

 

[davenn]

Let's get this straight. If you put 18V across a 4 ohm speaker you will draw 4.5 amps. period. End of story. Assuming the supply will handle it, the current will peak at 4.5 amps as it clips.

if you put 18V DC directly across a speaker, its likely to have its ( the speakers) lifespan measured in milliseconds

Dave

 

[jim hardy]

Maybe the point is this

That amplifier IC is capable of almost fifty watts, which in my days of vacuum tube hifi was a real sledgehammer.
A sledgehammer needs a substantial grip.
Driving even a single TDA2050 with a measly 400 milliamp supply is trying to swing a sledgehammer head with a toothpick handle.

On the bright side - probably that's why OP's speakers are not burnt up.

 

[Tesladude]

I meant to say 9-0-9, sorry. I also tried it with a 2amp transformer into a 4 ohm speaker. And it's pumpin just like before, little difference in current draw, and even if I am sopposed to measure negative AND posotive current, it's still coming out as only 100ma.

I am not calling you guys wrong or anything, I understand that this should be drawing a LOT more, it's just a little freaky that it doesn't seem to be! Although It's pushing my coaxial high quality speaker to it's near MAX excursion when I turn the volume up! And the chip is getting hot when I do turn the volume up high so It seems like it's drawing a significant amount of current but it is easily running off of such a tiny supply.

Also when I meant distortion I meant simply having your audio input too high for the amp, it sounds loud and fine until I get around max input, this happens with all audio ics that don't have a resistor or voltage divider or somthing to cut down the input. I was just saying that I was running this chip at MAX, and still barely any current draw.

but I will check my multimeter and mess around, do you guys want pics or anything?

 

[sophiecentaur]

I don't think anything is amiss. It's just that you aren't measuring what you think you are measuring. Do you have an oscilloscope? What is the peak to peak excursion of your output, measured to a pure resistive load? Even without a scope, you may be able to work this out with your DVM on peak hold, measuring from each rail in turn.

 

[Tesladude]

I just put may amp meter in seriece with ground and the speaker and was reading about .3A,

by excursion I meant on the speaker voice coil.
No I don't have a oscilloscope, I do this at home on a small table in what used to be my living room.
And I'm really tired of working on this, can anyone point me to an 18v transformer of switching power supply module or SOMETHING that will run all this well?

If i have 2 bridge channels with 4 tda2050, what current will the power supply need to have?

 

[sophiecentaur]

How do you know the actual impedance of your speaker? It is easy to obtain a few cheap, medium power resistors and make up a proper load to measure power with. (Series and parallel connected to spread the power over several resistors).
There are many ways that you could be misleading yourself - not least because 'loud' may not involve much power, actually. Make sure that what you are measuring is what you think it is. Being without a scope is quite a handicap when you are chasing problems like this one.

 

[Tesladude]

From where I got it says 4 ohm
I just need to know the current of my power supply must be able to handle if I have 4 ics in 2 channel bridge mode.
I'm done fiddling around, I found these inexpensive switching power supplies

http://www.ebay.com/itm/261239950815?var=560210347426&ssPageName=STRK:MEWAX:IT&_trksid=p3984.m1423.l2649

They have an adjust voltage trimpot and I am going to ask them if it can go up to 18v, if so then I have my voltage and all the current availability I would ever need for this.
If it can't go to 18v I was thinking bout a dc-dc converter boosting from 12 to 18.

Any thoughts?

 

[jim hardy]

And...

you might measure the AC voltage across your speaker...

go ahead and put up that picture you offered. We have only a vague idea as to your setup.

You said it's a 9-0-9 transformer, so your supply is more likely around +- 10 or 12 volts.

Looking at the datasheet fig 10 (the one at ST is superior to that third party one)
that amp will make about ten watts with that supply voltage- a substantial amount of sound.

Put your meter across the supply, from + to -
if your supply is from an 18V centertap transformer you should read something like 22 volts plus or minus a few.
Write down what it is at low volume
then increase volume until it starts dropping noticeably.

Which comes first - distortion or power supply sag?

Is your speaker mounted in an enclosure or is it in free air?
Is its surround made of rolled foam or paper??
Does it have a power rating, or can you post a datasheet for it?
Do you have the amp firmly bolted to a good sized heatsink with silicone grease as recommended?

We're trying to get you pointed at some basics but are not sure where to start.
First is to find out what's distorting - ipod, amp, power supply or speaker
any of them might be running into a limit.

Aha I see you posted...

Set your meter to lowest ohms scale and measure your speaker. A decent 4 ohm should read around 3 ohms .You can get 18 volt supplies easily, and that amp could handle 24's.
I ususlly look to see what good ol' Marlin Jones has on closeout. http://www.mpja.com/24-Volt-Power-Supply-62A-Potrans/productinfo/30467 PS/

 

[Tesladude]

I'm sorry to be hard to talk to guys! I've been doing this stuff for a year in a half now but somtimes I think I've been doing it for 30! talking to such smart people makes me realize that although I know A LOT more than the average person, I know NOTHING compared to you guys!Jim: I will try what you asked me to ASAP,

Its a 30 watt 4 ohm speaker, so I really am getting max from it as well, I might switch to a 12" woofer just so I don't blow the little guy.

I'm testing it Free air just sitting on my table.

I fibbed when I said it was a 9-0-9, it is actually a 6-0-6 (the 400ma transformer) , but it was cheap so I'v always gotten about 18v from it.

It's a small heat sink but I do my tests quickly so if anything it normally just gets a bit warm.

Trust me, its the amp. The ipod is pretty quiet even with headphones so normally I have to turn it up all the way and it is crystal clear! I have run this entire circuit on a 500watt 15" woofer in a cabinet and it sounded fine until I turned the volume up too much, If I can I will try and make a video.But it really just seems like the normal thing cause it happens with all the amplifiers i'v ever used plus someone told me it's normal. But idk lolHere is what I was thinking! Top schematic

http://www.diyaudio.com/forums/chip-amps/195315-bridged-tda2050-single-supply.html

Accept 2 of that circuit so I may have left and right ear.

Single supply bridge amplifier using either the supply you sent me or this one I found, depending on how much 4 chips in bridge will draw,
http://www.ebay.com/itm/170855720740?ssPageName=STRK:MEWAX:IT&_trksid=p3984.m1423.l2649

 

[meBigGuy]

1. Put in a sine wave and measure the RMS (AC) current or voltage. (is your meter in AC mode?) If you are measuring the current, short the meter to see if the series resistance is causing a problem. If it get louder or you hear a click, then there is an issue there. (some cheap meters have significant resistance). Note that your meter measures RMS AC which is 0.707 * peak (not peak to peak) for a continuous sine wave. So measure the RMS (AC) voltage across the speaker when you drive it with a sine wave and then calculate the currents.

2. It's still not clear what you are doing power supply wise. Are you testing with 1 amplifier now? Is it connected to 1 9 volt DC supply? You keep talking about a transformer. How are you connecting the rectifier(s) and filters? Or do you just mean a wall wart supply?

Regarding your possible 24V supply:

3. A 24V power supply means a peak power of 24*24/4 = 144 watts. PERIOD! NO DEBATE!
Peak current is 6 amps. PERIOD! The only way this is not true is if you don't drive to full output.

4. 48V p-p from a bridge amplifier (24 in 1 direction, then 24 in the other) means an RMS voltage of 24 x 0.707 = 17V RMS, so the RMS power from a peak to peak sine wave will be 17*17/4 = 72 watts RMS.

5. Do the above calculation for a 12V supply and you may be surprised.

If what you measure is different that what I described, then you have measurement errors or I have bad data.

 

[Tesladude]

Powering this on my bread board I am still using a small center tab transformer putting out 9-0-9.
This is powering just 1 tda2050. For the purminent power supply though I will need to run 4x tda2050 as 2 bridge channels on 8 ohms. I read that in bridge each chip sees only half of your speakers resistance so you shouldn't go bellow 8 ohms.

The 24v supply I found gives a total of 10A on 24v which I am hoping is enough current but I also found a 15A version too. 24v is just under the MAX rating of the chip so I was planning on putting a small voltage divider or a series of diodes or something just to drop the voltage down to maybe 20v.
So will 10A be enough running 2 bridge channels at 4 ohms?

 

[Baluncore]

Because it is differential, a bridge configuration produces twice the effective voltage of a single ended amplifier. But at twice the complexity. That is why they are used when supply voltages are restricted, such as in 12V car audio.

I read that in bridge each chip sees only half of your speakers resistance so you shouldn't go bellow 8 ohms.

By restricting speaker impedance to 8 ohms, you are defeating the only advantage of the bridge amplifier. You really do not need the bridge configuration because you have a choice of supply voltage.

If you cannot calculate that, then you should stop trying to design an amplifier and just buy one.

 

[Averagesupernova]

Something I believe is worth pointing out here. Power (watts) in this thread has been referred to as RMS. There is no such thing. I believe this probably started out in the car audio industry in marketing. Voltage can be defined as RMS and current can be defined as RMS, but never power. Power is always average. Even when measuring the power coming from an AM transmitter commonly referred to as PEP (Peak Envelope Power) it is average.
-
It may seem like nitpicking but if we were to say RMS power it implies that we could take Peak voltage and multiply it by peak current and then take that product and multiply it by .707 and we would come with the right answer. Of course this is not true.

 

[Averagesupernova]

By restricting speaker impedance to 8 ohms, you are defeating the only advantage of the bridge amplifier.

More power is not the only reason we choose bridge mode. Using one 8 ohm subwoofer in bridge mode is cheaper than using two 4 ohm subwoofers and puts the exact same load on the amp. Tesladude is right in that each channel of an amp running in bridge mode sees only half the impedance of the speaker. So if we have an amp that is only 4 ohm stable putting a 4 ohm speaker on it in bridge mode will overload each channel.

 

[Baluncore]

Tesladude is right in that each channel of an amp running in bridge mode sees only half the impedance of the speaker.

You are really confusing things here by referring to channels and to stability under load.

Using one 8 ohm subwoofer in bridge mode is cheaper than using two 4 ohm subwoofers and puts the exact same load on the amp.

But it requires not one, but two amplifiers, (plus a phase splitter), which is “half as cheaper”, so the advantage is cancelled.

The point I am trying to make is that there is no advantage in operating a bridged pair on 24 volts when one amplifier can be run on 48 volts. It is clear here that Tesladude is not trying to maximise power.

 

[Averagesupernova]

You are really confusing things here by referring to channels and to stability under load.

I can't possibly see how. Point it out and we'll clear it up.

But it requires not one, but two amplifiers, (plus a phase splitter), which is “half as cheaper”, so the advantage is cancelled.

A lot of commercially available car audio amps are 2 channel amps for stereo mode with for instance 50 watts per channel at 4 ohms. Suppose a person wants to run an 8 ohmsubwoofer. A 2 channel amp that is not bridgeable cannot run a single 8 ohm coil subwoofer at 100 watts where a bridgeable amp can. To do this would require a dual voice coil subwoofer with 4 ohm coils. So the manufacturers build multi-channel amps that are bridgeable to cover the applications where people want to just run stereo mode to run speakers in the kick panels or doors, or in bridge mode to run a single subwoofer. My '65 Ford has a 4 channel amp for front and rear speakers and a 2 channel running bridged to run a sub. As far as I know, single channel amps in car audio are geared towards running only subwoofers and are class D and are considerably more than a measly 50 watts. So don't bring it up. For that matter, my home theatre powered sub is class D also. It's not limited to car audio.

 

[jim hardy]

The TDA 2050 datasheet says in table 2 that output peak current is internally limited to 5 amps.
Again, the datasheet at ST's site is more legible than that third party copy ( I have a real resentment against those for-profit third party sites barging into search engines and cluttering things with their ads - what good are they doing anybody? I avoid them like the plague.)...

So I think I agree with others, one should ask 'what's the advantage of bridging? '

Most of you guys know the sound terms in next paragraph but i'll try to explain each as it's introduced anyway.

Bridging would let you push 5 amps through 8 ohms instead of 4 ohms..
That should double power output.
But the ear is logarithmic and that's why we have decibels(db) to measure sound power level(spl).
"spl" is usually usually expressed in 'db', which refers to some number of pascals of sound pressure;
db = 10 X log(Pmeasured/Preference)
and doubling the power adds only 3 db of spl,,,, which most listeners would not notice.

Would somebody who's still young and sharp please check my arithmetic here ? I'm reaching back to 1963 for this...

If we take a speaker that's rated 90db spl at 1 watt and drive it with 50 watts
we'd add about 10log(50/1) = 17db to its 90db of spl
and get 90+17=107 db of sound (quite a lot).

Bridging the amplifiers and buying an 8 ohm speaker instead, and driving it to 100 watts
would similarly add 10log(100/1) = 20db
and give us 90+20=110db

Now that's twice as much power, but twice is barely enough for the average ear to notice.
It'll sound like it got louder by the ratio of 110/107 = 1.028, not quite 3%.
All that work for 3%? Bah, humbug!

That's why I tell people "If fifty watts isn't enough for you then you need five hundred."
And that's because of the human ear's logarithmic characteristics. You have to admit - 107 vs 110 isn't worth a lot of effort unless you're into those car stero competitions where they actually measure sound pressure level with expensive instruments.

But to tesladude's question -
a ten amp supply for your +24 and another for your -24 will deliver all the current that a stereo pair of TDA2050 based amplifiers should allow to pass, regardless whether the individual amps are single TDA's (2 total) or bridged TDA's (4 total).

If you bridge you should go to 8 ohm speakers.

old jim

 

[Averagesupernova]

Now Jim, 3 db can mean the difference between hearing the music and feeling it. :)
-
Some amps are 2 ohm stable, and with good reason. A lot of car audio speakers are 4 ohm. Take a 4 ohm speaker (or run two 8 ohms speakers in parallel) and run bridged and you have 4 times the power as compared to running it from a single channel. Each channel sees 2 ohms.

 

[jim hardy]

So long as your amp is not current limited...
If it IS current limited then to maximize power you want to push that amount of current through as many ohms as you can.

.....

:) :):) Ahh that feel - track 12 of the Titanic soundtrack CD... :) :) :)
I was so proud of my Electrovoice 12's...

here's some speakers my kids built - array of cheap woofers..

They compared well with the Altecs in background...

 

[Tesladude]

Ok so would running bridged tda2050 on an 8 ohm speaker give me a noticeable increase in volume?When I first started all this I was going to use a simple linear power supply to run 2 tda2050s giving me left and right ear at a reasonable volume on an 8 ohm speaker. I tried it, and It didn't seem loud enough for me.
I wanted to continue useing the tda2050 because of it's simplicity, low voltage, and price.
So, having little experience with bridge, I figured I would just have 2 bridge circuits to give me a lot more power for each ear. I read if you have 100 watts into 4 ohms, then bridge you will have 200 into 8.
If that is so then bridged I will have 60 watts into 8ohms since this chip does 30 watts into 4ohms.
Then came in a suitable power supply for this and unless the bridged plans are changed, then that's where I am with this.

Now from what I think I am hearing up above is that this will not give me more volume?(virtually) And it is not worth the effort?