tda1562 power is not even near 70w...

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
I recently built an amp using tda1562 70w chip, currently feeding it from 350w old pc power supply, getting sound signal from my cellphone.
At max volume I'm getting only around 8 volt peaks, which according to formula V^2/R is like 16 watts (using 4 ohm speaker).
What could be the problem? I really want to get the full power.
Btw using 470nf input and 22000uf power capacitors.
 
Administrator
Joined 2007
Paid Member
This all depends on a couple of things.

1. How much voltage does the cellphone put out. Possibly not as much as you think.

2. How much voltage gain does the amp circuit have. That can be determined easily by the value of feedback resistors.

So 70 watt into 4 ohm (RMS ? we always deal in RMS values :)) is 17 volts RMS approx. Thats 24 volts peak or 48 volts peak to peak.

Now I've just looked at the data sheet for the IC.... and see its a bridge output that can make use of a "lift" capacitor to momentarily boost the output. Presumably it adds that voltage to the main rail for transient peaks.

Hmmm....
 
You can't get 70W continuous sine wave power out of this chip.

You say you're getting 8V peaks, but how are you determining this? You'd need to look at the output with a scope, and capturing the peak output might not be straightforward.

Best thing really is to ensure the gain is set correctly and that the drive voltage is adequate, then if the circuit is built correctly you'll get something close to the advertised output, given that you're running it a @ 12V rather than the recommended 14.4V (something to be considered).

Otherwise you'll need to give some consideration to exactly how you're going to test it accurately.
 
1. Well I measured during a song with a lot of bass and got around 0.32 volts max.
This probably isn't the right way to measure tho, is it?

2.Don't know what do you mean exactly but datasheet says min - 25 typ - 26 max - 27

I'm not 100% sure if it's 70w RMS or max, don't see datasheet mentioning it, but even if it was the max power, I shouldn't be getting only 8 volt peaks, should I?
It would be fine even if it was 50w rms...
 
well depends if the status pin is at high or not.
This chip is a quite sophisticated calss H amplifier, that can work in class b.

Allso notice, 70 watt at 10% thd, at 14.4 rail voltage.
The computer psu is 12 volts to begin with.
Allso, it is quite likely as mooly said, the cellphone fails to make the required input voltage.
Gain is fixed internally in these chips.
I would like to see the schematic of Your build, and the actual component values used.
These chips are -as i mentioned- quite sophisticated, if the heatsink is not big enough, and it gets hot it will switch back to low power mode.
Fancy stuff, but actualy it is a verry roboust chip.

the rail lift supply solution used in the chip IS a sensitive one on the otherhand. High ESR lift capacitors kill it. Would not be able to charge in time to discharge when in demand.
 
Administrator
Joined 2007
Paid Member
You have to measure using a continuous tone and have some means of accurately measuring the value. A DVM can be used at low frequencies (up to 400hz and many will cover the audio bandwidth but don't count on that unless specifically stated).

The 25/26/27 are DB values for gain. So 26 DB is a voltage gain of around 20. So 0.1 volts RMS input would give 2 volts RMS output.
 
It's circuit from http://web.argus.lv/shop/download/425132/hwindex.htm
capacitors C4 and C5 are 22000uf, input capacitors C1 and C2 are 470nf, C8 is 4700uf and the rest is the same. E, F are speaker outputs, P is voltage source.

Btw that power supply is outputting even slightly less than 12v (11.something), could it really make that big difference in power?
Also I tried using computer as signal source (seems to have like twice higher voltage), I seem to get more power, but the led of amp starts blinking as soon as volume is more than half
 
Last edited:
if that pin is floating then its in class b.
8v output voltage at 12 v ish supply rails seems pretty decent for class b.
at class H you will get around 14 volts.
get a decent supply of DC 15 volts (hint, a computer psu can be modified with a pot to obtain that) and you will have Your 70 watts to 4 ohm load in class H, probably without clipping.
 
You have to measure using a continuous tone and have some means of accurately measuring the value. A DVM can be used at low frequencies (up to 400hz and many will cover the audio bandwidth but don't count on that unless specifically stated).

No, you can't measure with a continuous tone. The chip is only capable of producing 70W for a very short duration. You could try making a track with Audacity that has a period of low volume continuous tone interspersed with very short bursts of high-volume tone at regular intervals. This way you might get something on a scope that will operate the trigger and have sufficient persistence to be visible. Or try a digital scope with capture. You could try to contrive some kind of sample-and-hold, but I'm guessing that's beyond you.
 
Administrator
Joined 2007
Paid Member
No, you can't measure with a continuous tone. The chip is only capable of producing 70W for a very short duration................

I referred to a continuous tone as a means of determining the true RMS voltage it was capable of delivering. It gives you a baseline to work to. Anything else becomes a little meaningless as this (admittedly interesting) "boost" capablility is totally programme dependent :)
 
Class H is an efficient design, but can not handle square wave input.
Actualy music does not have any square waves.
Until it is a sine wave, it will work well enough to be un-noticed by the listener, mesurements on the otherhand might -not sure, just might- reveal if it is a class b or class h amplifier.
The key point is that the lift capacitors only supply the amplifier if the singal is too large for the rails to handle. In otherwords, only when the sine wave peaks have to be playd, will the capacitors get discharged. For the rest of the signal, the amplifier will work as class b and the capacitors will be charged.
supposedly one can use a set f bigger capacitors, but it has to be considered that the capacitor must be able to charge in order to supply sufficient voltage. Therefore lift capacitors must have low ESR rating/impedance, and must be as close to the lift supply as possible, with the shortest leads possible, and even those have to be high cross sectional area conductors. This is required to obtain low series resistance, allowing faster charge of the capacitor.
Allso, a powerfull powersupply is allso a requirement. the voltage rail must not drop.
I would highly advise to not go close to the minimum required powersupply voltage.
I would keep it close to 16-17 volts insted of 14 ish.
The reason is simple, a higher supply voltage will allow the amplifier to stay in class b for a larger amplitude, resulting in more time for the lift capacitors to charge, and will reduce the amplitude section where the lift capacitors must release charge in class h operation mode.
Longer charge time allows to use larger capacitor as lift supply, making sure that the class h lifted supply voltage does not collapse do to the capacitor running out of charge.

usual kits sold with this chip include not so good capacitors, the higher voltage rail does help that quite a bit.
Sadly a computer psu is only safe to mod to 15 volt output voltage without major modification.
Guide: Power Supply Sense Wire Modding
the at/atx psu has capacitors rated for 16v max, while with the modification it is possible to obtain higher voltage level. So be aware, that anything past 15V on the 12v rail is a sure way to blow up something.
Mind me, i em absolute not responsible for anything anyone can think of if anyone trys this.
Infact, to save my stinkhole from any lawsuit, i highly not recommend the mod, opening a powersupply, not even looking at one is advised. Anyone reading this agrees that i em (and anyone associated with the linked material) can not be held responsible for any kind of outcome, be it positive or negative, be it real or imaginary.
 
So I tried connecting pin 16 (status) to 'M' so amp is forced in class H, but makes no difference... And afaik it should go into class H automatically when needed so I shouldn't force it to work that way all the time.
Anyway, the thing is, I can get more power using higher voltage signal (from computer), but when I turn it louder, the led at diagnostic pin lights up, so that kind of scares me lol
so I don't really know if it's safe or not to keep it loud, don't want to burn this chip.
 
i at the moment does not know what is wrong with the amplifier it self, but th chip can do the 70 watt power to an 4 ohm load. I do not recall automatic switching to class h. i do recall automatic switching to class b from class h.
i allso do not think Your powersupply manages enough voltage to make it work properly as per datasheet.
in anycase clipping in class h (indicated by the led) should happen at a higher speaker output then at class b. You can test thisway if it is realy in class H mode.
 
high input signal yield high output signal.
an output signal that gets close to rail voltage yields clipping.
the led flashes when the amplifier is clipping.
if the output signal gets close to supply rail voltgae, this ic when class H is enabled uses the lift capacitors to power the amplifier temporarly from a higher supply voltage rail, so clipping does not occur.
IF, that is, IF everything works in order.
otherwise the amplifier needs revision.
the IC is feature rich and decent. I would not realy think the IC is the suspect if things do not work.
on the otherhand one must understand that this chip has very sensitive internal mechanisms.
and a pc powersuply is so to say, at least noisy. Just a tip, but it can even have problems with the quality of the supply voltage. on the otherhand a car alternator/generator too is allso noisy.
But at least the supply voltage is close to what the datasheet mentiones, and one must allso understand that supply voltage and available output power have a nice relationship.
till death takes them apart.
 
Very soon I'll be able to try it with higher voltage...

Anyway, in datasheet I read this:
"Mode select input (pin MODE)
This pin has 3modes:
1.LOW, ‘standby’: the complete circuit is switched off,
the supply current is very low
2.MID, ‘mute’: the circuit is switched on, but the input
signal is suppressed
3.HIGH, ‘on’: normal operation, the input signal is
amplified by 26dB."

Its pin 4. Is it possible that somehow its working in 'mute' mode, not 'on', that's why I'm not getting enough power?
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.