max voltage for lm3876 - need clarification

[dvdwmth]

I was just reading this post

Gainclone vs Sony TA1055

where many people were taking issue with the transformer rating this guy had used for the gainclone he was trashing.

So I misunderstood the max rating specs on the national spec sheet by the looks of it.

Someone was explaining the voltage thing to a poster who was suffering from the same confusion that I was. Apparently the max voltage is on the data sheet is 80something and this is for one rail with the other at zero, so split evenly its -40somthing +40something.

He calculated Xvolts per rail plus %9 for something to do with regulation.

Can someone explain this very simply? How do I determine the actual max transformer rating as printed on the side of the transformer. And what is this about adding %9 because of regulation. My power supply is just transformer -> rectifier and then the 1000uf cap next to the chip. Is this %9 (or some other %) relevant.

I have two 30v 5A CT transformers that I thought were far from the max rating and now it seems I completely misunderstood what this max rating means. Are these tranformers out of the safe operating range?

Im aware that a few posters at the discussion I referred to say that 24v seems to sound best so its not necessary to point that out. Im hoping the existing transformers will work REASONABLY well. When people say something "sounds better" its hard to tell if they mean "after a-b testing for half a day I think there was a little more air around the left side of the string section" or "it sounded like total crap when i did it this way".


And one other thing.

I dont understand the equation to calculate the correct value for resistor Rm on Nationals typical application schematic for the LM3876. Im not going to use a switch for the mute anyway but I get the impression I should still put a resistor in there. Can someone tell me the correct value.

Thanks for your patience and assistance

[Nuuk]

You take the maximum supply voltage and divide it by two if you are using a split supply, ie positive and negative rails.

Next, you must understand that this is DC not AC which is supplied by the transformer.

To convert one to the other, divide the DC requirement by 1.414. That is the maximum voltage that the trasnsformer should supply.

You should also consider that your mains supply may be higher than the supply quoted on the transformer. For example, I have a tooroidla transformer that is quoted as supplying 25 VAC with an input of 230 VAC. But often, the mains voltage in my house is nearer 250 volts. That is 8% higherr so the 25 VAC output becomes 27 VAC.

The Current rating of the transfomrer is not critical as long as it is enough. It is the voltage rating which will cause problems if it is too high.

So, if we take a chip that has a maximum supply of 80 volts DC:

Divide it by 2 = 40

Divide that by 1.414 = 28

Allow a couple of volts for mains fluctuations and we arrive at 26 VAC.

The nearest standard transformer will be have 25 VAC secondaries.

So yes, your 30 VAC secondaries are too high.

[Sci]

The 2x30V transformer can still be used if you put regulators in there, but the heat dissipation would be rather high I suspect.

Erik

[janneman]

The reason for the numbers given by Nuuk is that the transformer rating is an RMS rating. This is something similar as average, but not quite the same. You can look it up on the net if you want to know exactly.
So, if your transformer has 30V AC, the 30V is the RMS but this voltage waveform is a sine wave, OK? Now if you retcify this, the caps will be charged to the peak value of that 30V RMS, which happens to be the square root of 2 higher (which is, you guessed it, 1.4 something). So your 30V AC RMS will charge the caps to about 42V DC.
Now, you lose about 1-1.5 V in the rectifier diodes, but if you do not fully load the transformer, it actually gives off MORE that the nominal 30V AC. That's the 9%, which is a usual number for mid-sized transformers. So, you are almost sure to get more than 2 x 40V DC, which is living on the edge of a cliff.

Jan Didden

[Nuuk]

Thanks Jan. I am always relieved when somebody who does know what he is talking about confirms what I have said!

[jackinnj]

and to add what Janneman and NUUK said, the line voltage can vary -- in the summer in the U.S. it can range from ~109 to 128 VAC. I am sure the situation is much better in Europe where the distribution system seems to be better maintained.

Thus, if you made your calculations assuming a "nominal" 117 VAC, but applied 128 VAC to the transformer primary you could see a secondary voltage 9% higher

I've actually put a data-logger on the mains panel for our house and experienced instantaneous voltage dips which are somewhate greater than those cited in paragraph 1 -- I was so annoyed that I contacted the engineering department of Jersey Central Power and Light and they did the same thing, only their data logger wasn't from Radio Shack -- the upshot is that they reinstalled the cable to our little part of New Jersey, positively affecting about 200 homes.

For Darkroom enthusiasts, the fluctuation in voltage is a real problem -- so much so that I developed an integrating englarger head -- it recalculates the exposure time based on the integrated light value (it's logarithmic btw).

[dvdwmth]

Thank you all for the excellent information. Im much clearer now.

Sure glad I stumbled on to that post.

One thing I find strange is that I have been using LM1875 chips with these transformers, thinking I was just on the edge, when in fact I was a mile out of the safe operating range. I would get some nasty static sounds if I turned it up to loud but for the most part it still sounded great. Now Im wondering if I have yet to hear any gainclone as it is supposed to sound.

Well, off to active surplus yet again

[janneman]

Originally posted by Nuuk
Thanks Jan. I am always relieved when somebody who does know what he is talking about confirms what I have said!

Hey, you're OK kid!

Jan Didden

[thomas997]

Just curious, but anyone know what the minimum voltage is?


The datasheet says 20V, but Im sure you can run a little less than that right?

Because I have about 18.66V



thx

[ir]

the minimum operating voltage is determined by two integrated design rules of the national chips

1. the negative rail MUST be at less then -9V to turn off the undervoltage protection

2. there MUST be a difference of at least 14V between positive and negative

so then, the minimum voltage would be +VE=+5V, -VE=-9V