Use two Transformers per chip amp?

[zenon]

I will be buying the audio sector kits.
Does that mean, when I look at the national .xls design sheet and it tells me the total power dissipated per IC = 52.31 Watts that it's actually going to be 104.62 Watts??

By cope I assume you mean heatsinking. I will be using large heatsinks, each chip will have it's own large aluminum plate.

[AndrewT]

Hi,
I am not sure you realise the implications of your decision to go bridged.
Bridging produces twice the power into twice the impedance.

If you choose 8ohm speakers then the amp MUST be designed for a 4ohm load. DO NOT choose 4 to 8ohm speakers for a bridged chipamp.

If the chosen +-Vrail allows 50W into 4ohm then the bridged amp will produce 100W into 8ohm.

The smoothing MUST be designed for each amplifier on 4ohm loading.
You should fit +-10mF on every amplfier. For stereo that is 4 times +-10mF = 80,000uF if that means more to you.

A plate of aluminium may not be enough. Read the data sheet on heatsink design.

[Nico Ras]

Originally posted by Nordic
Isn't this true for just about any transformer... I have yet two see a transformer with 2 exactly matched windings...

Where do you buy your transformers Nordic? If they not the same take them back.

It is equally simple to make them right than to make them wrong.

Low voltage transformers with equal windings should be within 50mV of each other.

Regards Nico

[AndrewT]

Originally posted by Nico Ras
Where do you buy your transformers Nordic? If they not the same take them back.

It is equally simple to make them right than to make them wrong.

Low voltage transformers with equal windings should be within 50mV of each other.

Hi,
it's even better than that with a bifillar wound dual secondary toroid.
The number of turns on the secondary are identical. That makes the voltage on the two secondaries identical.
There is room for a small difference in resistance between the two secondaries. Since they are bifillar they will be wound from different spools and subject to production tolerances on wire diameter. The tension on each winding may be set up differently and this could theoretically produce a difference in wound diameter, but I suspect this error is negliable.
The next biggest resistance error comes from the different installed length on each winding. This could give rise to upto a 1% or 2% difference in resistance if one winding were consistently following a longer route over the underlying windings. But 1% of a secondary resistance will lead to a difference of just a few milliohms.
Effectively the bifillar wound toroid has identical voltage under load from each winding.

There is one area where significant voltage difference could occur.
If the winding termination were badly implemented allowing the turns number to vary between windings. eg. 35 T vs 35.2T will give a voltage error that cannot be removed other than to reterminate the winding ends to ensure identical loops and wire lengths.

[head_spaz]

richie00boy says:

I would choose 11-0-11 transformers and wire secondaries in series rather than buy a pair of 22-0-22 and wire parallel. When paralleling, any slight voltage mismatch will create large currents to flow which will impact the quality of your supply.

I fully agree. Good thinking there.


It ends up looking like this :

[jaycee]

Right layout but showing the wrong voltages Remember that you multiply by 1.414 after the rectifiers. When suitably smoothed you will have 15V and 31V supplies (a bit less under load)

In this configuration you could regulate the 15V supplies to 12V and use this for a preamp.

[john65b]

After smoothing caps you will see the rail *1.414 voltages, correct?

[Nico Ras]

Originally posted by john65b
After smoothing caps you will see the rail *1.414 voltages, correct?

Only under no load, then the average voltage read on a meter is the same as the peak voltage since the capacitors are charged to the peak voltage. When you draw current then the average voltage decreases, i.e. ripple occurs.