Chip amp power supply- a beginners guide

[Redshift187]

Excellent work tent. Now you should connect the new toroids to the LM1875 PSU, check for light, then also the LM1875 amp and check for light. Then you'll know if it was entirely the transformers fault

[tent]

Originally posted by Redshift187
Excellent work tent. Now you should connect the new toroids to the LM1875 PSU, check for light, then also the LM1875 amp and check for light. Then you'll know if it was entirely the transformers fault

Hi redshift,
well I did it! I have now two sets of new toroids, 12V and 24V but my question is exactly that now I'm not seing any light coming from the lamp if I connect only the trafos (do they seem ok now) but do see some very very dim one with the LM1875 psu and even more with the loudspeakers, is that normal? I also did some mixed test with PSU from LM3886 and amp LM1875 and also there a small orange filament can be observed, but now on a full LM3886 version. Could I still have something maybe in the amp of LM1875? or the speakers? or is this perfeclty normal dissipation somehow?

Thanks,
tent:wq

PS: the summary was:

Toroids (two NEW 2x12V, 80VA) = NO LIGHT AT ALL

LM1875 psu = ALMOST NOT NOTICEABLE ORANGE FILAMENT
LM1875 psu+amp = ALMOST NOT NOTICEABLE ORANGE FILAMENT
LM1875 psu+amp+speakers = ORANGE FILAMENT A BIT MORE NOTICEABLE
(offset 14mVdc) so should I worry about the filament?

LM3886 psu = NO LIGHT AT ALL
LM3886 psu+LM1875 amp = ALMOST NOT NOTICEABLE ORANGE FILAMENT
LM3886 psu+LM1875 amp +speakers = ALMOST NOT NOTICEABLE ORANGE FILAMENT
(offset 107mVdc) maybe related to the fact that GN+ and GN- go to the same point on the amp?

Other thoroids (two other NEW 2x24V, 85VA) = NO LIGHT AT ALL
LM3886 psu+amp = NO LIGHT AT ALL
LM3886 psu+amp+speakers = NO LIGHT AT ALL
(offset 4mVdc !!) this sound great!

[AndrewT]

check the output offset and output noise with and without a muted signal source connected.
If this is OK then connect your speakers and listen to the music.

[Redshift187]

Sorry, I didn't realize those results were only with the new transformers.

Are both PSUs chipamp.com PSUs, but one is for the LM1875 and the other is for the LM3886?

There would obviously be a little current at idle, how much it takes to make a 60W bulb glow very dimly I don't know. As AndrewT said, if DC offset on the output is low with a source connected and the volume not at minimum, it should be fine.

[tent]

Originally posted by AndrewT
check the output offset and output noise with and without a muted signal source connected.
If this is OK then connect your speakers and listen to the music.

Originally posted by Redshift187
Sorry, I didn't realize those results were only with the new transformers.

Are both PSUs chipamp.com PSUs, but one is for the LM1875 and the other is for the LM3886?

There would obviously be a little current at idle, how much it takes to make a 60W bulb glow very dimly I don't know. As AndrewT said, if DC offset on the output is low with a source connected and the volume not at minimum, it should be fine.

Hi,
yes exactly, both PSU from chipamp.com each for the corresponding LM.
I measured with a muted source ad volume halfway the following:
. 13mVdc on the left channel of LM3886
. 32mVdc on the right channel of LM3886 (is something wrong with that channel?) and
. 112mVdc on both channels of the LM1875 (so is there here something wrong probably?)
The lamp turns dimly only on the LM1875 but somehow noticably only with connected speakers..

Kr,
tent:wq

[AndrewT]

Originally posted by tent






Hi,
yes exactly, both PSU from chipamp.com each for the corresponding LM.
I measured with a muted source ad volume halfway the following:
. 13mVdc on the left channel of LM3886
. 32mVdc on the right channel of LM3886 (is something wrong with that channel?) and
. 112mVdc on both channels of the LM1875 (so is there here something wrong probably?)
The lamp turns dimly only on the LM1875 but somehow noticably only with connected speakers..

Kr,
tent:wq

In general the bulb tester is used for checking that the amplifier is properly wired up.
This should be completed and then the bulb tester should normally be removed.
Then connect your speakers to your thoroughly tested amplifier.

The output offset is checked with no load on the output.
If you connect an 8ohm speaker with R~6r0 then 112mV of offset will cause ~19mA to flow through the load. If the offset changes when your source is connected or when you alter the volume control then the offset current through the load will change.

This could account for the dim glow you are seeing in the bulb tester filament.

Look at the datasheet for these chipamps. You will find a large variation in quiescent current (no load connected). This too will account for differences in the filament temperature.

[neeraj]

Originally posted by AndrewT
two channels of 68W+68W require a transformer between 136VA and 272VA.
Your 250VA is perfect.
The 5.8Apk only applies to sinewave into a resistive (non reactive) load.
The transient peak current into a 4ohm reactive speaker load could be around three times this, i.e. ~17Apk.
That is way in excess of the chipamp specification. The transformer can easily meet this transient requirement.

Hi AndrewT,
what is the VA for a trafo you recommend for an 8 ohm load.

Is 225 VA X 25-0-25 V is fine for single channel?

[Nuuk]

Neeraj, a lot depends on the speakers involved, particularly how efficient they are.

For instance, I can get a reasonable performance form 12v rails with some 12 inch full-range drivers, but I would use that type of supply for more demanding drivers.

So it would be very useful to know the exact drivers that your amp will be powering, and how they are connected up, ie passive crossovers etc.

[AndrewT]

Originally posted by neeraj
what is the VA for a trafo you recommend for an 8 ohm load.

Is 225 VA X 25-0-25 V is fine for single channel?

two channels of 68W+68W require a transformer between 136VA and 272VA.

do the arithmetic.
225VA can support between 112.5W and 225W of maximum total output power, economically.

I have a pair of 3886 running on a 225VA 25+25Vac toroid. It works adequately with a 1.3C/W sink.

[neeraj]

Thanks Nukk & Andrew T,

Originally posted by Nuuk
So it would be very useful to know the exact drivers that your amp will be powering, and how they are connected up, ie passive crossovers etc.

Speakers Are 8 ohm 3 way 100 W RMS Sealed Enclosure (DIY)
Woofer 100 WRMS
MID 90 W RMS
Dome Tweeter 60 W RMS
With 12db /Octave crossover network

Originally posted by AndrewT

do the arithmetic.
225VA can support between 112.5W and 225W of maximum total output power, economically.

I have a pair of 3886 running on a 225VA 25+25Vac toroid. It works adequately with a 1.3C/W sink.

As i have read somewhere that 47 labs used 170 VA /channel in Original Gainclone ( 8 Ohm,50 W Model).
Plz confirm A=(VA*PF)/V then 170 VA= 4.08Amp./Rail is it accurate?