I'm considering building a small powered speaker, and I'm working through building a power supply for it.
It will need to be a dual power supply, and I've read unregulated supplies are typically used for audio amps because they're cheap and they work. I may add a regulator but thats for another thought.

Vin 120v AC 60hz,
Transformer ratings I havent settled on, but i'm looking at something rated for at least 2A in parallel configuration, with 12v output (should be a 10:1 windings)
Anticipating about 2A output, I started playing around in LTSpice, and found to get a reasonably smooth waveform at about 2A output the smoothing capacitor must be over 4000uF, and the current is huge.

Looking at a handful of example circuits, this is about what the circuit looks like, but I don't think this will work. The transformer will burn itself up wont it?

Any thoughts or links that might help me with smoothing this (or pointing out what i've done wrong)?

I've attached a screengrab of the sim.


I understand a 12v 2A powersupply can be bought for something like $5, but this can be scaled upwards, to something closer to 100W or more. Its a learning experience.

 

Have a look at the pdf here:
https://usermanual.wiki/Document/HammondRectifierDesignGuide.146793413/html

 

hi T,
The very low transformer LTSpice inductances are not realistic for a mains transformer.
E

 

you can actually quite precise measure your TF by the RMS vlotage~current
PS! the mains voltage varies and the resulting XL also has a systematic drift along with the V.RMS mains
. . . so take 10 to 20 samples at different voltage values
(might be a hell to wait for the stable isle in time to be able to read your "drifting" meters)
. . . for low power 240VRMS TF the primary has . . . (see fig.)
. . . for "sclaing" the coils L.new = L.old · (U.new/U.old)²
-- actually it may be less than square dependence of amp turns versus inductance
about Single-Layer Coil Inductance Calculator, Electrical, RF and Electronics Calculators, Unit Converter

 

What would be realistic inductances? I can't find them specified on the Transformers I'm considering?

 

hi T,
This link is a good source of information on mains transformers.
https://sound-au.com/xfmr.htm

High H values for the primary and milliH for secondary

E

 

What power output to the speaker are you seeking?
If it is a linear amplifier you do not need to regulate the power supply output.
The size of the capacitor will depend on how much ripple is acceptable at the output stage. Make sure that the power supplies at the preamp stages are properly filtered.
As you increase the capacitance, the peak diode current will increase, as you already know.

 

What power output to the speaker are you seeking?
If it is a linear amplifier you do not need to regulate the power supply output.
The size of the capacitor will depend on how much ripple is acceptable at the output stage. Make sure that the power supplies at the preamp stages are properly filtered.
As you increase the capacitance, the peak diode current will increase, as you already know.

I'm looking for something in the neighbourhood of 5-10w per channel stereo, I don't recall exactly but when I last worked out the numbers I had peak output current to the speakers approaching 2A. My present concern is the power supply, and the capacitors sucking on 10-20A or more. What limits the charge rate?

 

hi,
With a high value 'recharge' current peak for the 4000uF , ensure that the rectifiers can handle that peak current.
BTW: the transformer must be capable of supplying the high current peaks

Check out this LTS typical circuit.
E

 

with the 12V 2A & 60Hz output the current is not "that" high but is expected to be high - just tile many 470µF 680µF 820µF or 1000µF in parallel

. . . ? 100W RMS → ← P=U²/R → U.rms=√(PR) → U.p-p=√(P·R·2) , U.p-p=±40V for 8Ω output , U.p-p=±28V for 4Ω output

IF it's a stereo , with 50W per channel then with 12VDC you got (without output transformer) ±6V.p-p →
→ R=U².rms/P=2·U².p-p/P=1.44Ω speakers ← it's about 3x 4Ω speakers in parallel

 

With a high value 'recharge' current peak for the 4000uF , ensure that the rectifiers can handle that peak current.
BTW: the transformer must be capable of supplying the high current peaks

That's exactly my concern. if the transformer is rated for 2A output, which is the output i'm looking for, but because of the smoothing capacitors the charging current is 20+A on the first cycle. When i reduced the capacitor to 470uF to keep the current down, the ripple basically becomes a full wave. I could parallel 10 470uF capacitors, but that still doesn't do anything to reduce the transformer current.

I'm going to be digesting and simulating some of the suggested information posted earlier. I hadn't considered adding an LC filter to the output (should have, but didn't), mainly because all the top hits on google searches were just capacitors shorted across the transformer outputs.

 

You Can't get 2A smoothed DC from a 2A transformer. From the link in post #2 the best you could expect with a bridge rectifier is 1.2A

 

In today's power crazy world, 5-10W per channel would be considered low. That is to your advantage.

Let's do some quick math.
±12VDC supply would be about 8V RMS.
8V RMS into an 8Ω speaker is 8V x 8V / 8Ω = 8W
That is right in your ballpark of 5-10W per channel.
Average current is 8V / 8Ω = 1A per channel or 2A for stereo. Hence your math seems right.

Now let us calculate the ripple voltage.
A good estimate for 50Hz fullwave rectification is
ΔV = 10 x I / C
where ripple voltage ΔV is in volts
Load current I in mA
Capacitance C in μF

For I = 2000mA, C = 10,000 μF

ΔV = 10 x 2000mA / 10,000 μF = 2V

For 60Hz line frequency ΔV would be 2V x 50/60 = 1.6V
Note that this ripple voltage occurs only on full load.

The peak diode current will be approximately 5 to 10 times the average current depending of the resistance of the transformer and the rectifier diodes. Hence make sure that the diodes can handle 10-20A peak current.

 

I think with a low power 2A transformer you will not reach your 8W per channel goal. However, it would still make a decent amplifier. A 4700μF on each side of the power supply would be adequate.

 

Have you selected and purchased the transformer already?
I would recommend getting a Hammond transformer. They are conservatively spec'd. You get what you pay.

 

Have you selected and purchased the transformer already?
I would recommend getting a Hammond transformer. They are conservatively spec'd. You get what you pay.

https://www.digikey.ca/product-detail/en/triad-magnetics/VPP24-1250/237-1078-ND/242522
I've been leaning towards this one, but haven't chosen yet as I'm still working towards understanding everything about this.
Also considering;
https://canada.newark.com/multicomp/mcta030-12/toroidal-transformer/dp/38K4875?st=toroidal power transformer

I guess it depends how far I want to go with it.
I'll keep your recommendation in mind.

edit;
https://canada.newark.com/multicomp...rmer/dp/38K4904?st=toroidal power transformer
This is on promotion for only a little more. Thats the thing, theres always something better for 15% more money. at every level.