Hello,

I am in middle of building Gainclone amp. It's based on two LM3886 op amps. Power requirement is +25v and -25v. This is minor problem, that I cannot figure out myself and there probably a thread here somewhere, but I failed to find direct answer.

Center tapped transformers are expensive and hard to get here, where I am from. Thus I'm interested in getting both - positive and negative voltages from dual rail transformer. I have attached a picture, that I found on internet (unfortunately I don't know the author). Could someone please comment, if this would provide the outcome I desire?

Additionally - thing that confuses me, is that each (positive and negative) side has it's own ground. (Here comes a amateur question) - should I treat these separate grounds somehow special? Or should I just wind them together and make common ground as usually?

 

hi k,
The transformer you have posted has two isolated secondary windings are you able to purchase a transformer of this type.?
If not, you could use two separate mains transformers to give + and - supplies.
E

 

Hi,

I thought of this as option, but two separate transformers cost about the same as one with two secondary windings. I have some old ones laying around, but I believe they could fail to offer the needed power needs.

Regards

 

your circuit is ok, just join the two PGND +/- together, this will give you a common ground and a positive and negative 25V.


http://diyaudioprojects.com/Chip/DI...er/TDA2050-Chipamp-Power-Supply-Schematic.png

 

your circuit is ok, just join the two PGND +/- together, this will give you a common ground and a positive and negative 25V.


http://diyaudioprojects.com/Chip/DI...er/TDA2050-Chipamp-Power-Supply-Schematic.png

Hi,

Thanks for thumbs up. Just to be sure, in this case 1R resistor is meant to be 1 ohm, correct?

 

Yes 1R is 1 ohm.

 

Yes 1R is 1 ohm.

Thanks. One more question about this circuit, - if I understand it right, then if the transformer gives 18vac, then after bridge and caps, the output will be something around 25vdc? Can I rely on that or will it drop when load is added?
Datasheet for LM3886 tells me, that range for power supply has to be within 20v - 80v. That's why I'm asking.

Thank you in advance!

 

Your output voltage will drop under load possibly to 20v or less, i would use a transformer with a voltage of 22-30Vac, then it should be in the range you need, or use voltage regulators for precision.

 

Note that you should derate the transformer RMS current rating at least 40% as compared to the needed capacitor-rectifier DC current output due to the high peak RMS current pulses the capacitor-rectifier draws from the transformer.

 

So why not just tie the two independent secondary windings end to end to make it into a center tapped transformer?

 

I told him that earlier in a pm, with this image...i think he's already made that circuit...

https://www.google.co.uk/search?q=centre+tapped+psu&prmd=inv&source=lnms&tbm=isch&sa=X&ved=0ahUKEwiZ6PuU147LAhXEWBQKHVETB0wQ_AUIBygB&biw=360&bih=511#imgrc=z6hSGK5ZlbCnWM:

 

Thank you for advice. I think I will go with joining the two transformers into one.

I am keeper and re-user, so I have all sorts of salvaged transformers. If I take two different transformers?
This would probably mean secondary voltages would differ slightly (within range of ~2v) and power rating as well. This would end bad, wouldn't it?

 

May do if the transformers are of different VA ratings, otherwise shouldn't make much difference.

 

I have recently found several DIY projects where people are making these Gainclones which all have two LM3886 op-amps. None of them actually tell, what transformer needs to be used aside from one author claiming that it should be at least 120 VA.
Based on amps datasheet power dissipation is 125 w, so I can assume, that I need at least 106.25 VA per amp and totally forget what he suggested?
So either, they are lucky that the transformers are still functioning (if of course they are) or without driving the amps to full capacity they just never hit that peak where it starts to overheat.

 

Very few audio amplifiers are ever working continuously anywhere near their maximum wattage rating so that's why they can get by with a slightly lower primary transformer VA rating than their output wattage rating would suggest they would need.

Add in the fact that if a transformer is rated for 100 VA it doesn't instantaneously burn up if a 101 VA load gets applied to it and that's how a lower VA rated transformer can live for years without burning up even in a hard run audio amplifier rig.

In the power supply end the peak demands only make up a small percentage of the average power usage which overall is well below the transformers continuous capacity.

 

Add in the fact that if a transformer is rated for 100 VA it doesn't instantaneously burn up if a 101 VA load gets applied to it and that's how a lower VA rated transformer can live for years without burning up even in a hard run audio amplifier rig.

In the power supply end the peak demands only make up a small percentage of the average power usage which overall is well below the transformers continuous capacity.

So can I provide supply based on average power usage instead of peak usage? If so, then for 4ohm speaker outputs, I'd need only 57.8 VA per apm, thus the originally suggested transformer of 120VA could actually be used?
*sort of glad I tried to overshoot rather than go under.

 

According to this, the rule-of-thumb for the average audio power to a speaker for a very loud music listening level (85dB or similar to a loud movie theater ) in a medium size (3000 cu ft) would be between 50W to 80W.
Although that may sound like a large range of power it only represents a 2dB difference in perceived sound intensity, which is barely audible.
If you listen at a more sedate level (but still fairly loud) of half that perceived sound level or 75dB (a normal piano level is 60-70-dB) then the average amplifier power would be reduced to only 5W to 8W.

 

Yesterday I disassembled old (sort of) radio and got transformer, that seems to be massifve. Problem is that I cannot find any official datasheet for it. After searching rtp2n3b013 I found the picture with the same transformer (attached), but all the info is in Polish. Despising that I got a hint, that it should contain 2x23vac and 2x40vac secondaries.
Problem, is that before this I haven't had to determine the windings, without any color reference. As I understand, then there should be pairs of terminals with continuity, correct? After a bit of googling, I found that the secondaries can be in phases and in such case, separate secondaries have to be determined by resistances in between. Can someone maybe explain this please? In some sort of simplified version.

 

Start out by measuring each pair of outputs, mark them up with the voltage,

For phasing, put the same pair of voltage windings in series and if correctly wired will give you twice the voltage, if incorrectly wired will be near to zero.

 

So can I provide supply based on average power usage instead of peak usage? If so, then for 4ohm speaker outputs, I'd need only 57.8 VA per apm, thus the originally suggested transformer of 120VA could actually be used?

Yes the odds are it will work just fine. To be honest for most average listening applications a 60 VA would probably still handle things well.

Build it and find out. At worst you find it can't handle your listening preferences and runs too hot so you then have to get a bigger one. If not life is good and you did a lot of worrying over nothing.