no your transformer has a 2.5Aac rating.
That is the bit written in stone.
If you use it to generate a 40Vdc supply then the AC rating is equivalent to (40+0.7)/sqrt(2) ~28.8Vac
The VA rating is equivalent to 144VA.
The transformer delivering 40Vdc will power amplifiers totaling 72W to 144W satisfactorily.
That is 36W+36W up to 72W+72W for a stereo build.
Two 50W to 60W 3886 will run satisfactorily from this transformer.
It makes no difference to the transformer, whether it is unwound to give +-40Vdc, or is regulated down to +-40Vdc.
BUT.
I personally do not recommend that any Beginner tries to build a regulated supply Power Amplifier.
Post10
offers in my opinion the BEST advice.
Last edited:
The heat load of a transformer comes from copper losses in the primary, copper losses in the secondary and iron losses in the core.
This heat load in general limits the VA rating of the transformer. Almost all "normal" transformers are designed to a limiting temperature. A very few are designed to a limiting regulation, but these are usually SPECIFIED by the customer and manufactured to the customer's requirements.
Now back to the three parts of the heat load.
If one reduces the resistance of the secondary, then the total heat load has dropped.
One could increase the current slightly to bring the total heat load back to the design value and to reach the design limiting temperature.
There is another way to reduce the secondary heat load:
ADD copper. That could be by adding a paralleling winding, or by adding extra turns.
adding 50% secondary windings and reducing the OUTPUT current to maintain the same VA will result in the same primary copper losses and the same iron losses, but significantly less secondary copper losses. 100%Vac*100%Iac = 150%Vac*67%Iac. The IsquaredR loss is 0.67²*1.5= 67% of the original secondary copper losses.
One could very slightly increase the OUTPUT VA to bring the total losses back up to the design levels.
The increase in VA, due to reducing the secondary copper losses, is quite small . So small it is probably not worth the effort of even trying to model and simulate for the effect.
Except that I have added more insulation and increased the thickness of the layers so trapping more heat inside the core.
I just accept that my "added copper" modified toroid may run a tiny bit cooler than standard.
This heat load in general limits the VA rating of the transformer. Almost all "normal" transformers are designed to a limiting temperature. A very few are designed to a limiting regulation, but these are usually SPECIFIED by the customer and manufactured to the customer's requirements.
Now back to the three parts of the heat load.
If one reduces the resistance of the secondary, then the total heat load has dropped.
One could increase the current slightly to bring the total heat load back to the design value and to reach the design limiting temperature.
There is another way to reduce the secondary heat load:
ADD copper. That could be by adding a paralleling winding, or by adding extra turns.
adding 50% secondary windings and reducing the OUTPUT current to maintain the same VA will result in the same primary copper losses and the same iron losses, but significantly less secondary copper losses. 100%Vac*100%Iac = 150%Vac*67%Iac. The IsquaredR loss is 0.67²*1.5= 67% of the original secondary copper losses.
One could very slightly increase the OUTPUT VA to bring the total losses back up to the design levels.
The increase in VA, due to reducing the secondary copper losses, is quite small . So small it is probably not worth the effort of even trying to model and simulate for the effect.
Except that I have added more insulation and increased the thickness of the layers so trapping more heat inside the core.
I just accept that my "added copper" modified toroid may run a tiny bit cooler than standard.
Last edited:
Thank you AndrewT and everyone else who chimed in.
Glad to read the math and now understand this transformer will work with modification or regulation.
Against recommendation I am going to at least try building a regulated supply since I do have a background in designing such circuits.
My plan is to use a LM317 and LM337 with some slightly overrated power transistors for the +- rail regulation and of course take safety precautions where needed.
I actually feel more confident going about it this way rather than mess with the transformer winding, unless the latter turns out to be more practical for some reason.
My previous chipamp I designed using a TDA7384 was built around a regulated supply and I swear it had tighter bass and detail than it did when I originally tested it with an unregulated one, but that may not apply with this current project. Time and testing will tell.
Suppose I will save that question for a new thread so not to get off topic.
Glad to read the math and now understand this transformer will work with modification or regulation.
Against recommendation I am going to at least try building a regulated supply since I do have a background in designing such circuits.
My plan is to use a LM317 and LM337 with some slightly overrated power transistors for the +- rail regulation and of course take safety precautions where needed.
I actually feel more confident going about it this way rather than mess with the transformer winding, unless the latter turns out to be more practical for some reason.
My previous chipamp I designed using a TDA7384 was built around a regulated supply and I swear it had tighter bass and detail than it did when I originally tested it with an unregulated one, but that may not apply with this current project. Time and testing will tell.
Suppose I will save that question for a new thread so not to get off topic.
Hello there
You could power the 3886 by single supply. So, 37V*1,41=52V aprox. In the datasheet it`s specify this thing. And no more complications.
Cheers.
I thought about that but it would limit the power output of the chipamps and require output coupling capacitors plus an extra transistor as shown in the single supply example in the datasheet.
For charge~4ohms it`s exactly what you need for, right near the action of limiting zone from inside of chip. Plus, you don`t need the protection module(extra relay+uPC1237+....).
One final question..
If I were to unwind some windings from the secondary should I then spread the remaining unwound windings out over where the removed windings were? Or should I just unwind what I need and leave the original windings where they are on the toroidal core?
I just realized the transformer is rated at 37-0-37vac at 115vac input. After measuring the secondary while plugged in it is actually 39.5-0-39.5vac which is even higher than I thought it would be!
I forgot the wall voltage here runs at 125v, even though many appliances are rated for 115 or 117.
So looks like I may have to unwind some regardless even if I later decide to add some regulation to it.
If I were to unwind some windings from the secondary should I then spread the remaining unwound windings out over where the removed windings were? Or should I just unwind what I need and leave the original windings where they are on the toroidal core?
I just realized the transformer is rated at 37-0-37vac at 115vac input. After measuring the secondary while plugged in it is actually 39.5-0-39.5vac which is even higher than I thought it would be!
I forgot the wall voltage here runs at 125v, even though many appliances are rated for 115 or 117.
So looks like I may have to unwind some regardless even if I later decide to add some regulation to it.
I'm going to go ahead and reiterate my suggestion that you use a lme49811 (lm4702 might be better) kit/pcb and have a go at that. Plenty of documentation on this site plus AN1490. (http://www.ti.com/lit/an/snaa031a/snaa031a.pdf)
The likelihood of it actually ending up better than your present machinations of unwinding the transformer is quite, quite high, and the end result *should* be better than any of the bog standard chip amps around.
The likelihood of it actually ending up better than your present machinations of unwinding the transformer is quite, quite high, and the end result *should* be better than any of the bog standard chip amps around.
Thanks for the suggestion however I am dead set on using the parts I have on my bench right now along with my sample LM3886s. Also no plans for using a PCB or kit form. I plan on either using point to point or perfboarding the project to keep costs bare minimal.
I just happened to have these parts and figured it would be a fun side project that could end up becoming something quite nice out of some scrap parts.
Hah, interesting idea.
However that would just add more resistance to the winding.
Not like removing some windings is really that hard to do.
I hope someone answers my previous question though since I need to unwind some to drop the voltage as now I realize there is no other reliable option.
Regulation is only an idea rolling around in my head and if I were to go that route I would run the transformer at almost top voltage the chipamps could handle by itself safely, and then regulate them down to a sane level. That way if the regulator were to fail and short it would not overvolt the chipamp beyond what it can handle.
I figure the regulation would give it some reserve for voltage fluctuations and help with low frequency audio, but that is beyond the topic at hand.
Last edited:
well there's another way , easiest of them all > use an DIY autotransformer on the primary ( some call it a bucking transformer IDK why tho )
find another much smaller* transformer where the secondary voltage matches the amount of primary voltage you want to reduce the mains by. *how small you ask, the secondary current rating the added device should be equal or greater then primary of the main transformer.
find another much smaller* transformer where the secondary voltage matches the amount of primary voltage you want to reduce the mains by. *how small you ask, the secondary current rating the added device should be equal or greater then primary of the main transformer.
Last edited:
A 100VA 115:25+25Vac transformer could be wired as an autotransformer to step down your supply voltage. The output of an autotransformer is NOT isolated so you still consider it's leads as LIVE.
115:50 (the two windings in series) gives a working autotransformer rating of 115/50 * 100VA or approximately 230VA and the maximum step down would be 115/(115+50) = 0.657times i.e. 125*0.657 ~ 82Vac
The VA rating of the autotransformer goes up as the total secondary voltage goes down.
a 115:15+15Vac 100VA would become 380VA and the step down would be 0.793times.
you choose the secondary voltage to give the step down you require and then use that to determine the VA rating you would need to buy.
To get a step down autotransformer:
wire the two secondaries in series.
wire the primary in series with the series connected secondaries.
Apply mains power to this three winding combination i.e. your 120Vac is connected to {115+sec1+sec2}
The output is taken from the top and bottom connections of the 115Vac Primary.
You MUST use a Mains Bulb Tester to power this up to allow you to measure the voltages and ensure you don't damage yourself or your transformer due to a wiring/phasing error.
REMEMBER:
the autotransformer is LIVE. All tappings are LIVE. be careful.
115:50 (the two windings in series) gives a working autotransformer rating of 115/50 * 100VA or approximately 230VA and the maximum step down would be 115/(115+50) = 0.657times i.e. 125*0.657 ~ 82Vac
The VA rating of the autotransformer goes up as the total secondary voltage goes down.
a 115:15+15Vac 100VA would become 380VA and the step down would be 0.793times.
you choose the secondary voltage to give the step down you require and then use that to determine the VA rating you would need to buy.
To get a step down autotransformer:
wire the two secondaries in series.
wire the primary in series with the series connected secondaries.
Apply mains power to this three winding combination i.e. your 120Vac is connected to {115+sec1+sec2}
The output is taken from the top and bottom connections of the 115Vac Primary.
You MUST use a Mains Bulb Tester to power this up to allow you to measure the voltages and ensure you don't damage yourself or your transformer due to a wiring/phasing error.
REMEMBER:
the autotransformer is LIVE. All tappings are LIVE. be careful.
Darklife, what about using a SMPS?
I have one here that's +/-28V unregulated. Bought a couple a few years ago and never put them to use. If you're interested I'll post a picture of it. Can be yours for the cost of shipping.
It's similar to this one...
POWER SUPPLY 500W +/-27V DC SMPS | Class D Audio Amplifiers
Unregulated half-bridge running at 80KHz. If you populate some missing components you can also get unregulated +/-12V outputs and there's spots for linear regs.
I have one here that's +/-28V unregulated. Bought a couple a few years ago and never put them to use. If you're interested I'll post a picture of it. Can be yours for the cost of shipping.
It's similar to this one...
POWER SUPPLY 500W +/-27V DC SMPS | Class D Audio Amplifiers
Unregulated half-bridge running at 80KHz. If you populate some missing components you can also get unregulated +/-12V outputs and there's spots for linear regs.
Oh brother
your world must be so dangerous yet all too complicated.
why would he need to pick a dual secondary to build a simple DIY autotransformer.
your numbers are too convoluted to follow, so no wonder you tell people to just buy more iron.
Last edited:
At least he's not dead! Safety here and promoting the understanding of electrical safety is of number one importance on DIYaudio. If you can't do it safely or understand what the potential risks are then do not do it all.
This is more important for us using 240V mains than you guys running on half of that.
Last edited:
I get that, the OP gets that.
the message cant be modulated with a sledgehammer?
If the OP expressed interest in the autotransformer method, any further instructions would naturally advise mains wiring ahead. but IMO Andrews machinations make it seem all too complicated fraught with dangers on every turn.
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.