By "offline dc power supply" do you mean an AC-DC switcher and the other a 60Hz transformer?
If so then the AC-DC is regulated (99.9% of the time) and the other will have a certain amount of ripple...unless regulated.
The AC-DC will have frequency components above hearing range. (IMHO...some people disagree). Whereas the 60Hz transformer will have frequency components at 120Hz and above (within hearing range)...
(But with good design neither of these issues with either power supply types will show up in the final output.)
With proper design the AC-DC switcher can draw current from a larger portion of the supply waveform...labeled as PFC type. Actually producing less of a current spike on the AC supply and less source losses. (small difference)
Biggest difference...overall efficiency...AC-DC switcher wins
Oops...and size...AC-DC switcher wins
...and heat...AC-DC switcher wins
simplicity...60 Hz transformer wins
IMHO
If so then the AC-DC is regulated (99.9% of the time) and the other will have a certain amount of ripple...unless regulated.
The AC-DC will have frequency components above hearing range. (IMHO...some people disagree). Whereas the 60Hz transformer will have frequency components at 120Hz and above (within hearing range)...
(But with good design neither of these issues with either power supply types will show up in the final output.)
With proper design the AC-DC switcher can draw current from a larger portion of the supply waveform...labeled as PFC type. Actually producing less of a current spike on the AC supply and less source losses. (small difference)
Biggest difference...overall efficiency...AC-DC switcher wins
Oops...and size...AC-DC switcher wins
...and heat...AC-DC switcher wins
simplicity...60 Hz transformer wins
IMHO
Ha yes i've come around too, nolonger are switchers an evil word for audio, the advantages now outweigh any possible negatives, provided they are well designed they can be quieter, more efficient and fully regulated for high power applications, which is a big advantage in my book
Thank you for the great information. This points me in the right direction to start researching. This may sound a bit dumb but would it be possible to bypass a car audio amp power supply with a home built solution with minimal side effects to sound quality assuming the built supply runs the amp rails at the same voltage and is able to supply the necessary current.
The advantages of swithcing PSU's over the linear ones have become more and more important so they are more and more used, this would be some of them:
-The production cost is less;
-The overall size is smaller;
-Theyr weight is less;
-Well Regulated voltage on the outputs on much higher eff.;
-Theyr better efficiency means less heat generated ( much less most of the times );
-Less influence from mains variations, meaning you can get the same power from much larger voltage range on the mains;
All of that comes at the price of much more complexity so a good design is much more tough to make.
PS: Yes, you can bypass a car audio power amplifier's PSU, you have to make sure the new unit generates the same VA characteristics and all should be good.
-The production cost is less;
-The overall size is smaller;
-Theyr weight is less;
-Well Regulated voltage on the outputs on much higher eff.;
-Theyr better efficiency means less heat generated ( much less most of the times );
-Less influence from mains variations, meaning you can get the same power from much larger voltage range on the mains;
All of that comes at the price of much more complexity so a good design is much more tough to make.
PS: Yes, you can bypass a car audio power amplifier's PSU, you have to make sure the new unit generates the same VA characteristics and all should be good.
I'll repeat the advantage of the linear PSU - Simplicity.
I believe there is an even bigger advantage.
Enormous short tern current capability. Easily 10times the rated continuous output current rating !!!!
Your 1kVA transformer with 60Vac secondaries is rated @ 8.33Aac and this becomes a continuous 4.1Adc after the smoothing capacitors. Short term current output can exceed 41Apk without any distress. And just as importantly:- the output voltage while delivering that transient, will hold up to within 90% to 95% of the rated output voltage.
Try to find an AC to DC switching PSU that can give +-90Vdc and give 41Apk while still maintaining >85Vdc on the loaded supply rail. I'll guess that a 4kVA switcher that has a transient capability of >=180% of rated continuous output current will meet the 1kVA spec of the linear supply.
I believe there is an even bigger advantage.
Enormous short tern current capability. Easily 10times the rated continuous output current rating !!!!
Your 1kVA transformer with 60Vac secondaries is rated @ 8.33Aac and this becomes a continuous 4.1Adc after the smoothing capacitors. Short term current output can exceed 41Apk without any distress. And just as importantly:- the output voltage while delivering that transient, will hold up to within 90% to 95% of the rated output voltage.
Try to find an AC to DC switching PSU that can give +-90Vdc and give 41Apk while still maintaining >85Vdc on the loaded supply rail. I'll guess that a 4kVA switcher that has a transient capability of >=180% of rated continuous output current will meet the 1kVA spec of the linear supply.
Last edited:
The 50/60hz power transformer can handle more load current, yes, but at what price i ask you? well the unswear is that the output voltage will drop verry much with increase load current, that peak current you speak of can happen only on verry big loads or more like shorts on the outputs, well aside from the fact that the bridg rectifier diodes would most probably die, the output voltage will by no means stay nowhere near 95% percent the rated output, it is almost impossible to get that, and all of that would be with a sorf of huge power transformer, and ofcourse a verry expencive one, thus the SMPS gains more ground every day.
Taking into consideration the great information you all have provided it would stand to reason that the switching supply is they way to go. Next question would be are any particular layouts preferred over others. I’m referring to the switching portion. (transistor configuration)
single- transistor forward
two – transistor forward
half bridge
full bridge
push-pull
I’ve noticed in reading that two-transistor forward and push-pull are common in computer psu’s
My gut instinct would lead me to believe a push-pull configuration would potentially be the easiest to layout and scale to meet power needs but I could be wrong.
single- transistor forward
two – transistor forward
half bridge
full bridge
push-pull
I’ve noticed in reading that two-transistor forward and push-pull are common in computer psu’s
My gut instinct would lead me to believe a push-pull configuration would potentially be the easiest to layout and scale to meet power needs but I could be wrong.
Switchmode supplies (SMPS) for that much power include power factor correction if they are made to comply with commercial standards for "office equipment" of the past decade or so. Other supplies (whether it's older or noncompliant switchers, or the conventional Big 60Hz Transformer) won't have this, and may blow a mains fuse or circuit breaker due to the current they pull, even though they're not pulling as much power as the mains circuit is rated for. Not having power factor correction at this high power level is also bad for other reasons (for one thing, the ugly current waveform could cause extra buzz in other audio equipment), so there's just about every reason to use a SMPS with PFC. Get a commercial off-the-shelf SMPS with the appropriate output voltage and current ratings, AND with PFC.
Most of the ATX units ( PC supply ) are half bridge, ofline push pull would need power switches with voltage ratings of 800V minimum, and with increasing ID/IC current theyr price flyes up.
Choosing the proper config is done by first stating the exact working conditions, voltage and current ratings on the output, where they get theyr power from, and why not, the buget.
Choosing the proper config is done by first stating the exact working conditions, voltage and current ratings on the output, where they get theyr power from, and why not, the buget.
That would mean 1,4KW of output power, and accepting an efficiency of let say 80% that would require about 1,7KW power from mains, and about 15A at 115Vac, that is quite allot, and at the upper limmit of the breaker, maybe you should change it with a higher current rating one.
But anyway why do you need that much current?
But anyway why do you need that much current?
....................why do you need that much current?
something in the range of 1000 to 1500 watts out of a class d amp for subwoofer duty
Ok i appologise for not seeing the first post in time.
Anyway i would think that for 35V per rail 20A is quite large current isn't it? i didn't work with class D amp before nore will i do in the near future but i doubt it would require that large current for just 35V rails... tell me please @AndrewT am i wrong in believing this?
Anyway i would think that for 35V per rail 20A is quite large current isn't it? i didn't work with class D amp before nore will i do in the near future but i doubt it would require that large current for just 35V rails... tell me please @AndrewT am i wrong in believing this?
You avoided my earlier question, before it was removed, but I consider it very important to this Thread, so I will repeat the question and give the answer.
Where does the amplifier get the current from to meet the demands of the speaker?
The transient current comes from the capacitors feeding the Power Rails. These capacitors are generally split into HF, MF decoupling and LF smoothing, when considering linear PSUs, (transformer+rectifier+smoothing).
The smoothing capacitance is also fitted to the SMPS, but it is generally of low value due to taking advantage of the high frequency of the supply into the SMPS smoothing capacitance, (some would refer to this as the output filter).
If the speaker is 2ohms and the supply is +-35Vdc then the maximum transient demand from the speaker can be around 35/2 * [fast transient factor]. This fast transient factor can approach 3times and has been measured at values slightly exceeding 5times what the nominal resistor value would demand.
Applying just 3times the peak transient current from the capacitors will be ~35/2*3 ~ 52.5Apk
A 56mF capacitance supplying that 52.5Apk for 1ms will discharge by ~0.94V
That 0.94V is ~2.7% of the original no load 35Vdc supply.
i.e. the 3times current factor has dropped the smoothing capacitance voltage from 100% of full charge to 97.3% of full charge in the 1ms of the transient demand.
Now what would a SMPS that has maybe just 100uF of filtering capacitance have to do to meet that 1ms transient demand of 52.5Apk?
BTW,
this is not an extreme example, 35Vdc & 2ohms speaker loading is only a 612.5W amplifier.
We are not into the kW range yet.
Where does the amplifier get the current from to meet the demands of the speaker?
The transient current comes from the capacitors feeding the Power Rails. These capacitors are generally split into HF, MF decoupling and LF smoothing, when considering linear PSUs, (transformer+rectifier+smoothing).
The smoothing capacitance is also fitted to the SMPS, but it is generally of low value due to taking advantage of the high frequency of the supply into the SMPS smoothing capacitance, (some would refer to this as the output filter).
If the speaker is 2ohms and the supply is +-35Vdc then the maximum transient demand from the speaker can be around 35/2 * [fast transient factor]. This fast transient factor can approach 3times and has been measured at values slightly exceeding 5times what the nominal resistor value would demand.
Applying just 3times the peak transient current from the capacitors will be ~35/2*3 ~ 52.5Apk
A 56mF capacitance supplying that 52.5Apk for 1ms will discharge by ~0.94V
That 0.94V is ~2.7% of the original no load 35Vdc supply.
i.e. the 3times current factor has dropped the smoothing capacitance voltage from 100% of full charge to 97.3% of full charge in the 1ms of the transient demand.
Now what would a SMPS that has maybe just 100uF of filtering capacitance have to do to meet that 1ms transient demand of 52.5Apk?
BTW,
this is not an extreme example, 35Vdc & 2ohms speaker loading is only a 612.5W amplifier.
We are not into the kW range yet.
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
- Home
- Amplifiers
- Power Supplies
- Power Supply Preference