Fortunately, they are also completely unnecessary. You need them just as badly as you need 100,000 horsepower in your car engine.
You could, indeed.
And whether bigger filter caps would make the shutdown situation better or not, is debatable. For all I know, it might make it worse, due to the huge current spike needed to charge up the big cap when you power up the circuit.
My suggestions:
- If you're getting enough loudness now without triggering the shutdowns, leave well enough alone.
- If not, save your money for a bigger power supply, rather than spending it on ridiculously oversized capacitors.
Do you really think a capacitor able to provide the missing 5 amps needed by the amp at full power for 0.05 second with 2.5 voltage drop is ridiculously oversized? Or you just never tried to calculate anything about it?
Maybe, if there is no appropriate soft-start. As I already told, it helps if it can start. After starting it stores charge available for the high peak current demands that can reach 20 A.
I usually buy 33000F 25V for 2 Euro.
Maybe, if there is no appropriate soft-start. As I already told, it helps if it can start. After starting it stores charge available for the high peak current demands that can reach 20 A.
I usually buy 33000F 25V for 2 Euro.
Well, the 120W aren't peak but continous power, so the calculation is "non optimal" to me.
Putting 100mF at the output will highly increase the stress for the internal psu components due to high peak recharge currents. I would rather suggest to have more primary capacitance as this translates to the secondary quite nicely without setting down the SMPSs regulation bandwidth.
100mF at 24V is ~290J
which is pretty much the same as
470uF at 340V (rectified primary) is ~320J
If you only have a simple 50/60Hz transformer powersupply, then yes, 100mF is adeguate.
It does, indeed. I'd made the unconscious assumption that you were intending both AC and DC operation.
In any case, if you're getting enough sound pressure level for your tastes from the existing power supply, there's no pressing need to upgrade it. Admittedly, you couldn't release a commercial product that way - someone would turn it up to full volume, and it would shut down, and they would be upset and want a refund. But this is isn't a commercial product, it's a personal product for yourself, so all you have to do, is not turn the volume up too far!
I expect your creation sounds significantly louder now, with the speakers in an enclosure?
The question of how much SPL is enough is a whole other story. Our ears have not evolved since the days when our ancestors lived in grasslands in a completely pre-industrial world, where it was very quiet, virtually all the time.
Never having been designed by evolution to cope with the astonishingly loud contemporary industrialized world, our ears are easily damaged by excessive SPL, and the damage tends to be cumulative, and it tends to be permanent.
Rock musician Mo Foster's book "17 Watts?" ( https://www.amazon.com/17-Watts-British-Musicians-Stories/dp/186074267X ) is an interesting look at how recent our taste for deafening (literally) loudness is. The 17-watt Watkins Dominator guitar amp became available in 1956; Foster and his band-mates got into an argument about this model in 1960, deciding it was absurdly loud, and far too powerful to be necessary or even useful. And this was a group of young rock musicians - not septuagenarian Sunday-school teachers.
Fifty six years later, musicians routinely lug 1200 watt P.A. systems into small coffee-shop gigs.
Insane amounts of audio power are cheaply available now - but our ears remain just as fragile as they were when the ice age ended eleven thousand years ago, and probably as fragile as they were when that 40,000 year old bone flute was being played for the first time in Germany. ( Bone Flute Is Oldest Instrument, Study Says )Sounds like fun!
This is where you will start to run into another engineering decision, the interplay between power, run time on a charge, and battery size, cost, and weight.
-Gnobuddy
Very true. I think the ears of the operator will prevent full-power use, both because of excessive distortion, and excessive SPL.
10% THD may have made sense as a rating method for small radios in 1930, but IMO it is a really deceitful way for chip manufacturers to rate power output these days, when even cheap amplifiers are expected to operate with a hundred times less distortion than that.
-Gnobuddy
Sigh. There are so many intellectual shoving matches on diyAudio, I really don't want to participate in one of them.
But to truthfully answer the first question, yes. There are thousands of successful power amplifier designs of similar wattage, stretching back fifty years, that use much less power supply capacitance. So yes, in my opinion, 0.1 Farad is ridiculously oversized.
As for the second question: yes, I know Q=CV, W = 1/2 C V^2, and i = C dV/dt. And I still think 0.1 F is way too much capacitance, and ridiculously expensive, for what is essentially a giant "adequate-fi" boom-box.
The cheapest 30 volt, 0.1F electrolytic cap I can find in stock at Mouser costs $81.50 (US dollars: http://www.mouser.com/ProductDetail...=sGAEpiMZZMtZ1n0r9vR22bNos12dWCDWi8d261o1oG8=
Part of the reason for my opinion is that 50 mS time duration you used, for a mere 10% rail voltage drop. Even a bridge-rectified old-fashioned 60 Hz power supply has a new voltage pulse arriving to charge up the filter caps about every 8.5 milliseconds. A switch-mode supply probably operates at several kHz or more, so power pulses come along every 100 microseconds or less.
So why on earth do we want a capacitor to store 50 milliseconds worth of charge, particularly when that single capacitor costs nearly four times as much as an entire new power supply, that will completely solve the problem? (Power supply: https://www.amazon.com/Switching-Tr...J26U/ref=sr_1_2?ie=UTF8&qid=1471925802&sr=8-2 )
I didn't want to leave your questions to me unanswered, which might lead others to believe I had no good reasons for my opinions. At the same time, I should stress that this reply isn't intended as an insult to you. It's simply my opinion, based on my view of the world, my technical knowledge, and the prices I would have to pay.
Heck, if I could find 33,000uF caps for 2 Euros, I might change my mind; I also wouldn't mind 500 more horsepower in my car if that only cost me two bucks!
-Gnobuddy
Then don't do it! Don't try to make something ridiculous just because you don't understand it!
With similarly undersized SMPS? Doubtfully.
And why is it ridiculous? Is calculating something ridiculous for you?
Why 1...10 F in a car not ridiculous?
The power pulses coming from this SMPS are incapable to recharge the cap while loaded with max power, no matter how much the frequency. An old fashion power supply is totally different, it is not hardly current limited. The difference is NOT the frequengy, but the overload characteristics.
So why on earth do we want a capacitor to store 50 milliseconds worth of charge, particularly when that single capacitor costs nearly four times as much as an entire new power supply, that will completely solve the problem? (Power supply: https://www.amazon.com/Switching-Tr...J26U/ref=sr_1_2?ie=UTF8&qid=1471925802&sr=8-2 )
I didn't want to leave your questions to me unanswered, which might lead others to believe I had no good reasons for my opinions. At the same time, I should stress that this reply isn't intended as an insult to you. It's simply my opinion, based on my view of the world, my technical knowledge, and the prices I would have to pay.
Heck, if I could find 33,000uF caps for 2 Euros, I might change my mind; I also wouldn't mind 500 more horsepower in my car if that only cost me two bucks!
-Gnobuddy
Here you are, it is 455 HUF = 1.5 Euro:
Webáruház
And the reason for 50 ms is: a typical bass beat lasts about this time. Nothing about the mains or SMPS frequency.
Then why dont you make an "optimal calculation"?
Do you know what the peak power for this PSU? I dont, therefore i dont calculate with it. But calculating with continuous power worth much more then not calculating at all, dont you think?
If the psu is current limited then its impossible to increase its peak current. If not, then recharge current is definitely lower than the peak current consumed by the amp. So your statement says the opposite of the trueth.
Only if the power converter stage was not the weak link. But it is definitely not the case here. And primary capacitor is not easy to reach, what is more definitely dangerous for somebody who can mix Volt and Watt. Please dont try to kill OP! ;-)
Wrong, again. Only transfer bandwidth is lowered by output cap, but it is a beneficial effect. Please try distinguish between line and load regulation!
At least this is correct.
What is the object of the restriction? "Only if...", or "only a simple..."? "Only have..." doesnt make sense. Nor a contrafactual hypothesys.
Gnobuddy,
Very true about "it's a personal project for yourself. So don't turn it up all the way."
Since the amp has separate volume pots for Sub and stereo, I have dialed both of those down such that I can dial the master volume pot "to eleven" without tripping the safety circuits.
Problem, if not solved, then, at least, "mitigated."
James
Very true about "it's a personal project for yourself. So don't turn it up all the way."
Since the amp has separate volume pots for Sub and stereo, I have dialed both of those down such that I can dial the master volume pot "to eleven" without tripping the safety circuits.
Problem, if not solved, then, at least, "mitigated."
James
Loud on 24V
If you want loud on 24V, efficiency is needed on the speakers. Tweeters are 112dB and woofers are 97.5. On a TPA3116 they go to a reasonable level before distortion. 48V and TAS5630 gets to party levels.
12" 4R JBL car sub (a great sub) does not quite keep up with these when the volume goes up on 48V.
If you want loud on 24V, efficiency is needed on the speakers. Tweeters are 112dB and woofers are 97.5. On a TPA3116 they go to a reasonable level before distortion. 48V and TAS5630 gets to party levels.
12" 4R JBL car sub (a great sub) does not quite keep up with these when the volume goes up on 48V.
An externally hosted image should be here but it was not working when we last tested it.
Last edited:
P.S. Polk car speakers are 93dB. Not too bad, but need twice as much(or more) power to get as loud.
My Morel speakers sound great, but at 89dB and 8R do not go "Loud" on TPA 3116. 24V simply can't produce enough oomph before clipping.
Normal listening is fine. Kickass requires more Volts and more efficiency, or many more Volts!
My Morel speakers sound great, but at 89dB and 8R do not go "Loud" on TPA 3116. 24V simply can't produce enough oomph before clipping.
Normal listening is fine. Kickass requires more Volts and more efficiency, or many more Volts!
As you have a DVC subwoofer, driving it via a separate stereo amplifier might be one way. You'd then need a separate LPF in the feed to it or LLXO.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.