PDA

View Full Version : latest Project - 220 WPC Amp



HipoFutura
10-19-2006, 04:33 PM
Started this months ago, but recently made some progress. The boards are done, and I roughed out the case today. Still need beau coupe work to make it pretty, but all in time. Hopefully, in a couple of weeks I'll have it running. I'm going to build two; one for each 2235 sub. The boards are designed and copy righted by Anthony Holton (Aussie Amps). It's 220W RMS at 8 ohms per channel. Don't know what the wattage is bridged, but it will be a scarry number!

Don

grumpy
10-19-2006, 06:57 PM
Nice :D If current capability isn't a limitation, that would make for nearly
900w into 8 ohms. Gonna need one heck of a PS to keep up with that.

-grumpy

HipoFutura
10-19-2006, 07:33 PM
Grumpy, the power supply is an Avel 625VA (2 - 50VAC secondaries) toroidal tranny with 40,000uf per rail. I'm building a surge board that will slow start the amp. This will keep it from tripping the house circuit breaker.

Don

grumpy
10-19-2006, 08:04 PM
running the secondaries in parallel? (guess I can look up Aussie Amps) :)

-grumpy

HipoFutura
10-19-2006, 08:15 PM
The PSU is symmetrical with +/- 50V rails.

Don

grumpy
10-20-2006, 08:01 AM
Hi Don,

Sounds like that will get you maybe 140W at 8 ohms, unbridged, if I understand correctly that your DC rail voltages will be +/- 50V ==> (<35Vrms**2)/8. 10V higher rails would
bump you up closer to 220W (if my coffee deprived brain is working this morning) :)
Cool project, regardless.

-grumpy

HipoFutura
10-20-2006, 08:23 AM
Grumpy, I hate to admit it, but you're in over my head. I need a bigger brain! Holton must be using 42V (peak to peak) - based on the 50V rails - for his 220W calculation. Why did you use 35V?

Don

grumpy
10-20-2006, 08:54 AM
I need a bigger brain! :rotfl:
Me too. :)

We're talking RMS watts (a way to express AC power), so it's not a straight conversion
using the p-p AC voltage squared over the resistance. One way to think of Wrms is as an equivalent
measure for how hot a load resistor would get if you put an equivalent amount of DC watts through it.
Since AC isn't "full on" all the time, to get an equivalent amount of "heat" into the load,
you have to introduce a factor. This is the peak-to-peak voltage (100V max, in your case)
divided by two (think of it so far as a rectified sine wave...the resistor doesn't care
what direction the current is flowing in... then adjust for the "dips" between the bumps (now swinging from 0 to +50V) to
get an equivalent power value (multiply by .707), which is the average value over time.

So your PSU is capable of supporting an output voltage of 35Vrms (or a bit less, if you include normal losses).
Now you can use V**2/R to get Wrms... something less than 150Wrms.

42Vrms would require +/-60V rails. The amp's output waveform has to stay within the
confines of your PS rails (ignoring output transformers & other weird stuff).

Hope I didn't make things worse. A picture would help (if I had one handy).

-grumpy

HipoFutura
10-20-2006, 09:00 AM
The '.707' multiplier was the piece I was missing. Now I get it. I'll email Holtan and see what he has to say.

Don

grumpy
10-20-2006, 09:49 AM
Ah... 50VAC secondary windings are assumed to be able to support something
like +/- 65V(+) dc rails. This may be what was missing. :) Makes sense now. -grumpy

(50Vrms transformer output gets you ~70V peak, rectify this... hold the peak values
with the humungo cap bank and you have your 65V PS rails)

HipoFutura
10-20-2006, 04:36 PM
Grumpy, Mr. Holton responded to my request regarding wattage. Since I can't cut/paste on this forum I'll have to paraphrase.

The 50V rails produce 70VDC after conversion and filtering. Full power at 8 ohms drops it to 66.6V. With your formula it works out to 221W RMS. This also accounts for a 6V drop across the output stage.

Well, there you have it! If I can find the time to finish up the case I should be able to fire this thing up in a few days. With it now looking like an amp I'm inspired to get it running sooner rather than later.

I'll use this and its twin to replace the Phase Linear 400 Series Two currently driving the subs.

Don

Rudy Kleimann
10-22-2006, 09:18 AM
625VA is approximately 625 Watts, but not really even that. You have to factor in the current-limiting introduced by the inductive reactance of the transformer coil windings. look at the back of audio gear: it will often state its power consumption in watts and Volt-Amps. The watts figure is usually around 20% less than the volt-amp figure. And all the watts of heat generated by the equipment is power consumed just running itself, and less to go to the speakers.

You may get 900W peaks out of it, but for what duration those peaks can last, it's hard to say. Depends on how the transformer acts as the output transistors drain the electrons out of those caps.

By its very nature, audio is extremely dynamic, so you shouldn't worry too much. However, there is a curve of power output vs. time period that will dictate the amplifier performance (and its sound) on high-power transients.

The +/-50V transformer outputs were, as I expected, RMS voltage ratings, which is the norm. That means 70V peak (RMS * 1.414, aka square root of 2, which is the same thing Grumpy is saying, only taking the equation in reverse). Your AC outlet is 120V RMS too, or about 170 Volts peak.

Don't worry about a soft-start circuit on those amps, unless the transformer or amp mfg. suggests the need for it, or your 120V source is already loaded heavily and you turn everything on at once. A 120V 20Amp circuit breaker will deliver 2400 watts continuous all day long without tripping, and your initial inrush current to those amps only lasts a few milliseconds.

BTW- how you like those 2405's on the D-75?

grumpy
10-23-2006, 09:33 AM
625VA is not 900W RMS!
Good points. :)
I don't think anyone was seriously expecting a sustained 900w from this amp.
(sorry Don, I've hijacked enough of this thread) -grumpy

HipoFutura
10-23-2006, 10:38 AM
Rudy, Grumpy has it right. I have yet to run my system over half volume. 200W RMS would actually be more than enough. But you know, bigger is better! If a few watts are good, then a whole bunch should be even better.

The Crown D-75 is great with the 2405s. I'm very happy with it. However, I can still hear a faint hiss when I get within three feet of the tweets (at head level). It's constant and doesn't change pitch or volume. Gone as soon as the music starts or you pull your head away from the speakers.

I have way too much hardware for a home stereo system. It's a sickness, but having bunches of fun with it!

Don

Thom
10-23-2006, 10:59 AM
Very often the difference in watts and volt amps on a transformer lable is only theoretical assuming a power factor of .8 . The amps will be constrained by the winding size the volts is assuming a certain input and for some reason .8 seams to usually be the assumed powerfactor. However the core can (I think) actually limit the power so if they skimped on the core the watts may be the limit. actually I haven't seen very many transformers with a watt rating. Also are you talking rms for a minute or two or for 4 hours. If any body out there is a transformer engineer and my facts are wrong then just say so and maybe they won't be next time but I think that's right. Watts will never be more than volt amps.

Ian Mackenzie
10-23-2006, 12:51 PM
I am sure it will be okay but you will need a soft start circuit.

Holton makes a nice amp kit!

Ian

Rudy Kleimann
10-23-2006, 06:34 PM
I have way too much hardware for a home stereo system. It's a sickness, but having bunches of fun with it!
Don

Your wife might hear...

Believe me, you'll be glad you have the power. There's an effortless quality to it. And believe it or not, you'll hit a few transients that are "up there", if only for a 1/2 cycle...:D

Rudy Kleimann
10-23-2006, 06:49 PM
...you will need a soft start circuit.
Ian


Why? Am I missing something? 900VA doesn't seem so large... I have plenty of power amps with much larger transformers and no "soft start" circuit(Mackie 1400's, Urei 6290, Crest 3001's, among others), and they never trip a 20Amp CB, even turning on two at once. Get 'em all going at once at a sound gig, well... that's another story:D

Regarding our xfmr's 900VA rating, how they really perform depends on how heavy the cores and the windings are. They may be capable of delivering more than 900VA for a few seconds or even minutes, but may fail at sustained power levels in excess of 900VA. Lord knows you won't be doing that at home -at least when the wife's at home:p.

grumpy
10-23-2006, 09:31 PM
Why? Am I missing something? 900VA doesn't seem so large...

Hi Rudy.

Reasonable question... more stuff to fail.

(I think Don's is 625VA... which is why I asked earlier about paralleling
output windings ... two would work into a lower impedance:)
...derating for current and expected
temperature rise are available from the mfg)

Other than playing nice with the lamps in the room (not everyone has
dedicated 20A breakers ... shameful, but true ;) ), even a brief slow start
(sub-second) is a normal mfg recommend for mid-size toroidial transformers
such as this and practically mandated for KVA+ units. Equipment I've used
toroids in had an expected turn on/off maybe once in 5 yrs, so we skipped the
soft-start (and power switches) after initially listening to the mfg :)

Power switch will likely last longer as well as other PS components.

-grumpy

Rudy Kleimann
10-24-2006, 05:25 PM
True, it would be "kind" to the switch and perhaps the PS components, but I would have a concern about it having a negative impact on the audio performance of the PS.

Do you have a diagram of a soft-start circuit? I'm curious.

andy11
10-24-2006, 10:15 PM
You will probably need the soft start... that's a huge cap bank and inrush current will be CRAZY!

BTW that's one hell of a power supply you got there! :D

Ian Mackenzie
10-25-2006, 12:01 AM
Why? Am I missing something? 900VA doesn't seem so large... I have plenty of power amps with much larger transformers and no "soft start" circuit(Mackie 1400's, Urei 6290, Crest 3001's, among others), and they never trip a 20Amp CB, even turning on two at once. Get 'em all going at once at a sound gig, well... that's another story:D

Regarding our xfmr's 900VA rating, how they really perform depends on how heavy the cores and the windings are. They may be capable of delivering more than 900VA for a few seconds or even minutes, but may fail at sustained power levels in excess of 900VA. Lord knows you won't be doing that at home -at least when the wife's at home:p.

The large cores react and result in inrush surge because of the hi flux in the core. Toriodals as a rule are more efficent than iron cores and tend to have a higher inrush surge. Just a thermister with a bypass relay and delay would work well.

Guido
10-25-2006, 03:42 AM
Just a thermister with a bypass relay and delay would work well.

Nelson Pass uses only a power thermistor in line with the toroid.
With my 2pcs. 1000VA transformers it works and I start them at the same time.

Ian Mackenzie
10-25-2006, 06:25 AM
Here attached a toroidal I specified for a Class A amplifier project for a forum member.

The primary windings are dual 120vac, dual seconday windings 21+21 volts rated at 12 amps each. The transformer is rated at 500 va, more than double the required load and core is oversized at 625 va for a low flux to reduce noise.

There are rectifier pcbs using the ST Micro BYW99W-200, which is
a dual "high efficiency, fast recovery" 35 amp diode at 200
volts.

According to Passlabs they did find a distinct
sonic improvement, and they also made it easier to meet
RF noise emission standards.

The capacitors filters will be 2 x 22,0000 50 volt @7.5 amps ripple current rating in parrellel in CLC topology per rail followed by local on board capacitors rated at 2 x 10,0000uf 50 volt / 2000uf 50 volt per rail.

Below are two internal views of the Passlabs X250.5. The transformer is enormous along with the significant number of filter capacitors.

Ian