Need low cost torrids

[Mike]

No more than probably around 36VCT (18-0-18)<p>the chip I use has a max voltage of +/-30VDC (60V) so the DC voltage output of the transformer after the rectifiers cannot exceed that voltage. Also, thats the absolute max voltage, so I would need less than that.

[jollyrgr]

Edd gives a very good description of rewinding a secondary of a transformer. I'll throw my two cents in on this subject. Insulate the Primary from the secondary with heavy "Kraft" paper. More than likely you do not have any of this. Something like white butcher's paper will work or the paper cut from a paper bag you bring your groceries home in will work. You want to somehow separate the primary and secondary wires so JUST IN CASE the windings rub against each other too much you don't get a short to the AC mains in the windings. As far as terminating the ends I suggest getting a piece of perf board or similar strong plastic form and drill four holes per extra winding in it. Put two holes side by side and leave some space then two more holes side by side. Then pull the wires through one of the holes and feed back through an adjacent hole. Sort of like sewing a button. But pull the wire out one final time and connect an insulated pigtail on the end of the windings. (A picture would probably be worth a thousand words here.)

[dacflyer]

i think i have what you need....but does it have to be 10 amps ? let me know what the minimum is that you need..ok,

[Mike]

No, whatever it takes to power 2 lm3886 and 2 opa549 (the 549 being bridged into 8ohm and the 3886s into single 8ohm each)

[Mike]

dacflyer, What transformers do you have that you think might work?

[dyarker]

The voltage you said in your first post (40V CT) was right.
60V * 0.707 = 42.42V, so a 36V to 40V transformer keeps you under the ±30VDC limit. (The transformer spec probably says 115V or 117V primary. But the power line can actually be 110V to 125V(?). Building a slightly lower voltage supply stops from exceeding ±30V when AC line is high.<p>-- current -----<p>The OPA549 spec says at 8A the worst case output voltage is 4.8V lower than positive supply and 4.6V above negative supply. Lets call it 5V.
(30V - 5V)peak / 8 Ohms = 3.126A, bridged so
3.125Apeak * 2 = 6.25Apeak, for two OPA549s driving one speaker.<p>The clipping volatge of LM3886 at 30V supply into 8 Ohms is 2.6V. Call it 3V.
(30V - 3V)peak / 8 Ohms = 3.375Apeak, for each speaker
3.375Apeak * 2 = 6.75Apeak
Itotal = 6.25Apeak + 6.75Apeak = 13Apeak
13Apeak * 0.707 = 10.01A RMS, so your original request for a 10A transformer is about right too! It wouldn't burn-up.<p>But I suggest a 12A to 15A (20A ?)transformer to reduce supply voltage droop at max load. A fullwave rectifier and capacitor power supply doesn't draw current from the transformer during the entire AC cycle. Current only flows when the rectified voltage is greater than voltage on the filter capacitor. With a 10A load, the current could be 20A to 30A pulses. The resistance of the wire in transformer causes the output voltage to decrease. An oversize transformer reduces this problem.<p>Check the fullwave rectifier too. With a 10A max continuous load, a 12A to 15A rectifier is okay for heat if heatsinked. The repetative current surge rating must be greater than the pulses going into the capacitor. The one-time surge rating must be greater than the large current that flows when you flip the switch to on.<p>--------------------------<p>How big a heatsink are you using?<p>Cheers & C U L -<p>[ September 03, 2004: Message edited by: Dale Y ]</p>

[High4Volts]

so i see this while about to toss out some old stepper motor index drive boxes at work and think so how much are these worth on ebay? I have about 3 or 4 Stealtite Power TO142 --- next line #2050.095.05 they came out of some Digiplan PKC80M/IFX drives.

[Mike]

Thanks, Dale.<p>Are you saying that the 3886 actually puts out more power (wattage) than the 549 since it puts out more current?<p>I thought the 549 was around 80W into 8ohm and even more into 4.<p>I have noticed it can put out much more volume without distortion, especially bass, which is suprising because it has less current output.

[dyarker]

(corrections in bold)
No. You said bridged which I take to mean each OPA549 drives one speaker lead and one OPA549 goes positive while the other negative. The 6.25Apeak flows through both OPA549s.
Ppeak = 50Vpeak * 6.25Apeak = 321.5Wpeak to speaker.
Pcontinuous = 321.5Wpeak * 0.707 = 240W <<Wrong! Should be:
Pcontinuousrms = 321.5Wpeak * 0.707 * 0.707 = 160.7W
120W 80W out of each IC.<p>The max power heating the amplifier IC occurs at about half of max output power. I'm too rusty on this to show the calculation, but 80W 50W, and makes more sense sounds about right. At zero out, ALL the supply voltage is across the IC, but current is 0A, so IC dissapates 0W. As output is increased, IC power also increases, but more and more of the supply voltage is across the load and less across the IC. At half out, half supply voltage is across the load and half across the IC, IC power dissipation is max. At max out IC dissipation is lower. The bold "about" above is there because you have allow for bias, minimum drop (4.8V), and waveform correction.<p>Each LM3886 outputs:
27Vpeak * 3.375Apeak = 91.125Wpeak
91.125Wpeak * 0.707 = 64.425375W RMS should be:
91.125Wpeak * 0.707 * 0.707 = 45.548740125Wrms
max IC dissapation under 40W about 25W (spec says 68W)<p>(I said I was rusty! How I goofed in my following post. If anybody is taking notes, fix them.)<p>[ September 04, 2004: Message edited by: Dale Y ]</p>

[dacflyer]

whewwwww...dales got my head spining...
him R A Smart guy !...<p>hey mike... heres what i got so far , they are regular core type transformers , not torroids.<p>1st one is..... a dual 16-0-16 you can bridge it for more current...but it already looks real heafty... looks easily 10amp single mode
dimentions are about 3.5"T X 4.25"W X 4.75"D<p>2nd one is..multi-point 46-33-0-33-46 its got a few other taps too..but they are isolated..but on same transformer...came out of a sony stereo
dimentions are...3.5"T X 4"W X 5.75D this is a big one,, i have some others but they are much higher voltage... let me know if these will work for ya...

[Mike]

Wait, so you are saying that peak power out of the bridged opa549 is 320W and RMS 240W?? if so, wow, thats a lot of power for a little chip to make.<p>Sorry to bug you again, but if I use that 16-0-16V 10A transformer with this amp, about how much power output can I expect with the bridged 549?<p>And, dacflyer, I am interested in the 16-0-16 transformer, as long as the input is standard 110-120VAC input. How much would you want for it including shipping it to Illinois?

[dyarker]

dacflyer,<p>Thanks, I think. But not too often please. It makes my head fat, then I make a big embarrasing mistake. I could just answer with just the numbers, but then others here (lurkers included) couldn't figure out how for themselves. It helps me too. I get to exercise skills unused since college.
-----------------------------------------
To All,<p>Speaking of mistakes, I made a beaut in previous post. On the 240W line I multiplied peak power by only one 0.707. That's the same as VoltsRMS times AmpsPeak. The units don't match, so wrong.
Ppeak = Vpeak * Ipeak
Prms = Vrms * Irms
for sine wave Peak_E_or_I = RMS_E_or_I * 1.414
1/1.414 = 0.707<p>Sorry, I'll go back and edit that when done here.
---------------------------------------
Mike,<p>filtered supply voltage:
32V RMS * 1.414 = 45.248V<p>minus rectifier and OPA549 clipping:
45.248V - 1.5V - 8V = 35.748V, call it 35V supply<p>current thru speaker:
35V / 8 Ohms = 4.375Apeak
4.375Apeak * 0.707 = 3.093125Arms<p>RMS voltage out:
35Vpeak * 0.707 = 24.745Vrms<p>RMS power out:
24.745Vrms * 3.093125Arms = 76.54Wrms<p>maybe I should work these on paper first, then type them!<p>Later,<p>[ September 04, 2004: Message edited by: Dale Y ]</p>

[dacflyer]

MIKE >>> i'll take $20.00 for it plus S/H
let me know....