I supply mine with +/-34v, take out 2 or 3 volts dropping on the chip and because being bridged, the speaker sees 31*2=62v.
V/R=I, so 62/8=7.75A
P=V*I, so P=62*7.75=480W
What am I doing wrong?
Take a look at this post: http://www.diyaudio.com/forums/group-buys/137430-bpa300-round-2-a-8.html#post1738771
It´s from BPA300 Round2.
It´s from BPA300 Round2.
when is the amplifier supplied with +-34Vdc?
when it is supplying no current to the load or when it is supplying maximum current to the load?
Allow a Vdrop across the amplifier according to the plot given by National in the datasheet.
Finally P (in watts) = Idc * Vdc.
If you are using anything other than constant DC voltage and constant DC current then you MUST use the appropriate conversion to find the predicted power output.
Question regarding Current Consumption for a mono block, and power supply requirements. Some help would be great.. im a green amateur without test equipment and havent built it yet.. still in planning.
1) How much current does one channel (2 bridged pcbs) consume at Idle, and how much does it consume during typical loud bass heavy playback? How much theoretical maximum?
im trying to determine if the schottkey diodes i have (6A) are within rating for this project. Also, need to know what kind of soft start to employ.
2) How much capacitance is enough? i'm targeting 45,000uF based on Alex's build... is this enough for stereo too?
3) im using the chipamp.com mini-aleph power supply board.
CCRC or CCLC for the smoothing caps? in a CCRC layout. What value R is sufficient.. 0.1ohm? 35watt? TO-220 -with a small heatsink - are these going to get very hot?? Should i just forget about CRC? Id consider an inductor to reduce power ripple, but i dont know how to calculate values, and dont have enough info to try
4) are bleeding resistors necessary, and if so, what size/rating should i be aiming at... 2.2k 3watt okay, or that laughable?
5) might Schottkey diode based rectifiers emit enough EMI to interfere with the ASP cross over for the Linkwitz Lx521 speakers? what if i seperated the transformer into another chassis, and included the linkwitz ASP in the same chassis as the amp and star grounded what i could? thoughts in EMI
1) How much current does one channel (2 bridged pcbs) consume at Idle, and how much does it consume during typical loud bass heavy playback? How much theoretical maximum?
im trying to determine if the schottkey diodes i have (6A) are within rating for this project. Also, need to know what kind of soft start to employ.
2) How much capacitance is enough? i'm targeting 45,000uF based on Alex's build... is this enough for stereo too?
3) im using the chipamp.com mini-aleph power supply board.
CCRC or CCLC for the smoothing caps? in a CCRC layout. What value R is sufficient.. 0.1ohm? 35watt? TO-220 -with a small heatsink - are these going to get very hot?? Should i just forget about CRC? Id consider an inductor to reduce power ripple, but i dont know how to calculate values, and dont have enough info to try
4) are bleeding resistors necessary, and if so, what size/rating should i be aiming at... 2.2k 3watt okay, or that laughable?
5) might Schottkey diode based rectifiers emit enough EMI to interfere with the ASP cross over for the Linkwitz Lx521 speakers? what if i seperated the transformer into another chassis, and included the linkwitz ASP in the same chassis as the amp and star grounded what i could? thoughts in EMI
The PSU rectifier charges up the smoothing capacitors.
The smoothing capacitors and the local decoupling capacitors supply ALL the CURRENT to the amplifier and speaker.
The datasheet specifies a maximum of 70mA of quiescent current for the 3886.
Simply multiply the number of chips by that maxima to find the highest continuous current draw from the capacitors.
The maximum peak transient current to the speaker is very high. It can be 10Apk to 40Apk depending on speaker and drive voltage and rise time of the fast transient.
The smoothing capacitors and the local decoupling capacitors supply ALL the CURRENT to the amplifier and speaker.
The datasheet specifies a maximum of 70mA of quiescent current for the 3886.
Simply multiply the number of chips by that maxima to find the highest continuous current draw from the capacitors.
The maximum peak transient current to the speaker is very high. It can be 10Apk to 40Apk depending on speaker and drive voltage and rise time of the fast transient.
That statement violates Kirchhoff's Law. It also violates conservation of energy.
It is true that the instantaneous current for the amp and the load is provided by the local decoupling/bypass caps, but that charge is replenished by the main reservoir caps. Thus, the average current drawn from the main supply caps is the sum of the average current drawn by the load and the average current drawn by the amp.
The output current of the LM3886 is limited to 11.5 A (TYP).
~Tom
As Andrew points out, the idle consumption can be found in the data sheet.
I would consider the worst case output current to be the worst case Vsupply/Rload. So calculate Vsupply for high mains voltage (+5~10 %) and use the lowest load impedance your amp will support. That gives you a back-of-the-envelope estimate for the peak output current. Note that the LM3886 limits the output current to 11.5 A (typical). Divide by sqrt(2) for RMS - assuming a sine wave signal.
Above estimate will lead to some over-design. In reality, few people drive their amps using sine waves at levels just below clipping for extended periods of time. Also, in reality, the power supply voltage will droop as more current is drawn by the load, which reduces the highest possible peak current.
For music signals, you need to apply another fudge factor, the crest factor. Even modern compressed music has a larger ratio between the peak and RMS value than a sine wave, so that sqrt(2) factor used for sine waves gets larger. This helps you out as the RMS current for a given peak power output gets lower.
All that is for output powers at clipping levels. In the last part of your question, you add more complexity. How loud is "typical loud bass heavy playback"? How efficient are your speakers? What's the size of your listening space? How far is the listening position from the speakers?
My listening position is about 2 m from the speakers. My speakers are 87 dB SPL @ 1 W, 1 m efficient. SPLs above about 85 dB for more than a few minutes will make my ears ring, so I consider 85 dB to be loud enough. 65 dB SPL at the listening position for comfortable critical listening is my normal mode of operation.
Doing the math: At 2 m, the SPL is 6 dB down from the SPL at 1 m. There are two speakers, hence, twice the power, so +3 dB for that. To produce 85 dB SPL at the listening position, each speaker must produce 87 + 6 - 3 = 90 dB SPL at 1 m. This requires 10^((90-87)/10) = 1.99 W.
My typical critical listening level is 20 dB down from that, so 20 mW. For background music, I often go 30 dB down - 2 mW - or 40 dB (200 uW) if I'm trying to carry a conversation while listening to the background music.
None of this math takes room gain into account. For a reasonably live room, you can figure 5~6 dB easily. So the actual power delivered to the speakers in my case is probably down another factor of four. So 25 uW for conversation background music, 250 uW for normal background music, 5 mW for critical listening, and 0.5 W when I really crank it. My amp is a 10 W 300B SET tube amp...
Having a volume control with dB readout is handy...
Depends... How much ripple are you willing to accept? I suspect 2 x 10000 uF or 2 x 22000 uF is probably fine.
Good luck finding an inductor of enough inductance to care that can handle the peak current without saturating.
P = I * E; E = I * R --> P = I^2 * R. I suggest derating the resistor at least a factor of two, so if you calculate a power dissipation of 10 W, use a 20 W resistor.
To get the best filtering, the R needs to be as big as possible. This also maximizes the power dissipation in the resistor. Tradeoffs, tradeoffs....
Personally, I wouldn't use RC filters. Just let the chipamp's PSRR to the job.
Depends on your rail voltage. I usually aim for a few mA of bleed current. Higher is better, but that burns more power in the bleed resistor. I'd aim to use a 2 W type. Given that it burns power continuously, I'd derate by a factor of 4-ish... So figure 500 mW dissipation. P = (E^2)/R --> R = (E^2)/P. E is the supply voltage. So for 28 V rails, you'd use: (28^2)/0.5 = 1.56 kOhm (round up to 1.8 kOhm - standard E-12 value).
I doubt it, but it depends on the layout and how much loop area the EMI radiating antenna has...
~Tom
Last edited:
Thank you Tom
Thank you for your Comprehensive reply Tom. As i dont have training in electrical engineering, and only limited University education in physics, your assistance is ->greatly< appreciated. Thanks for taking the time to share that info with me.
It will help me keep my planning within ball-park reason.
Thank you for your Comprehensive reply Tom. As i dont have training in electrical engineering, and only limited University education in physics, your assistance is ->greatly< appreciated. Thanks for taking the time to share that info with me.
It will help me keep my planning within ball-park reason.