opc, I just had a chance to have a better look at the measurement graphs, I don't know if this reflects badly on me or not but ... I have one huge smile after looking them over. Look forward to bringing a pair to life.
One question I had was, why does the THD fall off a cliff at around 10kHz in the measurements at 10W when at other output powers it increases with freq.? Is it just because you're well into that un-measurable realm?
One question I had was, why does the THD fall off a cliff at around 10kHz in the measurements at 10W when at other output powers it increases with freq.? Is it just because you're well into that un-measurable realm?
as i understand it, keeping the usable power of this unit up requires keeping the load impedance up to reasonable levels, the buffers are already current limiting at 15-20W into 8-9ohms (as a group); bridging wont bring more power in this case because there is no more current to give unless you add more buffers in parallel without bridging.
Last edited:
To answer a few questions:
Can the amp drive a 4 ohm load? - Yes, but you get roughly half the power or about 8W.
Can the amp drive a 2 ohm load? - Yes, but you get roughly 1/4 the power or about 4W.
Why not bridge two amps for more power? - Exactly as qusp said... bridging gets you more voltage swing, but the current capability remains exactly the same, and in this case, the current is what is lacking. If you read back a few pages you'll see my post about bridging... you need 40 buffers for a proper bridge configuration to double both the current and the voltage, and then you're into over $400 just for the buffers for one channel.
Why does distortion drop off sharply at 10kHz? - There is a high-pass filter built into the AP that can be set to filter out HF noise. In this case I think it's set to 30kHz, so by the time you get up 10kHz the HPF is filtering out the 3rd harmonic, which in this case is the dominant harmonic, and the distortion measurement drops accordingly. Based on that, if the distortion drops sharply above 10k, then you can assume the measurement is accurate up to 10k and no good above that.
The other two measurements were run without a HPF in place, so the distortion measured in those is indeed accurate right up to 20kHz.
As for the boards, if you have an outstanding power amp or headphone amp order and you want these, then I'll see what I can do to ship everything together. I'll compile a proper list tonight, and post it up.
Keep in mind there are only 40 boards total (or 20 pairs) since I'll be keeping the last 10 for myself.
Cheers,
Owen
Can the amp drive a 4 ohm load? - Yes, but you get roughly half the power or about 8W.
Can the amp drive a 2 ohm load? - Yes, but you get roughly 1/4 the power or about 4W.
Why not bridge two amps for more power? - Exactly as qusp said... bridging gets you more voltage swing, but the current capability remains exactly the same, and in this case, the current is what is lacking. If you read back a few pages you'll see my post about bridging... you need 40 buffers for a proper bridge configuration to double both the current and the voltage, and then you're into over $400 just for the buffers for one channel.
Why does distortion drop off sharply at 10kHz? - There is a high-pass filter built into the AP that can be set to filter out HF noise. In this case I think it's set to 30kHz, so by the time you get up 10kHz the HPF is filtering out the 3rd harmonic, which in this case is the dominant harmonic, and the distortion measurement drops accordingly. Based on that, if the distortion drops sharply above 10k, then you can assume the measurement is accurate up to 10k and no good above that.
The other two measurements were run without a HPF in place, so the distortion measured in those is indeed accurate right up to 20kHz.
As for the boards, if you have an outstanding power amp or headphone amp order and you want these, then I'll see what I can do to ship everything together. I'll compile a proper list tonight, and post it up.
Keep in mind there are only 40 boards total (or 20 pairs) since I'll be keeping the last 10 for myself.
Cheers,
Owen
It can indeed handle up to 24V or so, and in this case, I would indeed suggest splurging on the LME49610 to get the higher voltage capacity.
The LME49990 won't run on anything more than +/-18V so if you really want to run things at +/- 22V then you'll need to either drop the voltage down on the LME49990 or find another op-amp that can tolerate +/-22V.
Running higher rails would let you get quite a bit more power into higher impedance loads. If you happen to have a loudspeaker that is closer to 10 ohms, you could probably get almost 20W of output. This would required a larger off-board transformer though.
Cheers,
Owen
The LME49990 won't run on anything more than +/-18V so if you really want to run things at +/- 22V then you'll need to either drop the voltage down on the LME49990 or find another op-amp that can tolerate +/-22V.
Running higher rails would let you get quite a bit more power into higher impedance loads. If you happen to have a loudspeaker that is closer to 10 ohms, you could probably get almost 20W of output. This would required a larger off-board transformer though.
Cheers,
Owen
Hi Guys,
Based on what I'm seeing, I'm going to eliminate the possibility of a kit for this PCB. That means boards only, and I'll compile a list this evening with everyone who has expressed an interest.
The cost per board is $20 and shipping should be roughly $5 per pair unless you have an outstanding amp order in which case it should not impact the cost of shipping.
If you want boards, put a post up here with the quantity you require.
Cheers,
Owen
Based on what I'm seeing, I'm going to eliminate the possibility of a kit for this PCB. That means boards only, and I'll compile a list this evening with everyone who has expressed an interest.
The cost per board is $20 and shipping should be roughly $5 per pair unless you have an outstanding amp order in which case it should not impact the cost of shipping.
If you want boards, put a post up here with the quantity you require.
Cheers,
Owen
Thanks!