Tripath simulation with 120 ohm load
This image shows the results of simulating the output circuity of the tripath chip while driving three different loads.
The top of this image shows three versions of the tripath reference circuit from the chip amp to the speaker. Each circuit has a different speaker load on it: 4 ohms, 8 ohms, and 120 ohms (AKG-K1000). The circuit values are for the 4 ohm configuration, which may or may not be exactly how Sonic Impact supplies their amps. (If they used values for 8 ohm loads, it wouldn't make a big difference to the simulation anyways.)
The frequency response for each load is shown on the bottom. Cyan = 4 ohms, yellow = 8 ohm, green = 120 ohms. Notice the huge peak at 53kHz at 1.5 volts on the 120 ohm load. It's three times "louder" than the volume from 20-20kHz at about 0.5 volts. It's about 10db higher than flat. You can't hear it, but this ultrasonic signal is going to the phones. Now, the phones may not have a frequency response that extends that high, but even at 40kHz (which many phones will play), the output level is close to double what it is in the audible band.
I also ran the simulation with a 300 ohm load (HD-650). The peaking was about 20% higher. Everything else applies.Beauregard#1:
Whether your phones will survive these high levels at ultrasonic frequencies is not something I can tell you: it will depend on the phones. Your risk tolerance is another factor. Personally, I wouldn't drive my HD-650s with this amp. But I am trying to redesign the filter for headphone use.Beauregard#2:
The ASL won't put out ultrasonics, just more THD than you'd like. Only the best output transformers extend far above the audio band, perhaps up to 60kHz or 80kHz.Jon L#1:
Maybe you could hook up your phones to the amp and then put a scope on them. Tell us if you see any ultrasonics. The simulation may not be 100% accurate, but this is what tech support at tripath suggested I do.Jon L#2:
The analogy is apt: you're asking a class D amp to drive something it's not intended to drive, just as a Porche isn't designed to hold 1000 lbs of bricks.NeilPeart#1:
I'm not saying the tripath is better or worse than a $3000 SET, just that the reports comparing them for headphone use are not appropriate. You can't drive a 120 ohm load from a SET amp and expect it to sound right.NeilPeart#2:
Tripath makes about a dozen chips. The Bel Canto may not be using the lowest-powered (and highest THD?) chip like the SI does. No doubt better parts DRIVING AN 8 OHM LOAD
would make the SI sound even better. You still have the ultrasonics to worry about if you're driving a 120 ohm (or 300 ohm) headphone load.NeilPeart#3
Originally Posted by NeilPeart
Just because an amp isn't expensive and doesn't feature expensive tube outputs doesn't mean it's a poor-quality amp. Who cares though - enjoy your ignorance, Sir.
That's totally uncalled for
I'll forgive you when you invite me over to listen to your amp
Maybe I'll bring over a high quality power supply and we'll see just how far this thing can be pushed
But with your phones, not mine...RonD2:
Putting a 4 ohm load in parallel with the headphones is one possibility: the amp will see an approx. 4 ohm load and should behave properly. The problem with this is that most of the power will be delivered to the resistor, not the phones, so you'll have to run the amp full out to get enough power to the phones. If you try this, make sure you use a high power resistor rated for the amp's maximum power output.Ferbose:
I've looked at the charging circuit in one Chinese battery booster at Walmart. The charger did not use the charge controller chip from Unitrode (U????) that most good chargers use, so the battery's life would be reduced from the ideal of 100-200 charge/discharge cycles. Perhaps not a problem if you can get a month of listening per charge. Of course, other boosters may use the magic charging chip.mikeg:
I'm glad that neither amp was clipping with the K1000. I'd have to look at the circuit for the ASL amp (and spend more time than I'm willing) to figure out if/when it would clip and the amount of THD you'd get at the new operating point with a 120 ohm load.