But did he try everything? IC op-amps that by necessity have the compensation capacitance referenced to one of the rails benefit from things like I proposed on the AD8010 datasheet. For its day this was a rather substantial load for an 8-pin DIP. 8 parallel 150 Ohm video loads at SOTA diff gain and phase (i.e. inherent linearity).
Last edited:
Is this enough? Measured at 3.54Vrms into 6.8 ohm load. With notch filter, fundamental notched of 61dB.
Attachments
Yes, you are right. Needed to get all parameters and measurements near to perfection.
John, You have to try a class-A type supply with only smoothing caps on the raw supply. Think of a low noise referenced supply directly supplying the load. No electrolytic on the output of the supply. Very fast in sourcing and sinking.
I don't understand the facination with drawing power off of "massive storage electrolytics", as aluminum is a really bad conductor and big caps are really slow. Once you you are dealing with only DC the electrolytics are a detriment. In my experience, and in most applications, preamps especially, you just don't require massive storage or high accuracy regulation. A stiff voltage referenced supply is much more musical sounding.
Mike
Hi,
Which applies to valve based designs just the same only less so.
Cheers,
Most often, because if the high supply rails, passive makes sense with tube RIAA's. For opamp based designs,you lose close to 20 dB on designs that have a high gain stage followed by the passive 2120 Hz filter. If you place the HF shaping stage around the first opamp and do the LF EQ later, you have to trade off noise.
For opamp stages (and and low supply rail design for that matter) it seems that single stage EQ gives the best of all worlds - overload and noise.
There are many op-amps that would measure worse (some much worse) in this circuit.
I doubt it. I would have 10ma or more Iq in the buffer alone. That is 20ma or more Class A, even more in Class A-B. This is not the optimum drive for loads less than 500 ohms, that takes bipolars running equally Class A, but then biasing has to be added, which complicates matters somewhat.
Hmmmmm. This is pretty much my experience also. Fast, low z source and sink supply sounds better. And master/slave also sounds better. It might not matter so much with steady current draw -- more like class A circuits would have. Have no experience to say. But adding a lot capacitance on the output of a regulated supply just doesnt work well for audio.
THx-RNMarsh
- Status
- Not open for further replies.