I know I saw Jakob post either here or elsewhere that the ad536 had a lot to do with it. The source I found will ship worldwide for the same price, you just have to buy 10 (group buy?) I'm willing to pay $12 for them (and that includes shipping). If you build 2 channels, it will only boost your overall cost by about $12 at most. Its not like the that's chips are way cheaper or anything. I don't understand why people freak out about that. Unless you're making 1000 or something, only a couple will only add a few bucks. And it will still be way cheaper than the commercial alternative.
Just my opinion though! ;D
After reading further on the topic of the ssl dynamics section, I have come up with a couple questions for the group to sort of get more info, and establish a preference.
From what I've read, of the 4000 versions, it seems like the E series was the more coveted. Is this correct? And it seems like the 4000E version used 2 different VCAs over its production span (1537 and then 2151?). Is THIS correct? I've heard mixed reviews of the 1537, and its i think almost impossible to find enough to make it worthwhile anyway, so maybe the 218x or a "can type" VCA would be best. Did the dynamics section EVER use a discrete VCA such as a 202C gold can? I know they were used in fader VCA's but not sure about dynamics modules.
Moving on, when it comes to the "best" 4000 version, would that be preferred over the newer 9000 version? Is the only difference added peak sensing circuitry and the updated VCA? The only reason I prefer the 9000 version is b/c I've never heard a 4000 version (horrible reason, lame, i know. aka willing to be swayed) and because of the added functionality of peak sensing vs rms. I used that function as much as the other.
I think if we nail down some of these questions, we will be able to provide a better, more versatile version of the circuit to make most people happy. So we will need schems for both. Is the 82E10 the 4000E version? And I think I saw a 9000J or K version in the gmail account.
Maybe we could borrow Jakobs idea of providing a 218x circuit for the VCA, with the optional space for using a "can type" vca (if he'll allow it! ;D). And we can then combine the best features of functionality with the best sounding version. It doesn't seem like adding the peak detection to the 4000 version would be too difficult.
And it seems like we'll have to add balancing/debalancing stuff for the I/O and it looks like some circuitry can be omited that involves the more console related stuff as opposed to stand alone operation. I need to study the schems way more b/c I have questions, I just have to figure out how to ask them, and get them all together at once.
Lets keep/get it rolling though!