[sdiy] Problems with linearized SSI2164 ?
modular at go2.pl
Thu Feb 7 06:50:20 CET 2019
This is only my wild guess, but since they took the freedom to replace V2164 with SSI2164, then maybe also misplaced 3k instead of 30k as R75? Markings look similar... Roman Dnia 7 lutego 2019 00:13 Danjel van Tijn <danjel at intellijel.com> napisał(a): Here is a link to the schematic block: ibb.co ibb.co Really just a standard Mike Irwin configuration that we have used on literally thousands of modules. Dave Dixon and I have tended to use 30k as the standard input resistance to the VCA cells but on more recent designs we have started using 100k. Unfortunately we just built a ton of modules with the SSI2164 and our factory did not tell me they had switched to the new chip (from the v2164). We can re-work these by replacing the three 30k resistors with 100k but I would still like to understand this better. It also unfortunately means the SSI2164 is not a blind drop in replacement for the V2164 and I am going to have to test out the rest of our circuits very carefully before any more manufacturing. Hopefully this helps some other folks avoid the same headache! On Wed, Feb 6, 2019 at 3:01 PM Neil Johnson < neil.johnson71 at gmail.com > wrote: Hi Danjel, Can you share the circuit you are building? Do you have some means of pulling the leakage current out of the summing node? This is usually done with a high-value resistor. And the issue Matthias raised is that with 30k you are sending more current into the gain cell, so the leakage current will be higher (well, it's not really leakage in the classic sense, but the fact that as this is a log circuit you can never achieve infinite attenuation, so *some* current will leak through. As to why there is a difference, I would need to make some measurements on the bench. One difference I do know is that the SSI2164 has much larger transistors in the gain cell, giving better log conformance, but they may exhibit different "off" behaviours. The 0.85V you are seeing - that would equate to about 26dB attenuation. In the Irwin linearised circuit an input CV of 0V would force the VCA CV to very positive (op-amp output pinned to the +ve rail) as it tries to attenuate the bias current to 0 with infinite attenuation. Cheers Neil On Wed, 6 Feb 2019 at 22:20, Danjel van Tijn < danjel at intellijel.com > wrote: Hi there, Has anyone else experienced issues with the SSI2164 used in place of a Coolaudio 2164 in a linearized VCA configuration? With no CV applied to the circuit control pin of the VCA should be close to 0 (and it is when I swap back in a coolaudio chip) but with a SSI2164 the voltage at the control pin is around 0.85V. I am using accurate voltage sources for the negative and positive voltage references (-5 and +5). and as aforementioned swapping in a coolaudio chip returns the circuit to working as expected. Opamp is a TI TL072CPWR. The only two differences from the datasheet version are that we are using 30K for Rin and 510R+560p for stability; both are within range of acceptable values according to datasheet. Any thoughts? possibly bad batch of ICs? -- cheers, Danjel van Tijn Intellijel Designs Inc. m. +001 604-773-2426 www.intellijel.com www.intellijel.com www.instagram.com www.instagram.com www.facebook.com www.facebook.com/intellijel www.youtube.com www.youtube.com/intellijel www.twitter.com www.twitter.com/intellijel ______________________________ Synth-diy mailing list Synth-diy at synth-diy.org synth-diy.org synth-diy.org -- www.njohnson.co.uk www.njohnson.co.uk -- cheers, Danjel van Tijn Intellijel Designs Inc. m. +001 604-773-2426 www.intellijel.com www.intellijel.com www.instagram.com www.instagram.com www.facebook.com www.facebook.com/intellijel www.youtube.com www.youtube.com/intellijel www.twitter.com www.twitter.com/intellijel ______________________________ Synth-diy mailing list Synth-diy at synth-diy.org synth-diy.org synth-diy.org
-------------- next part --------------
An HTML attachment was scrubbed...
More information about the Synth-diy