Sphere Console VCA upgrade

Where we discuss new analog design ideas for Pro Audio and modern spins on vintage ones.
Westrek
Posts: 4
Joined: Tue Mar 25, 2014 8:57 pm
Location: Kansas City, Kansas
Contact:

Sphere Console VCA upgrade

Post by Westrek » Wed Mar 26, 2014 10:41 am

I'm working on a Sphere Eclipse C console that uses the DBX Black Can 202 VCA modules with the -6mv/db control law. I'm upgrading to a design that uses the SSM2018 vca. The modules that I want to use on the upgrade use the 2018 vcas that were designed in the '90s to replace the DBX 202C Gold Can modules in MCI 500 and 600 consoles that used the -50mv/db law. Although the 2018 states it uses -30mv/db, these worked perfectly in the MCI systems with no real changes in the control circuit scaling.

I've been a bit stumped ( I'm a decent tech, but no designer) on how to alter the Sphere fader pack to achieve the proper scaling. The console uses an automation system ( looks similar to the Allison Fadex) but it won't be used. Currently, putting the fader pack in the "Local" mode and placing the console in the "Write" mode, allows manual operation.

Here's links to the drawing of the fader module and the retrofit vca design. The vca itself is located on a another board and has no resistors in the control line.

http://westendstudio.com/wp-content/upl ... Module.pdf

Image

A couple of component differences I've noticed on the actual module:

R5 is 6.81K
R7 is 54.9K
R29,30,31 is 1K

Any thoughts would be appreciated.

Thanks!

Mike Miller

User avatar
JR.
Posts: 2599
Joined: Sat May 24, 2008 7:21 pm

Re: Sphere Console VCA upgrade

Post by JR. » Wed Mar 26, 2014 11:05 am

While that schemo was not that easy to follow it looks like the send to the VCA is through an X kohm resistor (10k?) and at the VCA there is probably an X ohm resistor to ground to form the bottom leg of a simple voltage divider. To convert from 6 mV/dB to 30 mV or whatever, you need to scale one or both resistors.

If the R to ground at the VCA control port is already different that might explain why they worked with no apparent modification.

JR

Westrek
Posts: 4
Joined: Tue Mar 25, 2014 8:57 pm
Location: Kansas City, Kansas
Contact:

Re: Sphere Console VCA upgrade

Post by Westrek » Wed Mar 26, 2014 11:38 am

At the original vca (202) there is no resistor from the control port to ground. On the upgrade design there is a R to ground (R5) which was 374 ohm. I removed that initially for this project.

What's the purpose of the 2 op-amp arrangement at U3 A/B and at U1 C/D with the diode?

User avatar
JR.
Posts: 2599
Joined: Sat May 24, 2008 7:21 pm

Re: Sphere Console VCA upgrade

Post by JR. » Wed Mar 26, 2014 12:17 pm

Is this a test? OK luckily for you I still had this schematic open,,, that opamp with diode connected across it can only conduct in one direction, so for the direction that the diode conducts the opamp will hold the - input to same voltage as the + input, for input conditions where the diode in not conducting, the input will look like a high impedance and vary freely (below 5V).

So when the diode is not conducting R24 looks like it is floating so does not change the gain/voltage at output of U3B. When diode is conducting (inputs above 5V) the gain of U3B increases so voltage rises faster.

These are generally called diode break amps, because the transfer function looks like different slope lines based on when the diodes "break" or change the gain.

[edit- ok the "purpose" is probably to provide more attenuation at extreme of fader travel. /edit]

JR

Westrek
Posts: 4
Joined: Tue Mar 25, 2014 8:57 pm
Location: Kansas City, Kansas
Contact:

Re: Sphere Console VCA upgrade

Post by Westrek » Wed Mar 26, 2014 2:57 pm

I had read that diodes were implemented with the original DBX 202 to help control "thumps" when un-muting the vca.

Thanks much for your reply and knowledge.

User avatar
JR.
Posts: 2599
Joined: Sat May 24, 2008 7:21 pm

Re: Sphere Console VCA upgrade

Post by JR. » Wed Mar 26, 2014 3:22 pm

I've already closed the schemo so I won't revisit.

It did not look like that to me, since it is making the gain rate of change larger above a certain voltage.

I can imagine a VCA getting a little squirrely when commanding silly amounts of attenuation, especially if using just one control port. VCAs are generally better behaved for extreme gain commands by driving the two control ports in opposite directions, half the distance each. For example -100dB would be -600mV at the control port or a whole diode drop. By driving the two ports with +/- 300mV you still have the same 600mV delta for -100dB gain command but the transistors inside the VCA is closer to normal operating voltages.

If you are serious about good performance I would look at the modern THAT/dbx VCAs. They really are quite good, while I am not familiar with 2018s.

JR

Westrek
Posts: 4
Joined: Tue Mar 25, 2014 8:57 pm
Location: Kansas City, Kansas
Contact:

Re: Sphere Console VCA upgrade

Post by Westrek » Wed Mar 26, 2014 3:49 pm

THAT no longer makes the retrofit 202 modules, but they do make similar vca chips to 2018. Actually the SSM2018 is going out of production at the end of the year. The 2018 modules I have built for the MCI application have around .007% distortion at 1K @ 0db. The old DBX202 has .4%.

Thanks again for your thoughts.

Mike

emrr
Posts: 504
Joined: Sun Mar 02, 2008 11:27 pm
Location: NC, USA
Contact:

Re: Sphere Console VCA upgrade

Post by emrr » Mon Mar 31, 2014 9:41 am

I think there are still PCB's available in DIY-land that will let you assemble a 202 retrofit. Might be wrong....
Best,

Doug Williams
Electromagnetic Radiation Recorders

vintagedesign
Posts: 18
Joined: Sat Oct 17, 2009 8:21 am

Re: Sphere Console VCA upgrade

Post by vintagedesign » Tue Apr 01, 2014 1:57 am

Isn't all info you need here? http://www.thatcorp.com/datashts/dn127.pdf
Thomas TK Kristiansson
Sweden
Vintagedesign / TK Audio

User avatar
mediatechnology
Posts: 3540
Joined: Sat Aug 11, 2007 2:34 pm
Location: Oak Cliff, Texas
Contact:

Re: Sphere Console VCA upgrade

Post by mediatechnology » Wed Apr 02, 2014 10:42 am

DN-127 is quite likely all you'll need.

If it were me I would use THAT2180A or B-grade "pre-trimmed" parts and not used the soon-to-be-discontinued SSM2018.

The THAT2181 allows external trimming and lower THD at a given ambient temp but based on tests that I've done (and I think Roger too) the 2180 will outperform it over even modest temperature variations since the TC of the on-chip trims tracks the TC of the error better. I would not recommend the servo trimming shown in another app note.

Post Reply