903-51X Comp / Limit

[[silent:arts]]

Update: few resistor values in the dbx schematic are wrong  :
2nd prototype seems to work correctly.
Some more tests to follow.

 

[culteousness1]

Very nice to hear! How did the pot orientation work out?

 

[[silent:arts]]

How did the pot orientation work out?

Odd. I mean good. All potentiometers = correct, trimmers wrong way (doesn't matter, but will be changed)

 

[[silent:arts]]

Received some Prototype Panels from Frank @ frontpanels.de:

Next I will test different VCA options, all is working good up to now.
Stereo (CV) link doesn't, however, I didn't expect it to do. Won't be in the final revision.

Audio would work with +/- 16V only, however, the Meter doesn't.

 

[emrr]

You'll have to shoot it out with the new DBX and the old 903. 

I see 903's can be had frequently in the $50-$75 range. 
900 racks are cheap, sometimes $100.  Same for the FS900 two space rack. 
I see several instances of a rack of nine 903's selling under $600. 

 

[TheJames]

One thing I always missed on the 903 vs the 160, is the hard knee.

 

[jensenmann]

I don´t know how close these 903 are to 160X or XTs. In case they are close, then it should be mentioned that the 1uF foil coupling cap in front of the VCA is too small IMHO.  I´ve modded several 160s and always found them to loose lowend in original state.  Increasing these caps to 6,8-10uF helped a lot and additionally cleared the lowmidrange. Though I used to like 160Xs for PA use I never dug them for recording because of the uninspiring audiopath. That changed after modding them, which involved a lot more than changing this cap. They can sound pretty darn good while sticking with the original compression behaviour which I really like.

Doug is right, prices of 160s and 903s are falling and they can be found for half the price than 2yrs ago. From an economical POV there´s no sense in DIYing one. But the fun in DIYing is something else, of course.

 

[[silent:arts]]

Hi Jens, in the 903 the cap in front of the VCA is 4,7uF.
1uF is in front of the dbx 146221 RMS chip.

 

[mitsos]

One thing I always missed on the 903 vs the 160, is the hard knee.

Is there a way to change the knee on the 903? That would be an excellent addition to this project!

Also, please, for those of us that, umm, don't have their 51x rack anymore, can you make the 24V pads on a tab that can be broken off? If I understand correctly the only thing running off 24V is the output amp, which could be modified easily to be 16V friendly, and then this project would be even cooler.  Thanks!

 

[[silent:arts]]

It tested fine with +/-16V only (lower headroom, however)

The new PCBs are ordered and are coming like this:

 

[mitsos]

Cool, 16V works better in my API rack.

Curious, why the doubled up diodes?

Any ideas on the knee change?  I wish I could help with that, I would already have done it to my original 903s. 

In for a couple PCBs for sure, faceplates would be nice too!

 

[[silent:arts]]

Curious, why the doubled up diodes?

to follow the original schematic as close as possible

Any ideas on the knee change?  I wish I could help with that, I would already have done it to my original 903s.

I changed nothing. I only added the option for impedance or transformer balanced output, and added a 202X VCA substitute circuit
(which is exactly the same 2150 / 218X VCA based circuit later revisions of the 903 used)

 

[mitsos]

Re: the knee change,  I didn't mean that you changed from the original, what I meant is, had you researched the possibility of implementing a hard knee in this design, instead of "over-easy" all the time. 

I (and someone else above) mention it because it would be nice to have a harder knee for faster transient stuff, and would make this project THAT much more versatile. (not that it already isn't, I like DBX stuff, obviously).  There are lots of posts about this all over the net, but as far as I've seen, no one's come up with a solution.  Maybe it isn't possible or is overly complex to implement in the original modules,  but, this being a new design, and with all the brainpower on this forum, this seems like the place to do it. 

Anyway, don't want to just give you extra work   so I'm willing to help by testing ideas on my original 903s, if you need. 

thanks!

 

[TheJames]

I agree with the knee comments.  I have a few 903's and really miss not having hard-knee as an option.  Over the years I've glanced at the 160 schematics in an effort to maybe see how I could make the 903 have the same hard knee function.  Ultimately I always got frustrated with trying to sort out the mess of the 160 schematics and think "another day maybe."

The new DBX 560a (160 in  500 series) gives you the option of "Over-easy" of hard knee compression.  I like the dual meters too (level and gain reduction).

Don't read it as discouraging, If I didn't have a 900 rack I'd be seriously eyeing this project.  However since I do have a 900 rack (and a few 903's) I'd be more interested in a 903 retrofit for hard knee.

 

[mitsos]

Ultimately I always got frustrated with trying to sort out the mess of the 160 schematics and think "another day maybe."

haha, me too!

The new DBX 560a (160 in  500 series) gives you the option of "Over-easy" of hard knee compression.  I like the dual meters too (level and gain reduction).

Good point, DBX is pretty good about sharing info, they're one of the only companies with so much free service info of legacy products online.  Might be worth an email to them?

However since I do have a 900 rack (and a few 903's) I'd be more interested in a 903 retrofit for hard knee.

I thought about this too and I think you'd sell quite a few of these to the tons of people who are still using 903s!  How's that for an incentive!


Of course, I still would like a couple PCBs of whatever version you end up with.

 

[[silent:arts]]

Well, if you want the hard knee option you should consider buying the DBX 560a.

Looking at the 160X schematic the side-chain is completely different to the 903, using two RMS converter, one for "over easy" and another for the hard knee and so on. It is not like changing a few resistor values or adding / leaving out a few components.

 

[mitsos]

Well, if you want the hard knee option you should consider buying the DBX 560a.

That is an option, but I'd rather build the Silent Arts 903 with switchable knee!  ;D 8)    Also, I just saw they have a de-esser based on the 902!

Looking at the 160X schematic the side-chain is completely different to the 903, using two RMS converter, one for "over easy" and another for the hard knee and so on. It is not like changing a few resistor values or adding / leaving out a few components.

I can't make much out of the 160X schem on their site, it's all cut up and there is a small sliver missing.  Specifically R83 and R84.  Do you have a full schematic?  Anyway, it seems to me, (someone smarter than me should look at this)  one RMS is only used for metering?  And the over easy switch is switching between two opamps following the U13 RMS detector.  half of U5 and half of U6, the ones surrounded by diodes.

Anyway, it's your project, like I said, I'll take some PCBs regardless.

 

[[silent:arts]]

902 to come later 

Attached the best readable of different 160X schematics I have.

 

[mitsos]

of course.. I was looking at the 160XT..

but wow, this one, the 160X, the schematic makes me dizzy.  I'll have to print it so I don't get seasick, hopefully others will chime in, and maybe the hard knee will work out...  That is, if you want to do it.

cheers!

 

[emrr]

My memory says a simple switch of knee changes the meter calibration, so one or the other meter setting will be wrong in the simple version.  DBX is probably using 2 SC and scaling them relative to one another; guessing.