DBX117 More compression please! DBX 117

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

OldMan

New member
Joined
Sep 4, 2006
Messages
3
Location
Switzerland
Hello

I would like to modify a DBX 117 to get a little more compression then 1:1,4

I have just the schematic that "Verbos" once have submitted.

Can (and would) anybody help me?

A better (bigger remember I'm an old man) schematic or a Manual woud also helps a lot ;-)



schemadbx117.jpg


Thanks and greetings from Switzerland
OldMan
 
Well bless old David Blackmer's soul, its been a while since I've seen one of these schematics.

All you have to do to get more compression and expansion is raise the gain of OA4 by increasing R39 in small amounts to taste - say 12K or 16K. This will amplify the output of the RMS level detector to a greater extent and drive the VCA harder. Maybe make this variable (with end limit resistors) and call it a "whack" control.
 
thank you!
You are so cool I have to where gloves to reply this!

@Michael you must have better eyes then me, but finally I found R39
I don't need more expansion just compression ;-)


What are the main differences (I mean pro and contra) of these two possibilities?

@Keith how mutch would be the compression after this modification? did you know that? Or have I just to try?

thanks!
 
thx

Version 1 (Keith) is working great! Except that all the useful compression rates are now in a very small range between 8 and 9 on the potentiometer.

OM
 
Determine what the lowest that you'll need is (say 8 on the pot) then what the highest you'll need is (say 9 on the pot) then take the pot out and measure it. I you have a 20k linear pot, I expect that 8 will be 16k into the rotation. I expect that 9 will be 18k into the rotation. The difference is 2k. What I'd do is replace it with a 2k pot, and add 16k below it and 2k above it.

Then replace it where you had it wired as per my suggestion, and you should now have the useful range that you wanted spread over the rotation of the new pot.

Keith
 
Got myself a DBX117 and I love this little unit.
I swapped the RCA for TRS Jack and I think I'll upgrade some caps...
I am trying to do some mods on it, I have found the following on gearslutz:

The compression factor may be extended to infinity by shorting out R35. The expansion factor may be increased to 3:1 by reducing R40. Doing both will result in a compression factor beyond infinity! The output will actually decrease when the output increases causing dynamic inversion. Recalibration of EXPRESSION control will be necessary.

Precisely doubling the epansion and compression slope is easily accomplished for a monaural source by routing the signal through the dbx 117 twice.

The NORMAL release ratio is presently optimum for most uses. Release rate may be increased by removing C13. Release rate may be decreased by increasing the value of C14 and decreasing the value of R42.

The DBX 117's level sensing circuit is responsive to the sum total of input spectrum from 20Hz to 20KHz. Should you wish to make it less sensitive to low frequency sounds, decrease the value of C11 to .033 mfd or even .01 mfd.

When wider dynamic level tracking is required the low level excursion limit may be removed from OA3 by removing D5 and shorting R33. The output for metering and alternative control systems may be taken from the junction of R34, 32, 35, 38. This signal may be fed to any external transfer function control circuit and the output of said circuit may be re-inserted by opening R40 and feeding into the junction of R40 and R41 with an external 220 ohm resistor
.


Right now I have bypassed R35 with a croco clip and I really dig how the comp is behaving (like crazy basically, the compression is so intense than with the large know you can blend some compression with some expansion which is what I was hoping for..) but when there's no audio, I have a HUGE white noise coming up like crazy.. When I poke around the large blue caps (C7, C11 and C17 I believe) the white noise volume kind of drops but doesn't disapear.

What could I do to avoid that?
 
instead of shorting across R35, perhaps you can just reduce it's value. Try another 1k resistor and another croc cable, and jumper across (2 x 1k in parallel = 500ohms) R35, if R35 is 1k if I'm reading this correctly, kinda hard to read... that should increase signal out of the detector and into the VCA. This is just a guess really.

I like keith's idea of moving the pot to the other side of R35. Maybe even replacing R35 with a 1K pot (linear?)...
 
I have tried that, but the more the value of that resistors leans to 0, the more I have that white noise coming up... I think this is some sort of hiss that I amplified like crazy. I believe it comes from the PSU that is not well regulated. On the interwebs I have found some posts from Jim Williams at Audio Upgrades. He tells that a good opamp replacement for the 4558 would be the National LME49720... but my unit is fitted with ML301AS (yes, ML301 not LM301s)

I have changed the main PSU caps for 2200uF instead of the original 470uF, as well as all the PSU caps. They were 22uF and I've put 100uF instead. My -10VDC dropped to -8VDC so I am trying to reduce the value of the -10V filtering cap right now as I post.

I wish to add a switch in the back of the unit for shorting R35 or maybe doing something else to get that inf ratio compression, as it's pretty interesting!, and avoid another pot. I like the controls on that comp, 2 pots, no VU and would like to keep something in the same spirit.

What would the pot do if it was on the other side of R35?
 
thomasdf said:
...
What would the pot do if it was on the other side of R35?

It's two "nested" voltage dividers... moving the pot would place it in the dominant position, i.e. the first v divider. If you look at where the "control voltage" (for lack of a better term) is tapped off from, the pot in the schematic has a very tiny effect on things as-is. Moving it would have a much greater effect on the signal going to OA4.

The hiss you're hearing with no signal going in is probably the normal noise floor + some HUGE gain. Kinda of like when you're listening to the radio and the DJ takes a little nap and you hear the noise floor creep up real loud when the station's Optimod or whatever goes haywire.
 
Thanks hymentoptera, this helps!
I'll try that moving that pot connexion to see how it sounds. If I understand correctly it will give me more compression and less expansion?

And yes about that white noise, I think you are right... I was affraid about that...! I need to find a solution to get rid of it, or at least lower it. Maybe opamp swap + tight wiring + aluminium foil on the wooden panels inside since the full recap + bigger PSU caps didn't change much.

Problem with the opamp swap is I couldn't find a proper datasheet for my opamps (ML301AS 7422) yet :)
Edit: I have found it, it is similar to the LM301AJ and the National LME49720 does not have the same pinout. So I'll have to to some leg bending... But I'll give it a try!
The only thing is that it's a dual opamp, so maybe it a bit dumb to use only one side of it?
Is there a single opamp that's as good sounding and fast as a the LME49720?
I quote Jim Williams here "Replacing the 4558 opamps with a National LME49720 really helps the sound and their -120 db psu rejection gives a hum free output."
It's very tempting!

Yes I hear you in the back, and no I am not getting a bigger DBX comp! This one rocks!
 
Wish I could help more, but I really know very little about this sort of thing. I can't make heads nor tails of that power supply... the bc109 and bc179 transistors giving (if I'm reading it right) +7vDC and -10vDC would seem archaic except I've never seen that done that way before. Perhaps gutting the entire PSU and replacing it with a custom regulated supply would do? Looks like there's plenty of room inside the box for a little perf board PSU with 4 x TO-220 or TO-92 (depending on power needs) regs. Or an external supply and get that transformer out of the box entirely.

Replace carbon films with metal film resistors in the signal path? Can't hurt to try.

This sort of thing is way over my head, but I'll be watching this thread for future replies and hoping to learn something. These things, and many of their cousins, still go for $50 all day long on ebay and look like there's plenty of room for improvement.

Can you show me what you found out about the ML301AS? Only thing I'm finding suggests it's another name for LM301 which is a single op amp. Maybe same as 741? Might have to tinker with the compensation caps if changing them out... Would love to see a datasheet, or at least a pinout to be sure what the heck it is.
 
I will answer properly and give you links tomorrow as I am on my phone right now.
Just wanted to give a short feedback on moving the pot on the other side of r35. The compression can be interesting indeed (maybe a bit too violent) but the f*in noise comes back like crazy again in the quiet parts...!
I will do further testing and maybe build a new PSU if the opamp swap doesnt help.
 
Ok, so the datasheet for the ML301AS can be found here:
http://www.datasheetarchive.com/dlmain/Datasheets-X2/DSA12000002430.pdf
Not the most convenient way, but this pdf has it's specs and pinout.
It says its substitute is the LM301AJ.

I have ordered a few National LME49720 and I will put them in, and will adapt the legs as the pinout is completely different.

When you say compensation caps, which one are you talking about? Is it the 33pF caps? C18 and CUnknonw on the top of the schematics? What value should I use for the LME49720? I have seen 27pF on some examples in the schematics... 6pF won't do much different right?
I was planing on wiring the opamps to V+, V-, Inv in, non inv in and out, as these are the only features used with the original opamps. The other pins are soldered but not linked to the rest of the circuit. Plus, some of the original opamps have a 33pF caps between pin 1 and 8... But I believe I won't need those anymore.
 
I'm no expert on these sort of things, so I don't want to lead you astray. I'm hoping someone else will speak up and correct any mistakes in our thinking here. I don't understand enough about upgrading opamps to say for certain what will work, what needs done, and what you can expect as far as performance, current and stability, etc.

Here's a link where they're talking about replacing some 301s, and if you can determine that those are being used in the same mode as in the 117, that that could be a place to start.
http://groupdiy.com/index.php?topic=42559.0

For instance, they're talking about using one of those little Brown Dog surface mount-to-DIP adapter boards to fit a LM49870 and strapping a 100nF cap across the V+ and V- pins 4 and 7... I've only done that sort of thing under the watchful eye of a more experienced man telling me what to do lol, so I couldn't say if this is right or wrong.

Opamps are still like voodoo to me!
 
Not sure if this is allowed in this forum, but here's a thread with upgrade options for 160.
https://www.gearslutz.com/board/geekslutz-forum/121807-dbx-160-vu-mods-caps-replacing.html

Not sure how different these are, but they seem to be based on the same principal and share a few parts...
 
No worries! I am waiting for some opamp guru too. I scratch built 4 of this adapters:
KTWICYkm.jpg

Not very fancy looking but they should do the job.
Basically what it will do is that it will use only one side of the National opamp and route the V+ and out to match the LM301 specs (and thus the PCB). I will bend the unused pins.
That 100nF caps looks interesting, I will do some further reading, thanks for the links.
 
Hello Folks
So I swapped the stock opamps for a few ones the past few days, didn't change much...EXCEPT with the LM49870 and with these, the noise is gone.
So that's for the good news! The only opamp that has less impact on the noise amplification is OA4. I can use whichever I want and it doesn't change much to my hears.
The only problem is with OA3 (sidechain opamp), here is why:
When the opamps are switched to LM49870s , no noise it's cool but with R35 bypassed, the compression amount is more than what I need and the output level is ridiculously low. My guess is that the LM49870 in the SC have too much gain and thus do to much gain reduction. Is there a simple way to lower the sidechain's opamps gain? My guess is yes, I need to find out... But a pointer would be helpful here :)
In the audio paths, LM49870 instead of the original ML301s sounds cleaner, less chunky / lofi but I can live with that, since the color of the unit mainly comes from the VCAs and the noise floor is much better.
It's just that crazy yet usable compression of bypassed R35 + ML301s in the SC that I'd like to get back to.
 
Maybe replace R39 (10k) with socket pins and try some lower value resistors there? 9.1k or 8.2k? I think the gain of the sidechain is finalized by that 10k resistor. A lower value increases the negative feedback I think (but I really don't understand how these compressors work! so caution please)

I wouldn't think a different opamp in the sidechain would make much of a difference. I hope it's not screaming in there and oscillating or something... that might explain the really quiet output, but I would expect it to have other harmful effects too. How has the control voltage changed? I'm guessing it's now much higher? Got a scope?
 
Back
Top