Modifying Auto Release Behringer MDX 1000

[Gertius]

Hi guys,
I have recently gotten a Behringer MDX1000 and really dig this thing!

After messing around for some time with adjusting attack/release times I´ve noticed that I dig the attack of the auto mode the most for punchy drums.
The only drawback is that I´d wish I could get a faster release, but the two are coupled in auto mode.

Do you think it´s possible to mod the circuit so that it gives auto attack like before, but manual release?
From looking at the MDX1200 schemo in the Technical Documents section, two things seem to happen when engaging the auto switch.
The timing capacitor at the RMS detector is being switched to a larger one (SW1C), and the attack/release network seems to be bypassed (SW1D).
Any ideas for modding the device for keeping the auto attack, but getting manual release time?

Cheers!
Chrisitan

 

[gyraf]

We'd probably need a mdx1000 schematic for that?

 

[Gertius]

Hi gyraf,
well yeah you´re probably right, but these seem to be impossible to come by.
I just assumed that it would be similar enough to the mdx1200 schematic.
Let´s say it is, how would you go about the 1200?

 

[analogguru]

We'd probably need a mdx1000 schematic for that?

Oh, thats easy:
- take an Alesis 3630 schematic.
- remove the "gate"-sidechain/(display).
- add a standard Behringer symm in.
- add a standard Behringer (superbal) symm out.

and voila, you´ve got it.

 

[Gertius]

Hi analogguru,
thanks for chiming in!

Interesting notion that the comps are identical.
I just compared the schematics of the Alesis 3630F to the MDX1200 (ok, it´s not the mdx1000) in the technical documents section, and they do seem different.
I´m aware that the MDX1200 has it´s RMS detector a little more discrete with the CA3046, but that´s not what I mean (I saw in the MDX2000 pics thread that the daughterboard with the CA3046 replaces an SSM2120). Also, I was ignoring the exciter or gate sections.

Sure, the general layout is pretty similar, but there are some differences, for example in the sidechain section (Threshold, Ratio, Attack, Release).
Are these differences maybe from the later revisions (mdx1200 vs. mdx1000)? Or are these differences meaningless (I´m  no compressor topology expert)?

Are you sure about this? From all the info I have read from you before I know that you know what you are talking about, but I also have a 3630 here, and I could have sworn that it sounds distinctly different to the mdx1000 (especially the action). Must compare again...

Cheers!
Christian

 

[analogguru]

Hi analogguru,
thanks for chiming in!

Interesting notion that the comps are identical.
I just compared the schematics of the Alesis 3630F to the MDX1200 (ok, it´s not the mdx1000) in the technical documents section, and they do seem different.

Exactly, and I was talking about the MDX1000.  The MDX1200 has similarities with the 3630 (esp. in the sidechain) but it is a different beast.  (The MDX402 you have  is also completely different to the MDX1000.)

Sure, the general layout is pretty similar, but there are some differences, for example in the sidechain section (Threshold, Ratio, Attack, Release).

Please elaborate where you can spot the differences (except that the "Peak/RMS" switch on the 3630 is called "Auto" on the MDX1000).

Are these differences maybe from the later revisions (mdx1200 vs. mdx1000)? Or are these differences meaningless (I´m  no compressor topology expert)?

I was talking about the MDX1000 (only).

Are you sure about this?

yes, because I traced both.

analogguru

 

[gyraf]

Are you sure about this?

yes, because I traced both.

;D Pretty good point!

Jakob E.

 

[Gertius]

Hah, yeah that is a good point
Thanks for clarifying, analogguru.

Please elaborate where you can spot the differences (except that the "Peak/RMS" switch on the 3630 is called "Auto" on the MDX1000).

Well I was talking about the sidechain section in the schematics of the mdx1200, that is different from the 3630 schematics (not too different, but still).
But as you said, the mdx1200 is a different animal then.

Ok, now that we have established the design of the mdx1000, how would you go about my original question on a 3630?
That is, keeping auto attack, but manual release. Is this possible?

 

[gyraf]

If the "Auto" function on the '1000 really is a rebranding of the RMS detector function, it won't be trivial to separate attack and release functions, as they're set with only one common component, the RMS averaging/charging capacitor.

Jakob E.

 

[analogguru]

Ok, now that we have established the design of the mdx1000, how would you go about my original question on a 3630?
That is, keeping auto attack, but manual release. Is this possible?

Not really, but you can play around with this:

http://www.thatcorp.com/datashts/dn114.pdf‎

http://www.thatcorp.com/datashts/dn03.pdf

see also the dbx 166A and 1066 schematic here:

http://groupdiy.com/index.php?topic=45603.0

The capacitor(s) to replace in the MDX1000 are C18 and C218.

Of course this circuit can we used too to modify an Alesis 3630.  ;D

If you play around with this circuit, then please report your results here - I didn´t have time to try it yet.

 

[Gertius]

Hi guys,
ok today I´ve done some intense listening comparisons between the mdx1000 and the 3630 and, like expected after analoggurus info, they sound nearly identical. It´s amazing how I got a very different impression a few months back when I listened with a couple of weeks in between, judging by the impression the devices left with me. The mdx1000 sounds cleaner though, probably because of the missing gate circuit. But the action is the same.

Thanks analogguru for the idea with the non-linear capacitor. It sounds very interesting and seems to be used in many dbx comps (I´ve also read 160XT somewhere, but no idea if that was accurate). I might try that later, but after listening today I had another idea. Basically I found that I would like both shorter attack&release times in RMS mode, so my plan is now to try out different values for the timing cap on the RMS detector (i guess C18 and C218 then).

I´ve looked at the datasheet for the THAT2252 (which seems to be a direct replacement to the dbx1253) and I have a question about the formula for the timing cap and the attack/release times here: http://www.thatcorp.com/datashts/2252data.pdf (page 4 on the right). I was able to calculate t for the 3630 with V=15V(is that correct?), C43=22u and R10=1M, so t=38ms. But now I´m stuck. Is that already my attack time? Or is that the time of the RMS averaging? How do I go from t to the att/rel times? I didn´t quite get that part of the datasheet.

Could anyone please help me understand?

Cheers!
Christian

 

[Gertius]

I just realized that the time constant is proportional to R (in case of 3630 the R10=1MOhm) as well as C.
Do you think I can just as well experiment with different values for R instead of C to change the time constant?
It just would be a lot easier, because I could use a pot, instead of a variation of Electrolytics.

Do you know of any drawbacks or limitations to this?