DBX 286A Modifications

drtechno

Member
This neat little channel strip seems to need some modding attention. Its not bad nor really that great, but I think its worth the try to make better.
 
Looking at the schematics, I see that this thing went through generations of revisions. it is here: dbx Professional Audio a zip file that contains all the versions.

even though I titled this thread dbx 286A Modifications, I will be modding the new version (286s)....
 
got the unit in today. It comes with a nice power cord. Looking at the specifications, the first thing I am going to look at is its mic pre. In the specifications is says that the input impedance is ~4.2K. that is a little high for microphones, and about twice the impedance of the transformer-less mic preamps used in the converters (which are not really that spectacular). So, I think first I should do is come up with the method of turning this transformer-less mic pre into a transformer input with an impedance that would be more agreeable to all of the microphones.
 
got the unit in today. It comes with a nice power cord. Looking at the specifications, the first thing I am going to look at is its mic pre. In the specifications is says that the input impedance is ~4.2K. that is a little high for microphones, and about twice the impedance of the transformer-less mic preamps used in the converters (which are not really that spectacular). So, I think first I should do is come up with the method of turning this transformer-less mic pre into a transformer input with an impedance that would be more agreeable to all of the microphones.

Not sure why you think 4k2 is high for a mic pre input Z? The modern 'fashion' is for higher impedances than the fairly standard 1k2 -2k2 Ohms. A few boutique pres have Zs over ten k!
The idea is that dynamic and especially ribbon mics get an improved HF response when lightly loaded (capacitors are mostly a purely resistive source and care not a jot) . Most authorities agree the effect is subtle.

Then, the higher the load on the mic the higher the output voltage will be improving noise performance. Do you have the specific schematic to YOUR strip? The maps I looked at seemed have Zins of no more than 2k balanced.

Transformer? Could be good for better noise (and brilliant for stopping RF!) but you want a step up ratio of about 1:4 (12dB boost) into an NE5534 or better*. GOOD transformers are of course costly and you have to be aware that they pickup hum.

*I would have a do with the LM4562. You could parallel the two sections (feed op via 2x10r) and get another 2dB or so noise improvement on an already very quiet chip.

Dave.
 
Not sure why you think 4k2 is high for a mic pre input Z? The modern 'fashion' is for higher impedances than the fairly standard 1k2 -2k2 Ohms. A few boutique pres have Zs over ten k!
The idea is that dynamic and especially ribbon mics get an improved HF response when lightly loaded (capacitors are mostly a purely resistive source and care not a jot) . Most authorities agree the effect is subtle.

Then, the higher the load on the mic the higher the output voltage will be improving noise performance. Do you have the specific schematic to YOUR strip? The maps I looked at seemed have Zins of no more than 2k balanced.

Transformer? Could be good for better noise (and brilliant for stopping RF!) but you want a step up ratio of about 1:4 (12dB boost) into an NE5534 or better*. GOOD transformers are of course costly and you have to be aware that they pickup hum.

*I would have a do with the LM4562. You could parallel the two sections (feed op via 2x10r) and get another 2dB or so noise improvement on an already very quiet chip.

Dave.

4.2K input z is very high impedance for mic inputs. The 20K balanced too seems high considering 10K balanced is the norm. If someone uses the operator's tolerance (called the 10 times rule) then the lowest microphone impedance can only be 420 ohms! so only some of the microphones fit in this range. I can't find the 286s schematic online. Only the 286A I could find. But that one doesn't look great either.

As far as the microphone getting a load from the mic pre, a higher termination impedance, the less the load is going to be for the microphone. Transformer-less output condenser microphones need an impedance that is low enough to stabilize its gain circuit. Transformer-less dynamics needs a low impedance input (ideally the same) so that it can develop the voltage evenly across its frequency response. Then any of the other mics that have a transformer. The most power transferred without acquiring insertion loss is when the output transformer of the mic has the same impedance as the preamp.

Ideally, a mic transformer should have input taps at 50, 150, 300, and 600 ohms. When I get off work today, I'm going to open the unit up, take pictures and amend the 286A schematic I posted below with the values soldered in the 286s.


Here is the spec sheet:

https://3e7777c294b9bcaa5486-bc9563...802_1324424228/dbx286s_DataSheet_original.pdf

And here is the mic pre schematic for the A model:286micpredetail.jpg
 
Last edited:
I don't know where you are getting these ideas from friend? You seem to be some 50 years behind modern practice, nobody 'power matches' audio signals any more (no, not even valve amps and speakers!) .

Dynamic mics at least are often not the actual source impedance they state as 'nominal. The SM58 for instance is actually 300 Ohms and were you to plug that into a 600R pre amp you would lose 3.5 dB of signal (with a corresponding noise penalty) for no good purpose. The suggested 300 Ohms would of course cut signal by 6dB. This is the reason for the times ten rule and all modern microphones are, AFAIK designed with that loading scheme in mind, even capacitors.

BTW that circuit actually has an input resistance of about 3k1 due to the loading of the phantom power Rs.

Twenty K is not an unusual input Z for a balanced line input. The classic circuit is formed around an op amp with four 10k resistors thus the full balanced input MUST be 20k*. You could use 4x5k but then the unbalanced input would be rather low for some sources at 5k.

You can of course mod your own kit any way you chose and rationalize the reasons but many people visiting the forum find matters of signal transfer and 'Impedance' very confusing and it does nobody any good to muddy the waters. You might of course have some very new data that turns upside down all that I have learned about audio electronics in the last 50 years or so but unless you can present that I shall stay with 'the herd'!

*It is actually LOWER than 20k when driven from a symmetrical source but the reason is subtle (sort of 'anti-bootstrapping') and makes buggerall difference anyway!

Dave.
 
Have fun Doc!!!! :D. I popped the 'bonnet'(British speak Dave might like) on my DBX286s and found pretty much what I expected. Personally I avoid/hate working on surface mount components, but if it's something that makes you feel warm and fuzzy, certainly have a go at it. I don't find the 286s to be bad by any means with the variety of mics I've used with it. Usually an RE20 or 320 straight in or an SM7b with a Triton Fethead(22kohms of the Fethead brightens up the SM7b). If I were you and didn't like the High-Z input of the 286s, just sell and get something more to your liking.

I didn't look very far, but the IC's marked 'ZFN115' didn't get any Google hits. The 4580 is pretty common.
 

Attachments

  • 286s_2.jpg
    286s_2.jpg
    507.9 KB · Views: 155
  • 286s_1.jpg
    286s_1.jpg
    2.9 MB · Views: 75
I keep a spreadsheets on some of my gear. This is a section of mic preamps showing input z from the manufacturers specs. Yeah, the DBX286s is a tad on the high side of others, but does help dynamics that get plugged into it.
 

Attachments

  • pre z.jpg
    pre z.jpg
    105.5 KB · Views: 66
Have fun Doc!!!! :D. I popped the 'bonnet'(British speak Dave might like) on my DBX286s and found pretty much what I expected. Personally I avoid/hate working on surface mount components, but if it's something that makes you feel warm and fuzzy, certainly have a go at it. I don't find the 286s to be bad by any means with the variety of mics I've used with it. Usually an RE20 or 320 straight in or an SM7b with a Triton Fethead(22kohms of the Fethead brightens up the SM7b). If I were you and didn't like the High-Z input of the 286s, just sell and get something more to your liking.

I didn't look very far, but the IC's marked 'ZFN115' didn't get any Google hits. The 4580 is pretty common.

Looks really different from the schematic. It looks like they went to a transistor based amp like what is in the Behringer ADAT and others. I'm personally not a big fan of surface mount and audio because its so easy to go cheap too far and the product suffers. The polymer electrolytic caps they are using for audio coupling might not be the greatest but there hasn't been enough data out there to just trust that they are ok either.

I think these newer generation of transformer-less mic preamps are not really engineered that good or there is something lost in sourcing the parts since its surface mount.

Btw I didn't realize this thing went surface mount. But that doesn't deter me. Its more of a challenge of a different flavor sourcing good surface mounts though...

But it does explain why no one as really attempt modifying them.
 
I don't know where you are getting these ideas from friend? You seem to be some 50 years behind modern practice, nobody 'power matches' audio signals any more (no, not even valve amps and speakers!) .
.

I think there is flaws in modern practices. Mainly because there is no real electronic law it is truly fallowing.
 
I think there is flaws in modern practices. Mainly because there is no real electronic law it is truly fallowing.

I don't really understand that. No matter. ALL electronic circuits have to follow the laws of Mr Ohm and Co. It might SEEM at times that something 'magical' is going on but if you have the equipment and chops you can always find a rational reason for circuit behaviour, at least for things as simple as mic pre amps!

I shall commend to you Small Signal Audio Design by Douglas Self. (V good section about electrolytic and other capacitors)

Dave.
 
I don't really understand that. No matter. ALL electronic circuits have to follow the laws of Mr Ohm and Co. It might SEEM at times that something 'magical' is going on but if you have the equipment and chops you can always find a rational reason for circuit behaviour, at least for things as simple as mic pre amps!

I shall commend to you Small Signal Audio Design by Douglas Self. (V good section about electrolytic and other capacitors)

Dave.

I do agree that all electronics circuits have to follow Ohm's and Kirchoff's law. I really think that the newer generation of E.E.s that are designing in the field don't necessarily look what the operator's needs are in pro audio. So I think they look for "it works" scenarios instead of "quality of design". A lot of E.E.s know how to build a preamp. But very few can do it on the same level as Mr. Neve and Mr. Putnam (rip).


I'll look at that book. I'll see what his view is on caps. One of the things I have noticed over the years that the construction materials and design of caps has changed. So one manufacturer seem to be better than another on one application, but doesn't work in a different circuit. A year or so ago I changed out a orange drop in a multi channel unit that the replacement orange drop made that channel sound different. I knew it was the part because swapped it onto a different channel and the problem followed the part.
 
Last edited:
BTW I was hopefully going to see an ssm2017 or an ina217 installed as the mic preamp instead of four SMT transistors. Its not big of a deal as I've heard an API DOA and it was really good. But something tells me that the transistors and resistors are not going to be matched in the DBX like the guy who built the DOA.
 
ya, you take a look and see if it's worth the effort. Like a caboose on my new Behringer mixer to house the discrete op-amps
 
A little mod for better performance is sometimes OK. Just depends how much room you've got to stuff the improvements into. :D
 

Attachments

  • vw.jpg
    vw.jpg
    186.2 KB · Views: 17
I didn't look very far, but the IC's marked 'ZFN115' didn't get any Google hits. The 4580 is pretty common.

Those are lm339's or ST's equivalent.
i noticed the changed to using that 4805's for the vca / rms cips.
also changed to a jrc4580

interesting changes.

I would be curious what the rest of the board uses. if you pop the hood again id love to see some more pics as i cant see the chip numbers for the glare... and even be interested in what they did on the "S" version to the power supply and transformer too.

it does looks like there is not enough chips there, id bet they have surface mount chips on the underside.
 
Popped the bonnet again.....

Full sized images....
IMG_1202.JPG - Google Drive
IMG_1201.JPG - Google Drive
IMG_1203.JPG - Google Drive
IMG_1204.JPG - Google Drive
IMG_1205.JPG - Google Drive
https://drive.google.com/file/d/0BySzwrcyKPNyeTNxZi0yZUo0OEk/view?usp=sharing
https://drive.google.com/file/d/0BySzwrcyKPNyd0IzRHhkUG5aa2s/view?usp=sharing
https://drive.google.com/file/d/0BySzwrcyKPNySVB5aDRTUHpOVlU/view?usp=sharing

There are a couple 4305 IC's that seem to be doing the VCA stuff > http://www.thatcorp.com/4305_Low-Cost_Analog_Engine_IC.shtml
339's for level comparators to drive the LED metering.
There are several transistors that appear possibly to be the front end of the preamp which then feed the 4580(?).
A few TL074C JFET OPAMPS.
LM3406M towards the output end of the board > http://www.ti.com/product/LM3406 (???)

Some similarities to its predecessor, but more different.
 
Back
Top