dbx 160xt mods (schematic?)

[Sean Sullivan]

Originally Posted by Jim Williams

A couple of 1N4148 diodes are used. Replacing them with 3mm green led's helps as it places the diode conductance at over 2 volts, it's under 2 volts with the silicon diodes. Remember to use a Sharpie to black out the LED's as they can conduct leakage current from external light sources.

Jim Williams
Audio Upgrades

Figured out which diodes to replace (CR8 and CR9), but I'm not sure what kind of LEDs to buy. I looked at the specs on the 1N4148 and these LEDs were similar:

Green LEDs

If I can just be sure these LEDs can work, I can order the parts and run them through passes! I've been working on a lot of bluegrass projects lately and I like these on banjo and mandolin, so I'm hoping that these improvements will make them even more useful.

[Sean Sullivan]

Hmm...from a user of the Prodigy Pro DIY forum:

"Replacing these with LED's will drastically change the behaviour of the non-linear time-constant.

In a steady-state condition, the timing capacitor is a combination of the physical cap C41 (10uF) that goes from pin 6 of the RMS to pin 6 of the LF353 (inverting input) and the virtual capacitance resulting from the active capacitance multiplier Ceq = C41/C45 . C47 = 10u/47n x 220n = 47u. Resulting in a total of 57uF.
Under large signal variations, the output of U9 swings and the diodes dynamic resistance is very low, reducing U9's gain to 0. The only capacitor left is C41 + C47 = 10.22uF. That means that the circuit is 6 times faster under large variations than at steady-state.

The difference of threshold between LED's and signal diodes makes the detector stay in its slower behaviour for much longer, making it less responsive to attacks. Is that really what you want? That's why dbx, in their great wisdom, have modified ALL the components around the circuit when they went for LED's instead of Si diodes."

Jim, have you found this to be the case?

[Sean Sullivan]

Hmm...maybe I should just try it and see what happens? That sounds a little dangerous

[Jim Williams]

I've seen several versions of that circuit. The version I used in the Basson Sound bass heads didn't respond the way I wanted.

I ended up going with a high quality 10 uf cap. For bass guitar uses, I didn't want the ultra fast attack time. String "Poppers" would create some pumping if it was too fast. I have it set now to grab those transients after the 5k hz and up transient slightly passes. That masked the pumping/breathing so it's extremely smooth for bass guitar. The energy levels at 5k hz and up are not enough to change percieved levels.

The players responded extremely well to those time constraints at Namm this week. It was a joy to hear the very best players do their thing and watch as their techniques were enhanced by the compressor design. They said it was the absolute best sounding bass rig at the entire show.

Me happy.

Jim Williams
Audio Upgrades

[Sean Sullivan]

Interesting. I'd be using them mostly mandolin and banjo, and I don't want to choke down on the attacks, I think that's what kills so many modern bluegrass recordings is losing the strike from mandolin chops and the plunk of a banjo. I love my Audio Upgrades Aphex 651 on bass

So, using the LEDs I listed and changing the 10uF electrolytic with something like a Panasonic FC or Black Gate STD 10uF capacitor would help? Do you prefer to use the LF353 or change that op-amp to something else? I was going to replace the LF353 used for the input to a AD8512, would you use the same for the LF353 used in the inverting input?

[dale116dot7]

Unregulated supply on the output of a push-pull emitter follower within a feedback loop does not normally contribute much to the sound - the PSRR (power supply rejection ratio) usually is pretty good. Pretty much all power amps work that way - it's not necessarily inherently bad. It can depend on the PC board layout, etc. If improving regulation to that output stage improves the sound, then there are inherent design problems with it. Perhaps there are, and maybe it'd help. But improving the supply before regulation might be a better first step.

Upping the main filter caps to reduce the ripple some wouldn't be a bad thing. The designer(s) probably did not want the output stage doing two things if it were connected to the regulated supply. First is overheating the main +/- 15V regs, the other is polluting the +/- 15V rail if the output is heavily loaded. I wouldn't necessarily move the outputs to the regulated supply - or if I was to dabble in regulation, I might regulate them to +/- 18V or something like that. If you up the output capacitors on the voltage regulators, make sure that you understand the powerup and powerdown ramp sequence. You can easily blow up those regulators if the output stays higher than the input. Conveniently, when these regs blow, they usually short input to output, so you can take out most of the IC's if you're not careful. Some designers put an inverted diode (from output of a 78xx/79xx reg) back to the input so if the output voltage is higher than the input (happens during turn-off), the reg doesn't poppencorken. I've blown up a few regs (and boards) doing this, and so I'm much more careful now.

[Sean Sullivan]

Just a quick question before I order, do I want AD8510/12 "ARZ" or "BRZ" The difference in the the voltage-input offset (80uV to 100uV)

[Sean Sullivan]

Well, I went ahead and order the op-amps, capacitors, and adapters from Brown Dog.

[Sean Sullivan]

Most of the parts arrived today, hopefully the rest will show up tomorrow.

I'm going to add the LEDs to one of the DBX's and keep the diodes in the other and see which I prefer.

[Sean Sullivan]

Well, I finished the mods today and the units sound really good. Here's a re-cap of what I did:

U1: AD8512 on Brown Dog Adapter
U2: AD8510 on Brown Dog Adapter
U3 and U4: LM6171
C3 - C6: 2200uF/35V Nichicon HE with Wima MKP-2 bypass
C7 - C8: 220uF/50 Panasonic FM with Wima MKP-2 bypass
CR8 and CR9: Green LEDs (blacked out with sharpie)
C16: 4.7uF/63V Wima MKS-4 with MulitCap PPMFX .01uF/600V bypass
C14, C14, C20, C25, C29, C44, C49: .01uF NPO ceramic
.01uF NPO ceramic across Pin 3 and Pin 4 of U2

[a1greatdane]

Originally Posted by The Beatsmith

couple of novice questions:

- what is the outcome of these mods, soundwise?

- is there someone in the UK that could do this for me?

Hi,
I would be more than happy to help You with these mods at CSW-Chilli Sound Works, which is London based.
Drop us a line if interested with possible phone no. on a private message and I'll be straight back in contact. I have the nessesary equipment incl. a good desolder station and experience with SMD replacements plus 25year pro experience.

Kind regards

'Dr' O aka a1greatdane

And yes - the potential improvements is worth the efforts...

[Sean Sullivan]

I love what these modifications did for my DBX 160XTs. Amazing on bass, and good on kick drums when you need to tame a drummer who's a little inconsistent.

[wildplum]

I want to put in a big thumbs up for Jim Williams upgrades. I recently had him do a 160XT for me and the before/after difference is amazing! I highly recommend Audio Upgrades.

[Rukas]

Id like to send Jim my DBX 160x which I LOVE for vocals right now, but Im worried maybe I wont love it as much as it has been modded? Maybe what I love about it will get modded away?

[Speeddemon]

Originally Posted by Sean Sullivan

Well, I finished the mods today and the units sound really good. Here's a re-cap of what I did:

U1: AD8512 on Brown Dog Adapter
U2: AD8510 on Brown Dog Adapter
U3 and U4: LM6171
C3 - C6: 2200uF/35V Nichicon HE with Wima MKP-2 bypass
C7 - C8: 220uF/50 Panasonic FM with Wima MKP-2 bypass
CR8 and CR9: Green LEDs (blacked out with sharpie)
C16: 4.7uF/63V Wima MKS-4 with MulitCap PPMFX .01uF/600V bypass
C14, C14, C20, C25, C29, C44, C49: .01uF NPO ceramic
.01uF NPO ceramic across Pin 3 and Pin 4 of U2

Sean, I've read that the NE5534's (U3 and U4) were already pretty good to begin with. Any reason to change them?

@All: how would putting BurrBrown OPA2134PA's in U1 and U2 change from the AD ones?

I have both a blackface DBX 166 and a DBX 160X sitting here, opened up.
Also, does one need the bypass caps for the 2200uF electrolytics and if so, where you place them? Parallel, or do you mean the 2 .01uF ceramic caps need to be replaced? (right where the wires from the transformer 'enter the board'?)

I don't have a good schematic for the 160X (my manual came without one), but the one in my 166's manual is actually readable, albeit very, very small. But nevertheless, better than the scanned ones on the DBX site.

The input coupling caps on the 166 are 3.3uF (NP?) electrolytics if I'm not mistaken.