Sunday, June 8, 2014

"Nigel" quad reamp and 24 channel summing mixer

This was a fun and laborious project, the offboard wiring was pretty intense, there's just a lot going on.  I still need to finish the last two DB25 snakes.  I used Edcor transformers, DB25 summing PCBs from the GroupDIY White Market, shipped from the UK, and a Par-Metal enclosure, along with Neutrik jacks.  My switches are a little cheapy but they basically work.

It's a 24 channel Folcrom-style analog summing mixer, the last 8 channels can be switched out when not in use, so down to just 16 channels, to minimize crosstalk.  Make up gain is provided by a stereo pair of mic preamps, any of your choice.

The outputs are switchable between Edcor trasformers, or 100uF Silmic II capacitors bypassed with small Wima caps for a little cleaner sound.

The unshielded transformers do not like to be close to the computer or other power supplies, but other than that they can be very quiet.  The Edcor transformers have a really nice sound, kind of "vintage."  The capacitors are quiet in any scenario.

The reamps are switchable between unbalanced 1/4" for standard re-amping, and balanced mic-level XLR for running into preamps for extra coloration of tracks, an idea I took from  Peterson Goodwyn, using H-pads.  Grounds can be lifted on the input, and there is RF filtering copied from the Jensen reamp schematic.

I used 6.8K resistors on the summing mixer inputs and other values for the 4 mono channels, see DIY Recording Equipment article for the math.  I went for about a 150 ohm output impedance when calculating the shunt resistor values.  The output Z climbs slightly to 160something when switched down to 16 channels.  I thought 4 mono channels was right for kick, snare, bass, and lead vocal, the rest of the inputs are L and R pairs for panned tracks or stereo busses.

My short review is, for rock music especially, or anything with lots of tracks, the summing mixer does indeed provide extra space and depth, dimension, to the sound vs the flat clean sound of DAW summing.  I'm all about it.  The analog summing section itself does something too that does not happen with just running a pair of preamps on a DAW summed mix, you can hear it.  The slight complication of setting up the mixdown is absolutely worth it for the sound.  The transformer vs capacitor output was a good design choice, as I prefer either one based on the program material.  I like the transformer for a little midrangier sound, with slightly rolled off high and low exteremes, and a little color.  I like the capacitor output for a very clean sound with deep sub bass.

When I plugged it in and listened to it on a rock band mix I had done, I knew that analog mixing was something of a missing link for me.  Just a little extra spice that helps things "fit together" in a pleasant way, sort of a picture frame around your careful mixing work.  It won't make or break a mix but I like the subtle detail it adds.  I cannot wait to add hardware EQ and compression to the stereo bus.  In the mean time Waves NLS and software comp/EQ are good stand-ins.  I really am curious to see what a real full-on mixing board will do now.  I've got my eyes on the Allen and Heath ZED series by the end of the year.


Preliminary diagrams/not used.  The one on top is the idea to use a 10K:10K line input transformer on every channel of the summing mixer.  I decided not to do this, the wiring, and cost, would have been outrageous.  I decided to get my mojo on the outputs using only a pair, which is what the second two diagrams are attempting to do.

Also if you look at the small handwriting you will see a list of more expensive and fancier transformers you could use instead of the cheap but decent Edcor.  These would be low impedance output mic matching transformers I think anywhere from 200:200 to 600:600 would probably work.  Not all manufacturers use the same impedance for this type of transformer.  I end up using the capacitor output most of the time anyway, and get my transformer sound from the make up gain preamps.  The 600:600 Edcor just offers "more color" if needed.  Kind of sounds "Vintage."  NOTE: I did not use any of these diagrams it was just brainstorming.  See below for what I think I did.

Here is the output section.  I'm super paranoid about phantom power so I used capacitors when bypassing the transformers.  Use two 4PDT switches, one to route the inputs of the transformers/caps from the summing mixer, and the other switch to rout the outputs of the transformers/caps to the XLR jacks.  Since this signal is balanced the caps are configured in +/- pairs to work in place of the transformers.  The transformers are 600:600 Edcors but there are many others that would work.  The caps are 100 uF Elna Silmic II bypassed with .01 uF Wima film caps (parallel wiring).  You can probably play with these values too, or even skip the small bypass caps altogether.  I was just sort of copying Jeff Steiger's VP312DI coupling cap arrangement.  The necessity of the small cap is debatable.  I find that these values and brands sound very good in real world use.

Here is the basic reamp idea of the Nigel.  Basically the ground lift scheme is from the Jensen reamp schematic which is much easier to read than this scribbling.  I also added an XLR output on a switch that uses an H-Pad to drop about 36 decibels, you can find calculators for the resistor values on Google search.

Edcor 600:600 transformer.

Elna Silmic II 100 uF capacitors bypassed with .01 uF Wima capacitors in parallel.  I think 100 uF is flat enough with the low impedance output to extend very low in frequency before any roll off.  I have not run it through a scope, but it certainly sounds plenty deep and low, much more so than the transformers.

This is the reamp section, I did 2 channels on each perfboard.  2 boards for a total of 4 channels.  The toggle switches between Hi-Z instrument level output (TS jack) and mic level output (XLR).  You can see the shapes of  the L-pad (TS jack) copied from DIY Recording's reamp unit, and the shape in sort of an H or a spider/crab shape of the H-Pad for the XLR output that I calculated.  The transformer is in the signal path on both jacks, the switch just changes the resistor pad networks along with balancing/unbalancing for the appropriate jacks.  Be sure to take into account the signal drop across the transformer when calculating pad values.

Reamp inputs with ground lift.  A plastic TRS jack is required to be able to lift the ground.  A metal jack will automatically be physically coupled to the chassis on the sleeve just by tightening it down, so plastic is needed here.

Here is a simple DC-blocking capacitor coupled Line 2 Mic Pad in a standalone box that you could use to run a stereo mix output into a pair of preamps.  It's not a summing mixer but it does give some extra color.  I used Vishay BC Blue 100uF electrolytic capacitors, and .01 uF worth of polystyrene caps in parallel/bypass.

BHL Phono Preamp / WAC -49- #2 with Pyramid capacitors

I built another one of these using basically the same design...see my older post for the layout and general plans if you want to do one.  This time I used vintage USA Pyramid paper-in-oil capacitors for the 1 uF coupling caps.  The sound is even fatter and smoother than the standard Russian caps.  Ever so slightly, of course, but I appreciate the difference.  I think cap choice makes a pretty big difference in a preamp like this, I've read other builders say the same.  One guy used cheap panasonic film caps I think and it didn't sound as good to him.  This is no place to skimp when buying parts.  After all, this is an audiophile piece of equipment.  I had to sand the leads on the big caps clean because there was wax and junk all over them.  They soldered up fine, and work perfectly, despite their old age.  The best part is how cool they look, obviously, and the nasty wax coating that's all over them.

The other difference this time around is I used 100 uF bypass caps on all four JFETs, in parallel with the 50 ohm resistors.  The output is nice and high and easily matches a CD or MP3 of the same material in volume.  I would say that this is an essential modification to the stock BHL unit, I don't know why he didn't design his kits with more volume.  The level with no bypass capacitors in place is just not adequate.  The nice part about working on stripboard like this is it's easy to fit new parts like that, although a redesigned layout with the 100 uF caps specifically placed would probably be more elegant and not so tight.

This thing truly sounds great, and is very quiet.  The next step will be to build a dedicated 20-30 volt regulated DC power supply, and to finally ditch the old Technics direct drive for a better turntable.  I've got my eye on the Audio Technica LP1240 and various Shure and AT cartridges.

I gave the old one to a friend in need.  Things are sounding good in here.