SumThing Passive Summing PCB - Schematic and Layout

[jtoole]

Hello all.
Here is a schematic and PCB layout for the passive summing project I've discussed previously (here).

Schematic
PCB layout

Key Features:

• - Passive balanced summing
  - 16 inputs into 2 outputs
  - 2-button bypass IN 1/2 to OUT 1/2 (for normal monitoring)
  - Mono/Stereo select per Channel
  - Channel config via either jumpers or pushbutton switches on PCB.
  - Input available on DB25, 0.1" header, or 0.2" screw terminal
  - Outputs on 0.1" header or 0.2" screw terminal (wired xlr, TRS, etc)
  - IDC pinout for off-the-shelf IDC->DB25 assembly (25-pin PC parallel port style)
  - DB25 pinout for tascam/protools
  - Optional signal ground cap, ground lift, or normal ground
  - PCB Pushbuttons - Assign OUT 1/2 for each channel (DPDT)
  - PCB Pushbuttons - Master Bypass IN 1/2 to OUT 1/2 (4PDT)
  - Pushbutton pad spacing is setup for ITT PHA2UEE switches (Digikey 401-1192-ND - $1.20ea @ 25+ qty)

The PCB can be configured from full-featured 1 or 2U setup (bypass and channel assign front-panel switches, XLR I/O (2U) or DB25 ) to very simple 1U setup (no switches, db25 i/o) - also could send output directly to makeup gain in same enclosure.

Big thanks to Ampfix who created the schematic and layout! Also thanks to others on the forum who've provide input and information. After making another set of revisions, we're going to get a few prototypes built and try everything out. My hope is to create a project both useful to myself and to others on the forum. Please point out any issues, ask questions, provide comments, and let me know if you think this PCB might be useful to you.

If everything goes well, I'll offer PCBs to the extent that there is interest, as well as provide a BOM, build guide, and possible front panel layouts once the project is completed.

Thanks in advance for the feedback! :thumb: Always appreciated!

 

[NewYorkDave]

Looks familiar...

http://groupdiy.twin-x.com/albums/userpics/10031/balancedmxr.pdf

 

[jtoole]

Dave -

Yes Indeed. I hope that wasn't implying your dissatisfaction in some way. I specifically thanked you and made reference to your topology of the assign switches in the previous thread about this PCB.

We put together a new schematic to add new features (bypass, IDC + headers) and to import netlist for board layout. Is that a problem?

jt

 

[needlz]

I would be down for a PCB or two... I was about to start a low brow p-to-p summing box with big ole copper bus wire but this would save me some aggrevation.

 

[fallout]

I'd take one or 2 myself....

-Jay

 

[waldorfcave]

Nice project... I take one ;-)

 

[jtoole]

Minor correction:

The IDC headers are setup to use a standard PC-style IDC26 to DB25F _Parallel Port_ cable, not serial cable (which is DB25M). These cables were used very commonly when PC were built w/o I/O on the motherboard. I've found a bulk OEM source for $1 each at 10+ quantities, so I may order some in bulk to offer w/ the PCBs. They range in price from $2-$8 in single quantities from various suppliers. Two will be required for each PCB (if using db25).

I don't use DB25 I/O (tascam/protools) but from my searching, it seems DB25 _female_ is what is needed on the unit - does that seem correct to anyone who uses those interfaces. What is the connector on the back of the protools I/O (m or f)? All the DB25->XLR cables I found are Male, so I'm assuming female. Hopefully a protools I/O could then be attached directly via a DB25M->DB25M cable... Need to think through the pin assignments there, but I think it will work.

Just trying to make sure we don't have anything 'backwards'...

jt

 

[nwsoundman]

PT's 192 has female DB25's.

 

[king-rb]

i would take one or two...
build the same without pcb a few days ago.. will check the difference.. and make you a wav...

 

[jtoole]

[quote author="king-rb"]i would take one or two...
build the same without pcb a few days ago.. will check the difference.. and make you a wav...[/quote]

That would be interesting - I'll take you up on that once we get boards going... thanks! :thumb:

[quote author="nwsoundman"]PT's 192 has female DB25's.[/quote]

Thanks! Double checked all the pinouts and it looks like everything will work out fine...

 

[synthi]

I`ll get 1 or 2 of your pcbs!

Synthi

 

[BR]

I'll take 2 price permitting.

Gil

 

[rbaker]

I definitely am in for 1-2+

 

[ioaudio]

----> black market.

 

[jtoole]

[quote author="ioaudio"]----> black market.[/quote]

The intention of the thread was hopefully that I could get some feedback on the schematic and layout, not to take orders. Realistically I'm probably 4+ weeks from having PCBs to sell here, because I will be ordering prototype PCBs doing testing before I order a run for the community.

I do appreciate that there is interest in the PCB, but I really was hoping to get some input on the PCB and schematic -> hence "Drawing Board".

jt

 

[fallout]

Should there also be a mute switch for unused inputs? When an input is not used, should the mix resistors for that channel be connected to ground via switch (mute)? I think i remember reading that somewhere here, might have been some info from NYD...

-Jay

 

[ioaudio]

[quote author="jtoole"]The intention of the thread was hopefully that I could get some feedback on the schematic and layout, not to take orders. [/quote]

ok, sorry ! i thought that design is finished. maybe nyd chimes in for more detailed discussion.

 

[jtoole]

[quote author="fallout"]Should there also be a mute switch for unused inputs? When an input is not used, should the mix resistors for that channel be connected to ground via switch (mute)? I think i remember reading that somewhere here, might have been some info from NYD...[/quote]

This design (thanks to the NYD toplogy) is setup such that when a channel is not assigned to either Left or Right output, the internal buss for each is effectively terminated by 2x 10k resistors (the summing Rs). The input themselves just see an open circuit.

It is my understanding that this effectively resolves the issue of unused inputs by terminating unused channels into the sum. I believe this will keep crosstalk and db loss consistent irregardless of # of inputs used. Does that make sense? (in case I'm missing something - always possible :grin

Thanks for the input!! :thumb:

[quote author="ioaudio"]ok, sorry ! i thought that design is finished. maybe nyd chimes in for more detailed discussion.[/quote]

No problem at all. I probably should have been more explicit about that in the first post - maybe why I haven't gotten much response... I'd love to hear nyd's comments - frankly I'm still a bit worried that I overstepped on his design w/ the schematic after his last response.

jt

 

[NewYorkDave]

I was only expressing amusement at the fact that the schematic was just a re-draw (with connector details added) of what I had posted. I suppose Justin Morse was probably similarly amused when he saw what I had posted a year ago and told us that it was essentially the same as his "Folcrom."

When given a set of input impedance, output impedance and loss requirements--especially when they happen to be in that rare happy combination that requires no tough trade-offs--the design of the network becomes pretty straightforward (first year EE stuff), which is certainly why he and I arrived at nearly identical results. He made a product, I made a diagram for DIYers.

Obviously, one of us has more street smarts than the other, but neither of us can claim ownership of the idea of a resistive mixing network with assign switches. These things predate all of us. I believe, though, that the idea of such a passive balanced box as a product--to use with a DAW and mic preamps to avoid the perceived shortcomings of DSP mixing--really comes from Morse. I don't know of anyone else selling them before he came along.

Anyway, I hadn't read that other thread. I really don't get why anyone would need a PCB for something like this, but... whatever.

 

[jtoole]

Thanks for clarifying Dave. I figured it was no problem, but wanted to let you know we appreciated the idea you provided, and weren't just trying to take it and run like it was our own.

[quote author="NewYorkDave"]I was only expressing amusement at the fact that the schematic was just a re-draw (with connector details added) of what I had posted. I suppose Justin Morse was probably similarly amused when he saw what I had posted a year ago and told us that it was essentially the same as his "Folcrom."[/quote]

But not priced like a one... and don't forget that bypass switch :grin: We wouldn't have bothered to redraw except to import netlist.

[quote author="NewYorkDave"]Anyway, I hadn't read that other thread. I really don't get why anyone would need a PCB for something like this, but... whatever.[/quote]

Well, thats a valid point depending on what you need, personally I want a box that will be very rugged - part of my setup is portable, and I thought a PCB setup would be more durable. With switches, at 2 dpdt per channel, thats 192 switch connections + 68 for resistors + 48 for xlrs = 308 points of failure... Plus a pcb provides other benefits like db25 w/o soldering (more robust again, others expressed this interest).

Most importantly its been great fun getting the project together :thumb: I promise I'll do something more exciting in the future :wink:

jt