Designing the pcb

Where we discuss new analog design ideas for Pro Audio and modern spins on vintage ones.
carlmart
Posts: 169
Joined: Mon May 19, 2014 6:11 am

Designing the pcb

Post by carlmart » Wed Feb 18, 2015 12:50 pm

OK, schematic is finished and now it's time to design the pcb. For that I will be using DipTrace.

As I'm building a portable design, that will enclosed in a small box, I will have to use two boards to hold the schematic, which is quite large.

It has an input stage (discussed on another thread and based on THAT's 1510 chip. Then there's passive high pass filter stage. Then there's a limiter stage, then there's the output stage. Besides that there's the headphone stage and LED meter stage.

Up front we will have most of the pots and switches, and all the LEDs.

So I'm using a large horizontal pcb, holding as many parts as I can, and a frontal, less wide pcb for the hardware.

The dilemma is how to come and go, to and from the frontal pcb, if possible using as few connections as possible.

User avatar
JR.
Posts: 3657
Joined: Sat May 24, 2008 7:21 pm
Contact:

Re: Designing the pcb

Post by JR. » Wed Feb 18, 2015 1:46 pm

When passing audio between separate PCB it is best to treat that audio as audio + and audio -, even though the audio - may be 0V, Then use a differential receiver on the other end. In a large circuit with different building blocks there may be several differential stages that can double as receivers.

If locating pots ir controls on a different board from the rest of the circuit send both signal and ground back and forth... trying to save pins on a connector can corrupt the audio.

JR
Cancel the "cancel culture", do not support mob hatred.

carlmart
Posts: 169
Joined: Mon May 19, 2014 6:11 am

Re: Designing the pcb

Post by carlmart » Wed Feb 18, 2015 2:13 pm

I'm not too clear about the "differential receiver on the other end" concept, practically speaking. You mean adding another active stage at the other end?

About sending both signal and ground together, always, back and forth, I agree with you mean. I suppose you mean the signal gos to the pot with its ground, and returns from the pot with is ground, that is four wires and not three.

On my other design I had used a flat cable, and a ground wire was always in between. Using wires I'd say like using separated twisted wires. Is that it?

That has always been the way I had done my DIY preamps and power amps, and the way I always wired and soldered my interconnects. Maybe that's why I never had a hum or noise ever.

You always have to be careful with ground loops doing pcbs too, and sometimes it's not that easy to think about it there.

What I always try to create is some sort of star ground, where all grounds return to a single point, which is not easy to accomplish.

I will not be using ground planes, as I think they are dangerous too.

My first concern is with the first stage, where the signal level is very low. My idea is to make the path from the input caps to the IC as short as possible.

The problem is the gain switch, which has to be on the frontal board. So I wonder which is the best way to connect the gain pins to the gain chain up front.

I have already of also putting the IC in the front board, but that might affect the noise getting into the chip. A big dilemma.

olafmatt
Posts: 37
Joined: Wed Oct 22, 2014 11:33 am
Location: Europe

Re: Designing the pcb

Post by olafmatt » Wed Feb 18, 2015 2:23 pm

carlmart wrote:I'm not too clear about the "differential receiver on the other end" concept, practically speaking. You mean adding another active stage at the other end?
You (most likely) don't have to add an extra stage. For example, if you run your signal into an inverting opamp, just build that opamp as differential amplifier. Take the "negative signal" from the point where the previous stage has it's ground connected. That way you just need two more resistors and an extra "signal" trace.

Olaf

carlmart
Posts: 169
Joined: Mon May 19, 2014 6:11 am

Re: Designing the pcb

Post by carlmart » Wed Feb 18, 2015 2:32 pm

Olaf,

Unfortunately I might not be able to do that on any part I'm interfacing between boards.

User avatar
JR.
Posts: 3657
Joined: Sat May 24, 2008 7:21 pm
Contact:

Re: Designing the pcb

Post by JR. » Wed Feb 18, 2015 3:22 pm

In theory all ground is 0V and you can send single legged signals back and forth. In the real world all audio is relative to some local audio 0V, This is true even if you stayed on one large PCB...

Not trying to complicate your world... build it and you will find out what is and isn't a problem.

JR
Cancel the "cancel culture", do not support mob hatred.

carlmart
Posts: 169
Joined: Mon May 19, 2014 6:11 am

Re: Designing the pcb

Post by carlmart » Wed Feb 18, 2015 4:26 pm

Very and sadly true. Now I'm fighting to not reduce the main board too much, but it has to leave place to the front board.

As it has to be a bit high too (14mm) so I can place two switches below, for phantom power, I better consider that on the front board's back too to make up space.

Sorry, I'm talking out loud now.

It's incredible how things start to shrink from what you thought you have.

On top of it, this prototype version will be TH, not SMD. Some will say it sounds better!

carlmart
Posts: 169
Joined: Mon May 19, 2014 6:11 am

Re: Designing the pcb

Post by carlmart » Thu Feb 19, 2015 10:28 am

As this version I'm doing for the preamp is TH, which cap type should I use for the input filter of the XLR. Values are 470p and 47p.

Film types are fine, but ceramics are smaller for that. Although no all ceramics sound good for audio.

On a different project I would go for polystyrenes, but they are not too small.

The options seem to be X7R or COG.

User avatar
JR.
Posts: 3657
Joined: Sat May 24, 2008 7:21 pm
Contact:

Re: Designing the pcb

Post by JR. » Thu Feb 19, 2015 11:52 am

COG
Cancel the "cancel culture", do not support mob hatred.

carlmart
Posts: 169
Joined: Mon May 19, 2014 6:11 am

Re: Designing the pcb

Post by carlmart » Sun Feb 22, 2015 7:53 am

This is being my first in many areas.

1) I will be using two boards for the main design, and probably add some piggy backs, e.g.: for the limiter and headphone amp.

2) Unfortunately this is a TH version, which makes fighting for every millimeter a continuous, and sometimes losing, fight.

3) At the same time there will be an SMD part: the THAT limiter chip. QSOP-16 .

One doubt I have is how to handle the grounds. My idea, which I would like your opinion on, is to have a supply ground and a separate signal ground.

The supply bypass on every chip would go to the supply ground. All other connection to grounds will be considered audio ground and go to a different bus trace.

Both grounds just join at one point. Would that be proper or nothing to worry much about?

Post Reply