-
Should I use a ground plane in my audio frequency PCB?
Posted by danieljfarrell on 16 Sep, 2021 11:44 -
I am thinking about designing an audio amplifier PCB.
At high frequencies ground planes are used because the return current tightly follows the signal track.
However, at audio frequencies, I think that the currents in the ground plane are more spread out. This makes ground plane for audio frequencies a bad idea because if significant current is following in the ground plane different parts might have different voltage drops causing ground loops and noise.
I’m really curious if there are any best practices to follow, books, articles or other resources which describe designing audio frequency PCBs.
-
A ground plane is not a bad idea but the problem you identify is real. At radio frequencies, return currents are localized but at audio frequencies they spread out causing ground loops. This can be minimized by using separate ground planes for each circuit function which is equivalent to cutting slots into the ground plane to direct return currents where they should go, or it can be thought of as using multiple ground planes with a single point ground. And if you are prepared to do that, then a careful layout without a ground plane also works.
Does it matter for audio? At the 0.001% distortion level in power amplifiers it does, just like careful layout to control return currents. Otherwise except for resistance to external electromagnetic interference, it would be an exceptional case were a simple ground plane caused any problem or measurable difference in performance.
-
Put simply, I would not use a ground plane. You need to separate high current (output returns) from low current (line level signal returns).
Get a copy of 'Audio Power Amplifier Design' by Doug Self. Look at Chapter 25.
Actually I just found a PDF copy by a google. Prefer to recommend you buy it though - it's well worth it. So is 'Small Signal Design' same author. -
Audio is not exactly my strong suit, although I have dabbled quite a bit. I have some anecdotal advice.
I think you are more likely to make a bad/problematic circuit if you try to avoid using a ground plane. I agree with the above advice...Priority one for a power amp layout is probably to keep the A's of current on the output stage away from the uA - mA of current in the input/pre-amp stage. This is easy enough if you adopt a hybrid ground-plane approach, where you have a small-signal plane and large-signal plane tied together at some central "star" point.
I tried designing a Fender 5F6A clone a few years ago with a purely single-point "star" construction (no planes), and all I got was basically an oscillating mess. I think tubes tend to be a little more prone to this based on high impedances and the high Q resonant circuits that follow.
I redesigned with a segmented ground plane approach...probably a little over the top, and probably didn't need to give each tube stage it's own little plane. But, never the less, the oscillation problems went away and it is dead quiet...well, more quiet than most guitar amps I've played through.
-
Generally I would go with a ground plane, but with some careful thought regarding component placement. Possibly some cuts in the plane around sensitive areas to control the return paths, but don't overdo it. TimNJ's example looks good to me.
-
Thanks for the well reasoned comments all.
I think I might end up designing a ground plane PCB and a star ground PCB just to try it out!
-
Bare in mind that with a star ground (you should still have at least 2 star points for high and low current returns) you can easily hack it to try different configurations and see the results to optimise it. Not very easy with a ground plane.
-
What ground loops and noise? Why are you designing your circuit so it picks up these things?
I think the disconnect is, those who teach "star grounding", stop just short of a general truth. It works for DC, sure, but real circuits are anything but DC.
Actually for that matter, I expect a lot of examples of "star grounding" aren't. As in, they don't actually accomplish the thing they set out to do in the first place, and just make things altogether harder, while inviting problems like oscillation and susceptibility.
The generalization, that works up to RF, is to use a star topology of transmission lines (TLs). Common mode voltage ("ground loop") is allowed between regions of the circuit, because there are no connections crossing between those regions -- except along transmission lines, which are routed through common nodes. At RF, common mode arises as induction along a ground path; since the TLs are well coupled to ground, minimal mode conversion occurs and immunity is high.
To be more specific, define a node as any subcircuit, group or local region, where the voltage drop or phase shift across that region can reasonably be considered as zero. That is, it's point-like, zero-dimensional; a local star point.
Define connections to nodes as ports. Ports are also locations in the circuit which can be treated point-like. Kirchoff holds in the usual way, and all that. Note that connections to transmission lines obey a modified Kirchoff's, where the signal current flows in one terminal of the port, and immediately out the other (image current). Whereas, along a given wire in a transmission line, the current is delayed by propagation, so Kirchoff does not hold in general.
The topology preferably shall be loop-free. If signals need to be looped back between nodes, they should be routed along existing connections (following the unique path between those nodes). The topology really doesn't matter much as long as shielding is good, but also common mode chokes or isolation transformers can be used to improve response.
Overall, the circuit can be understood as a network of transmission lines between nodes, and the crosstalk / shielding of those connections is what determines the performance of the circuit, and what if any remediation shall be used to deal with it (e.g. keeping distance between transmission lines, adding CMCs, etc.).
At AF, we don't have the luxury of well coupled grounds: copper foil rolls off (i.e., is around a skin depth) in the 10-50kHz range, below which it's more or less just another resistive conductor. Note this is no excuse: we must still follow RF practices, as we might be using devices that operate into the 100s of MHz. If we follow the RF methodology exactly, we run into the problem that our transmission line grounds carry some DC supply current from the circuits they're routing between; and this manifests as an error added to our signal at one end or the other -- mode conversion, classical ground loop. What to do?
It's convenient, then, that we built our circuit with transmission lines. We can separate their grounds from circuit ground, lifting one end at DC, so that we can sense the DC/LF error, and subtract it from the signal. At high frequencies, the shielding remains good (or, it can be, at least) and we don't need any special precautions; therefore we can bypass the DC-lifted end to GND. We get what looks like an ordinary transmission line at AC, but a differential pair at DC.
We could go one step further and make the situation symmetrical: route a full differential trace pair between locations, and use a differential receiver to pick up the signal. This can improve CMRR (particularly around the crossover frequency), but has the downside of requiring an amp with enough CMRR bandwidth to cancel out the noise.
Applied to traditional circuits, full differential design isn't the simplest. If you've got op-amps handy, not really a problem, but for single discrete devices (tubes, transistors), you have the restriction that power ground is also signal ground. This can be addressed to some extent by separating them with bypass caps (e.g. bypassing the cathode to signal GND, leaving the resistor going to PGND alone). You're still stuck by problems like PSRR, where a fraction (triode) or whole (pentode) of the B+/VCC supply's ripple is added onto the output signal. This can be solved in precisely the same way: prefer topologies that are differential (so the supply noise can be canceled out), or ground-referenced (e.g. CCS or gyrator load, with shunt feedback to set gain and reference output to common).
Note that amplifiers don't need anything fancy, as far as ground topology. The main load is the output, so connect the rectifier (and filter) to that. Route power and ground back to the driving stage, and so on, in a linear route, to the input stage. No ground loop to be found, no hum, no crosstalk, no susceptibility. The only interesting part comes when considering stereo, from the same power supply: now a U topology is had, with the power supply being common. Ah, but then the inputs must be split, and in general, they'll just be coming from a common ground source, won't they? Indeed, so we must consider some differential treatment at the input. The input grounds don't need to be hard-grounded, they only need to be referenced to their respective input amplifiers. I can think of some ways to apply a common-cathode stage for this purpose, or there's always the long-tailed pair.
Tim -
I think it's also worth remembering that you don't connect your inverting op-amp's non-inverting input to "ground." You connect it to a signal that happens to be at 0 V.
How do you ensure it's a clean quiet 0 V (because it's a reference)? Aye, there's the rub, as the bard said. -
There is a middle ground between ground planes and star grounds. Its a ground plane with cuts.
There are benefits to ground planes in analog because they can shield sensitive high impedance signals from each other. It even helps if the ground plane is under the tracks, does not have to be between then since the E field lines will close themselves down against the ground plane.
The point of cuts in the ground plane is to constrain high current signals along a path that does not include any sensitive ground connections. Tho in the case of audio amplifiers the power is usually in the positive and negative power rails, the ground inside the amplifier circuitry is usually just a signal, so in that case you can just put down a big ground plane and not worry about it. -
A significant part of the challenge is that audio amps don't operate in isolation. They are connected to other equipment. The input signal negative 99% of the time is connected to earth. So you have an earth connection on your amplifier from the power supply and another from the input signal. You need to be carful where they join
-
Sorry. I missed the attachment when reading the first time.
This approach really interests me because it seems so pragmatic!
Can you comment a bit on the ground plane cuts. I can see they are isolating different parts of a the circuit and they all seems to be funnelled towards a point in the centre. But the “star” point is not really a point, it’s more a region. I’m try to visual what the ground plane currents looks like.
I am also coming from the guitar amplifier design perspective! -
The cuts just make the ground plane act like sort of a star ground. So they go outwards from the star ground location.
Another way to look at it is making multiple ground fills for certain things, then making a connection between two ground fills using a narrow piece of copper. The result of this is that large currents within one ground fill wont spread into some other more sensitive ground fill. The idea is the same as star ground. Each branch of that star is sort of like a ground fill section. The currents inside one branch does not send currents into other branches. -
The generalization, that works up to RF, is to use a star topology of transmission lines (TLs)....
Tim
I can't understand your description. Would you be able to share some examples to demonstrate what you describe? -
My topology lecture might be helpful?
https://www.eevblog.com/forum/beginners/help-designing-emi-filter-for-12v-170v-nixie-power-supply/msg3684994/#:~:text=The%20general%20truth%20is%20topological
Extending the topological deformation idea: we can deform a box into a long, thin section, and neck it down; now we have a shielded cable between boxes. In the same way that an almost arbitrarily-large hole can be opened on a face of a box (given the limitations), we could splay open the shield of that cable, keeping the ground plane close to the traces.
Further, suppose we "bud" another box off the cable. We could make it inline as a node chain (a linear network, (box 1)---(box 2)---(box 3), cable from 1-2, cable from 2-3), or we could tee it off (cable from each box to a central tee where signals are exchanged between them). We can do either process in general, and we will produce a loop-free graph.
If we wish to route a signal between two boxes that are spacially adjacent, we should have some sort of isolation, or at least common mode immunity, because we can only do so by violating the shielding/grounding assumption -- that signal must penetrate the shield at (or near) one box, travel through space, and penetrate the shield of the other box.
...Box? Node, I wrote a definition for that, didn't I.
Ah well, such is the price of informal forum-postings. And the lack of figures. Anyone wants to start drawing here, please don't be shy...
Note there's nothing strictly wrong with loops, at AC, as long as they are well shielded. We can extend shielding over the above penetrating signal and all is well; the loop allows circulating currents on the shield, but it stays outside where it belongs; and anyway, we can in general expect circulating currents due to resonances between cable lengths, nodes, etc. as various kinds of antennas.
But this is not something you can approach at LF, where the voltage drop along even a very thick, wide conductor may be significant. We use topology to confine ground currents to local paths as much as possible, then use differential sensing where we cannot avoid loops.
Tim -
I spend just a few hours throwing this board together so please excuse its faults! It is a single stage for a valve instrument amplifier. You could chain multiple boards together to make up a full amplifier
Would you focus on the way the ground plane is cut into three areas: power, audio 1 and audio 2 (the valve has two amplifiers in one package). The ground currents are funnelled towards the area in the top right where the power ground and both signal ground meet.
Is this kind of approach a reasonable half-way house between full ground plane and full star ground?
This is an audio circuit so max. frequency 20kHz (less because it is an instrument amplifiers not HiFi). I know cuts like these are generally discouraged in high frequency circuits for EMI reasons but that should not be a problem for audio.
Happy to get some feedback!
-
The inimitable Bruno Putzeys had the right idea as described here https://www.hypex.nl/img/upload/doc/an_wp/WP_The_G_word.pdf
basically you have a ground plane, but treat the signalling as balanced right the way down the line, with connections to the ground only used to define common mode voltages, and only at a single net tie between the plane and each inter section line.
It does cost you a few extra passives, and makes inverting stages and MFB filters the preferred topology rather then non inverting sallen and key, but that generally does no harm as it also removes the common mode swing from the opamp diff pairs, so is generally the way to play anyway.
I can confirm that this works really rather well, having redesigned a few things this way and them abused them by running a few amps of harmonically rich hash thru the ground plane to confirm it worked as intended, mush better and easier then faffing about with star points and other such magical thinking. -
That’s was an interesting read, and I’m only halfway through!
I total agree that differential inputs make a lot of sense for audio. Problem is for a valve/tube based instrument amplifier it’s impractical, or at least I need to read up one what would be required. I think it might increase the BOM too much.
The part when he discusses the 180-degree out of phase differential inputs vs one signal wire and one ground reference was interesting.