Wondering about MPP Beach REV-E ground plane pads

[Polymer]

I have been reading a lot about having a solid ground connection and no daisy chaining power rails.

What has me wondering are the piddly little connections to the ground plane from the pads.
There are usually four of them to a pad. Is this enough for a "solid" ground?

Is this intentional? Why not have the pads solidly connected to the ground plane, I mean complete copper connection all round the pad?

The ground plane also looks to be interrupted by many traces. The way back to "mother" ground looks like a run through a maze.

Is there maybe a third PCB layer of contiguous ground inside the PCB?

Here a close up picture with the piddle pads and also possible return paths for Tx coil electrons.

[Mdtoday]

I have been reading a lot about having a solid ground connection and no daisy chaining power rails.

What has me wondering are the piddly little connections to the ground plane from the pads.
There are usually four of them to a pad. Is this enough for a "solid" ground?

Is this intentional? Why not have the pads solidly connected to the ground plane, I mean complete copper connection all round the pad?

The ground plane also looks to be interrupted by many traces. The way back to "mother" ground looks like a run through a maze.

Is there maybe a third PCB layer of contiguous ground inside the PCB?

Here a close up picture with the piddle pads and also possible return paths for Tx coil electrons.

[ATTACH]45576[/ATTACH]

Hi Polymer,

It is standard board design practice to connect pads like this to large copper areas (generally ground planes) and it is done mainly to enable reliable solder joints.
Without them, the amount of heat required to solder to ground plane areas increases quite dramatically and then you run the risk of overheating some components.
The MPP board shown is layed out quite well..

When it comes to signal grounding and low noise design, shortest route is best and the so called "star" grounding method is often used, the current from each ground path meets at one point only.
Then you have partitioning of ground area methods too, all this helps control current paths to ground in an effort to keep noise down, among other things.

On the flip side, things like the TX Tank cap also require short & fat traces to help reduce the nasty high current switching noise

Here is a snippet of part of a star ground on one of my boards (not MPP)




cheers

Mdtoday

[godigit1]

Mdtoday
I like the way you did the star ground,
Got me thinking.

[Mdtoday]

Mdtoday
I like the way you did the star ground,
Got me thinking.

Hi Godigit1,

Thanks and yes, it really does help and in particular around the front end amp stages, of course, it becomes hard to get all of the signal chain to same point (and not all needs to be) but then that's where partitioning of the ground plane helps.
It is an extension of the star ground method in a way.

Cheers

Mdtoday

[Polymer]

Hi Mdtoday,

Ha, you're a Star! Thank You for your concise description and your star example.

It now makes sense to me. Easy solderability is important, especially to beginners like myself.

Would putting in a wire from Tx Return connection directly to the cap benefit much?

[waltr]

Hi Godigit1,

Thanks and yes, it really does help and in particular around the front end amp stages, of course, it becomes hard to get all of the signal chain to same point (and not all needs to be) but then that's where partitioning of the ground plane helps.
It is an extension of the star ground method in a way.

Cheers

Mdtoday

That is an interesting PCB method for a Star Ground. Nice.

Yes, it is very important to keep the high current return paths and the low current signal ground paths separate with 'star' connection.
This also goes for the power distribution.

I did my own version of the Hammer Head and keep the TX ground/power separate for the signal circuits with a star at the battery connection.
Very quiet without many of the issues I read about the this forum with other PCB layouts.

[Mdtoday]

Hi Mdtoday,


Would putting in a wire from Tx Return connection directly to the cap benefit much?

You could do this but in this particular design, it is ok, you can't get the cap any closer to MOSFET, although I would have had thicker tracks and separate return right back to power connector.
Having said that, the layout is fine for the Tx pulse current in this design, I would not go hacking the board.
if you want to look at experimenting, lift the tank cap leg that is connected to return and connect a wire from it directly to power input connector area...I don't think you will see a lot of difference but this is the way to think when designing layout.
A design With higher Tx currents, larger tank cap and higher Tx freq, you would...

cheers

mdtoday

[Mdtoday]

That is an interesting PCB method for a Star Ground. Nice.

Yes, it is very important to keep the high current return paths and the low current signal ground paths separate with 'star' connection.
This also goes for the power distribution.

I did my own version of the Hammer Head and keep the TX ground/power separate for the signal circuits with a star at the battery connection.
Very quiet without many of the issues I read about the this forum with other PCB layouts.

Hi Waltr,

Yes, certainly makes the difference, every little thing helps..
I also separate or partition the ground planes where required, that makes a difference too.
I haven't actually built a Mini-pulse..been doing my own designs but I might lay out an SMD version using 0805 parts for fun when I get some time.
Probably go with PIC micro for timing

[waltr]

Hi Waltr,

Yes, certainly makes the difference, every little thing helps..
I also separate or partition the ground planes where required, that makes a difference too.
I haven't actually built a Mini-pulse..been doing my own designs but I might lay out an SMD version using 0805 parts for fun when I get some time.
Probably go with PIC micro for timing

Sounds good and look forward to seeing this.

I have been using a PIC16F88 and starting a thread in the "Programming" sub-forum for discussion.
Please join in.

[Polymer]

You could do this but in this particular design, it is ok, you can't get the cap any closer to MOSFET, although I would have had thicker tracks and separate return right back to power connector.
Having said that, the layout is fine for the Tx pulse current in this design, I would not go hacking the board.
if you want to look at experimenting, lift the tank cap leg that is connected to return and connect a wire from it directly to power input connector area...I don't think you will see a lot of difference but this is the way to think when designing layout.
A design With higher Tx currents, larger tank cap and higher Tx freq, you would...

cheers

mdtoday

The observation of "the flow of things" is sticking to my mind like velcro ... thanks for the inspiration.
I will be playing around with this on the MPP with a direct connection from coil return to GND.
Maybe I will spot a difference on the scope, maybe not. I'm curious and learning.

What would be a ballpark figure for "higher Tx frequency"?

[Polymer]

Found an old banged up Sony stereo on the side of the road from the 80's. Couldn't resist.

The insides have lots of nice components for experimenting and in sizes I can handle without a magnifier ...

Anyway, I found a lovely star ground connection on one of the PCB's. What a work of Art!

Here's a picture:

[waltr]

Wow, now that is a star.
Great find for usable parts.

What would be a ballpark figure for "higher Tx frequency"?

I run 1500Hz (666.6usec period) on my HH2.
Higher frequency helps the integrator to change faster thus- more sensitivity.