How do I establish 2 grounds...Sewer ground and signal ground?

[momo]

I read another thread where R.G states this....However, noise issues in layout often devolve right back to having to know the circuit intimately. If you know which sections are noisy and in what fashion, you can devise a strategy for keeping the noise where it ought to be and not where it shouldn't be. The same principles apply: don't use sewer ground for reference ground, don't make later stages share the ground return wire with earlier stages, put inputs as far from outputs as you can, don't make overlapping circuit loops. The particulars of how to do that really depend on the circuit.
.
Now ive been building the Ruby amp with an added tremolo inline before the Ruby. I was having oscillation on the Ruby and I think its either PeteMoore or MarkM that gave me the hint of putting Ruby ground before the tremolo connection point(use sewer ground for the Ruby.) That solved a bit of osscilation but not all...

Now im trying to understand this, if Im not supposed to put everything on the signal ground, how do I find 2 grounds?, to me the gound on the PCB, is the same as the battery ground?
The way I understand this, you use the battery ground as the sewer ground and you ground the effects on a common piece of metal not connected to battery?
How am I supposed to power the tremolo if I cant connect the battery ground to the effect?, does plugging the effect ground on a common metal piece seperated from battery ground work to get power to the effect?

Im really confused about this, I cant seem to find 2 seperate grounds....I guess that would solve the oscillation on the Ruby?
Ive got to figure this out, it would be a shame to put to trash this 1 month project because of uncurable oscillation!
thanks y'all for the help!

[R.G.]

Here are some pointers.

A pipe is just a pipe, until you put some liquid through it. It's the USE of the pipe that makes it a water supply main or a sewer.

Same with a wire. A wire which is carrying "used electricity" back to the power supply for recycling is a sewer. A wire which is carrying essentially no current through it has essentially no voltage across it, so the voltage at both ends is very nearly identical. That's a reference ground, because the voltage at the end away from the power supply is the same as the end at the power supply.

The power supply is the ground against which all others are measured. From here, wires go to other places. Ideally, one wire goes from here individually to each place where ground is used. That isolates any possible mixing of sewer ground and signal ground. This is called star grounding because the grounds radiate from the power supply point in a star. The ugly opposite of star grounding is daisy chain grounding; one wire comes from the power supply and every other ground is just hooked to it. Sometimes people do this with the chassis, figuring it's a big metallic wire. Sometimes it even works.

In an amp, an example of a sewer ground is the speaker return. The speaker return wire should always go on a separate wire directly to the power supply ground. It's a high current sewer, and the wire which carries it develops a voltage proportional to the currents. If your other circuitry connects to the wire as though it was a reference ground, you're actually feeding speaker voltage back into the circuit - unwanted feedback. Getting oscillation from this is easy.

An example of reference ground is the ground wire to the input jack. It carries almost no current, only the small signal currents.

Sometimes deciding which grounds are sewer grounds and which are reference grounds is hard. If you want to post your schemo, I'll try to identify them for you.

[momo]

R.G, thanks for the quick and always pertinent info.
Now if I understand this, I should put the negative wire from the spkr directly to batt ground and not go through the provided Ruby ground pad?
So, if I understand this, here is the signal flow....

Star grounding from the battery ground, all grounds(Hearthrob tremolo,booster, and Ruby would directly come from the batt, and plug the negative spkr wire directly back to batt ground bypassing the ground pad on the Ruby layout?

Signal flow would be, Guitar-Tremolo-booster-Ruby, all in line without having individual bypass switches. So Guit input in tremolo,that output to booster input,booster out to Ruby in.All having grounds start from battery including negative spkr wire?

I dont know how to draw a shematic and post it, but my effects would be the following
http://aronnelson.com/gallery/album76/Hearthrob_Tremolo_LAYOUT
http://aronnelson.com/gallery/album76/AMZ_Mosfet_Boost_wBias_Cntrl_LAYOUT

Thanks for all the help!

[MikeH]

Just to check; is your Ruby oscilating when there is no guitar plugged in?  If you don't use a shunt tip input jack on just about any amplifier, even one as small as a ruby, it will oscillate like crazy without something plugged into the input.

[momo]

No Mike, it starts to ocsillate at full volume and then Id say at about 90% of the gain, this has something to do with using the tremolo with the ruby, as individually they do fine...the AMZ booster was not done when I did these tests. Ive come up with a diagram of how I think it should be, I just dont know yet how to upload a gif on a thread!
thanks for the help.

[John Lyons]

Momo
To upload a pic you need to upload it to photobucket or some other web host.
Click the insert image (picture icon) and type in the address of your pic in between the and then click Preview to make sure the link works.
If it look right then click Post and you are good to go.
600 pixels across or smaller is fine.

John

[momo]

Alright, I finally figured out how to draw something, save it as a pic.....now I was drawing with a mouse so it looks like a 5 yr old kid did the thing!
So this is how I would wire it, any problems you see?, from what ive read, I think this would be the way to go to cure oscillation.

[R.G.]

The grounding looks better.

You know, the same issues arise in power supplies. If you daisy chained your +12V lines like you show, the power drawn by the Ruby causes a voltage drop on the power supply wire to the tremolo, back at the first of the chain. The mosfet booster could have the same issue, as probably neither circuit has any power supply rejection.

It could also be capacitive pickup as you turn the gain up.

I would
1. run a separate wire for +12 to each of the tremolo and booster
2. If that didn't fix it, I would insert a resistor in series in the power wire to each of the tremolo and booster and add a power supply bypass to ground on the circuits. 100uF paralleled with 0.1uF ceramic should do it.
3. If that doesn't fix it, start moving wires around with a wooden stick or selectively shielding signal input leads.

[momo]

OK, im progressing with this, I plugged everything as in the diagram,exept for the grounds, I now run seperate grounds for the tremolo and AMZ to the battery ground,... im not getting any output from the Ruby.....so I took my sound probe and well at least im very happy to report that I got the Tremolo and Amz booster working together all on 12v.If I sound probe the input of the Ruby, I get what im supposed, very cleanly boosted signal that I can wet with the tremolo if I wish.

My problem now is I cant get any sound of the spkr wire using the probe and I think this is normal, so how can I debug this?, the Ruby works fine isolated.
So since im getting absolutly no signal or noise out of the spkr, I think its a negative spkr grounding issue, just as if it was not plugged.

I tried to plug the negative of the spkr directly on the Ruby ground(I know thats not the way to go)instead of on the battery, instant ocssilation. So at least, when I plug the spkr negative direct to Ruby ground and ocssilation, I know that its making the connection.

In the diagram, the only way the Ruby is connecting to ground is via the battery, this should be good right?, both spkrs give a 4 ohm total.

So So close on this one!

[momo]

So here is a new diagram to be clear, RG if you are reading this, Im not familiar with the resistor and cap tech, can I just try  the resistors? and if so what value....thanks

the arrow points to the ground that makes the Ruby work, if I take that one out, the spkr negative to batt ground does not make the ruby work.
If it is plugged then the Ruby oscillates when using the drive pot past half way.

http://i150.photobucket.com/albums/s114/momoo_bucket/NEWDiagram.jpg[/img]]

[momo]

If that didn't fix it, I would insert a resistor in series in the power wire to each of the tremolo and booster and add a power supply bypass to ground on the circuits. 100uF paralleled with 0.1uF ceramic should do it.

So anyone can tell me how to do this?, Can I do one of these on a perf board and link it to the 2 effects?
thanks

[momo]

Well, I just hit a BUMP in the road , I did a search on power supply bypass to ground, did not find anything that I can follow to build this,
Can anyone suggest how to build this power supply to ground bypass to cure ocssilation on my Ruby....
again, thanks for the help.

[momo]

Ha, got it!....ya, so im posting the trick for anyone interested that made everything work......R.G has all the right info and well if I had just done exactly as he said, it would of worked!....
So the thing is I did put the sewer ground directly to the battery,....and it was not working ocssilating very bad, it was almost directly, heres the trick, I had the sewer ground and the rest of grounding on what i thought was direct to the battery. The battery lead wire which is about 2 inches long is not direct on the battery!!!
So I put the sewer groung DIRECTLY on the battery connection, the rest of the circuit grounds went 2 inches further on the battery lead wire and that was it!

No more ocssilation, all cured even at full level.
So I learned something today, a bit more than yesterday,but not as much as tommorow.....
Im so happy with this build!...ill be posting a report as soon as the finishing touches are done
Cheers

[Solidhex]

Yo

  A little note to this thread. You can use a battery holder like this one:  http://www.mouser.com/Search/ProductDetail.aspx?qs=1rIBfDHV7ieiur1fZ%2fJFJQ%3d%3d
  It has solder lugs that connect directly to the battery snaps making the whole star ground and star power supply thing a lot easier...

--Brad