Echo Processor IC PT2399

[linguist]

I have a Echo Processor IC PT2399, on the data sheet they have A Ground & D Ground Symbols on one of the Schematics. ATTACHED.

Could someone help me out with this as I don't fully understand the grounding for the circuit.

I have breadboarded the circuit & have The D Ground symbols to Battery Ground & the A ground Symbols to a rail of their own--basically all connected together but not to the Battery ground.

The circuit doesn't appear to work the way I have it, when I input a sinewave I get an amplified sinewave out but no echo to the output signal.

I was expecting to see a sinewave output with another following shortly after it, but I have not viewed this particular signal before so I don't really know what I should get when it is working correctly.

Cheers

 

[KeepItSimpleStupid]

Think of A ground as a clean reference. Think of D ground as a dirty ground where the absolute reference is less important.

A ground an D ground should be kept separate until they are joined together where the power supply reference is. They can be isolated a bit using LR filters. If the chassis is the ground, then the signal inputs and output grounds can be connected to chassis via small resistors (2-10 ohms). This helps to reduce ground loops.

 

[linguist]

Thanks KISS,

So I just connect the Input & Output grounds to the Chassis ground using a 2-10 Ohms resistor like you mentioned, is that right?

Do I do the same with the Circuit A Grounds, at the moment I have them on a rail of there own--not connected to anything, do I connect these to the Chassis ground also using a suitable resistor as well?

Cheers

 

[KeepItSimpleStupid]

I failed to pay attention to the schematic. The input and output don't show the A grounds.

A and D grounds must stay separate, but be connected at one point. That one point can be chassis, bur it should be a single point on the chassis where they come together,

IF and only IF you use isolated phono plug inputs AND use a metal chassis, the isolated input ground to the gizmo can have the (2-10) ohm resistor to chassis.

For now, two resistors max assuming 2 channels.

 

[linguist]

Thanks Again,

Ok, I have attached a quick drawing with the grounding to make sure I have it right.

The attachment shows a rough square representing the circuit & shows the D ground Battery-Chassis ground top Left, the bottom Left shows the Circuit A Grounds.

I have drawn a line to the Left that is not connected, is this where I should connect the A grounds to the D ground. I think this is what you meant?

Just need to make sure.

Cheers

 

[linguist]

If possible could someone else have a look & see if I have the grounding worked out correctly, looks like KISS has left the building.

Cheers

 

[ronv]

I only see the one digital ground. The important thing is that the digital ground current doesn't flow through the analog ground. So.. At the point where the power enters the box make a point called ground. Pin 4 (digital ground) should have its own wire to this point. Hook all the analog grounds together trying to use pin 3 as a reference point and bring another wire back to the same point called ground. The .1 and 100ufd caps should be close to pins 1 and 3.

 

[KeepItSimpleStupid]

Yep, went to sleep. There are 2 digital grounds. One on pin #4 and one near pin #6. As ronv said, currents through that ground should not flow through any of the analog ground paths.

I'm not sure if your using a metal enclosure with a wall type power supply or is this circuit going to be battery operated?

I'm also not sure if your using isolated jacks for the inputs?

The low value resistors that I spoke of can help prevent hum. When small currents try to flow though nearly zero resistance, large voltages develop.

On an amp that I built all the references terminated on a single lug bolted to the chassis.

On an instrument I built which had 6 power supplies and measured small currents, the references were critical.

 

[linguist]

Thanks for the replies,

AHH I see, I think?

I drew the A & D grounds back to front.

So, I have the A Ground Plane & D Ground Plane seperate & I connect these together at a single common point.
There is a 100uF PS Cap, is this a good place to join both ground planes?

When I make up the circuit board, what distance should I have between the two ground planes to avoid any issues?

I will be using an Aluminium enclosure.
Battery & Wall type power supplies, provisions for both.

I have to admit I have only been using NON isolated jacks for everything, could you explain why & where I should be using Isolated jacks.

Cheers

 

[KeepItSimpleStupid]

Take a look at this article: Ground loop problems and how to get rid of them

Join the grounds at power supply common. Two wires to a lug on the case is somewhat ideal.

With a wall power supply, non-isolated jacks should be fine.

When you have to deal with protective ground and signal ground with a single ended signal there can be problems as the article states.

 

[linguist]

Thanks for your time,

I will, have a read of the link you supplied.

I just connected the circuit as suggested but it still doesn't seem to work when I input a sine signal, all I seem to get is a perfect sine output?

Pin 2 voltage is around 2.5V & pins 12 to 16 are also 2.5V.

Like I mentioned I thought I would see another sinewave appear some mS later than the first for the echo but I am not sure about that as I haven't viewed this before.

I dismantled the circuit & rebuilt it with the exact same results & tried another IC?

Not sure what to try now?

Cheers

 

[ronv]

I could be wrong, but I think it just adds delay to the input so you need to feed the output into your amp as well as the original input.

 

[KeepItSimpleStupid]

Exactly what I was going to say. Use a dual trace scope and look at the relative delay between input and output.

In order to get an echo, you need the original signal and the output (delayed signal).

 

[linguist]

Ah ha that makes sense, i'll try it out.

Thanks for the help!

Cheers

 

[linguist]

Just an update, I tried mixing the input signal to the output signal as suggested with not much luck.

I have been testing this circuit with a stable 1kHz sinewave & noticed that when I switched to 500Hz another signal appeared momentarily.

I then used the manual sweep mode & swept the sinewave between 250Hz & 500Hz to start with & sure enough there is an echo signal. I disconnected where I mixed the signal & just had a look at the output itself without the input being mixed.

It has a very noticable echo when sweeping the frequency up or down & you can adjust the delay time & how much it repeats. I suppose it makes perfect sense when you think that Audio signals etc are harmonic signals with multiple frequencies & not stable sinewaves.

It's a very clean echo signal, above 3kHz there is not much in the way of an echo but that's a good thing.

Nothing worse than that reverb sound with higher frequencies.

It looks very good.

Thanks KISS & ronv, your suggestions made me look closer & ultimately I have it working nicely now.

Cheers

 

[linguist]

I have just tested the circuit with a guitar, it appears to work ok but there is noise associated with the echo & it can sound a little fuzzy depending on what adjustments are made.

Obviously i'm no electronics wiz, so could someone be kind enough to point me in the right direction as to filtering or fixing the noise on the echo, the main sound is clean, it's just the echo can get a little fuzzy.
It's possible to get it almost clean with some settings with just a hint of noise with it but other settings get fuzzy.

KISS mentioned LR filtering but I am unsure where to start or if this is what I require?

The shorter the delay time the cleaner the echo is.

Can capacitor type cause this issue?

The 560pF caps I have used ceramic dics caps & for the 10uF I used Tantalum Caps, I'm not certain I should have used these, all the rest are poly caps & electrolytic.

Cheers