Author |
Message |
rogerlatur
Joined: Dec 22, 2012 Posts: 118 Location: france
|
Posted: Mon Jun 24, 2013 9:59 am Post subject:
different circuits on the same board. How to... |
|
|
If you look at this illustration, you see 2 circuits A & B that I want to put on one small unique board.
CURRENT:
Can I do like in 3 or will I have a problem because both circuits will interfere when I pick the current from the same main nodes (blue is neg, red is pos, green is gnd).
On 1 & 2, I first thought I should pick the current for each circtuit separately, but as I am going to build more bigger boards with more separate circuits on a board, I would prefer to start with the good habits, this is the reason why I ask here what you think.
I am even not sure if I should not have duplicate in 1 the decoupling principle from circuit A on top to circtuit B after the second current inputs [after GND (green) / MINUS (blue) / PLUS (red) in B]. I am not sure if one can use one input for multiple circuits on the same board.
I am also realizing while writing this that I should not change the order (gnd/minux/plus) on the same board as it could be a source of error! But this is another problem.
GROUND:
When you compare 1, 2, and 3, you see how I started in 3 and ended up, not being sure if having R2, trimmer and out (a sleeve of a jack) on the same node is ok or not (like in 3).
It is actually the same principle like for the current, that I am not 100% sure about, wanting to know if I can do (I mean if it makes sense) as shown in 3 or not.
Thank you ! |
|
Back to top
|
|
|
Uncle Krunkus
Moderator
Joined: Jul 11, 2005 Posts: 4761 Location: Sydney, Australia
Audio files: 52
G2 patch files: 1
|
Posted: Mon Jun 24, 2013 1:22 pm Post subject:
|
|
|
I would just copy the exact same layout three times across a board. Any connections which are common between them (ie, +V,0V,-V) just run right through on unbroken tracks. All other tracks are cut between units. I don't think you'll need separate decoupling caps for each section. (top right) Although you may want to triple the value just to make sure the one set are fine.
With stripboard, I've always found that accuracy is paramount to your joy. Copying the exact same layout three times, reduces confusion, and increases accuracy. _________________ What makes a space ours, is what we put there, and what we do there. |
|
Back to top
|
|
|
rogerlatur
Joined: Dec 22, 2012 Posts: 118 Location: france
|
Posted: Tue Jun 25, 2013 10:21 am Post subject:
|
|
|
Uncle Krunkus wrote: | I would just copy the exact same layout three times across a board. Any connections which are common between them (ie, +V,0V,-V) just run right through on unbroken tracks. All other tracks are cut between units. I don't think you'll need separate decoupling caps for each section. (top right) Although you may want to triple the value just to make sure the one set are fine.
With stripboard, I've always found that accuracy is paramount to your joy. Copying the exact same layout three times, reduces confusion, and increases accuracy. |
Thank you for helping me !
I am sorry I do not understand your post (english is not my first language). Mostly the part with "triple the value" or "copy the exact layout 3 times".
I guess I did not explain what I was looking for the right way.
For this example, I want to build 1 board with 2 circuits. Circuit A, circuit B.
Between all 10n caps and both opAmps, the conducting lines are cut in the middle, so I can connect both sides without interfering, but in my post here I concentrate on the conductive lines and the nodes where to pick the current, more than the circuit (aim) itself.
I am showing here 3 illustrations, because I am not sure I am doing right in number 3, because I consider it as more efficient, and this is the board I would like to buiƶd if what I am doing with the connections is fine.
The question being:
Can I use 1 decoupling part for multiple circtuits and connect different opAmps and components from different circtuis to the same conductive line (+/0/-) and as I understand you say it is ok to do so.
I thought maybe connecting both opAmps (one TL071 and one TL072) might make them interfere when I pick the current from the same line. Or even connecting the sleeve of a jack to the ground line where the ground also goes to a trimmer, to a cap and to a res. Once I understand that it is not a problem and that connection "order/chronology" does not matter, I guess I will be ready to go further in DIY.
I ask here showing 3 different ways I could build the board, not being sure if I really can do like in 3, and at the same time I have the feeling it is stupid from me to think that it is safer to pick the current 2 times (splitting +, 0, - coming from the distro board into 2, like shown in version 1 and 2). As I do not want to kill any component/module doing something wrong (mostly playing with current and voltages), I thought I would ask here. |
|
Back to top
|
|
|
Uncle Krunkus
Moderator
Joined: Jul 11, 2005 Posts: 4761 Location: Sydney, Australia
Audio files: 52
G2 patch files: 1
|
Posted: Tue Jun 25, 2013 11:21 pm Post subject:
|
|
|
Sorry rogerlatur,
I didn't read your post properly.
The caps and beads are all part of the power conditioning, so make sure you take any extra power feeds from the point past the beads.
You've got two wasted tracks up there, so I'd shorten your bead leads, and use the extra 4 tracks for distributing power for this whole board.
You can even cut the tracks which feed the beads and mount them horizontally, thus saving another two tracks.
Yes,.. all points connected only by tracks or links are electrically equivalent, (most of the time! lol!) so it doesn't really matter what order things come in contact with that point.
Any other questions,.... just ask. _________________ What makes a space ours, is what we put there, and what we do there. |
|
Back to top
|
|
|
PHOBoS
Joined: Jan 14, 2010 Posts: 5591 Location: Moon Base
Audio files: 705
|
Posted: Wed Jun 26, 2013 5:09 am Post subject:
|
|
|
ok here's my 2 cents or bitcoins or whatever currency you prefer.
First of all you should place the beads directly at the point where the power enters, so before the capacitors. And if you really want to
make it perfect then use beads and capacitors for each circuit.
But with these simple circuits on such a small board I'll doubt you will have any problems and 1/2/3 will all work. You don't have to use
those long leads as you do in 3 (allthough it's a good method). With the capacitors on each chip it'll work fine if you get the power from
the closest point available. So the negative supply (minus) for circuit B could be connected to the trace next to pin 4 of the chip of circuit A.
What you want to avoid is chaining power TOO much through circuits,. So if you have a bunch of them it's better to all have them powered
from one point instead of chaining them from circuit A to B to C etc. But again in depends on the circuits. It's especially useful to do it
this way if you have audio circuits combined with circuits that create/use short pulses (LFO's, Sequencers etc.)
If I have a big circuit board I usuall make some powerails on one side and then tap those for each circuit with it's own decoupling capacitors.
As for your Ground (GND) it's similar to the powerlines (it's actually part of it) and best way is to wire most of it to one point, which is the
GND on your powersupply. You don't have to do this for each GND connection of course, but it's a good practice to do this if you have
speakers or LED's or anything else that uses some amount of current. Also GND connection for in/outputs should be wired to this point
if possible. _________________ "My perf, it's full of holes!"
http://phobos.000space.com/
SoundCloud BandCamp MixCloud Stickney Synthyards Captain Collider Twitch YouTube |
|
Back to top
|
|
|
rogerlatur
Joined: Dec 22, 2012 Posts: 118 Location: france
|
Posted: Wed Jun 26, 2013 11:52 am Post subject:
|
|
|
THANK YOU for your precious help here !!! |
|
Back to top
|
|
|
rogerlatur
Joined: Dec 22, 2012 Posts: 118 Location: france
|
Posted: Wed Jun 26, 2013 12:41 pm Post subject:
|
|
|
PHOBoS wrote: | With the capacitors on each chip it'll work fine if you get the power from the closest point available. So the negative supply (minus) for circuit B could be connected to the trace next to pin 4 of the chip of circuit A |
Okay. I understand. I made the confusion between getting the power from the closest point available and getting the power, like I did, the closest to the opAmp power terminals !
I realize now I even made a mistake with the neg connection on circuit B.
But I will do anyway like you adviced. |
|
Back to top
|
|
|
rogerlatur
Joined: Dec 22, 2012 Posts: 118 Location: france
|
Posted: Fri Jun 28, 2013 4:59 pm Post subject:
|
|
|
I made a huge mistake: I did not test the circuit after I made the modifications with decoupling and the 2 circuits on the same pcb ! I thought as it looks very simple, nothing could go wrong. Really stupid !
Now, nothing works like it should, like if both circuits were interacting.
On the output the voltage is not only totally wrong, it is even getting higher when I move my hand toward the circuit !
Before I start debugging and unsoldering (starting with deconnecting Circuit B to see what happen), I would like to have your opinion:
No matter how the circuit should react, do you see something wrong with the beads/caps ?
Here a picture of how I built it:
|
|
Back to top
|
|
|
rogerlatur
Joined: Dec 22, 2012 Posts: 118 Location: france
|
Posted: Mon Jul 01, 2013 12:26 pm Post subject:
|
|
|
I will desolder and debug. Learning by doing. And I will ask if stucked, |
|
Back to top
|
|
|
PHOBoS
Joined: Jan 14, 2010 Posts: 5591 Location: Moon Base
Audio files: 705
|
Posted: Tue Jul 02, 2013 3:43 am Post subject:
|
|
|
you should place the capacitors (and circuits) after the beads,.. so the power enters through the beads, to the rest of the circuit.
as for it changing when you move your hand closer, I noticed you don't have anything connected between pin 6/7 of the TL072. I expect you have a
pot there to control the gain, which might be picking up noise. You could try shielded wire, but it might work better if you set the gain high with a fixed
resistor and place an attenuator at the input. _________________ "My perf, it's full of holes!"
http://phobos.000space.com/
SoundCloud BandCamp MixCloud Stickney Synthyards Captain Collider Twitch YouTube |
|
Back to top
|
|
|
rogerlatur
Joined: Dec 22, 2012 Posts: 118 Location: france
|
Posted: Tue Jul 02, 2013 4:14 am Post subject:
|
|
|
PHOBoS wrote: | you should place the capacitors (and circuits) after the beads,.. so the power enters through the beads, to the rest of the circuit |
I thought I did so in the last picture I posted. compared to my first pictures. Then there is something I don't understand with "after".
PHOBoS wrote: | as for it changing when you move your hand closer, I noticed you don't have anything connected between pin 6/7 of the TL072. I expect you have a pot there to control the gain, which might be picking up noise. You could try shielded wire, but it might work better if you set the gain high with a fixed resistor and place an attenuator at the input. |
The 3 points represent a trimmer I am sorry I should have mention it (I do not have any icon for this in the freeware).
Tested before, on the breadboard, without decoupling (caps/beads) and without circuit B together, all was working fine.
Same with Circuit B. |
|
Back to top
|
|
|
rogerlatur
Joined: Dec 22, 2012 Posts: 118 Location: france
|
Posted: Tue Jul 02, 2013 4:15 am Post subject:
|
|
|
Oh ! I understand while reading again and comparing posts !
The caps should begin after the bead legs... I did so. I hope I understood well what you meant.
It is going to be difficult to unsolder and change on the kind of pcb I use, but it is the price to pay to make it work.
I am though really not sure this is the main problem. I will see. |
|
Back to top
|
|
|
PHOBoS
Joined: Jan 14, 2010 Posts: 5591 Location: Moon Base
Audio files: 705
|
|
Back to top
|
|
|
rogerlatur
Joined: Dec 22, 2012 Posts: 118 Location: france
|
Posted: Tue Jul 02, 2013 4:58 am Post subject:
|
|
|
PHOBoS wrote: | yes you got it connected right now
and if it's a trimmer soldered directly on to the PCB it should work. |
I will make the changes.
Thank you or your help + patience !!! |
|
Back to top
|
|
|
Nardu
Joined: Feb 28, 2011 Posts: 62 Location: Braila, Romania
|
|
Back to top
|
|
|
rogerlatur
Joined: Dec 22, 2012 Posts: 118 Location: france
|
Posted: Fri Jul 05, 2013 1:22 pm Post subject:
|
|
|
Nardu wrote: | Hello ! @rogerlatur
- I hope that the image below will help you to optimize your circuit. |
THANK YOU VERY MUCH !!!
I will then make the changes.
I see now what Uncle Krunkus did mentionned about the bead which might be positionned horizontaly. I understand now, it is possible to do so because there is no polarity matter here, it is good to know for the other circuits I am planning ! |
|
Back to top
|
|
|
rogerlatur
Joined: Dec 22, 2012 Posts: 118 Location: france
|
Posted: Sat Jul 06, 2013 6:10 am Post subject:
|
|
|
@Nardu:
Again, it is a very interesting approach (also in the opAmps area) and it also helps me and adds clarity to the way connections might be done: THANK YOU !!!!!! |
|
Back to top
|
|
|
rogerlatur
Joined: Dec 22, 2012 Posts: 118 Location: france
|
Posted: Sat Jul 06, 2013 12:18 pm Post subject:
|
|
|
I rewired a bit because of some mistakes (like the voltage divider in B, where I did not even pick the signal between the 2 resistors while building on the board !) and, even if I could not do exactly like Nardu because all was soldered on a quite small board, I also made all the important changes (bead/caps section, ground etc): all works now. Thank to all your kind support here !
There is however one issue which I would like to solve.
On the output of circuit A, which function is to amplify the incoming signal by 20%, there is a voltage fluctuating when no incoming signal is plugged in.
it goes from 0 down to -x.xx volts more in +/-0.01 volt steps (switching between 2 values, up/down) very slowly. Sometimes fluctuating/switching between 2 values (up/down 0.01v), but always climbing and climbing (I mean decreasing at it is negative). In 8 minutes it went from 0v to around -3.8v. And after it remained around -3.92v. Turning between -3.88v and 3.92v.
If I turn the circtui off and on after a while, the voltage goes directly to -2.8v and much more rapidly to -3.x even -4.x volts In a few secs only.
The voltage goes crazy, changing much more rapidly and in 0.1v or 1v steps (not 0.01v) when touching the board.
Maybe you have an idea what it could be ? I thought it was related to circuit B but I think it is not (I removed opAmp on B without any changes).
PHOBoS wrote: | as for it changing when you move your hand closer, I noticed you don't have anything connected between pin 6/7 of the TL072. I expect you have a
pot there to control the gain, which might be picking up noise. You could try shielded wire, but it might work better if you set the gain high with a fixed resistor and place an attenuator at the input. |
Could the trimmer be responsable for this ?
If yes, I am sorry I do not understand what I should do with the "fixed" resistor you mention. |
|
Back to top
|
|
|
gdavis
Joined: Feb 27, 2013 Posts: 359 Location: San Diego
Audio files: 1
|
Posted: Sat Jul 06, 2013 6:08 pm Post subject:
|
|
|
I've never used strip boards so I'm not totally sure I'm reading the pictures correctly, but is input A floating when nothing is connected? That would cause erratic behavior. The input needs a ground reference, when it's floating there is leakage current inside the chip that will charge the parasitic capacitance of the input. The high gain of the opamp makes it very sensitive to small voltages on the inputs, even that caused by waving your hand over it if it's floating.
Something like a 1M resistor between the input and ground would provide a ground reference with nothing connected while still maintaining a high input impedance when something is connected. That gives the leakage current a place to drain out instead of charging the input and should stop the output from drifting.
You should also ground the unused opamp inputs as again, the floating inputs will drift possibly degrading performance of the rest of the chip. You can leave unused output floating. _________________ My synth build blog: http://gndsynth.blogspot.com/ |
|
Back to top
|
|
|
rogerlatur
Joined: Dec 22, 2012 Posts: 118 Location: france
|
Posted: Sun Jul 07, 2013 4:24 am Post subject:
|
|
|
gdavis wrote: | I've never used strip boards so I'm not totally sure I'm reading the pictures correctly, but is input A floating when nothing is connected? That would cause erratic behavior. The input needs a ground reference, when it's floating there is leakage current inside the chip that will charge the parasitic capacitance of the input. The high gain of the opamp makes it very sensitive to small voltages on the inputs, even that caused by waving your hand over it if it's floating.
Something like a 1M resistor between the input and ground would provide a ground reference with nothing connected while still maintaining a high input impedance when something is connected. That gives the leakage current a place to drain out instead of charging the input and should stop the output from drifting.
You should also ground the unused opamp inputs as again, the floating inputs will drift possibly degrading performance of the rest of the chip. You can leave unused output floating. |
Input A is floating (thank you for pointing that out. Now I know what "floating" means). I will try that.
Each step learning something new. What a journey !
Thank you so much !!! |
|
Back to top
|
|
|
rogerlatur
Joined: Dec 22, 2012 Posts: 118 Location: france
|
Posted: Sun Jul 07, 2013 10:51 am Post subject:
|
|
|
The module works now as it should.
Many thanks to:
gdavis, Nardu, PHOBoS, Uncle Krunkus and, last but not least, elmegil. |
|
Back to top
|
|
|
PHOBoS
Joined: Jan 14, 2010 Posts: 5591 Location: Moon Base
Audio files: 705
|
|
Back to top
|
|
|
|