How do I "isolate" voltage drops in my circuit?

Thread Starter

bignobody

Joined Jan 21, 2020
97
I'm still relatively new to this, so this might be a stupid question or I might have some terms wrong.

I've been experimenting with Voltage Controlled Oscillators, mostly CD40106 based.

Let's say I have a circuit with 2 Voltage Controlled Oscillators. One is used as a clock oscillator to drive another IC, and the second is an audio oscillator which will go to an amplifier.

To have an indicator LED for each oscillator, I buffer each oscillator's output and connect each buffered output to an LED and then to ground.

Doing this I have noticed the voltage drop across my Clock oscillator's LED is affects the voltage of the Audio oscillator, which is altering the sound produced in an undesirable way. If I remove the Clock oscillator's LED the interference stops, because the voltage drop is gone.

Is there a way to "isolate" the clock oscillator LED's voltage drop in my circuit so it doesn't affect the audio VCO?
 

Thread Starter

bignobody

Joined Jan 21, 2020
97
@joeyd999 Thanks for the link. Grounding could be the issue. I will do some more research into best practices.


No schematic and no power supply voltage.
This is true, I was trying to keep my question general.
You said you have a buffer to "isolate" the LED so how can the oscillator be affected?
I said no such thing. I said I buffer the oscillator outputs and connect those outputs to an LED.


What are you doing to "buffer" these outputs?
Using two other input pins on the CD40106.
Do you have the circuits properly de-coupled?
Maybe, maybe not. Another area for me to research. Thanks!
 

Delta Prime

Joined Nov 15, 2019
1,311
I'm still relatively new to this, so this might be a stupid question or I might have some terms wrong.
I have read your past inquiries of your projects do not self deprecate yourself you are wise in seeking knowledge but you must understand from my experience you're asking questions too.. ( in no way I include myself. )To experts who have been there and done that, and can be quite short and blunt. Which quite frankly is an asset to this site. I'm 20 years old I take this as an opportunity as I hope you do as well to have fun with electronics you're not asking questions to a teacher they are experts in every sense of the word.keep on asking questions and try to wear them out just like I'm trying to do. By the way you have gotten a response from mister audio guru again. You have acquired the brass ring that I have been seeking. you know what they say ( If you have not gone duck you're out of luck.)
 
Last edited:

crutschow

Joined Mar 14, 2008
34,280
I was trying to keep my question general.
No, you are asking specific questions about a specific circuit.
How is that "general"?
So to answer your specific questions we need your specific circuit.
Otherwise the general answer is: "It could be anything".
 

ebeowulf17

Joined Aug 12, 2014
3,307
I'm still relatively new to this, so this might be a stupid question or I might have some terms wrong.

I've been experimenting with Voltage Controlled Oscillators, mostly CD40106 based.

Let's say I have a circuit with 2 Voltage Controlled Oscillators. One is used as a clock oscillator to drive another IC, and the second is an audio oscillator which will go to an amplifier.

To have an indicator LED for each oscillator, I buffer each oscillator's output and connect each buffered output to an LED and then to ground.

Doing this I have noticed the voltage drop across my Clock oscillator's LED is affects the voltage of the Audio oscillator, which is altering the sound produced in an undesirable way. If I remove the Clock oscillator's LED the interference stops, because the voltage drop is gone.

Is there a way to "isolate" the clock oscillator LED's voltage drop in my circuit so it doesn't affect the audio VCO?
We need to see your complete schematic in order to give good answers, but here are two questions in the meantime:
How much current are the LEDs drawing from your buffer circuits?
Do you have power supply decoupling caps at the power input to every chip you're using?
 

Alec_t

Joined Sep 17, 2013
14,280
The CD40106 output voltage will drop significantly if you load it with an LED directly, with no current-limiting resistor. You say you are buffering the oscillator, but don't say how. If the buffer is a gate in the same IC as the gate used for the oscillator then there might be some undesirable interaction if the LED current is more than a few mA.
 

Thread Starter

bignobody

Joined Jan 21, 2020
97
If the buffer is a gate in the same IC as the gate used for the oscillator then there might be some undesirable interaction if the LED current is more than a few mA.
Thanks, Alec_t! I am indeed buffering the LEDs on the same IC, so this is something I'll look at.

How much current are the LEDs drawing from your buffer circuits?
Do you have power supply decoupling caps at the power input to every chip you're using?
I've got a few different colours that I'm playing with, but generally between 10 to 20 mA.

I have been using a 10 uF cap across the power and ground just before my voltage regulator and a 1 uF cap across the power pins of each IC, though I am reading that .1 uF is the one usually recommended.



No, you are asking specific questions about a specific circuit.
How is that "general"?
I am working on a specific circuit, this is true. But I am trying to gain a more general understanding of this type of problem that I can learn to apply for myself, which is why I tried (not entirely successfully) to keep my question general.

Otherwise the general answer is: "It could be anything".
Too general. I've also received answers of "grounding", "de-coupling caps", and "buffer overloading". These were the type of general responses I was looking for, but I appreciate and understand your response.

Then the buffer outputs driving the LEDs might be a little overloaded and the oscillator outputs are perfect.
This could be the case. I will try some more experiments. Thanks!


@Delta prime: Thanks for your response! I don't mind short, blunt answers at all. Saves time for everyone involved ;)


For people who regularly make schematics to post here, what software do you recommend? Trying to hack one together from bits and pieces in image manipulation software feels really tedious and inaccurate...
 

MisterBill2

Joined Jan 23, 2018
18,167
I'm still relatively new to this, so this might be a stupid question or I might have some terms wrong.

I've been experimenting with Voltage Controlled Oscillators, mostly CD40106 based.

Let's say I have a circuit with 2 Voltage Controlled Oscillators. One is used as a clock oscillator to drive another IC, and the second is an audio oscillator which will go to an amplifier.

To have an indicator LED for each oscillator, I buffer each oscillator's output and connect each buffered output to an LED and then to ground.

Doing this I have noticed the voltage drop across my Clock oscillator's LED is affects the voltage of the Audio oscillator, which is altering the sound produced in an undesirable way. If I remove the Clock oscillator's LED the interference stops, because the voltage drop is gone.

Is there a way to "isolate" the clock oscillator LED's voltage drop in my circuit so it doesn't affect the audio VCO?
Why connect an indicator to an audio output? That is my first question. BUT if you must, then add another op-amp connected as a unity gain non-inverting stage between the output and the LED circuit. That will provide isolation.
 

ebeowulf17

Joined Aug 12, 2014
3,307
Thanks, Alec_t! I am indeed buffering the LEDs on the same IC, so this is something I'll look at.


I've got a few different colours that I'm playing with, but generally between 10 to 20 mA.

I have been using a 10 uF cap across the power and ground just before my voltage regulator and a 1 uF cap across the power pins of each IC, though I am reading that .1 uF is the one usually recommended.




I am working on a specific circuit, this is true. But I am trying to gain a more general understanding of this type of problem that I can learn to apply for myself, which is why I tried (not entirely successfully) to keep my question general.


Too general. I've also received answers of "grounding", "de-coupling caps", and "buffer overloading". These were the type of general responses I was looking for, but I appreciate and understand your response.



This could be the case. I will try some more experiments. Thanks!


@Delta prime: Thanks for your response! I don't mind short, blunt answers at all. Saves time for everyone involved ;)


For people who regularly make schematics to post here, what software do you recommend? Trying to hack one together from bits and pieces in image manipulation software feels really tedious and inaccurate...
I'm with @Alec_t on this one: LTspice is a free and powerful circuit simulator. If you have any interest at all in simulating circuits, this is the way to go. It's not the most user friendly or intuitive for beginners, but it's well worth the effort to learn!

That said, if you don't want the simulation abilities, I recommend DipTrace. Although there are professional level versions of it, the free-for-hobby use version of it is quite powerful, more than enough for anything I've ever needed. It makes much nicer looking schematics, and also provides PCB layout capabilities (which LTspice doesn't.)

Between those two tools, I'm totally covered. There are several alternatives to DipTrace on the schematic and PCB layout side. Eagle is more popular in the hobby world, so there's more user support online, but I hate it's interface more than I can say. It's the worst in my opinion (just personal preference though - many here love it!) KiCAD is the other one that comes up a lot. It's open source and apparently quite powerful. I didn't love it, but don't really have anything bad to say about it either.
 

Thread Starter

bignobody

Joined Jan 21, 2020
97
Why connect an indicator to an audio output? That is my first question. BUT if you must, then add another op-amp connected as a unity gain non-inverting stage between the output and the LED circuit. That will provide isolation.
Honest answer is I like flashing lights, haha. But more practically I like the visual feedback they provide when adjusting the potentiometers that control the oscillators. I'm not using any op-amps (which you wouldn't know, because I haven't provided a schematic) but I appreciate the advice.
 

Thread Starter

bignobody

Joined Jan 21, 2020
97
@ebeowulf17 Thanks for your suggestions! I have at least installed LTSpice and had a poke around a little. I think Eagle is out for me now that they've moved to a subscription model, but I will have a look at DipTrace and KiCad. Cheers!
 

ebeowulf17

Joined Aug 12, 2014
3,307
@ebeowulf17 Thanks for your suggestions! I have at least installed LTSpice and had a poke around a little. I think Eagle is out for me now that they've moved to a subscription model, but I will have a look at DipTrace and KiCad. Cheers!
One thing you may notice early on with LTspice is the seemingly odd component list. An awful lot of common, almost standard, components are missing. That's because it's a free utility provided by a commercial company, and so it's promoting their product line.

Honestly, you can get a huge amount of work done with the built in library of parts - you just have to find the Analog Devices (and formerly Linear Technology) part numbers for equivalent parts.

That said, if you need very specific parts, one of the members here, @Bordodynov, has generated a stunning library of parts that can be added to LTspice to make it more complete. I don't remember where the links to it are, but I believe there's one on this site somewhere, and there's also a popular user support group on Yahoo which includes these libraries.
 

Audioguru again

Joined Oct 21, 2019
6,671
Since you did not post your schematic I guessed about what you are doing. I made a simple Schmitt Trigger oscillator, not your unknown voltage controlled one. Then I used a buffer to power the LED.

The datasheet of a CD40106 shows a maximum allowed power (heating) of 100mW for each output and shows graphs of typical output currents at 3 different supply voltages.

If your unknown supply voltage is 5V then the inverter that is driving the 2V red LED in my schematic has 3V across it and the "sourcing" graph shows a typical current in the LED of only 3.5mA. Then the power (heating) in the inverter is 3V x 3.5mA= 10.5mW so it is safe.

If your unknown supply voltage is 10V then the output current is shown to be about 18mA then the inverter heats with 8V x 18mA= 144mW and will overheat so a current-limiting resistor is needed in series with the LED. We can have the resistor heat with 70mW then the inverter heat with the remaining 74mW and will be safe. Use a 220 ohm resistor.

Maybe you problem of the LED affecting the oscillator is caused by you overheating the inverter or the entire IC.
 

Attachments

Thread Starter

bignobody

Joined Jan 21, 2020
97
@Audioguru again Very informative post, thank you. I appreciate the extra trouble you have gone to for providing it based on the limited information I gave. I admit, I hadn't even considered the heat factor here. I appreciate the breakdown, too, as I am still getting accustomed to understanding how to interpret all the information in data sheets.
 
Top