Advanced Beginner interested in learning more especially abt analog synthesizers

[slick8086]

Hi,

So I say I'm an "advanced" beginner because I've been doing really basic electronics for a long time.  I know what most components are like resistors, capacitors, diodes, transistors, LEDs etc.  I've made a bunch of electronic projects before.  I can solder well.  I have and can use a multimeter.  I've even designed and had made a simple PCB. https://oshpark.com/shared_projects/Riu5kmba.  I can "read" a schematic, but my understanding of it is very limited.  Like if it was a sentence I could sound out all the words phonetically, but I wouldn't understand what some of the words meant and I'd only have a vague understanding of what the whole sentence meant without some help.  I can assembly pretty complicated kits, and get them working properly, but it is more like painting by numbers. 

Things I know that I don't know.  I know there is a thing/concept called and RC circuit, but I don't know how to employ it.  I know there are things called op amps, but I don't know how to use them.  There is more that I'm not thinking of right now for sure.   

Obviously I don't know, what I don't know.  Where do I go from here?  I know that I need to deepen my understanding of the basic things I already know but I'm easily not finding the resources online to learn? 

My goal is to someday be able to design and implement an analog synthesizer on my own.  I've been looking at schematics for VCOs, VCAs, ADSRs, and so on, but unfortunately they all go right over my head.  Like I could probably breadboard some of them and get them to work, but I wouldn't know why they were working, and if they schematic had an error, I wouldn't know.   I couldn't adapt a schematic that works on +15v -15v to work on +12v -12V instead.

What is my next step?  Are the resources available online or do I have to sign up for EE classes at a real school?

Advice is much appreciated.

[rstofer]

Things I know that I don't know.  I know there is a thing/concept called and RC circuit, but I don't know how to employ it.  I know there are things called op amps, but I don't know how to use them.  There is more that I'm not thinking of right now for sure.   

A series RC circuit can be either a low pass filter or high pass filter depending on configuration (whether the signal goes in the resistor or the capacitor).

https://www.electronics-tutorials.ws/filter/filter_2.html
https://www.electronics-tutorials.ws/filter/filter_3.html

When we talk about a 'low pass' filter we mean that at high frequencies, the signal is attenuated.  We pass low frequencies, we attenuate high frequencies.

Op amps are a separate course!
Google for 'w2aew op amp' to find 3 excellent videos
Google for 'eevblog op amp' to find 3 more excellent videos

Op amps are important building blocks for all kinds of analog circuits, particularly audio circuits (mixing boards, synthesizers, etc).  They can be used to create all kinds of active filters (low pass, band pass, high pass) as well as adding signals.  I don't do audio so I'm not real clear on how synthesizers work.  Google might help...

https://noisehackerspace.com/amazing-op-amp-powerful-synth-diy/
 
https://www.electronics-tutorials.ws/filter/filter_5.html
https://www.electronics-tutorials.ws/filter/filter_6.html

You really need to know op amps.  There is a great book "Op Amps For Everyone" that might be helpful:
http://web.mit.edu/6.101/www/reference/op_amps_everyone.pdf
But only after understanding the w2aew and eevblog videos above.

In fact, you are going to need to understand Ohm's Law, Kirchhoff's Laws (particularly the current law) and a wee bit of matrix algebra (just a tiny bit, no big deal).  Or some other way to solve simultaneous equations.  Low level stuff...

Oh, and you can use op amp integrators to build up analog computers.  An integrator is a low pass filter (of sorts).

[3db]

I agree with the great advice given by RSTOFER.
You should also check out the late Ray Wilson's Youtube channel  https://www.youtube.com/channel/UCp6OK4J97aPWccFpUgKj_Mg
His website is well worth a visit and you can buy his book which will help you understand more about synths.

3DB

[rstofer]

Obviously I don't know, what I don't know.  Where do I go from here?  I know that I need to deepen my understanding of the basic things I already know but I'm easily not finding the resources online to learn? 

Khan Academy Electrical Engineering program
https://www.khanacademy.org/science/electrical-engineering

Digilent Real Analog class
https://learn.digilentinc.com/classroom/realanalog/

There are literally thousands of tutorial sites on Google.  You need to refine your search and look for specific topics.

You will need some math.  Maybe you already have plenty but, if not, Khan Academy is a good place to start.

My goal is to someday be able to design and implement an analog synthesizer on my own.  I've been looking at schematics for VCOs, VCAs, ADSRs, and so on, but unfortunately they all go right over my head.  Like I could probably breadboard some of them and get them to work, but I wouldn't know why they were working, and if they schematic had an error, I wouldn't know.   I couldn't adapt a schematic that works on +15v -15v to work on +12v -12V instead.

Due to op amp design, it might very well be impossible to make it work.  The signal may have a high enough voltage swing to drive the +-12V op amp to saturation whereas the +-15V design had enough headroom.  This is a datasheet issue!

You will also want to convert dual rail designs (like +-15V) to single rail (+5V?) designs.  The book I recommended above can help with that.  It should be obvious that signal swings will be more limited.

Rail-to-rail op amps try to minimize the issue of not being able to swing the input or output signals close to the rails.  But read the datasheet!  Some can go to the + rail but not the - rail, some are rail-to-rail on input but not output, and so on.

What is my next step?  Are the resources available online or do I have to sign up for EE classes at a real school?

Advice is much appreciated.

Buy some op amps and play with them on a breadboard.  I would suggest starting with +-15V amps (see videos for op amp selection) and moving to single supply op amps later on.

The LM358 is as good as any.  Note that it can only get to about 1.5V from the + rail but apparently it can get all the way to the - rail.

This clipping phenomenon is described in w2aew's videos.  He shows it on a scope.

Got a scope?  Got a dual output supply?  Signal generator?  If you really want to do this thing and you don't have enough equipment, please consider the Digilent Analog Discovery 2 and the associated parts kit.  It will allow you to track the Real Analog course.  The AD2 has a dual output supply (low voltage, +-5V), dual channel oscilloscope, dual channel signal generator, 16 bit digital IO and logic analyzer.  None of which begins to show you the tools in the Waveforms software.  You want to see how an RC low pass filter REALLY works?  Fine, use the Network tool.  It will give you attenuation and phase shift versus frequency in a beautiful Bode' Plot.

If I were just starting out and wanted to LEARN electronics, the AD2 would be my first tool.  It is truly a lab in a backpack (with a laptop) and totally capable of replacing a bunch of test gear.  Read the instructions carefully, there are voltage limitations.  Don't even think about using it on some kind of high voltage switching power supply.

https://store.digilentinc.com/analog-discovery-2-100msps-usb-oscilloscope-logic-analyzer-and-variable-power-supply/

I bought the Jameco 20626 Dual Output Supply the other day.  I haven't assembled it yet but I'll get around to it sooner or later.  I had planned to use it as +-15V for an analog computer I built some time back.

https://www.jameco.com and search for 20626  The link is too ugly to post.

[rstofer]

You didn't say what equipment you had and, of course, I jumped in with the Analog Discovery 2.  It is remarkable how many people think this is a toy!  In terms of accuracy, the scope on the AD2 uses a 14 bit ADC.  My Rigol DS1054Z, and many much more expensive scopes) uses an 8 bit ADC.  Yes, the frequency is limited but includes all of the range of interest for audio and most digital circuits.

The Pro bundle plus the analog parts kit is the way to go
https://store.digilentinc.com/analog-discovery-2-pro-bundle/

Add the analog parts kit for $55
https://store.digilentinc.com/analog-parts-kit-by-analog-devices-companion-parts-kit-for-the-analog-discovery/

You're looking at $354 which is a TON of money, until you compare it to an equivalent set of individual pieces of test equipment.  In any event, education has never been free!  Dave did a video on the Analog Discovery



It is for the original version, replaced by the AD2...

You can download the Waveforms software and play with the tools even without a device.  Just use the Demo device and explore all the features.

I know, I sound like an evangelist, but I really believe that the AD2 is ideal for learning electronics.  Everything you need is in one place.  You can, of course, omit the Pro Bundle and the analog parts, the AD2 is $279 by itself.  You can buy your own chips and add the BNC adapter board later on.  The advantage to the full package is that it matches the Real Analog course.

ETA:  The video shows a very old version of the software.  Download and play with the newest version:
https://reference.digilentinc.com/reference/software/waveforms/waveforms-3/start

[slick8086]

My electronics bench is pretty sparse.  I have a couple good soldering irons, a Fluke 114, 4-5 breadboards with accompanying jumble of wires, various hand tools (pliers, side cutters, crimpers, etc.), and a crap ass HQ-POWER 0-30v 0-3A bench power supply.  I need a DMM that can measure current though.  In a couple weeks (next paycheck) I was planning on buying the Siglent SDS1202X-E, but now I think you've convinced me to get the AD2 and parts kit.  Especially if it will help with their educational materials.  I had looked USB scopes, but I've seen them poopooed a few times here and other places and it gave me pause. 

I'm actually up to the Kirchoff's Laws section in the Khan Academy Circuit analysis course (about 15 lessons in), I started last week, but even though I'm learning it I wasn't sure that it would apply towards my end goals.   I should have just accepted that this fundamental stuff would be necessary for any and all future EE projects.  But it is nice to have some reassurance that I'm going in the right direction. 

I'm going to find and watch the op-amp video next. 

thanks again for your advice!

[rstofer]

There's nothing wrong with choosing the Siglent SDS1202X-E.  It's a 200 MHz dual channel scope and will serve very well.  OTOH, in the digital world, I rather prefer the SDS1204X-E to get 200 MHz with 4 channels.  I would actually buy the SDS1104X-E and unlock the 200 MHz capability - details elsewhere.  But I want 4 channels, regardless of how I get them.  On a scope...  It clearly isn't possible on the AD2.

At some point down the road, your equipment list will be much longer.  But, in this thread, I am talking about LEARNING electronics as opposed to just copy and paste of other people's projects.  If you watch Dave's video, you will see he spends a lot of time on the Bode' plot.  That's because we are always interested in gain and phase versus frequency.  Every tuned circuit, every amplifier, they all have some kind of gain (even if negative) and phase shift.

You just about can't do a Bode' plot with a scope.  You can bash something together but it isn't the same.  You can do a decent FFT with a scope and we are often interested in harmonics and their level.

It seems you are interested in the EE and LEARNING part of electronics.  That's good because if you plan to design stuff, you're going to need to know more than just copy/paste.  You're traveling down the right path, in my view.

I think you can get a better LEARNING experience out of the AD2 but for general scope work, the Siglent will always be a better choice.  Dave mentions this in his video - there aren't any knobs!

Before you make your final decision, why not look through some of the lab sessions for Real Analog?  See how they use the AD2.  I wish I could say I have done this but I haven't.  Make sure they are using the capabilities.  See what it might take to do the experiments if you didn't use the AD2.

It's not my money but I think I would start with the AD2 and work through the lab exercises.  See how far you get with the EE part of the game.  Later on, you might want to revisit the scope, power supply, signal generator choices.  They aren't really all that important.  I have ALWAYS (as in, since I built one from the ARRL Handbook when I was 13 back in '58) up until today, had a scope of one kind or another.  I added the power supply, signal generator and DS1054Z within the last couple of years.  Power supplies were wall warts or batteries and the signals were almost always digital.

[rstofer]

It would be wise to have one or two (or three) DMMs separate from the AD2.  If you promise to NEVER use them on mains voltages, I highly recommend the Aneng 8008.  It's a darn nice meter for $27 (at Amazon).  Nobody trusts the CAT rating so mains work should probably be reserved to a higher cost meter, probably Fluke <whatever>.  Dave did a review on this DMM.



You mentioned knowing how to use a DMM so I suppose you already have at least one.  It is not uncommon to measure 3 things in a single experiment.  The 8008 also has a square wave output!

Be very careful with voltage limits of the AD2.  If you use the BNC adapter and scope probes things are better.  Just select x10 and leave them that way.  That will also give you the best frequency response.  It does remove the differential capability of the analog inputs and provides earth ground to the probe ground lead.  This leads to an entire new set of issues.

See Dave's video on "How Not To Blow Up Your Scope".


Remember, you are limited to +-25V.  I prefer to use battery packs for breadboard experiments if I don't have access to a current limiting supply.  Down the road, I would think a real bench supply would be the next item.  The AD2 is weak in the power supply department but the levels are adjustable from 0V to 5V and 0V to -5V.  I would put the power supply ahead of the scope.  The AD2 makes a pretty decent scope.

Real design work, down the road, will be modestly heavy in math.  It is engineering, of course, and engineering of just about any variety comes down to a ton of math.  Electrical gets seriously heavy into math when we get to signals and signal processing.  Don't worry about it now!

Sure, if you have the opportunity to spend 5 years getting a degree in EE, go for it.  Then spend another year or so getting a Master's.

Search for Electrical Engineer at www.bls.gov and drill down by location to see what the job opportunities and pay scale look like.

Check out computer engineer or software engineer - they pay as well and the job market is much brighter.

[slick8086]

You mentioned knowing how to use a DMM so I suppose you already have at least one.

I have a Fluke 114, so I need to get another one that can measure current too anyway. 

Before you make your final decision, why not look through some of the lab sessions for Real Analog?

Yeah I'll do that, but I think I'm still leaning towards the AD2 because I do have a lot of learning to do and for the synth stuff I want to do I think it will suffice until I can afford a better 4 channel scope.

Down the road, I would think a real bench supply would be the next item.

So my crap ass HQ-POWER 0-30v 0-3A bench power supply is that bad huh?

Sure, if you have the opportunity to spend 5 years getting a degree in EE, go for it.

I wish I had the time and money for that.  If I were to take any classes they would just be for the info, at this point, degree program is not for me.

Check out computer engineer or software engineer - they pay as well and the job market is much brighter.

I was a Unix/Linux Sysadmin for 15 years, and I decided I needed a change.  Fewer offices and desks.  For the last couple years I've been running CNC carbide grinders making end mills. It barely pays the bills, but I feel healthier not sitting down all day long.  Electronics has interested me since I was a kid, not sure I have time to start it as a career now though, but it would be fun to have as a side gig If I could make some projects that others might appreciate.   

[rstofer]

So my crap ass HQ-POWER 0-30v 0-3A bench power supply is that bad huh?

It will do just fine!  Since you can get to 30V, you can power +-15V op amps.  You will need to create a virtual ground by connecting two resistors in series across the supply and using the center point as 'ground'.  This is done all the time with single supply op amps and, in fact, what you would be doing is using a dual rail op amp with a single supply.  Look at circuits around the Interweb.  A couple of 1k resistors is probably adequate or you can use much higher values (less power) if you use an op amp voltage follower.  All of your signals will be with reference to the virtual ground.

Google for 'op amp virtual ground' and you'll find stuff like this:http://www.swarthmore.edu/NatSci/echeeve1/Ref/SingleSupply/SingleSupply.html

If I were to take any classes they would just be for the info, at this point, degree program is not for me.

There are plenty of online programs.  You can also check out a community college.  The problem is, there are prerequisites for most technical courses. Maybe you can 'audit' the course (take it but not for grade or credit), likely not.  Off to the Interweb...

One side thought:  I run my AD2 through a powered USB hub.  I am hoping to protect my high dollar Surface Book (my main computer).  There are also USB isolators but I don't know if they are 'fast' enough.  The idea would be to power the AD2 from a wall wart and break the ground connection to the computer.  The computer ground connection is connected to earth ground and this can be problematic depending on how the project is powered.  Clearly, you wouldn't jump the negative connection of your single supply dual rail power source to the green PE terminal.  That would shift the virtual ground up by 15V.

Grab up some dual rail op amps (LM358?) and get started.  You have the power source, you can create the virtual ground (w2aew video also discusses this) so you can get started with all kinds of op amp experiments.  I would set the current limit VERY low - a few dozen milliamps at most.  The op amp likely takes very few mA and the short-circuit current is about 40 mA.  The PS should limit somewhere below 40 mA I think.  I would want to keep the magic smoke inside the plastic container.

Since you are looking at dual rail, it doesn't matter much which op amp you choose, the classic is the old 741 but they cost too much in the metal can and twice what a LM358 costs in PDIP but still under a buck at DigiKey. You will still see clipping when the signal gets near the rails and you will probably see inversion when the output signal gets below the bottom rail.  Again, w2aew talks about this.  Inversion isn't at all the same as clipping, the output, instead of stopping at the lower rail, flips to the upper rail.  Weird!  Note 6 page 4

http://www.ti.com/lit/ds/symlink/lm158-n.pdf

BTW, when you scope these dual rail op amp projects, the scope (-) wire (and the signal generator ground lead) connect to the virtual ground, not the negative supply.  You will almost always have a resistor between your signal source and the op amp input, 10k is not too high (usually) and helps limit any fault current from the signal generator.  I would put that in FIRST so that even if I did an 'oopsie' short circuit to the power rail, the current would be limited.  The AD2 has a lot of built in protection, I'm just paranoid.

Have fun!

[rstofer]

I decided to practice what I talked about with the 30V supply and virtual ground.  I have attached a .pdf of an LTspice schematic.

Important things to note:

• The virtual ground is created by a voltage divider consisting of R1 and R2
• Everything is referenced to this virtual ground.  Note the signal input and scope output connections
• I drew the power supply as a simple 30V floating (ungrounded) supply as this is what you will be using
• I forgot to show the power connections at the op amp.  Connect 0V to the V- pin and +30V to the V+ pin.
• I have also omitted the decoupling capacitor across the op amp power pins.  0.1 ufd 50V
• It may be useful to add a capacitor between virtual ground and one of the power rails.  0.1 ufd 50V
• I have indicated the leads from the Analog Discovery 2 by color
• Given the matching values of R3 and R4, we expect a gain of 1
• I actually built the circuit and it works fine, with a gain of 1
• IMPORTANT: The black wire from the signal input will bring an Earth Ground to the virtual ground point.  The significance of this may seem nebulous but it's vitally important to understand how USB and the computer Earth Ground mess with this kind of thing.
• Depending on the op amp, 30V is right at the upper voltage limit.  If the circuit works with the PS cranked back to 24V there would be more margin against the Max Voltage spec.  Nevertheless, a bazillion op amps are running at 30V.

There are some excellent tutorials and a forum on the AD2 over at Digilent.

Make absolutely sure the virtual ground works before hooking up the AD2.  You are close to the point where the ESD protection kicks in if you should somehow get the entire supply across the AD2 inputs.  Nominally, the limit is +-25V.  Under the exact wrong circumstances, you can get as high as 30V if the black lead is on 0V and a signal lead is on +30V.  There's not much that can go wrong with a virtual ground as long as the circuit isn't shorted.  This would be a good thing to check with a DMM.  The virtual ground should be 15V away from the rail voltages.  Whether it is + or - 15V depends on which way you hold the probes.  Just make sure the virtual ground is centered between the rails.  Then hook up the AD2.  They also make rail-splitter chips...

BTW, that's why people use the BNC adapter.  Yes, it earth grounds everything you connect to the probe ground (it would connect to virtual ground in this case) but the 10X on the probe leaves a lot of voltage margin.  I didn't use my adapter for this experiment.

There's nothing magic about the TL081, it just happens to be the first op amp I picked up.  It has very high input impedance because of the JFET input.  This is critical for analog computers because input current upsets Kirchhoff's Current Law and the integration is less than perfect.

*********************
It is vitally important that your power supply be floating.  There can be no electrical connection between the (-) terminal or the (+) terminal and the earth ground terminal.  Most power supplies are floating.  Turn the supply off and test continuity between each output terminal and earth ground.  My DMM reads OL on all outputs - as I expect.

That virtual ground is the 'ground' for the entire system and will be earth grounded the moment you hook up the black wire from the AD2.  It wouldn't do to short the power supply which could occur if the power supply wasn't floating.

This is so much more straightforward with dual 15V supplies.  Then we ground the center point to earth ground and everything works fine.
*********************