Greetings! Intro and looking for advice.

[Mukrakiish]

Greetings to all, been floating around and lurking as a non-member for a while...but haven't had much to say yet until I knew a little something first.

I'm an M.E. by trade, but I've always liked tinkering with electronics. We took some in school, so I got a feel for many aspects ranging from how to build a multimeter, PLC's, motor controls, some discreets, op-amps and so forth. I bought myself a mixed bag of various parts, stack of resistors, capacitors, diodes of all types, some IC's that are basic logic styles, handful of op-amps and a few 555's.

As for equipment I've got a decent DMM, a low-cost linear DC-DC power supply for up to 30v for playing around with. I've got a nice solder iron, supplies to suit, experience in soldering and basic electrical. I've tinkered some...but I want to move up. I've got the 2nd ed of Art of Electronics, Electronics from Make: Magazine and assorted others. Admittedly, I haven't gotten into AoE quite yet and I'm considering getting the 3rd edition + the Learning with AoE. It looks really good, but I wonder how far it goes without a scope.

Now, here's the clincher. I know that I need a bench supply, ideally a scope, signal generator, so forth. Problem is...I don't have a big budget since I'm still trying to pay off the student loan and support the family 

I'm really on the fence about spending big dough on a lab supply and scope, I know that I need them for serious work and for serious learning...but I'm also the kind of guy that would prefer to build something that does something, rather than building say...some fancy load bias audio frequency something or other specifically geared towards the EE crowd you know?

So, any advice from you seasoned veterans out there? I just don't have the kind of dough (and honestly I'm worried that I won't get the 'value' from a nice digital scope compared to someone who is in EE). How far can I go down this road of learning before I hit roadblocks? I've seen 100x over that the USB style scopes are just 'toys' so...yeah, there's my backstory, I'm sure a handful of you can relate. I'm not scared of building things from scratch, but I do like to have things like current limits and short prevention. I've zapped myself enough times to know it's not much fun. (Learned many a lessons from building a laser cutter at home). I'm willing to spend a few bucks to get in...but if you say you need $600-$700 just to have a working 'bench' then I get gun-shy.

Thanks, looking forward to any comments! You all seem like a fantastic crowd. 

[Rick Law]

Welcome!

ME as in Medical Examiner or Mechanical Engineer?

(It would be interesting to get view points from one who spends the day dealing with the dead if you are a Medical Examiner)

[Mukrakiish]

Mechanical Engineer, though being a Medical Examiner...that would prove to be an interesting crossing of hobbies. Heh heh.

[RoGeorge]

Any basic sound card can be both an oscilloscope and a signal generator for audio spectrum (i.e. https://www.zeitnitz.eu/scope_en). Even with a decent scope and signal generator on the bench, I still found myself using the sound card here and there.

To learn, you don't need fancy expensive instruments.
Happy tinkering!

[rstofer]

Power Supply, you say...  Well, I've been messing around with electronics for 60 years or so and I still don't have one.  Don't have a decent signal generator either!  I recently bought a DS1054Z scope to supplement (not replace) my Tektronix 485 that I bought 12 years ago for $200.  I have had other incantations since I built one when I was about 10 years old from plans in the ARRL Handbook.  A scope is required!

How ever do I get along without a power supply?  Most of my projects are microcontroller based and those plug in via USB or 5V wall wart.  I use batteries for small robotics projects and I use wall warts to feed 7805 regulator type devices.  Yes, I'm going to buy a Rigol DP832 one of these days.

I can generate a lot of square waves with a microcontroller, CPLD or FPGA so I don't spend much time missing a proper signal generator.  Nevertheless, I plan to buy the Rigol DG1032Z at some point.

I can't say much about LTAOE, I haven't worked through the book although I plan to.  I bought the recommended Global Industries PB-503 Prototype Board but it is really expensive. It does have 3 power supplies, a signal generator (sine,triangle,square) plus TTL along with switches, knobs and dials.  There is a generous breadboard area.  I don't know what to think about the tool but I suspect it will be very handy.  There are clearly other ways to do this.

Why buy this stuff at my age? Well, my grandson just started EE school!  I hope he will get a lot of use out of these tools over the next few years.

I almost always recommend that newcomers get the Digilent Analog Discovery device because it has two scope channels (full differential, very interesting!), two signal generators (arbitrary waveform), two power supplies (very limited current and voltage), 16 bits of digital IO which can be used as a logic analyzer as well as simple IO and some neat features like a network analyzer.  FWIW, the network analyzer scans across a frequency range and plots the output of the circuit versus frequency.  Handy for filters and such.

For $279, there a lot of electronics that can be done with that gadget.  Serious electronics!  This is not a toy!  http://store.digilentinc.com/analog-discovery-2-100msps-usb-oscilloscope-logic-analyzer-and-variable-power-supply/

If I were just starting out, I would seriously consider the Analog Discovery (and, yes, I have one) and scale the circuits in terms of voltage such that the AD could handle it.  There are limits to the device and they need to be heeded.  There'll be no measuring mains voltages or higher voltage ampllifier circuits.  Troubleshooting PC power supplies is pretty much out of the question (or at least it is an advanced topic).

There are junk USB scopes out there and there are also some very nice tools.  The AD is one of the better devices.  There's nothing quite as nice as a 27" scope screen!

Just starting out, get a decent DMM and the Analog Discovery.  That lab is complete and capable for a long time to come.






ETA:  The power supply capabilites of Ver 2 are seriously expanded.  It can also be used with an external wall wart to provide up to 750 mA (max 2.1W) per output.  And the output is variable.

[rstofer]

These will never replace a true bench power supply but for breadboarding, they may be perfect.  Simply plug in a USB cable (I would do that from a powered hub, but that's just me...) and you can daisy-chain more than one device:

http://store.digilentinc.com/powerbricks-breadboardable-dual-output-usb-power-supplies/

The 12V is out of stock for the next 3 weeks but it would be great for op amp circuits because all of these modules provide both +V and -V.  The +-5V would be great for more modern op amps and, of course, the +-3V would be handy for microcontroller projects (ignore -V).

The +-12V may be a little low for the very old opamps that can't get their outputs anywhere near the rail (741, etc).  Change to a more modern device or scale the experiment.

[edy]

I know some may frown upon old CRO's (cathode ray Oscilloscope) but Dave has a great video on how useful they still are and how cheap you can get them for. I have picked mine up from eBay or Kijiji , Craigslist for $1/MHz.

[CatalinaWOW]

I think the answer to your question will depend a lot on you.  What will it take to keep you fascinated?  If it is just learning the theory behind robotic controls or industrial controls you can do almost everything with a sound card and the various software that is available (for free).  Oscilloscope, spectrum analyzers, even dynamic system analyzers.  The limitation to the near audio frequency range won't limit you much if any.

If making and testing strange gadgets is what interests you, you may need more sophisticated instruments. 

So identify what you are interested in now and think about what you will need to do to define your gear needs.  Don't worry about identifying what you will be interested in five years from now, it is likely to change anyway.

[GreyWoolfe]

Try visiting hamfests if there are any in your area.  I got a 100Mhz Tektronics 2235 scope for $45 condition unknown at the end of the last day of a hamfest because the seller didn't want to pack it back up.  Some contact cleaner and vigorous manipulation of the horizontal and vertical controls and I had a working scope.

If you have schools in your area that teach EE, see if they have any old scopes that they plan to get rid of, you might score a deal there.

[Mukrakiish]

Thanks all the replies and thoughts...lots to ponder over. You're definitely right Catalina, I really need to evaluate what will make me interested now.

I've found some fairly interesting power supply options that I think will satisfy that, I will have a look at the sound card scopes (I didn't even know that was possible, too cool!), being free you can't go wrong on that front. I was having a look at some lower priced DSO's ($300 shipped, lower memory Hanteks/Rigols) which seem like a fairly attractive price point considering what they do. I also look at the scopes as a means to learn more, "see" more of what is going on inside the system. Much like in Mech. Engineering, I like to see how it works rather than pure math (not that there is anything wrong with the math). The Discovery looks very attractive with the exception that with shipping and accessories it pushes me beyond an actual desktop scope since I'm not in the US. Shame.

The area I live in, the hard part with the CRO's is the shipping that kills. Very little EE activity in my vicinity though in neighboring provinces it's blossoming. Cest le vie.

Does anyone have any thoughts on the new AoE or the Learning AoE? I like AoE for it's reference and teaching level but to have a "course" to walk through sounds quite interesting.

[CraigHB]

How far can I go down this road of learning before I hit roadblocks?

Pretty far actually.  The most basic equipment is a benchtop DC supply and an oscilloscope.  Of course you can spend thousands on that stuff, but you don't ~have~ to.  If you want to do RF stuff you need a pretty good scope, but for most digital stuff you don't need anything fancy.  Below is a link to some really inexpensive stuff you can use.  It's cheap enough that if you run into limitations, you can buy more expensive equipment as you need it.

http://www.banggood.com/Wholesale-Arduino-Compatible-Kits-and-DIY-Kits-c-3091.html

[nanofrog]

FWIW, you might want to consider used equipment, particularly for the PSU & signal gen. given you're location. 

For example, you could opt for a Rigol DS1054Z oscilloscope and upgrade/hack it to push it to 100MHz, double the memory, and add a bunch of software features at no cost.  Then opt for quality used gear in good or better shape for say a triple output PSU and a used sig. gen. that suit your needs. 

[Mukrakiish]

You know it's interesting, I had a look on Kijiji and I found a handful of decent looking old scopes, a couple in the $/mhz range. Maybe I should consider just saving myself the money and buying an older high quality one and put the remaining budget into the books/parts for learning.

I can get a Rigol 1102E for $400 even...or some older ones like a Tektronics 7613 for $100. The Tektronic looks good for the value from what I've been reading. I'll admit, I do love the idea of buttons and dials over the software version like through a soundcard. That and scopes just look awesome as a piece of equipment on the bench.

[RoGeorge]

With 400$ you can buy a new Rigol DS1054Z, which is way much better then a 1102E. DS1054Z can work up to 100MHz if you unlock it. 50 MHz for DS1054Z is only a software limitation. If you can buy a DS1054Z, do it. At its price, it's a steal. It is very well built, 4 channels/100Mhz and it has tons of features.

[nanofrog]

I can get a Rigol 1102E for $400 even...

You can get a 1054Z for that, which is a much nicer oscilloscope (newer design w/ 4 channels). Then hack it to unlock the higher BW, memory, and software features. 

[vk6zgo]

You know it's interesting, I had a look on Kijiji and I found a handful of decent looking old scopes, a couple in the $/mhz range. Maybe I should consider just saving myself the money and buying an older high quality one and put the remaining budget into the books/parts for learning.

I can get a Rigol 1102E for $400 even...or some older ones like a Tektronics 7613 for $100. The Tektronic looks good for the value from what I've been reading. I'll admit, I do love the idea of buttons and dials over the software version like through a soundcard. That and scopes just look awesome as a piece of equipment on the bench.

If the Tek 7613 is working OK,depending upon what plugins are fitted,it can be a very good instrument.
It is an analog Storage 'scope,but in some cases the Storage function may no longer work.
It can still,in that case,be used as non-storage unit .

My 7613 cost me $A135 some years back,& came with a 7A18,& a 7A12 vertical plugin.
The horizontal plugin is a 7B53A.

This set up provides 4 channels if required.
The Storage function works perfectly on my Oscilloscope.

You can only store one waveform at a time,so you would need to photograph the display with a digital camera if you wanted long term storage.
The analog Storage function will allow you to see things like switch-on transients.

That said,over many years,I have very seldom felt any great need for Storage.
I think the mindset of DSO users is different---it is so easy,so why not use ir?

[tggzzz]

Don't buy something until you have a specific use for it - any equipment costs money and time (to learn what it can/cannot do). That implies defining your project and what the project needs.

You can do a heck of a lot with a multimeter, a switch, a potentiometer, an LED, and printf() statements - provided you also have understanding and imagination. OTOH, you while can't buy understanding and imagination, you can buy equipment

An old analogue scope may be all you need, but make sure it is a working scope - because it will be difficult for a beginner to safely get one working. Ignore any scope where you can't see waveforms on each channel. Ideally look for >=100MHz. Many places are simply thowing out 20MHz analogue scopes. Don't forget the cost of scope probes.

The Analog Discovery (q.v.) is a good starting point for audio/mechatronic/simple digital projects. (Simple digital => no signal integrity problems). Doubly so if you can get the academic discount. I use a Tekronix 485 for signal integrity problems, an AD for capturing single-shot transients and as a waveform/pattern generator.

If you haven't "got into" TAoE2, then don't spend money on TAoE3. Most of TAoE2 (and particularly the important semi-unique parts) is still valid, and you don't need the updated information yet.

PSUs: earth connections are the big issue here. If you can sort those out, then you may be able to get away with wall-warts and/or variable voltage switching PSU modules from fleabay.

[Gyro]

You can do a heck of a lot with a multimeter, a switch, a potentiometer, an LED, and printf() statements - provided you also have understanding and imagination. OTOH, you while can't buy understanding and imagination, you can buy equipment

Come to that, whatever happened to the good old logic probe? It used to be a very common item on the bench to compliment an analogue scope. The ones which included pulse and memory function were very good for detecting pulse trains and edges that you couldn't see on the scope and you could even get a reasonable indication of duty cycle from the level led brightnesses. Easy to build too.

[tggzzz]

You can do a heck of a lot with a multimeter, a switch, a potentiometer, an LED, and printf() statements - provided you also have understanding and imagination. OTOH, you while can't buy understanding and imagination, you can buy equipment

Come to that, whatever happened to the good old logic probe? It used to be a very common item on the bench to compliment an analogue scope. The ones which included pulse and memory function were very good for detecting pulse trains and edges that you couldn't see on the scope and you could even get a reasonable indication of duty cycle from the level led brightnesses. Easy to build too.

Hmmm. I've had one for many years, but have never found a use for it.

An LED's brightness is sufficient for the duty cycle. Presence of a transition can be determined by using it to trigger a slow-ish sweep on a scope, even if you can't see the pulse width.

[Gyro]

I used to use mine quite a lot... apart from that overpriced HP one, with the single flashing bulb in the end, that floated around the lab.

I guess loading isn't so much of an issue these days compared to TTL and 4000, LEDs are much more sensitive now too.

[tggzzz]

I guess loading isn't so much of an issue these days compared to TTL and 4000, LEDs are much more sensitive now too.

Yes. Nowadays, with fast transitions, the capacitive load dominates the resistive load. Beginners don't realise that at 100MHz a "low" impedance *10 Z0 probe presents significantly less of a load than a so-called high impedance 10Mohm *10 probe (and the Z0 probe's capacitance doesn't resonate with ground lead inductance at ~100MHz).

[rstofer]

Logic probes had a place.  I had an HP and now I have 3 or 4 Radio Shack units.  None of which ever get used anymore.  The logic probe will tell you that something happened but it won't tell you when or in what relationship with some other signal.  With decent scopes being inexpensive, why bother?

The thing about the new DSOs, especially the low end units like the DS1054Z, is that you can get a basket full of capability for a very modest price.  Not that $400 is insignificant, it isn't!  But the decoding option is extremely handy when you are trying to do anything with serial protocols.

Long story:  Say you have a Tektronix 4006 Vector Graphics Display Terminal from the '70s that you have been carting around for 40 years.  Further, let's say you run across the Tektronix Plot10 library of graphics routines written in Fortran.  So now you want to hook gfortran to a Windows VCOM port so you can write ASCII to a USB->RS232 adapter by way of a C library.  Well, first you need to get the serial link working and that's where the DS1054Z comes in.  It's really neat when the decoded message on the display is exactly what was supposed to be sent!

Sure, I could see the wiggles with my analog scope and, given time and patience, I could probably decode it but why bother?  Yes, I could also loop the stream back to the PC and into a terminal emulator like PuTTY but the scope solution is cool!  Besides, the loop-back only works for RS232, not SPI or I2C.

I like my Tek 485 and I used it the other day for a little troubleshooting on my CNC mill.  It's still a workhorse but the DS1054Z is getting a lot more use.

[Gyro]

I guess loading isn't so much of an issue these days compared to TTL and 4000, LEDs are much more sensitive now too.

Yes. Nowadays, with fast transitions, the capacitive load dominates the resistive load. Beginners don't realise that at 100MHz a "low" impedance *10 Z0 probe presents significantly less of a load than a so-called high impedance 10Mohm *10 probe (and the Z0 probe's capacitance doesn't resonate with ground lead inductance at ~100MHz).

No argument there  ...

[tggzzz]

Neatly done. What is the spear ground's wire, and what type is the resistor? Have you published traces or frequency-domain measurements?

I haven't made my own Z0 probes, since I have 3 HP 1.5GHz probes, but I have made a lower-inductance add-on for a standard *10 probe. See
https://entertaininghacks.wordpress.com/2015/04/23/scope-probe-accessory-improves-signal-fidelity/
and
https://entertaininghacks.wordpress.com/2016/09/17/scope-probe-accessory-higher-frequency-results/

I used 0.8mm piano wire for the spears, and they can pivot radially. I "cut" them by repeatedly flexing the wire until it word-hardened and broke. The ends contain sharp edges that cut through oxide layers nicely, and slip less easily.

[Gyro]

Thanks.  The ground wires were just the stiffest wire I could find in the sweepings on the bench, I'm pretty sure they're plated steel. Cutting with bevel edge cutters (well I wasn't going to use my flush ones!) gave a good 'spear'.

No I don't have the facilities to do frequency characterisation these days (sob). The best I could do was parallel measurement alongside Tek P6139A with a 100MHz squarewave logic source on a 250MHz analogue bandwidth scope. I could clearly see waveform distortion on the Zo channel as the Tek probe was attached / removed, but only slight amplitude change on the Tek probe channel when adding / removing the Zo probe.

The resistors are bog standard 1k metal film (0.25W)... 20:1 with 50R through terminators. They are socketed so that I can swap resistors as needed.

There's a thread on here about self-rolling probes (I see you get a mention for your EDN article. 

https://www.eevblog.com/forum/projects/lo-z-probe/

Thanks for the links, reading now.