How do I prevent errors that keep my electronic circuits from working?

[nshin31302]

As an electronic hobbyist, I've soldered various circuits of interest from the internet. However, around 80% of the circuits that I've soldered fail to function properly. My usual workflow for making circuits involves following the schematic and soldering the components while using a highlighter or pencil to keep track of what was soldered. However, upon powering the circuit, many of my  circuits do not function. Troubleshooting the circuit involves determining which section(s) work, and which ones do not. Once that is determined, I check for wiring errors with the schematic, check for continuity between solder joints, check for solder bridges, check component orientation, and check for correct part values. After hours of troubleshooting, some of the circuits still fail to work properly. I've also tried using various types of circuit boards including breadboard, perfboard (both pad per hole and breadboard style layouts), and even diy PCBs, which I assumed would give more success. However even circuits on PCBs have failed to work properly. So far, I've had the most success with breadboard. What are ways of avoiding and preventing errors while making circuits? Thanks in advance.

[danadak]

If circuits involved, several stages, consider -

1) Do one stage at a time and do a simple test on finishing each stage.
2) Use DVM and look for short at each circuit "interior" node.
3) Check each R for value before soldering in.
4) Double check polarized cap orinetation before soldering in. Same
for diodes, zeners.
5) When first powering up set power supply in constant current mode,
setting current at a representative value. So if there are shorts parts
do not get burned out.

In general sounds like you are doing the right thing marking off schematic one
component at a time.


Regards, Dana.

[ataradov]

This depends on the type of stuff you try. Circuit diagram is not a recipe for 100% reproduction, it is a documentation for a specific implementation. Actual physical design and component type/placement matters a lot. It is like sheet music, it does not really tell you how to play the piece, just what the piece is.

And also, what is your source of schematics? Many places have approximate schematics to get the idea across, but not readily usable for replication.

It also depends on complexity. More or less complex stuff always requires tuning/debugging, that's part of the gig.

[nshin31302]

This depends on the type of stuff you try. Circuit diagram is not a recipe for 100% reproduction, it is a documentation for a specific implementation. Actual physical design and component type/placement matters a lot. It is like sheet music, it does not really tell you how to play the piece, just what the piece is.

And also, what is your source of schematics? Many places have approximate schematics to get the idea across, but not readily usable for replication.

It also depends on complexity. More or less complex stuff always requires tuning/debugging, that's part of the gig.

I usually breadboard the circuits first, which almost always tends to work. But once I try to make the circuit more permanent, the circuits tend to not work. Some examples of the schematics are:

http://uzzors2k.4hv.org/index.php?page=flybacktransformerdrivers
and

http://brohogan.blogspot.com/2010/02/geiger-counter-part-1.html

[xrunner]

I usually breadboard the circuits first, which almost always tends to work. But once I try to make the circuit more permanent, the circuits tend to not work.

Well that's interesting ...

That seems bass-ackwards to me. If you are sure you didn't mis-type that, then we'd have to be scientific about the reason. Two things that come to mind are that breadboards would tend to have higher resistance contacts than a good solder joint if they have been used a lot or are of low quality, and can have some capacitance there between rails, albeit not a lot probably in the low pF range. I think Dave did a video on how much.

Perhaps you have had a streak where some higher resistance in certain places made the circuit work, and when you soldered it the lower resistance made it not work? I don't know I'm just brainstorming. 

[ataradov]

In both cases check that 555 actually starts and generates frequency.

How exactly do you do a permanent assembly? On a protoboard? Can you show some pictures?

[Brumby]

After hours of troubleshooting, some of the circuits still fail to work properly.

The first question that comes to mind is: For the circuits that you DID get working ... What was the problem you fixed?

[Brumby]

The second question comes from the difference between construction on a breadboard and on other substrates....  How are your soldering skills, what equipment do you use and what settings do you use (if there are any)?

How exactly do you do a permanent assembly? On a protoboard? Can you show some pictures?

Pictures could be very helpful.

[rstofer]

There are solderable boards that exactly mimic prototype boards.

https://www.amazon.com/Busboard-Protot-Breadboard-Tie-Point-Breadboards/dp/B01KHWGLJ4

I have never used these so I can't really say too much about them.

[Audioguru]

Are you buying real name-brand parts from a local electronic parts distributor or cheap Chinese fake parts from ebay?

[nshin31302]

Perhaps I should have rephrased my question better. My main question is on how to avoid making mistakes while soldering circuits in order to save hours of troubleshooting,


 

I usually breadboard the circuits first, which almost always tends to work. But once I try to make the circuit more permanent, the circuits tend to not work.

Well that's interesting ...

That seems bass-ackwards to me. If you are sure you didn't mis-type that, then we'd have to be scientific about the reason. Two things that come to mind are that breadboards would tend to have higher resistance contacts than a good solder joint if they have been used a lot or are of low quality, and can have some capacitance there between rails, albeit not a lot probably in the low pF range. I think Dave did a video on how much.

Perhaps you have had a streak where some higher resistance in certain places made the circuit work, and when you soldered it the lower resistance made it not work? I don't know I'm just brainstorming. 

I didnt mistype the statement, I've had more success on breadboard than on perfboards. My guess is that breadboard is easier to work with than perfboard, as im not focused on soldering. However, the PCB that I made was already laid out, yet it did not work.

In both cases check that 555 actually starts and generates frequency.

How exactly do you do a permanent assembly? On a protoboard? Can you show some pictures?

To use the geiger counter circuit as an example, I check if the 555 timer circuit is functioning properly by measuring if a high voltage is present or not. But my probelem with the griger counter circuit is always with the tone generation/counting part of the circuit.
Also, some examples of my attempts at making a geiger counter circuit are in the attachments.

After hours of troubleshooting, some of the circuits still fail to work properly.

The first question that comes to mind is: For the circuits that you DID get working ... What was the problem you fixed?

The probelms that I fixed were mostly wiring errors.

The second question comes from the difference between construction on a breadboard and on other substrates....  How are your soldering skills, what equipment do you use and what settings do you use (if there are any)?

How exactly do you do a permanent assembly? On a protoboard? Can you show some pictures?

Pictures could be very helpful.

I would say my soldering skills are fairly decent, (images of my soldering are in the attachments). The soldering iron I use is a Hakko FX-888D soldering station, of which I usually solder at 680-720°F.

Are you buying real name-brand parts from a local electronic parts distributor or cheap Chinese fake parts from ebay?

For most of the resistors and capacitors, I use resistor and capacitor kits off of ebay (I make sure to check their values eith a multimeter), but for semiconductors and some special resistors and capacitors I use digikey to order the components.
Also thanks to all who have replied.

[ataradov]

Perhaps I should have rephrased my question better. My main question is on how to avoid making mistakes while soldering circuits in order to save hours of troubleshooting,

Same way as you avoid any mistakes - by being careful and getting more experience.

In case of the sound part - check that you have that exact part number for the inverters. You do need an IC with schmitt trigger inputs. Using any plain inverters may work, or may not.

PS: there are no attachments.

[nshin31302]

PS: there are no attachments.

Problems with attachments... (File size too large?)

Geiger Counter Attempts:
https://drive.google.com/file/d/0B6FEd1tXKqNzbUdvQUttMjloVUE/view?usp=drivesdk
https://drive.google.com/file/d/0B6FEd1tXKqNzZ3pGOFM2TFlxaVE/view?usp=drivesdk
https://drive.google.com/file/d/0B6FEd1tXKqNzREwzZjNCM2tjVW8/view?usp=drivesdk
https://drive.google.com/file/d/0B6FEd1tXKqNzTndaeks5ai1YZTQ/view?usp=drivesdk

Also here's a video of the geiger circuit working on breadboard.
https://drive.google.com/file/d/0B6FEd1tXKqNzTVptQUVkb3lDVG8/view?usp=drivesdk

[ataradov]

Soldering needs more flux, but otherwise looks ok. Find a way to verify if lower inverter generates, and if not - pay closer attention to that part  of the circuit.

[rdl]

This may or may not have anything to do with problems you've had, but with the 555 it's always a good idea to put a decent size capacitor (0.1 - 1 uf) from pin 8 (power) to ground, as close to the pin as possible, and always connect pin 5 to ground with a small capacitor (.01uf or so) if it's not actually being used.

[xrunner]

I didnt mistype the statement, I've had more success on breadboard than on perfboards. My guess is that breadboard is easier to work with than perfboard, as im not focused on soldering. However, the PCB that I made was already laid out, yet it did not work.

It's just human error in the wiring. On the protoboard you prolly make a lot of mistakes but you really don't think much about it because you just pull the wire out and put it back where it needs to go. When you are soldering it's harder to correct the error and the topology is not the same, so it causes errors. Everybody has gone through that. Just take your time really is all it amounts to. 

[KL27x]

Perhaps I should have rephrased my question better. My main question is on how to avoid making mistakes while soldering circuits in order to save hours of troubleshooting,

For me, part of the solution was to stop doing the parts thru hole. Flipping a board to work on it from the bottom is very difficult. It is fine if you using a PCB, but actually building it as you go increases your chances of making mistakes, IME. From the bottom, all you see are solder blobs and everything is the wrong way around. Being able to see the entire circuit while working on it is much easier for me. I use the proto/veroboard copper side up. All the DIP parts go on like giant smd parts. When you use a breadboard, all the parts and jumpers are on one side and you are looking at it all at once without mental gymnastics. This is a big part of the reason why you have less trouble with breadboard, in my estimation.

Jumpers: when you are using such methods, you will use a lot of jumpers. You want to avoid building ratsnests. These are difficult to work on and to debug. By using very thin, malleable, solid core wire, you will keep jumpers nice and compact and they will stay where you push them. 30AWG kynar wire is highly recommended.

Yes, your circuits will be incredibly ugly. But beautiful when it comes time to rework or debug. When you stop worrying about pretty and focus on function, you will remove a lot of sources for error. If it really needs to be small, then you might have more success and less time to build a larger, sprawled prototype, first. Then revise it to make it smaller. Eventually you may make the jump to PCB CAD.

Why your PCB doesn't work? Who made it? If there are electrical errors, they are easiest to find in the software. On a toner transfer board, you have no thru holes. Make sure components are soldered both top and bottom where necessary. The pcb is the easiest way to make a circuit. Try it again. An oscilloscope and a microscope are two of your best friends, if you get more serious.

[orin]

PS: there are no attachments.

Problems with attachments... (File size too large?)

Geiger Counter Attempts:
https://drive.google.com/file/d/0B6FEd1tXKqNzbUdvQUttMjloVUE/view?usp=drivesdk
https://drive.google.com/file/d/0B6FEd1tXKqNzZ3pGOFM2TFlxaVE/view?usp=drivesdk
https://drive.google.com/file/d/0B6FEd1tXKqNzREwzZjNCM2tjVW8/view?usp=drivesdk
https://drive.google.com/file/d/0B6FEd1tXKqNzTndaeks5ai1YZTQ/view?usp=drivesdk

Also here's a video of the geiger circuit working on breadboard.
https://drive.google.com/file/d/0B6FEd1tXKqNzTVptQUVkb3lDVG8/view?usp=drivesdk

Nothing stands out as an obvious problem.

When I was young, nothing worked properly for me, however I built it until the early days of microprocessors when I actually managed to make a simple SC/MP system (go look it up) that worked - drew the PCB with an etch resist pen, etched it myself etc..

These days, most things I build do work... eventually.  I'm still surprised when I apply power and whatever it is actually works!

I'd say keep trying - some times you get a run of bad luck.

BTW - is that toner transfer you are using for the PCB and 'silk screen' on the top?  Is it conductive?

Flux was mentioned.  Are you using water soluble flux either separately or in the solder?  If so, clean it off.  It can (more likely will) become conductive in a humid environment... oh, that geiger counter circuit has some high impedances - even more important to get everything squeaky clean - or air-wire the 'high' voltage nodes.

[Shock]

The only way to prevent mistakes is to practice, be methodical, and do fault analysis (learn from the cause of the problem).

Stick to circuits that others have working well. After you have completed soldering components in your vero/perfboard etc, make a copy of the schematic and mark off each individual component when you confirm it's connected and orientated correctly. Don't overheat the component when soldering.

When powering up for the first time fuse or current limit the circuit. One thing that may not be obvious is power ratings of certain components and when heat sinks (cooling) is required.

Check that the supply voltages are getting to the correct parts of the circuit and measuring correctly on your multimeter. Use math to calculate expected results and if required simulate the circuit.

If the circuit doesn't work troubleshoot and don't throw in the towel till you discover the cause, then you will learn from it and the experience gained will speed up next time.

[CJay]

It really is as simple as planning, practice and attention to detail.

You can get stripboard layout packages but I've never needed one, seems to me that stripboard is simple enough that you really shouldn't need one and that they could actually stop you developing skills that will be useful when you become more advanced in the hobby.