Attached is the schematic for this circuit:

U4 3.3V regulator is dropping out and resetting the processor.

Sequence of events:
1. Microcontroller receives command from processor via wireless UART (blue trace)
2. Processor sends response back to PC via wireless UART (red trace)
3. Shortly after that the 3.3V regulator drops down to less than 2V for 16ms, resetting the processor (green trace).

I could 'fix' the issue if I added a 2200uF capacitor from 3.3V to ground.

This first capture is without the 2200uF capacitor, note the green trace is the drop out of the 3.3V regulator:

This second trace is showing the 5V regulated input to the 3.3V regulator. It appears stable at 5V, doubt it is the cause of the issue:



Finally the output of the 3.3V regulator with a 2200uF capacitor from 3.3V to ground:



According to the datasheet for the 3.3V regulator, in order to operate properly it needs a minimum of 0.47uF ceramic capacitor on the input and output (or greater). Per the schematic I have a 10uF cap on both the input and the output.

Also: This circuit is currently built up on a solderless breadboard. The SMD IC chips are mounted on SMD to 0.1" adapters. So the electrical wiring could be having EMI issues and or inductance issues for the regulator.

The input power is from a linear power supply putting out 9V and current limited to 100mA.

The 5V regulator is rated up to 180mA and the 3p3V regulator is rated up to 200mA. I know I am not exceeding these limits because the power supply does not drop into constant current mode during the drop-out time of the 3.3V regulator. Not to mention the 5V that feeds it is stable as a rock.

I am wondering if there could be something wrong with the 3.3V regulator or it is just the fact that this is not the final PCB and is having other issues?

Thanks in advance to all input!

 

For a start, get rid of the current limiting on the 9V supply to rule that out for sure.
And breadboards often introduce problems. They are good for prototyping but definitely not for real use.
There may be problems with the layout causing ground bounce so try to keep the leads short, or even build the power supply part on a piece of PCB with good grounds.
Once again, it just may be the breadboard introducing funnies.

 

Please send a schematic with IC part numbers.

 

For a start, get rid of the current limiting on the 9V supply to rule that out for sure.
And breadboards often introduce problems. They are good for prototyping but definitely not for real use.
There may be problems with the layout causing ground bounce so try to keep the leads short, or even build the power supply part on a piece of PCB with good grounds.
Once again, it just may be the breadboard introducing funnies.

Thanks for the input! I had the problem even with the external current limiting set for 300mA to eliminate that as a possibility. I am kind of thinking like you are about it. I have a PCB already designed, but I did not want to purchase the PCB without first testing a prototype. So you may be correct.

 

When interpreting the Data-Sheet for a Regulator,
Transient-Response-Graphs will usually be provided.

There are 3 general types of Bypass-Capacitors, each having their own characteristics.
Electrolytic,
Tantalum,
and Ceramic.
If You see the Output-Capacitance being specified at ~2.2uf, that's usually
referring to a Tantalum-Capacitor,
which is "fast enough" for most low-Frequency Applications.

If it specifies anything larger than that, it's usually referring to an Electrolytic-Capacitor,
often referred to as a "Bulk-Storage-Capacitor".
These are somewhat "slow" and can't keep-up with high-Frequency-demands,
so,
they are usually specified along with a Ceramic-Capacitor,
which type is much more capable of handling high-Frequency-demands.

If the Data-Sheet specifies an Output-Capacitor that is less than ~2.2uf,
they are normally referring to a Ceramic-Capacitor only.

Keep in mind that Capacitors of any type have specifications that will
determine their suitability and performance in a given situation.
Sometimes a variety of different types are used in parallel for maximum performance.

You MUST HAVE Ceramic-Bypass-Capacitors,
preferably situated as close to the Power-Input-Pins as is possible.
Bulk-Capacitors "may be" optional,
that is, if the regulators can easily supply the "worst-case" maximum-Current.
.
.
.

 

Please send a schematic with IC part numbers.

? The IC part numbers are are displayed on the attached schematic in the original post. It is included as a pdf at the bottom of the message?

 

When interpreting the Data-Sheet for a Regulator,
Transient-Response-Graphs will usually be provided.

There are 3 general types of Bypass-Capacitors, each having their own characteristics.
Electrolytic,
Tantalum,
and Ceramic.
If You see the Output-Capacitance being specified at ~2.2uf, that's usually
referring to a Tantalum-Capacitor,
which is "fast enough" for most low-Frequency Applications.

If it specifies anything larger than that, it's usually referring to an Electrolytic-Capacitor,
often referred to as a "Bulk-Storage-Capacitor".
These are somewhat "slow" and can't keep-up with high-Frequency-demands,
so,
they are usually specified along with a Ceramic-Capacitor,
which type is much more capable of handling high-Frequency-demands.

If the Data-Sheet specifies an Output-Capacitor that is less than ~2.2uf,
they are normally referring to a Ceramic-Capacitor only.

Keep in mind that Capacitors of any type have specifications that will
determine their suitability and performance in a given situation.
Sometimes a variety of different types are used in parallel for maximum performance.

You MUST HAVE Ceramic-Bypass-Capacitors,
preferably situated as close to the Power-Input-Pins as is possible.
Bulk-Capacitors "may be" optional,
that is, if the regulators can easily supply the "worst-case" maximum-Current.
.
.
.

The manufacturer recommended 0.47uF ceramic capacitors on both the input and output or greater value. I have 10uF ceramic capacitors on both the input and output of all regulators on the board. The PCB layout is much better as these capacitors are right next to the regulators. I have them as 'close as possible' on the solderless breadboard but it is far from optimum. I am thinking it could just be the bad layout for the prototype and may be fixed with final PCB but not sure about that.

 

All traces will introduce "some" Inductance and Resistance,
so a Bypass, Capacitor is quite often required
as close as possible to the Power-Input-Pins of any
device capable of demanding high-Frequency Current-Spikes.
.
.
.

 

Note on schematic:
The schematic shows the 3.3V regulator part number S1318A33-M5T1U4. These were rated for 100mA these had to be backordered so I am currently using a pin compatible same manufacture part number that is 200mA rated with this part number:
S1313D33-M5T1U3. These are from the same manufacturer and belong to the same family. I have not updated the schematic with that new part number. I have attached the datasheet.

 

All traces will introduce "some" Inductance and Resistance,
so a Bypass, Capacitor is quite often required
as close as possible to the Power-Input-Pins of any
device capable of demanding high-Frequency Current-Spikes.
.
.
.

The PCB meets those requirements. I am not sure that will 'fix' the issue so I plan to add a 2200uF electrolytic cap to the final PCB layout and will try the circuit without that installed first. Thanks for the advice. I have a feeling the solderless breadboard layout is the cause of the issue, but not 100% certain.

 

All traces will introduce "some" Inductance and Resistance,
so a Bypass, Capacitor is quite often required
as close as possible to the Power-Input-Pins of any
device capable of demanding high-Frequency Current-Spikes.
.
.
.

Where you able to look at the schematic? It is attached at the bottom of the original post as a pdf. It has bypass capacitors shown on page 1. They are shown at the top and are applicable to each of the high speed IC chips.

 

Please send a schematic with IC part numbers.

Did you see the schematic? It is attached at the bottom of the original post. Easy to miss. It is a pdf file.

 

Yes but unfortunately a Schematic doesn't show real-World physical layouts.
.
.
.

 

Yes but unfortunately a Schematic doesn't show real-World physical layouts.
.
.
.

The prototype solderless breadboard is the current 'layout'. I have a PCB layout ready. I wanted to test for any major issues by building the prototype on solderless breadboard prior to purchasing the PCB. You may be correct. The protoboard 'layout' is terrible. Attached is the breadboard as you can see it is not a pretty site:

 

Yes but unfortunately a Schematic doesn't show real-World physical layouts.
.
.
.

For the final layout I have a Kicad layout, I can send it if you want. But you and the community reviewed that layout before. This is just a prototype. So I am thinking the issue is terrible layout on solderless breadboard. I am just not 100% sure about that.

 

Here is another picture with the LDO and the 10uF ceramic bypass caps pointed out:

 

The datasheet for the low dropout regulator used says the capacitors are (must) be ceramic.

 

The datasheet for the low dropout regulator used says the capacitors are (must) be ceramic.

Thanks, they are ceramic 10uF on both input pin and output pin in 0805 package. I had to add the 'bulk' 2200uF capacitor to the board to 'fix' the drop out. I did not think I would need bulk capacitors if following manufacturers recommendations?

 

I've never seen this type of Prototyping done before,
but then again, I only dabble around the edges of Digital Stuff.

My estimation is that this type of layout is a complete and utter disaster,
but what do I know, it might be common practice.

Some things just can't be expected to perform correctly when
implemented in this "patchwork" fashion.
I estimate that You may never get this project to operate reliably in this form.
I think You will be chasing your tail trying to diagnose "the next" problem.
This is one of the reasons that I don't think much of SMD Devices when it comes
to "Maker-Projects" or DIY / Hobbyist type projects that will never be
put into Commercial-Automated-Mass-Production.

This Project would be easier to implement on a Large Perf-Board,
with mostly Through-Hole-Parts.

Figure on designing 2 or 3 or more proper, professionally manufactured,
Multi-Layered, SMD Circuit-Boards,
before You end up with a reliable product.

But remember,
I may have no-clue of what I'm talking about,
this is just my take on it,
and more experienced People should be consulted.
There are plenty here, and I'm sure they will chime-in.
.
.
.

 

I've never seen this type of Prototyping done before,
but then again, I only dabble around the edges of Digital Stuff.

My estimation is that this type of layout is a complete and utter disaster,
but what do I know, it might be common practice.

Some things just can't be expected to perform correctly when
implemented in this "patchwork" fashion.
I estimate that You may never get this project to operate reliably in this form.
I think You will be chasing your tail trying to diagnose "the next" problem.
This is one of the reasons that I don't think much of SMD Devices when it comes
to "Maker-Projects" or DIY / Hobbyist type projects that will never be
put into Commercial-Automated-Mass-Production.

This Project would be easier to implement on a Large Perf-Board,
with mostly Through-Hole-Parts.

Figure on designing 2 or 3 or more proper, professionally manufactured,
Multi-Layered, SMD Circuit-Boards,
before You end up with a reliable product.

But remember,
I may have no-clue of what I'm talking about,
this is just my take on it,
and more experienced People should be consulted.
There are plenty here, and I'm sure they will chime-in.
.
.
.

Thanks! I really appreciate your input. Interestingly enough this is the only problem I am running into and the 2200uF bulk capacitor 'fixes' it. You are probably correct, this is just a protoboard. I have the final PCB and layout complete in KiCad. I just did not want to order PCB boards until I know the circuits all worked. Yea, solderless breadboards are not the best. I have to agree. They are a great way to get a 'feel good' feeling though before you spend the money on a PCB.