Pico Prototyping - Building a "Pico Uno"

@inst-tech Thanks for the document; I suppose I am "math challenged." Despite that, have located and read multiple 1117 datasheets and there are significant differences among them. Would you please post a link to the one you referenced for the calculations you performed? Thanks again.

Here is the video about my project.

If you have comments or questions, please post them on the video. Thanks.

@tracecom EDIT upon further examination I see it does at least include the MINI360 and reset button, I have all the other parts in any case so I ordered 5 with slow shipping it works out to about $7 CDN all in each.

I have never used PCBWay, does the $5 include all the parts?

@tracecom

Posted by: @tracecom

Here is the video about my project.

If you have comments or questions, please post them on the video. Thanks.

 Nice work! 😉

If anyone is still interested in the answer to the problem I was having with the AMS1117-5 modules, it is actually quite simple. The first batch of 10 modules that I bought all contained defective AMS1117-5 IC's. The failure mode is that when the IC senses that it is overheating, it does not shut itself down; instead, it goes into thermal runaway, releases smoke, and then connects the input voltage to the output pins. In my case, the input voltage was over 12VDC. I ordered another batch of 10 modules from Ali Express, and tested three of the ten; they all failed in exactly the same way. My conclusion is that all 20 of the modules I had were built using counterfeit AMS1117-5 ICs. My next video will include this information; in the interim I suggest that you test any modules that you have before you use them.

Hi Charles @tracecom,

  It seems you got a very poor batch of chips, and maybe they were fakes or rejects. This is a risk when buying in a cut-throat market. Your advice to check out new acquistions is indeed timely.

--------

However, whilst your experience is extremely poor, over several decades I have noted that failed linear voltage regulators have appeared as one of the most common problems, and that the design of the equipment meant that the failed regulators were expected to run 'hot'   ... and this goes back to types like LM7805 in an era in which 'fakes' were not an everyday concern.

My own impression, for which I have only flimsy anecdotal evidence, is that the 'thermal shutdown' feature of linear voltage regulators has been over played for all of those decades, and hence by most of the well known manufacturers. The data sheets seem to imply 'thermal shutdown' as a normal feature that can be relied upon for everyday switching use, like the room thermostat on a central heating system. My personal opinion is that it should be treated more like a car safety belt ... hopefully never needed, it might save you when the unexpected happens, but the safest strategy is to minimise the chance of a crash.

It is well established that increasing the operating temperature of an electronic component (relative to say 20 degrees Celesius (70F)) will rapidly (exponentially) increase the rate of failure. A rough rule of thumb sometimes mentioned is doubling the failure rate, for every 10 degrees Celsius (18F) increase. So clearly, even if the shutdown mechanism is reliable, there is a strong incentive to minimise the operating temperature.

I would therefore recommend anyone using a board with a small regulator like the AMS1117 takes care to ensure it is not running hot in normal operation. A simple finger test maybe good enough for most amateur needs ... if it is uncomfortable to touch, consider reducing the thermal stress on it. Remember you are touching the plastic case; the actual silicon chip temperature will be much hotter as the plastic is a thermal insulator.

Feeding a 5V UNO board with 12V means 7V must be resistively dropped across the regulator. For a current demand of (say) 100mA, this means 700mW of heat to dissipate, which is considerable, particularly if it is not directly connected to a large area of thick PCB copper.

The current demand depends upon the actual circuit, including peripherals, and even the software program. Thus, in some cases, the current demand maybe relatively low, say less than 50mA, and the AMS1117 can cope easily, but a small change software or peripherals, etc. could result in current demand of 100mA or more, and the AMS1117 at risk of overheating.

If your AMS1117 is in danger of overheating, the simplest solution (assuming use of the 12V supply is a fixed requirement) is usually to add a small buck regulator to the system, that will drop the 12V to about 7V (assuming a 5V AMS1117), so that  the AMS1117 only has a 2V drop to cope with, reducing its heat dissipation by 7/2 = 3.5 times.

Just a thought. Take care my friend, Dave

@davee Thanks for your post. Prior to my recent problems with the AMS1117-5 modules, my experience was primarily with linear regulators in TO-220 cases. And although I understand that copper foil under SMD regulators can provide some heat sinking, it has always "appeared" to me to be inadequate when compared to a TO-220 with a finned heat sink attached with a bolt and nut. And even then, I have benefitted from the thermal shutdown feature of an LM7805, for example. I have a preference for linear voltage regulators because of the lower noise when compared to switching regulators, but I recognize that they have their limitations.

@tracecom I live for now full time in an RV so my main power supply is 12V. I almost always use a Buck converter to get from 12V down to 7ish then a LVR to get to 5. Similar for 3.3V. I have not noticed any noise issues yet (using LM2596). When I move back into an apartment next year I will be doing mostly outdoor stuff so battery power in the 3.7V and 7.4V form factor. I might add a LVR to go from 7.4 to 6 because the boards usually have their own LVR to take it down to 5. BTW, this gives 3.3V a big edge for me. Please keep us updated on what you find re the 1117 boards.

Hi Charles @tracecom,

  Thanks for the reply ... to which I largely concur.

  I too have reservations about switchining noise associated with small buck regulators ... especially when they are selling at about 2 for the equivalent of US $1 at Aliexpress, etc.

For some applications, say a low noise audio pre-amp, then I might expect problems using a buck regulator, but driving a small microcontroller board, combined with the filtering effect of the AMS1117 on the microcontroller board, suggests that adding a pre-regulator buck, is certainly not the only possible source of noise, and is probably not going to be the dominant source of noise. Of course, certain sensors etc. may produce very low level signals and need extra care.

This all includes a degree of guesswork and luck, so it was only a suggestion, and I would certainly recommend adding a few odd extra decoupling capacitors in the usual strategic places.

Good luck with your project, Dave

Here is a link to my video documenting my difficulties with the AMS1117-5 modules.

@inst-tech Would you please post a link to the 1117 datasheet you referenced for the calculations you performed and the document you posted? Thanks again.

Posted by: @tracecom

↑

@inst-tech Would you please post a link to the 1117 datasheet you referenced for the calculations you performed and the document you posted? Thanks again.

Hi Charles,(@tracecom)

Please refresh my memory as to the documentation you are requesting as I've been away for the past couple of months, and I'm not sure of what post you're referring to.

looking forward to seeing how you progress with your pojects..

regards,

LouisR

@inst-tech It was earlier in this thread when you posted this document. https://forum.dronebotworkshop.com/wp-content/uploads/wpforo/attachments/4280/5225-2022AMS1117-5-linear-voltage-reg-temp-calc.doc

Thanks.