Power

A "new fashioned" televisor, using an Arduino to drive the motor and display.

Moderators: Dave Moll, Andrew Davie, Steve Anderson

Power

Postby Andrew Davie » Mon Jun 26, 2017 11:34 pm

My first "power brick" arrived from AliExpress today. It's rated 12V @3A. Had a good look at it on the oscilloscope today at work (no load) and it was showing 11.98V with no discernible "ripple", however there appeared to be some semi-random noise amplitude 100mV. Overall it looked pretty good. It's very light, obviously a switched mode power supply. Anyway plugged it into my televisor when I got home and it was happy powering the Arduino and the LCD display, but when I tried playing anything the motor briefly started and then everything reset. In other words, it was unable to handle the power requirements of my televisor (and I hadn't even asked it to drive the LEDs yet). Expected, but a tad disappointing. I have three or four other power bricks on the way, but I expect similar results from them.
User avatar
Andrew Davie
"Gomez!", "Oh Morticia."
 
Posts: 1590
Joined: Wed Jan 24, 2007 4:42 pm
Location: Queensland, Australia

Re: Power

Postby Klaas Robers » Tue Jun 27, 2017 1:42 am

Did you measure the voltage of the power supply while you switched on? Did it fall back to zero? And did it stay zero? If it returned to 12 volts, I would expect that everything started up again......

Can you power the motor from something else for the time being? Can you switch a resistor of several ohms in series with the motor? Do some experiments and you will see that you get feeling of where the problem hides.
User avatar
Klaas Robers
"Gomez!", "Oh Morticia."
 
Posts: 1656
Joined: Wed Jan 24, 2007 8:42 pm
Location: Valkenswaard, the Netherlands

Re: Power

Postby Andrew Davie » Tue Jun 27, 2017 2:10 am

Klaas Robers wrote:Did you measure the voltage of the power supply while you switched on? Did it fall back to zero? And did it stay zero? If it returned to 12 volts, I would expect that everything started up again......

Can you power the motor from something else for the time being? Can you switch a resistor of several ohms in series with the motor? Do some experiments and you will see that you get feeling of where the problem hides.



I measured after switching on, with everything "trying" to run. That is, after the motor was started (but it wasn't running - the disk kind of twitched). At that point, the power supply was measuring 11.87V so its voltage was OK, but the motor wasn't going. I could power the motor from the original power supply, but I'm kind of abandoning this new power brick as I don't see how it's in any way suitable...? Edit: Yes, everything started up again but not so perfect - the program seemed to be "locked up" in that I could no longer select a track, so I expect the momentary voltage drop left the Arduino in an undefined state.
User avatar
Andrew Davie
"Gomez!", "Oh Morticia."
 
Posts: 1590
Joined: Wed Jan 24, 2007 4:42 pm
Location: Queensland, Australia

Re: Power

Postby Andrew Davie » Tue Jun 27, 2017 6:34 pm

Another power brick arrived today, from AliExpress. This one was marked 15V @2A. I had a quick think (not too much of a think!) about the consequences of plugging this directly into my circuit and thought "what's the worst that can happen". Well, the Arduino fired up (it's rated at "up to 20V" but nominally no more than 12V). Also, the LCD - in fact everything was working OK. I selected a track, and the motor started spinning just fine - but just then the band fell off so I couldn't see a picture. And I started to smell something burning. Uh oh! So, power off and start to feel the wires - nothing obvious. Transistors cold, resistors cold. Arduino - ouch!! Red hot. And no longer functioning. So, take-home from this is the Arduino doesn't like 15V (it's supposed to take 5V so 15 is pushing it, I admit). But the good news is the power brick seemed to be doing just fine with the motor and powering everything but the LED matrix. The Arduino - well that's a bit sad, but I now have 10 or so spares so no biggie. I'll put in a DC converter -->5V to power the Arduino before I try the power brick again.
User avatar
Andrew Davie
"Gomez!", "Oh Morticia."
 
Posts: 1590
Joined: Wed Jan 24, 2007 4:42 pm
Location: Queensland, Australia

Re: Power

Postby Steve Anderson » Tue Jun 27, 2017 9:57 pm

Quoting the Arduino Uno datasheet...

"The board can operate on an external supply of 6 to 20 volts. If supplied with less than 7V, however, the 5V pin may supply less than five volts and the board may be unstable. If using more than 12V, the voltage regulator may overheat and damage the board. The recommended range is 7 to 12 volts."

From what you have been saying in the past you're using a lot of resources on the Uno which as said above could overheat the regulator - and this appears to be the case.

I suggest using something like a 7808/7809 or a LM317 to pre-regulate the input voltage into the Uno. This should keep the regulator on-board from frying. I'm somewhat surprised that the say it can handle 20V...but I presume that's if the device is doing little and in sleep/shutdown mode most of the time. A rather poor and vague datasheet in my opinion. Not written for those who may not understand the consequences of excessive thermal dissipation.

In addition a pre-regulator should get rid of that 100mV of noise which is far more than I would tolerate releasing a design into the world. Chinese junk...not the sea-going vessel type.

Steve A.
User avatar
Steve Anderson
"Fester! Don't do that to 'Thing'"
 
Posts: 5360
Joined: Fri Mar 30, 2007 10:54 pm
Location: Bangkok, Thailand

Re: Power

Postby Andrew Davie » Tue Jun 27, 2017 10:16 pm

Steve Anderson wrote:Quoting the Arduino Uno datasheet...

"The board can operate on an external supply of 6 to 20 volts. If supplied with less than 7V, however, the 5V pin may supply less than five volts and the board may be unstable. If using more than 12V, the voltage regulator may overheat and damage the board. The recommended range is 7 to 12 volts."

From what you have been saying in the past you're using a lot of resources on the Uno which as said above could overheat the regulator - and this appears to be the case.

I suggest using something like a 7808/7809 or a LM317 to pre-regulate the input voltage into the Uno. This should keep the regulator on-board from frying. I'm somewhat surprised that the say it can handle 20V...but I presume that's if the device is doing little and in sleep/shutdown mode most of the time. A rather poor and vague datasheet in my opinion. Not written for those who may not understand the consequences of excessive thermal dissipation.

In addition a pre-regulator should get rid of that 100mV of noise which is far more than I would tolerate releasing a design into the world. Chinese junk...not the sea-going vessel type.

Steve A.



Just a few corrections :) ...
1) I'm using an Arduino Micro (not a Uno). an Arduino Micro datasheet.
2) I'm using a different power brick than the one with the 100mV noise - this one I haven't measured yet. Doesn't matter really as I agree on the regulator bit.
3) Thanks for the advice on regulator.. I'll see what's available on AliExpress as a standalone package I don't have to design/solder.
User avatar
Andrew Davie
"Gomez!", "Oh Morticia."
 
Posts: 1590
Joined: Wed Jan 24, 2007 4:42 pm
Location: Queensland, Australia

Re: Power

Postby Andrew Davie » Tue Jun 27, 2017 10:46 pm

I went digging into my box of tricks to see what I already have in-hand.

I found the following that I believe are voltage regulators...

MC78T12
LM317T
L7912C-V

I think the first two, but particularly the second one, might fit the bill. I want to convert 15VDC --> 5VDC (ideally) but if not, then --> 12VDC. I'm thinking one regulator "protecting" the arduino from excessive voltage, but leaving the motor and LED circuit with the original 15V feed from the power brick.

Here's a tutorial on using a LM317T as a variable voltage power supply. I can do this!
User avatar
Andrew Davie
"Gomez!", "Oh Morticia."
 
Posts: 1590
Joined: Wed Jan 24, 2007 4:42 pm
Location: Queensland, Australia

Re: Power

Postby Steve Anderson » Wed Jun 28, 2017 11:23 am

I suggest using the LM317 (with a heatsink) and setting the ratios of R1 & R2 to produce around 8V into the Micro. For some reason I thought you were using the Uno. This will cut down the dissipation in the on-board regulator by a factor of around three. Instead of losing 10V (15V-5V) it's only dropping 3V (8V-5V) which at the same load current is about a third of the power.

The exact same sentence is used in the Micro datasheet too.

The heatsink should not be required to be overly large.

Steve A.
User avatar
Steve Anderson
"Fester! Don't do that to 'Thing'"
 
Posts: 5360
Joined: Fri Mar 30, 2007 10:54 pm
Location: Bangkok, Thailand

Re: Power

Postby Andrew Davie » Sat Jul 01, 2017 12:58 pm

Steve Anderson wrote:I suggest using the LM317 (with a heatsink) and setting the ratios of R1 & R2 to produce around 8V into the Micro. For some reason I thought you were using the Uno. This will cut down the dissipation in the on-board regulator by a factor of around three. Instead of losing 10V (15V-5V) it's only dropping 3V (8V-5V) which at the same load current is about a third of the power.

The exact same sentence is used in the Micro datasheet too.

The heatsink should not be required to be overly large.

Steve A.


This is what I went with, thanks. I now have two LM317 regulators. One produces 8V for the Arduino, and the other 12V for motor and display. I decided to "daisy chain" in that the input to the 8V LM317 comes from the output of the 12V LM317 (rather than from the source 15V for the board). The thinking is that the LM317 is then dropping less voltage to go from 12V to 8V instead of 15V to 8V and it's going to run cooler.
User avatar
Andrew Davie
"Gomez!", "Oh Morticia."
 
Posts: 1590
Joined: Wed Jan 24, 2007 4:42 pm
Location: Queensland, Australia

Re: Power

Postby Steve Anderson » Sat Jul 01, 2017 10:29 pm

Yes, correct. But all you are doing is moving the heat from one device to the other. That excess energy has to be got rid of somewhere! Always a compromise as ever.

Steve A.
User avatar
Steve Anderson
"Fester! Don't do that to 'Thing'"
 
Posts: 5360
Joined: Fri Mar 30, 2007 10:54 pm
Location: Bangkok, Thailand

Re: Power

Postby Andrew Davie » Sun Jul 02, 2017 10:47 pm

power3.jpg
power3.jpg (321.18 KiB) Viewed 16331 times


Here's my current thinking for rev 2. The power input can be variable - say 17V - 25V. That's immediately converted to 15V though a LM317 regulator.
That 15V is used by the motor and LED array.
The 15V is converted to 8V by another LM317 and fed to the Arduno.

I'm thinking this way because I realised that with my earlier design "feeding" 12V to the LED array wasn't really suitable for 3 LEDs. Yes I know everyone has been telling me this.
Anyway, thought I'd throw up this design for review.
User avatar
Andrew Davie
"Gomez!", "Oh Morticia."
 
Posts: 1590
Joined: Wed Jan 24, 2007 4:42 pm
Location: Queensland, Australia

Re: Power

Postby Andrew Davie » Mon Jul 03, 2017 1:38 am

That TIP122 bottom right, running the LED display matrix: since the input is PWM @5V I can replace it directly with an IRL540 - right? - and it will run in saturated mode with the PWM signal, thus reducing the voltage drop from the 1.4V of the TIP122 to 0.7V of the IRL540. Does this seem reasonable?
User avatar
Andrew Davie
"Gomez!", "Oh Morticia."
 
Posts: 1590
Joined: Wed Jan 24, 2007 4:42 pm
Location: Queensland, Australia

Re: Power

Postby Steve Anderson » Mon Jul 03, 2017 11:39 am

The IRL540 if 'typical' should be OK but you may hit a 'worst case' example. It may be worth exploring alternatives, it all depends on what you can get. In my case I find out what the local supplier has and hope to find a device suitable.

I'll have a look through some datasheets and see if there's a better device. I can't recall the total LED current.

Before ordering the PCBs check for differences in the pin-out between the TIP and the final choice of FET.

Steve A.

See if you can source an IRF540 (with an 'F' not an 'L'). You might consider an IRFD220 which is in a 4-pin DIL package, but not until you've sorted out this overheating problem.

Tried to attach two pdfs, both got stuck part way through the upload, I'll try again later.
User avatar
Steve Anderson
"Fester! Don't do that to 'Thing'"
 
Posts: 5360
Joined: Fri Mar 30, 2007 10:54 pm
Location: Bangkok, Thailand

Re: Power

Postby Andrew Davie » Mon Jul 03, 2017 3:12 pm

Steve Anderson wrote:The IRL540 if 'typical' should be OK but you may hit a 'worst case' example. It may be worth exploring alternatives, it all depends on what you can get. In my case I find out what the local supplier has and hope to find a device suitable.

I'll have a look through some datasheets and see if there's a better device. I can't recall the total LED current.

Before ordering the PCBs check for differences in the pin-out between the TIP and the final choice of FET.

Steve A.

See if you can source an IRF540 (with an 'F' not an 'L'). You might consider an IRFD220 which is in a 4-pin DIL package, but not until you've sorted out this overheating problem.

Tried to attach two pdfs, both got stuck part way through the upload, I'll try again later.



According to this datasheet for the IRL540 the continuous drain current is 20-28 A so if I'm looking at the right thing, then I'm pretty sure my LED array will be a fair bit below that ;) The array is about 240 mA. I just have a ton of IRL540 on hand (as well as TIP122) I was more thinking if I swapped the TIP122 out then I'd have more voltage for my LED array, right?

I think you missed my post - the overheating problem is not there in the replacement board. There must be some fault in a component or the traces in my original board, because I replaced the IRL540 with a different manufacturer one and it still heated up. But in the second board I built it is "stone cold" and operating the same as my prototype design. So, I consider the "overheating problem" kind of sorted.

I do have some IRF540 but I was just musing that I could possibly save 0.7V extra for the LED array if the TIP122 were replaced with a logic-level FET running in saturation mode like with the motor. I couldn't see any reason why not.
User avatar
Andrew Davie
"Gomez!", "Oh Morticia."
 
Posts: 1590
Joined: Wed Jan 24, 2007 4:42 pm
Location: Queensland, Australia

Re: Power

Postby Steve Anderson » Mon Jul 03, 2017 7:20 pm

Yes, a logic-level FET is what you need, hard enough at 5V, even worse at 3.3V! These general purpose power FETs usually need around 10V on the gate to fully saturate them, though it does depend on the load current and the individual device.

Using a MOSFET is a good move, though I doubt you notice any real difference in LED brightness just because the voltage loss has gone from 1.2V down to less than half that.

I assume you have a series resistor in the base lead for the TIP device, with a MOSFET that can be removed. i.e. zero Ohms.

"But in the second board I built it is "stone cold" and operating the same as my prototype design. So, I consider the "overheating problem" kind of sorted."

I would investigate why this happened, not just hope it doesn't return.

Steve A.
User avatar
Steve Anderson
"Fester! Don't do that to 'Thing'"
 
Posts: 5360
Joined: Fri Mar 30, 2007 10:54 pm
Location: Bangkok, Thailand

Next

Return to Andrew Davie's Arduino Televisor

Who is online

Users browsing this forum: No registered users and 2 guests

cron