Hot capacitor with 18m2 low power board (CHI030)

premelec

Senior Member
@ jensmith25 -IF- you haven't replaced the hot cap please do that - there seems to be indication that it's still installed... it may or may not help... but it's a logical first step...
 

jensmith25

Senior Member
Do you have an oscilloscope? If the capacitor isn't being DC voltage stressed, and isn't installed backwards, the MP3 unit may be inducing a severe radio frequency (RF) oscillation, either in itself or the voltage regulator. That could also make the decoupling cap overheat very quickly. There should be a 0.001 uF ceramic cap in parallel with the electrolytic to swamp that kind of thing, but different regulators are very sensitive to capacitive loading, and to the equivalent series resistance (ESR) of the decoupling caps. The exact device data sheet is the bible.
But if that is the case, why would it be fine during programming and testing, which literally took hours or it being in varying states of off and on but always the music was being played because we needed to sync the lighting sequence to the music. Surely, if there was such a problem it would have occurred during the testing and not in the 10-15mins of the customer having it on?

There isn't another capacitor. Only the electrolytic and the standard decoupling capacitor for Picaxe chips.

My husband has an oscilloscope but what would I be looking for in the wave?
 

jensmith25

Senior Member
@ jensmith25 -IF- you haven't replaced the hot cap please do that - there seems to be indication that it's still installed... it may or may not help... but it's a logical first step...
As I discovered yesterday, it's only hot when the MP3 player is linked so I don't think there's necessarily anything wrong with it specifically and I can't risk using the board anyway so will just do a new board.
 

jensmith25

Senior Member
UPDATE:

Soldered up a new CHI030 board and tested the mp3 board and it all seemed to be working fine. Plugged the Mp3 into the CHI030 board and to start with it was all working fine until a few minutes in and the speaker started emitting a loud buzzing noise so I quickly unplugged.

The regulator felt hot I think. It is a bit hard to tell between the two as they're so close together but pretty sure it was the regulator.

I've tested the Mp3 player on the 14m2 audio board AXE171 (running off a 4.5v battery box) and it's working fine so there must be some sort of problem when it's being plugged into the CHI030 board. Maybe it is an oscillation problem with the regulator.

If I have one I can try the 0.001uf ceramic capacitor.

EDIT: Had a 1nf ceramic cap so soldered that onto the pins of the electrolytic cap but the regulator is still getting red hot after a few minutes.

Not sure what else to do. Didn't wait for the speaker to start buzzing.
 
Last edited:

hippy

Technical Support
Staff member
I've tested the Mp3 player on the 14m2 audio board AXE171 (running off a 4.5v battery box) and it's working fine so there must be some sort of problem when it's being plugged into the CHI030 board. Maybe it is an oscillation problem with the regulator.
Or maybe it's just drawing too much current through the regulator. The DF Player module will draw 20mA in standby and more when accessing the SD Card and/or driving the speaker.

Perhaps disconnect the regulator, power the CHI030A from the same battery box you powered the AXE171 from, and see if that solves the problem.

Admittedly it would seem odd that the regulator did not overheat during previous testing but does now.

It would be worth connecting the DF Player module back to the AXE171 to check that it is still working.

You could also power the AXE171 and DF Player through a regulator to see if that also overheats when you do.
 

jensmith25

Senior Member
Or maybe it's just drawing too much current through the regulator. The DF Player module will draw 20mA in standby and more when accessing the SD Card and/or driving the speaker.

Perhaps disconnect the regulator, power the CHI030A from the same battery box you powered the AXE171 from, and see if that solves the problem.

Admittedly it would seem odd that the regulator did not overheat during previous testing but does now.

It would be worth connecting the DF Player module back to the AXE171 to check that it is still working.

You could also power the AXE171 and DF Player through a regulator to see if that also overheats when you do.
Yeah, I just don't understand how it was fine when I was testing it and now it's not.

I did replace the DF player but not the SD card but it seems fine as it's working perfectly in the AXE171 setup with 4.5v. I played it through twice.

I can try it off 4.5v but the problem is that I need 12v for the LED strips so I can't run it off 5v for the customer.

I could see if a voltage regulator board was able to cope better. I guess could try that with the AXE171 board initially and see what happens.

It all works fine off 12v until I connect the mp3 player.
 

jensmith25

Senior Member
Just tried running it through the AXE171 with a LM78XX PCB regulator circuit (one of the cheap ones off the bay) and it seems to be running fine. Played through once with no issues. Regulator only mildly warm. It has a big heatsink too.

I can try hooking that up to the CHI030A board and see if it's okay with the LED strips and MP3 player but what's the best way to connect it given I've got the regulator already on the board. I can remove the regulator but I hate desoldering as it takes forever on multi pin components so I'd probably just snip it off but will it need a wire link to connect the two halves of the PCB again or ???

EDIT: Just thinking, would it be easier just to separate the two devices and leave the regulator as is on the CHI030A board and have another wire from the 12v power supply to the 5v LM78XX regulator and have that suppling the power to the MP3 player so only the data is coming from the CHI030A board to the MP3 player?
So power is split to both the CHI030 board and the MP3 board, both through 5v regulators and hopefully both are happy?
 

hippy

Technical Support
Staff member
Just tried running it through the AXE171 with a LM78XX PCB regulator circuit ... Regulator only mildly warm. It has a big heatsink too.
If a TO-220 style regulator with a heatsink tab is getting warm then a transistor style regulator is likely to get even hotter. It might be worth measuring the current being drawn.

You could just replace the transistor regulator on the CHI030A board with a TO-220 style regulator , power the CHI030A and DF Player through that, but check the pinout first. Viewed from top, legs downwards, I think this is right ...

Code:
  TO-92         TO-220
  .---.        .-------.
 /     \       | O O O |       
| O O O |      |-------|
`-------'      `-------' Tab
Out 0 In       Out 0 In
 

jensmith25

Senior Member
If a TO-220 style regulator with a heatsink tab is getting warm then a transistor style regulator is likely to get even hotter. It might be worth measuring the current being drawn.

You could just replace the transistor regulator on the CHI030A board with a TO-220 style regulator , power the CHI030A and DF Player through that, but check the pinout first. Viewed from top, legs downwards, I think this is right ...

Code:
  TO-92         TO-220
  .---.        .-------.
 /     \       | O O O |       
| O O O |      |-------|
`-------'      `-------' Tab
Out 0 In       Out 0 In
I assumed mildly warm wasn't anything to worry about?

I'll try and check the current. Current drawn by what? The MP3 player?

I could do that and just swap them - I wasn't sure if the pinout was the same but this other board has 4 diodes, 2 beefy capacitors and a resistor so I assumed it might be better at coping if the MP3 player is drawing more current.
 

hippy

Technical Support
Staff member
I assumed mildly warm wasn't anything to worry about?
Mildly warm is probably not anything to worry about. The TO-220 style of regulator is designed to carry larger currents and get somewhat hot. Smaller transistor style regulators aren't.

I'll try and check the current. Current drawn by what? The MP3 player?
I would guess so.

I could do that and just swap them - I wasn't sure if the pinout was the same but this other board has 4 diodes, 2 beefy capacitors and a resistor so I assumed it might be better at coping if the MP3 player is drawing more current.
You can use that board in place of an actual regulator on the CHI030. To save unsoldering and soldering it in to where the regulator is supposed to go you can connect it between V1 and V2 power connections at the bottom of the board.
 

jensmith25

Senior Member
You can use that board in place of an actual regulator on the CHI030. To save unsoldering and soldering it in to where the regulator is supposed to go you can connect it between V1 and V2 power connections at the bottom of the board.
Can you explain further? Do you mean gnd goes to V1 and positive goes to V2?
 

hippy

Technical Support
Staff member
If you look at the circuit diagram you will see V2 goes to the on-board regulator input, its output goes to V1.

So you can remove the on-board regulator, take a wire from V2 to the off-board regulator input, take its output to V1, and connect its 0V to either V1 or V2 0V.
 

westaust55

Moderator
A TO-220 voltage regulator is rated for 1 amp. However, with no added heat sink that will get warm at around 200 mA.
So if the TO-220 LM7805 regulator is getting warm the then the TO-92 78L05 rated at only 100 mA will be overloaded and hence getting very hot. In fact the 78L05 is likely shutting down or cycling on and off.
 

jensmith25

Senior Member
If you look at the circuit diagram you will see V2 goes to the on-board regulator input, its output goes to V1.

So you can remove the on-board regulator, take a wire from V2 to the off-board regulator input, take its output to V1, and connect its 0V to either V1 or V2 0V.
Thanks Hippy.
 

jensmith25

Senior Member
A TO-220 voltage regulator is rated for 1 amp. However, with no added heat sink that will get warm at around 200 mA.
So if the TO-220 LM7805 regulator is getting warm the then the TO-92 78L05 rated at only 100 mA will be overloaded and hence getting very hot. In fact the 78L05 is likely shutting down or cycling on and off.
Okay, I thought the regulator wasn't affected by the current of the outputs, or is this specific to the MP3 player where it is drawing current from the regulator?

It would be good to have a warning in the instructions for the CHI030 board if there is a problem with current with the 78L05 as that's the specific one you are told to use.
 

hippy

Technical Support
Staff member
Okay, I thought the regulator wasn't affected by the current of the outputs, or is this specific to the MP3 player where it is drawing current from the regulator?
All linear regulators are affected by current drawn. It is basically 'output current times voltage dropped' which has to be dissipated and that goes as heat. So 100mA drawn from a regulator dropping 12V to 5V would be 0.1 x (12-5) = 0.7W.

It would be good to have a warning in the instructions for the CHI030 board if there is a problem with current with the 78L05 as that's the specific one you are told to use.
The 78L05 is recommended for providing current to the PICAXE on the CHI030A board when powered by greater than 5V so does not take into account current drawn by other boards using 5V from the CHI030A board.

I would guess that was presumed to be general electronics knowledge and anyone using a regulator would know that its current rating should not be exceeded and a regulator capable of supporting the current being drawn should be used if more were required. Much like not drawing too much current out of an I/O pin.
 

jensmith25

Senior Member
All linear regulators are affected by current drawn. It is basically 'output current times voltage dropped' which has to be dissipated and that goes as heat. So 100mA drawn from a regulator dropping 12V to 5V would be 0.1 x (12-5) = 0.7W.



The 78L05 is recommended for providing current to the PICAXE on the CHI030A board when powered by greater than 5V so does not take into account current drawn by other boards using 5V from the CHI030A board.

I would guess that was presumed to be general electronics knowledge and anyone using a regulator would know that its current rating should not be exceeded and a regulator capable of supporting the current being drawn should be used if more were required. Much like not drawing too much current out of an I/O pin.

I knew the maximum of the regulator but the difference here is that another board was being connected. That obviously changes things to an extent that I wasn't aware of. I assumed, presumably incorrectly, that the MP3 player wasn't drawing that much current since there's no info on that in the MP3 instructions when it's being played and it was being powered off the AXE171 board quite happily off an I/O pin!

I guess I don't fully understand the difference between powering it off 4.5v and why it's suddenly a problem powering it off 12v.

EDIT: it looks like the MP3 player is drawing 21-22mA seen a high of 30mA? Measured on MP3 player. But it wouldn't play when the multimeter was connected in series. That doesn't seem like it should be a problem?

The regulator is warm after playing through the whole lot which is a few minutes. I could get one of these to be sure??: It's 3Ahttps://www.ebay.co.uk/itm/Raspberry-Pi-DC-DC-Converter-Regulator-12V-Stepdown-to-5V-3A-15W-Module-Power/142752974294?epid=1274078478&hash=item213cbd85d6:g:JXQAAOSw5cNYcSC7
 
Last edited:

hippy

Technical Support
Staff member
That obviously changes things to an extent that I wasn't aware of. I assumed, presumably incorrectly, that the MP3 player wasn't drawing that much current since there's no info on that in the MP3 instructions when it's being played and it was being powered off the AXE171 board quite happily off an I/O pin!
It is powered off the AXE171 supply rail, 5V or 4.5V. If you are powering it off an I/O pin on the CHI030A then that might also explain things.

It wasn't easy to tell where power was connected to in your earlier photo. I assumed it was to to the V1 connection on the CHI030A.

I guess I don't fully understand the difference between powering it off 4.5v and why it's suddenly a problem powering it off 12v.
In normal use there usually isn't really much difference. In one case it's getting its power from the batteries, in the other it's getting its power through the regulator.

Batteries will supply as much current as they are capable of supplying, regulators will supply as much as they are rated for but will get hot as that is approached and they may shut down when too much current is drawn or they overheat.

EDIT: it looks like the MP3 player is drawing 21-22mA seen a high of 30mA? Measured on MP3 player. But it wouldn't play when the multimeter was connected in series. That doesn't seem like it should be a problem?
Around 20mA-30mA would seem as expected for the setup when not playing. Not sure why it won't play with the meter in-line. I use a bench supply which shows current drawn so am not familiar with the pro's and con's of measuring in-line.

The regulator is warm after playing through the whole lot which is a few minutes. I could get one of these to be sure??: It's 3Ahttps://www.ebay.co.uk/itm/Raspberry-Pi-DC-DC-Converter-Regulator-12V-Stepdown-to-5V-3A-15W-Module-Power/142752974294?epid=1274078478&hash=item213cbd85d6:g:JXQAAOSw5cNYcSC7
That should work but switched mode regulators can sometimes put high-frequency ripple on the power rail which may interfere with audio circuits. There's no way to tell without trying it.

But there is probably no need. A linear regulator getting warm is entirely acceptable and to be expected. Getting too hot is the problem. Replacing a smaller TO-92 regulator with a TO-220 regulator is usually considered perfectly acceptable.
 

jensmith25

Senior Member
It is powered off the AXE171 supply rail, 5V or 4.5V. If you are powering it off an I/O pin on the CHI030A then that might also explain things.

It wasn't easy to tell where power was connected to in your earlier photo. I assumed it was to to the V1 connection on the CHI030A.
It's connected to the V+ and G of the strip of connection points between the Picaxe chip and the darlington driver chip. I didn't think that was off an I/O pin.

Okay, thanks Hippy. Will test thoroughly with the LM7805 regulator over the weekend and hopefully all will be well.
 

westaust55

Moderator
When using a linear voltage regulator, the greater the voltage difference/drop between the input and output the more heat is generated.

Let’s say you power the 5 Vdc regulator from 8Vdc and the total current is 200 mA (0.2 Amps).
The regulator losses as heat is:
Heat = (Vin - Vout) * Amps = (8 - 5) * 0.2 = 3 * 0.2 = 0.6 Watts

Now for the same load current but we Power the regulator from 12 Vdc
Heat = (12 - 5) * 0.2 = 1.4 Watts

So the extra input voltage is 50% higher but the voltage regulator must dissipate and extra 0.8 Watts which is 2.3 times the original heat dissipation.

Adding the MP3 module we do not know the power demand when playing a tune but there is indication it is around 30 mA when idle (not playing)
Let’s say the output to drive the speaker is 0.25 Watts.
That means an extra 0.25 / 5 = 0.05 Amps or 50 mA.
We add that to the idle current and the MP3 is 75 mA
It could be more
Now add that the the CHI-030 current load and you are likely very close the the 100 mA of the 78L05 rating hence it will get hot.

By comparison a 7805 regulator in TO-220 case will barely be discernibly warm.

Recommendation: change the 78L05 with a 7805 1 amp regulator in a TO-220 case and likely no heat sink will be required.
 

geezer88

Senior Member
Around 20mA-30mA would seem as expected for the setup when not playing. Not sure why it won't play with the meter in-line. I use a bench supply which shows current drawn so am not familiar with the pro's and con's of measuring in-line.
The effect of an in-line amp meter is the series resistance it adds to the circuit. Using my cheap Craftsman DVM, in the 20A range the resistance is only 0.4 ohms. The milliamp range adds 4 ohms, and the microamp range adds about 100 ohms. Whether this matters in a particular circuit will depend on many factors, but in all cases the effect will be to reduce the voltage to the circuit under test. This reduction will equal follow Ohm's Law. V=R*I

tom
 

premelec

Senior Member
Well noted tom... however have you tried 20 amps on that meter? 20 x 20 x .4 = 160 watts.. should keep your hand warm using it!
 

AllyCat

Senior Member
Hi,

in the 20A range the resistance is only 0.4 ohms.
I make that an 8 volt drop at 20A and a dissipation of 160 watts. :confused:

Sounds like it should release the magic smoke rather well. ;)

To endorse what others have said; Ohms Law and Power = Volts x Amps are the first things to consider in almost any electronics design.. When Power > 1 watt things might start to get quite warm (and many components are only good to around 250 mW). :(

Cheers, Alan.
 

geezer88

Senior Member
AllyCat and Premelec: thanks for pointing out what happens when I don't question a reading with a meter. If the 20A range really had that much resistance, you are totally right that magic smoke and more would result!

The error of my measurement was the test leads and connectors I used. My good meter was used to measure the cheap meter, but I was too lazy to do a four wire measurement, and while still in a lazy mood, I didn't "reasonableness check" the results.

I don't have time now, but in the next day or two, I'll take apart the cheap meter and properly measure and/or examine resistors for value.

thanks for the posts,
tom
 

hippy

Technical Support
Staff member
I believe one trick for high current reading meters is to provide a low resistance wired or tracked path from in to out capable of handling the huge amount of current, then measure the voltage drop across that, and that's proportionate to current.

As said, 20A through 0.4R, would be quite a lot of power. Probably okay for AC if it heats up as electrons go one way, cools down when they go the other :)
 

premelec

Senior Member
@Hippy - now a discourse on peak, RMS, etc current - and thermoelectric effects ["cools down when they go the other"] is in order... ;-0 perhaps comment on 'do electrons actually go anywhere or just bump on each other down the wire - or up the wire or in the wire' also... Thanks for all your sage remarks and humor!
 

geezer88

Senior Member
OK, I'm back. With better data. This time I loaded each range with full scale current, except the 20A range. For it, I only put 1 amp. The readings are the voltage drop across the meter as it is reading the indicated current. The voltage readings were done with my good meter, a Keithley 196.
Code:
Range     Current      Delta Volts      Calculated R
  uA         0.004          0.411            103
  mA         0.400          1.560            3.9
   A         1.00           0.250            0.25  on this range, the meter has a note on the face:  15sec in 30min
The note is referring to the issue raised. My meter would become a handwarmer if left measuring high currents for very long.

Anyway, my post was intended to warn folks that their measuring instruments can give them false readings, depending on the effect the meter has on the circuit under test. That concern is still valid.

tom
 
Last edited:
Top