Struggling to power audio module

Thread Starter

Torbek

Joined Apr 19, 2019
61
Hi, I have a circuit with a ucontroller (16F1825) which controls a well known sd-card amplifier known as a DF player.
The battery supply is 4.2v max (li-ion battery) and at that voltage the DF player doesn't work well, if at all, despite the specs suggesting otherwise., and I have tried a few!

I added a small boost converter to boost the voltage to 5v for the DF player, and with this it works well.

So, controller is at battery voltage, DF player at 5v. I used a IRL540 to power the DF player on/off, as the standby current was unacceptable even when in sleep. Everything worked well on a breadboard. Now I have put it all on stripboard I am struggling to saturate the Fet I think, and I do not know why... well, I think I have an idea, but I am surprised it worked well on an old breadboard but not on stripboard! When I say saturate, when I measure the ground of the dfmodule switched on the meter is reading maybe 25mV above ground, so suspect noise is the issue between ground and data, perhaps from the boost reg.

I find that in order for the DFplayer to be turned on, the pic isn't quite saturating the transistor enough even at 4.2V. If I remove the output pin of the pic and connect the gate to +V the dfplayer is then able to respond to the pics comm request and turn on.. But with the meter, I am not measuring any significant difference between the pic pin and going straight to the supply.

Iknow switching the ground is not ideal, but not sure what else to try tbh! I have no P fets to try, and to be honest they never work as well as N-type in low side switching, I would rather not add to the circuit but realize I might have to?

The circuit has to be able to go in to standby with minimal current consumption this is why I want to switch the sound module.

Can someone please give me some guidance as to how I can reliably switch the sound module on/off with the difference in voltages between pic and dfmodule please?

1632067047654.png
 

ElectricSpidey

Joined Dec 2, 2017
2,757
The first thing I would look into is why it worked on the BB and not the stripboard, the usual answer there is a mistake in construction, but your bypass test probably eliminates this.

Have you considered powering both the PIC and the module from the booster?

And, of course there are all of the usual suspects...proper bypassing, proper ground paths...etc.
 

Thread Starter

Torbek

Joined Apr 19, 2019
61
ok folks will update tomorrow. I think it is simply the fet impedance, and the data to ground that is the main issue...

I think it is a good point about using the buck converter for the whole circuit, it doesn't seem to use much itself. It does generate noise but I think I need to craft a little pi filter to eliminate it entirely.

I have been careful to route ground and positive together and used lots of decoupling caps both large and small to no change...

There is no fault with the Veroboard layout, yes I had to make some changes but everything is routed as it should be.

I just wonder if the jumble of wires and other things meant the breadboard was right on the edge of operating....
 

MisterBill2

Joined Jan 23, 2018
18,167
Any ideas folks?
On some occasions the punched holes in the strip board leave a very skinny strip of copper that can ADD A REAL VOLTAGE DROP IF ANY CURRENT FLOWS. That could be why the strip board version does not work.
To locate that kind of voltage drop problem the too to use is a volt meter with a 0.6 volt full scale range. Then you can easily see each section's voltage drop.
At least once, in an experimental test setup, the problem was a high resistance in some clip leads.
 

Thread Starter

Torbek

Joined Apr 19, 2019
61
On some occasions the punched holes in the strip board leave a very skinny strip of copper that can ADD A REAL VOLTAGE DROP IF ANY CURRENT FLOWS. That could be why the strip board version does not work.
To locate that kind of voltage drop problem the too to use is a volt meter with a 0.6 volt full scale range. Then you can easily see each section's voltage drop.
At least once, in an experimental test setup, the problem was a high resistance in some clip leads.
I have tried coating the tracks in solder though to try to reduce voltage drops.

I think what I will try is boosting the pic up to 5v, and trying that. Not happy as it will consume more power but I am not sure what else to do!
 

MisterBill2

Joined Jan 23, 2018
18,167
I have tried coating the tracks in solder though to try to reduce voltage drops.

I think what I will try is boosting the pic up to 5v, and trying that. Not happy as it will consume more power but I am not sure what else to do!
Solder coating is not enough in many instances. You may need to jumper the strips that carry power. Consider that if the design circuit worked once, then it is not a circuit problem, it is a hardware problem.
 

Tonyr1084

Joined Sep 24, 2015
7,852
On some occasions the punched holes in the strip board leave a very skinny strip of copper that can ADD A REAL VOLTAGE DROP IF ANY CURRENT FLOWS. That could be why the strip board version does not work.
To locate that kind of voltage drop problem the too to use is a volt meter with a 0.6 volt full scale range. Then you can easily see each section's voltage drop.
At least once, in an experimental test setup, the problem was a high resistance in some clip leads.
Something I've seen in my own journey was use of the wrong flux and failure to clean it away. It was conductive to some extent and caused my project to fail. That was many years ago. I'm just throwing this out there in conjunction with what MisterBill2 is saying. There can be stray currents for numerous reasons.

The wife decided to deep clean the microwave oven. In doing so she managed to get grease to flow onto the PCB. When she plugged it back in - it didn't work. She thought she ruined it. I took the PCB out and flushed it with alcohol, not a recommended cleaning method, but it was all I had. Then with compressed air (yet another non-recommendable method) I blow-dried the PCB. Put it back into service and it's still working to this day. That was maybe 5 years ago ? ? ? In short, a clean board is a happy board.
 

William Ketel

Joined Jan 23, 2018
16
What is different from the breadboard that did work to this version that has a problem? That is what to look at first. And measuring the voltage drop at the ends of connections that should be at the same voltage is the way to locate voltage drops..
 
Top