7805 trouble

[Resistanceisbutyl]

Hey folks,

I'm troubleshooting an industrial control set, there seems to be a power spike when the downstream of a 7805 is loaded, the voltage spikes from a stable 5v to almost 7 when the load is applied.

(1) The problem I am troubleshooting is erratic behavior in the microcontroller circuits downstream during and only during the load condition, it is reasonable to assume this may be the cause, correct?

(2) Is this spiking likely the fault of the 7805, or could it be the filter caps? They appear to be oversized since they are both 100uf.

Thanks!

 

[alec_t]

What load current? Resistive, inductive or capacitive load? Does the load share a common ground line with the micro?

 

[Mosaic]

Verify the 7805 centre pin ground. If it is weak then the current load will elevate the 5V ground reference locally resulting in the spike.
A simple continuity check will not tell. I t may just be simpler to hardwire a backup ground to the pin, or use a milliohmeter. If u have a 'scope u can dual trace the spike and the 7805 ground pin to see any correlation.

 

[WTP Pepper]

The 7805 has a series pass transistor that under normal regulation is forward biased. However depending on the input voltage and the speed of the transients caused by the load can go reverse biased and stop regulating. i.e. It can lift the supply. This can cause all sorts of problems to things like a micro relying on a steady supply voltage from the same source. A transient of 7v could kill a micro faster than you can blink.

To those amateurs who tried to contradict me previously as a possible cause of a regulator not regulating, here it is in action.

Transorbs could be an option but a proper over voltage current dump config would be better. You could also try regulating the micro off its own reg.

100u on a 7805 should be fine, but you are relying on the cap to cope with the dv/dt which may work once or twice but may fail over time and tolerance.

Designing regs that cope with transients on their output is more than dropping in a chip like a 7805.

 

[Mr RB]

...
...the voltage spikes from a stable 5v to almost 7 when the load is applied.

... during and only during the load condition ...

What is the load? It's possible it is nothing to do with the regulation, your load might be inductive and unfiltered and making positive spikes on the 5v rail.

Can you please show the full schematic or at least provide info on "the load"?

 

[ericgibbs]

What is the load? It's possible it is nothing to do with the regulation, your load might be inductive and unfiltered and making positive spikes on the 5v rail.

Can you please show the full schematic or at least provide info on "the load"?

hi,
I would agree with Roman, you should consider that a 7805 regulator must be used with adequate capacitance on its output in order to cope with transient currents draws by the load, which can momentarily exceed the current the rating of the 7805.

Likewise if the load has a switched inductive component the value chosen for the capacitor on the 7805 regulator output must able to absorb transient positive or negative overshoots on the regulated of the 5V rail.

To consider a 7805 in isolation without the appropriate associated capacitors that are required to suit the load conditions is meaningless.

E

 

[ben7]

Just to add to the conversation - what about bypass capacitors on the microcontroller itself? How far is this load away from the regulator? Wire gauge?

 

[dr pepper]

100u on the input sounds high, do you have the schem so you can check to see if the caps are original?
Presumable you are testing this on the bench, your setup might be affecting the unit, are you using your own dc supply or the units mains powered supply?

 

[alec_t]

I think we've lost the OP

 

[dr pepper]

Maybe he's holidays

 

[tvtech]

Hi Guys

Sticking my neck out here but anyway...think about this.

1. Probably 80% of EEPROM failures I have ever dealt with on TV's are attributed to "the bad bunch" from Atmel years ago...specifically 24CO8 series..
2. However, in particular, one specific chassis here fails regulary with this EEPROM. Even ST fails lately. (Thomson)
3. This chassis does things differently to others out there. If has it's own little Linear Transformer that alllows the SMPS to shut down when in Standby...and always provides 5V. Really nice reliable SMPS saving design. I do not believe I have ever fixed a blown SMPS on one of these sets. Ever...
4. Problem is, these particular sets eat EEPROM's........and guess what, they have a 7805 keeping them powered through thick and thin.


Now back to the original 7805 post. Yip, I believe the 7805 is the one damaging the EEPROM somehow when crap comes along. 7805 never blows or goes faulty. But, somehow, it lets a little spike or something through that the Chassis designers never catered for.

For example, we had a set come back last week. Same EEPROM eating chassis. Open it. EEPROM was replaced July 2012. Look at Job Card. Cannot possibly be EEPROM again...

Defying what my mind tells me...program another EEPROM. Set up and running. All good again.

Something to do with the perfect 5V Regulator. Customer not charged a cent.

Regards,
tvtech

 

[Legardo]

Hey folks,

I'm troubleshooting an industrial control set, there seems to be a power spike when the downstream of a 7805 is loaded, the voltage spikes from a stable 5v to almost 7 when the load is applied.

(1) The problem I am troubleshooting is erratic behavior in the microcontroller circuits downstream during and only during the load condition, it is reasonable to assume this may be the cause, correct?

(2) Is this spiking likely the fault of the 7805, or could it be the filter caps? They appear to be oversized since they are both 100uf.

Thanks!

 

[Legardo]

• It sounds some short circuits in microcontroller it draws too much current.

 

[Mosaic]

Leonardo, the thread is 3.5 years old.