So I blew a 7805, no big deal?

Started by camel85kv November 17, 2003
Hello,

I've been working on a circuit that uses the BasicX with these:

Accelerometer (measure pulse width)
Angular Rate Gyro (measure analog voltage)
Pot (measure analog voltage)
Dual I2C motor drivers (send I2C signals)

My program was working fine until last week when my voltage regulator
died. I was running a 7805 (A 5v regulator, for the less informed
amoung you) from a 12v source, making it dissipate way too much
heat. My program stopped working, so I checked around... not
surprisingly the 7805 burned my finger and was putting out a whopping
2.35 volts. Keep in mind the BasicX is set up to use its own
regulator. The 7805 supplied power to the sensors, NOT the BasicX
itself.

So I replace the 7805 and all should be well... but no. The sensors
work fine according to my voltmeter, but the ADC and pulse measuring
functions of the BasicX are jittering. For example, when I read the
value of the pot without moving it at all, the BasicX reports numbers
like:

.615835
.611925
.607038
.615835
.610948
.613880

So what's going on? Did the bad 7805 spike and kill something? Is
the onboard voltage regulator bad? Is the crystal shot? Could it be
a power supply issue? Any suggestions at this point would be great!
Thanks guys.

-Kyle Vogt




camel85kv wrote:
So what's going on? Did the bad 7805 spike and kill something? Is
the onboard voltage regulator bad? Is the crystal shot? Unlikely, but possible. Usually, a power supply problem will kill the
BX24. I'd test the bx24 with a pot accross the 5v supply and NOTHING
else connected. Write a simple program to look at one A/D pin at a time,
and see if the jitter is real or not.
camel85kv wrote:

> Hello,
>
> I've been working on a circuit that uses the BasicX with these:
>
> Accelerometer (measure pulse width)
> Angular Rate Gyro (measure analog voltage)
> Pot (measure analog voltage)
> Dual I2C motor drivers (send I2C signals)
>
> My program was working fine until last week when my voltage regulator
> died. I was running a 7805 (A 5v regulator, for the less informed
> amoung you) from a 12v source, making it dissipate way too much
> heat. My program stopped working, so I checked around... not
> surprisingly the 7805 burned my finger and was putting out a whopping
> 2.35 volts. Keep in mind the BasicX is set up to use its own
> regulator. The 7805 supplied power to the sensors, NOT the BasicX
> itself.
>
> So I replace the 7805 and all should be well... but no. The sensors
> work fine according to my voltmeter, but the ADC and pulse measuring
> functions of the BasicX are jittering. For example, when I read the
> value of the pot without moving it at all, the BasicX reports numbers
> like:
>
> .615835
> .611925
> .607038
> .615835
> .610948
> .613880
>
> So what's going on? Did the bad 7805 spike and kill something? Is
> the onboard voltage regulator bad? Is the crystal shot? Could it be
> a power supply issue? Any suggestions at this point would be great!
> Thanks guys.
>
> -Kyle Vogt > <http://www.netflix.com/Default?mqso`178356&partidA16732" target="_blank" rel="nofollow">http://rd.yahoo.com/SIGcbp4rqn/M&7637.4116732.5333197.1261774/D=egroupweb/S06554205:HM/EXP69192238/A53619/R=0/*http://www.netflix.com/Default?mqso`178356&partidA16732 >
> ">http://docs.yahoo.com/info/terms/>.



> So what's going on?

Is the new 7805 oscillating? Do you bypass both sides? Tom Becker
--... ...--
www.RighTime.com
The RighTime Clock Company, Inc., Cape Coral, Florida USA
+1239 540 5700



It's great to see suggestions coming in!

Tom, when you say "do you bypass both sides" I'm assuming you mean
put a capacitor from power to ground both before and after the
voltage regulator. No, I have not done this, but the circuit worked
fine before without these capacitors. I will try adding them anyway.

Neil, I am unable to disconnect anything but the gyro, because this
is built onto a pcb.

I can see how the ADC may be jittering, but I just really don't
understand how the pulse measurement capability of the bx-24 would be
messed up unless the clock is unstable. Would an unstable power
supply do this? If this problem continues much more, I suppose I
could unsolder the bx-24 and proto this circuit until I find the
problem. But I'd rather not!

-Kyle Vogt



> ... I will try adding them anyway...

It's certainly true that a regulator often works perfectly well without
capacitors surrounding it but good engineering practice usually puts a
small smoothing electrolytic immediately upstream and a smaller
transient bypass cap immediately downstream of the IC. For 7805s, as I
recall, something on the order of 0.5uF or larger upstream, and 0.1uF or
so downstream, should be sufficient. The smoothing cap can be larger,
but the transient cap should remain small.

You can skip the upstream cap if the power supply filter cap is nearby,
but if you're running leads from a bench supply or a wall wart, you
should add it.

Another possibility: could you have also partially smoked your power
supply? Tom Becker
--... ...--
www.RighTime.com
The RighTime Clock Company, Inc., Cape Coral, Florida USA
+1239 540 5700




> Another possibility: could you have also partially smoked your power
> supply?

I'm using a brand new 9v battery to eliminate that possibility right
now.

Turns out that a few extra caps like Tom suggested fixed the
problem. Stupid mistake on my part, oh well. The ADC now outputs
steady numbers. Yay! Thanks to Tom and Neil for the help!

-Kyle Vogt


I know this is probably unrelated except when it comes to
filtering. For example, I know sometimes in IC design you add a
small value cap (like .1uf) across the gnd/vcc pins of an IC. Is it
ever necessary to add thise to a bx24?

Thanks,
j32768

"Tom Becker" wrote:
>
> It's certainly true that a regulator often works perfectly well
without
> capacitors surrounding it but good engineering practice usually
puts a
> small smoothing electrolytic immediately upstream and a smaller
> transient bypass cap immediately downstream of the IC. For 7805s,
as I
> recall, something on the order of 0.5uF or larger upstream, and
0.1uF or
> so downstream, should be sufficient. The smoothing cap can be
larger,
> but the transient cap should remain small.
>
> You can skip the upstream cap if the power supply filter cap is
nearby,
> but if you're running leads from a bench supply or a wall wart, you
> should add it.
>
> Another possibility: could you have also partially smoked your
power
> supply? > Tom Becker
> --... ...--
> GTBecker@R... GTBecker@S... www.RighTime.com
> The RighTime Clock Company, Inc., Cape Coral, Florida USA
> +1239 540 5700





> ... sometimes in IC design you add a small value cap (like .1uf)
across the gnd/vcc pins of an IC. Is it ever necessary to add these to
a bx24?

Depending upon one's view, a small capacitor from an IC's power pin to
ground serves:

- to provide a fast reservoir of current to supply to the IC when it
briefly demands high current due to internal switching, and/or

- to provide a low-impedance path for fast noise spikes - glitches that
the IC generates when internally switching - to ground before they have
a chance to propagate along the power bus to other circuitry where they
can cause problems.

To help keep the power bus clean, bypass capacitors are often liberally
applied; some IC sockets, in fact, have bypass capacitors built in. The
BX-24's pin 24 feeds a small onboard regulator whose output is already
bypassed so it is probably not necessary to add one. Tom Becker
--... ...--
www.RighTime.com
The RighTime Clock Company, Inc., Cape Coral, Florida USA
+1239 540 5700



I do anyway

Tom Becker wrote:

> > ... sometimes in IC design you add a small value cap (like .1uf)
> across the gnd/vcc pins of an IC. Is it ever necessary to add these to
> a bx24?
>
> Depending upon one's view, a small capacitor from an IC's power pin to
> ground serves:
>
> - to provide a fast reservoir of current to supply to the IC when it
> briefly demands high current due to internal switching, and/or
>
> - to provide a low-impedance path for fast noise spikes - glitches that
> the IC generates when internally switching - to ground before they have
> a chance to propagate along the power bus to other circuitry where they
> can cause problems.
>
> To help keep the power bus clean, bypass capacitors are often liberally
> applied; some IC sockets, in fact, have bypass capacitors built in. The
> BX-24's pin 24 feeds a small onboard regulator whose output is already
> bypassed so it is probably not necessary to add one. > Tom Becker
> --... ...--
> www.RighTime.com
> The RighTime Clock Company, Inc., Cape Coral, Florida USA
> +1239 540 5700 >
> <http://www.netflix.com/Default?mqso`178338&partidA16730" target="_blank" rel="nofollow">http://rd.yahoo.com/SIGcrtgg29/M&7637.4116730.5333196.1261774/D=egroupweb/S06554205:HM/EXP69272775/A53618/R=0/*http://www.netflix.com/Default?mqso`178338&partidA16730 >
> ">http://docs.yahoo.com/info/terms/>.