Weird start up issue.

[JohnRoberts]

I recently rigged up a (cheap) 5.1 system for my living room. The decoder accepts light pipe input and 6 analog audio channels out.

My weirdness is if I start up the decoder using the switch on an outlet strip (also switching 3 power amps) the decoder goes into a locked up mode... All input LEDs lit.  Switching the outlet strip on and off again does nothing, but if I unplug the 5v power jack and plug it back in it boots up properly. The decoder does not have its own power switch.

My suspicion is that the external 5v ps (wall wart)  has a slow start up ramp or bounces that prevents the micro inside the decoder from  cleanly reseting (often the reset is generated from the rail voltage coming up sharply).

This is annoying but it is a cheap ass decoder.  It works reasonably well after I get it to properly boot up. It runs hot so I don't want to leave it running 24x7 but maybe I will, since it is too much work to mess with the power jack every time.

I wonder if a RxC on the 5V would help (or maybe just a large C)? It's a 1A PS so I might need a big ass cap... I doubt the decoder draws the full 1A but it does get pretty warm.



JR

PS: I could probably design a delayed start up circuit so the 5V snaps up, but I don't want to redesign everything.
 

 

[moamps]

My suspicion is that the external 5v ps (wall wart)  has a slow start up ramp or bounces that prevents the micro inside the decoder from  cleanly reseting (often the reset is generated from the rail voltage coming up sharply).

Why the "slow start up" works if you  pull the wall wart  off and plug it on, and doesn't work if you switch the strip off and on?
What's the difference the wall wart experiencing?

IMO, you should use a delayed main switch for the wall wart.

 

[bluebird]

I would try another 5V wall wart. maybe a 2 amp. or a 1 amp linear. I bet you could find one at a thrift store.

 

[JohnRoberts]

My suspicion is that the external 5v ps (wall wart)  has a slow start up ramp or bounces that prevents the micro inside the decoder from  cleanly reseting (often the reset is generated from the rail voltage coming up sharply).

Why the "slow start up" works if you  pull the wall wart  off and plug it on, and doesn't work if you switch the strip off and on?
What's the difference the wall wart experiencing?

The difference is what the decoder experiences.

If I try to leave everything plugged up and switch it on with the outlet strip, my speculation is that the 5V supply built into a small wall wart, doesn't come up in a fashion the decoder likes. 

OTOH if I unplug and replug the 5V PS jack into the back of the unit while the 5V supply is already up to a stable 5V, the decoder boots up and works properly. 

IMO, you should use a delayed main switch for the wall wart.

I do not expect changing how the mains power is applied will make any difference.

If my speculation is correct a cheap and simple solution is to wire a switch in series with the 5V power cord, and just leave the wall wart plugged into the mains, but that is one more switch to turn on and off.

I hate having to fix other peoples designs (but it could always be worse) .

JR

 

[sahib]

.................

My suspicion is that the external 5v ps (wall wart)  has a slow start up ramp or bounces that prevents the micro inside the decoder from  cleanly reseting (often the reset is generated from the rail voltage coming up sharply).
.................

My suspicion is not. I bet it is an ordinary, mill of the run psu even if it is smps.

The micro will lock-up when it is under power. When you plug in the psu you are essentially re-setting the micro as well as providing power.

Obviously fitting an on/off switch on the power supply is not possible due to lack of space. I would then put it into a larger plastic case and fit the switch.

Otherwise fit a capacitor on the  wall socket switch, fit another one on the mains input of the psu, and  finally fit something like 1uF poly on the psu DC output. If these do not help  the first idea of a larger psu case with a switch becomes more practical than designing/fitting a soft start.

 

[moamps]

OTOH if I unplug and replug the 5V PS jack into the back of the unit while the 5V supply is already up to a stable 5V, the decoder boots up and works properly. 

I understand now. You can try a small 5V SPST relay at the wall wart's output delayed (or not) with a RC network.

 

[JohnRoberts]

.................

My suspicion is that the external 5v ps (wall wart)  has a slow start up ramp or bounces that prevents the micro inside the decoder from  cleanly reseting (often the reset is generated from the rail voltage coming up sharply).
.................

My suspicion is not. I bet it is an ordinary, mill of the run psu even if it is smps.

Probably smps but that doesn't matter, what does matter is whatever the PS start up profile is, it causes the the surround processor to lock up.

The micro will lock-up when it is under power. When you plug in the psu you are essentially re-setting the micro as well as providing power.

When I un-plug are re-plug the 5V power jack with the 5V supply already stabilized, it resets and runs properly.

Obviously fitting an on/off switch on the power supply is not possible due to lack of space. I would then put it into a larger plastic case and fit the switch.

If I add the power switch it will be inline on the 5V power cord. 5V is not much of a shock hazard, so electrical tape should be adequate insulation, I think I have some shrink tubing I could use. While I have the cord open I may experiment with adding a cap across the 5V to see if that makes it start cleanly enough to work.

Otherwise fit a capacitor on the  wall socket switch, fit another one on the mains input of the psu, and  finally fit something like 1uF poly on the psu DC output. If these do not help  the first idea of a larger psu case with a switch becomes more practical than designing/fitting a soft start.

This is pure speculation informed by the fact it starts up properly if I plug the already stable 5V supply cable in. So if anything it needs to start up faster, and/or stay up after it starts going up.  I guess i could always carry this to my bench and put a scope on it (the 5V PS output).

JR

 

[JohnRoberts]

OTOH if I unplug and replug the 5V PS jack into the back of the unit while the 5V supply is already up to a stable 5V, the decoder boots up and works properly. 

I understand now. You can try a small 5V SPST relay at the wall wart's output delayed (or not) with a RC network.

I actually have some tiny SMD power mosfets around that might work. These don't require much gate voltage so 5V should turn them on adequately, and being high impedance delay is trivial. What are the odds I have the right polarity switch on hand? (No, the mosfets on hand are low single digit on resistance so too high Z for 1A supply.)

Back to the mechanical switch.

JR

 

[Gene Pink]

(also switching 3 power amps)

Charging all the caps in those amps at the same time, how far does it sag down that particular branch circuit in your house?

And for how long? Might be time to blow the dust off of the ol' trusty Simpson 260 analog meter, to find out.

With the amps turned off, does the gizmo boot up using the switch on the power strip?

Gene

 

[JohnRoberts]

(also switching 3 power amps)

Charging all the caps in those amps at the same time, how far does it sag down that particular branch circuit in your house?

And for how long? Might be time to blow the dust off of the ol' trusty Simpson 260 analog meter, to find out.

With the amps turned off, does the gizmo boot up using the switch on the power strip?

Gene

Thanx but no it still latches up... I still suspect the start up profile of the SMPS wall wart.  Might behave better with a decent cap on the 5V line inside the surround processor, but I am not going to waste any more time on this... for now it is just plugged into an un-switched outlet.  I don't believe it is really drawing 1A and if it wears out sooner , so be it.

JR

{edit- I sent an email to the manufacturer in china, sharing my diagnosis but since I don't speak mandarin I don't expect an answer.  maybe they can check their current PS for start up compatibility. [/edit]

 

[JohnRoberts]

I actually got a response back in better english than I speak chinese.

They sez "Our power output is fixed with 5V,generally U.S. voltage is 110V, so occasionally shortage of voltage occasionally short of voltage." 

I measure 119.8VAC  at my outlet.  8)

I may buy a new 5v PS just to test my theory... only  $6-7 for onseys.

For now I will just leave it plugged into an un-switched outlet.

JR

 

[mattamatta]

I have an eq that does a very similar thing - when powered off a Radio Shack switcher wall wart I can hit the front panel power button on the unit and it turns on fine, but if I leave the unit power on and turn on the power at the rack it goes into a fault state.  I don't believe there's any microcontroller action on there but I believe there's internal switched PSU architecture that has some of its own smarts (probably in a controller IC).  I haven't used it much, so I never looked into it any more.  It's possible I have an old downrev engineering version and I know it has its quirks.

 

[Gene Pink]

...but I believe there's internal switched PSU architecture that has some of its own smarts (probably in a controller IC).

Yeah, some of those switchers are a little too damn "smart" for their own good.

If there absolutely has to be a wall-wart involved, just gimme a hefty chunk of iron and copper that would make CJ proud.

Going "green/efficient" is generally a good idea, but not at the expense of functionality, where JR has to leave his gismo on all the time. Much like low-flow toilets, where's the efficiency, if you have to flush it three times before it will gobble down everything you threw at it?

Gene

 

[nashkato]

i guess you did have your switched outlet in mind for causing this ?
those outlet strips often enough don´t switch both neutral and live .
don´t know if this may be a cause .

 

[JohnRoberts]

Today my replacement wall wart PS arrived ($6.50 from digikey).

As I speculated the (cheap) 5.1 decoder now boots up properly when plugged into the same switched outlet as my amps using the new switching PS lump.

Another benefit of turning the decoder off when I am not watching movies is that the decoder lost sync (?) several times over the last few weeks.  Sound output would be low, distorted or missing from some channels. The signal would clean up (reset) if I changed the mode, or input selector. This may be another symptom from the faulty PS.  The voltage of the PS measures correct, but it may have noise in it, or some other flaws. We'll see if it ever loses sync with the new PS.

JR

 

[CJ]

they might have used a RC time constant scheme on Pin 1 reset, maybe a 3 legged voltage detector on pin 1 would clean up the reset voltage, could be a hassle splicing into micropitch SMD boards,

 

[JohnRoberts]

I guess I wasn't clear, the new PS fixed the start up problem so there is nothing left to fix.

JR

 

[JohnRoberts]

Spoke a little too soon...

The new PS fixed the start up problem but the 5.1 decoder got stupid (sync?) at least one time with the new improved PS. It clears up if I change inputs or mode but inconvenient. Clearly a cheap crap design (too cheap to work right all the time), but good enough to get my system sorted out.

I need to look for a used AV preamp that accepts the optical input with master volume control and other useful features.

JR