How good are you at troubleshooting??

[JohnRoberts]

Preface... I believe I am above average at troubleshooting stuff, and this one has got the better of me.

Ok, a simple consumer gadget using a microcontroller to switch an outlet on and off at programmed times, a glorified lamp timer...

Interface is a small LCD screen with 4 front panel buttons for setting clock and program on/off times, A top on/off override button and a top slide switch to select between off auto mode, and a random auto mode that is +/- 15 minutes of programmed times.

Almost everything works.

The clock works and keeps reasonable time, The manual on/off switch switches the relay on/off, and with the front buttons I can program in 6 sets of on/off events.

But what doesn't work, is in either of the two active mode switch positions, no programmed on/off...

So I opened it up... kind of surprising how much stuff I found inside considering I bought this for <$13.00 at Walmart.

The power board was cheap single sided bakelite with through hole parts. A relay and even a back up battery, so it doesn't get stupid every time the power blinks off.. then there is another SS board for the front buttons, the processor sits on a small 4 layer SMD PCB but the processor  is direct bonded to the PCB under a lump of epoxy.. but I can probe the pins that fan out to relatively large by comparison 44 pin square SMD package where a normal micro would fit.

Since everything works but the modes, the mode switch is the obvious candidate to check. It's a single pole triple throw, with the wiper tied to ground, and selectively shorting one throw at a time... I trace these three switch throws to three consecutive input lines on the micro controller.  As expected the grounded pin is at 0V and the other two are sitting up at 2.5V the PS voltage for this puppy. In each position of the mode switch only one of the three lines is held low and the other two at logic high.

So as far as I can confirm the controller is running fine..the clock sets and keeps good time, it also accepts and displays 6 sets of program times. The manual button works, which also confirms that the micro controller is able to talk to the relay switching the outlet power on and off when it wants to.

So my list of possible faults is pretty slender.. Since I don't see any pull up resistors, I am ASSuming that the logic pull ups are built inside the micro controller, which confirms also that the switch logic is making good connection to the controller.

Short list also very unlikely is faulty program code inside controller, or damaged input ports on the controller. Actually two input ports need to be damaged since it has two different on modes and neither works.

When I asked the company who made these what they through the problem was they answered, me opening it up. 

At least they have a sense of humor.... I told them if it worked I wouldn't be opening it up....  8)

OK, I'm open for suggestions, I am truly stumped on this one... It's simple, everything looks like it should work and most is, but it doesn't do the one thing it is supposed to do.

JR

 

[mnats]

Have you tried removing the backup battery, essentially performing a reset of the micro?

 

[keefaz]

Maybe it works  I mean some timers are not super intuitive to operate
See first review :
http://www.amazon.com/Intermatic-DT620-Heavy-Indoor-Digital/dp/B004TGO6RY/ref=pd_cp_hi_3

 

[12volts]

Maybe it works  I mean some timers are not super intuitive to operate
See first review :
http://www.amazon.com/Intermatic-DT620-Heavy-Indoor-Digital/dp/B004TGO6RY/ref=pd_cp_hi_3

Beat me to it, though another timer I once tried.


Frank

 

[MagnetoSound]

I've got nothing useful to add here, just a question ...

- why are you spending time on this instead of just buying another timer??

 

[zayance]

Why Waist if its reparable, or even think its faulty.....

 

[JohnRoberts]

Thanx for the feedback.

The unit has a reset button, so I reset it at least once... I expect pulling the battery would be redundant.

+1 to the possibility of secret handshakes to make it work.

The programming seems logical enough. 6 possible events, with separate on time and off time  for each event.  The programs are persistent, I can revisit and confirm what they are set for.

The instructions even go so far as to give an example for overlapping program times, it turns on at the earlier time in overlapping segment and off at the earlier time. This is logical, getting a turn on program command when already turned on is ignored, and turning it off when already off is ignored.

Now for yet another puzzler, I bought this to turn my electric blanket on at night before I retire, and  off again early the next morning. After giving up on getting it to work properly, I left it plugged in and used the manual on/off to switch my blanket. Last night inexplicably, when I went back to turn it on, it was already on... It had actually worked, and this morning it turned off give or take on time.. It was in random mode so that was even correct +/- 15 min...

So I programmed another turn on time for a half hour later and that worked... still in random mode, I switched it to normal exact time mode, and it didn't turn off...

So now this is maybe working but only in random mode... arghhh

This is still far from working 100% as described in the instructions.

We'll see... It seems to do better if I don't mess with it. At least it is doing something.

JR

PS: The reason I just didn't get another one is the store is a 50 mile round trip, and I am still not completely certain there isn't some operator error, or software quirks involved. Still not certain. 

 

[MagnetoSound]

PS: The reason I just didn't get another one is the store is a 50 mile round trip, and I am still not completely certain there isn't some operator error, or software quirks involved. Still not certain.

Well that's totally understandable - and you know I wasn't being rude, just asking.  8)


(So it works in random mode, huh? Maybe you could rig it to, uh ... I dunno, like, ring a bell or something when it turns the blanket on so you know how soon to go to bed?  ;D )

 

[Rochey]

Why Waist if its reparable, or even think its faulty.....

If it'll take longer to repair (your hourly cost * repair time) , than it'll cost new - and if there isn't much to learn from fixing it - replace with new, and strip the old parts from it and use them for DIY!.

E.g. I have a broken DVD player at home that I plan to take all buttons and PSU from

 

[JohnRoberts]

As a manufacturer I have an interest in seeing how inexpensively products can be designed and manufactured.

Yes, after a few more trials I have confirmed that random mode is working and normal mode is not.  Random mode is +/-15 minutes, so eminently usable for my application.  I am still a little disappointed that my VOM measurements showed the mode selector switch providing all 3 correct switch closures to ground, but operational evidence suggests that the switch is not working after the case is closed up, or the controller is not responsive to that one input line. 

On the subject of cost effective manufacturing..  they bond the controller chip directly to the PCB which is apparently cheaper than buying the controller bonded inside a larger SMD package and then soldering that to the PCB. The PCB has a footprint for the normal sized 44pin SMD package so that was how they developed it and perhaps support lower volume manufacturing.

the relay to switch power doesn't seem that cheap but I guess it was,,, 4 layer PCB seems like a lot, but the bottom of the PCB interfaces directly to the LCD so probably only way to get it routed in that small of a footprint.

But I've learned all I am going to learn.... time to move on...  One final amusing tidbit, the product warranty is for the "product lifetime" .... WTF is the product lifetime?  I warranty this product for as long as it is working? After that you're on your own.  ;D.. I suspect they are counting on consumers not bothering with pursuing warranty claims for cheapo products.

JR

 

[mnats]

E.g. I have a broken DVD player at home that I plan to take all buttons and PSU from

I had a cheap DVD player that stopped reading discs, so I did the same - stripped out the power supply and some switches/buttons thinking I could use them down the road. Later I found that one rail of the switch mode supply was faulty so my hasty assumption that there was something wrong with the laser/spindle motor was incorrect.

Electric blankets are the thing that got me into messing around with microcontrollers. I thought building a timer with a nice display would be a fun project. Then one day I saw a set of two mechanical timers at Ikea for $6.99. No regrets though - micros are a lot of fun.

 

[PRR]

> operational evidence suggests that the switch is not working after the case is closed up

So jump-solder the one mode you want.

I've been having such bad luck with electric blankets that I put a 240V 1KW electric heater in my sleep cubby. But powered from 120V... hardly gets warm, but that's warm enough. And it does not creak/clank as lectric baseboard is prone to do. Then for temp and time I use a better-grade baseboard thermostat. This one modulates 25%-50%-75%-100% so when it is just a degree cold it simmers hardly-warm, but will kick-up 100% with more temp difference.

The downside is an un-lit low-contrast LCD and very clumsy program sequence, also 2+5 day mode. (All days are the same to me.) I have the M-F nights working but weekends need a re-re-programming which is just not worth the squint and press-press-press-press.

> How good are you at troubleshooting?? 

Sometimes not good. My plow-truck stalled-out. I spent a LOT of time replacing all the fuel lines, because they broke when I went to replace the fuel filter. Then drop the tank to replace the fuel pump. Whole sender unit too ugly to re-use. Put it tgether, won't run, and gas gauge interacts with pump cycle. Pull tank out again. (BIG tank.) Works fine on ground. Put it in, and now it won't crank or power the gauges. Spent much time finding a blob of tape hiding a circuit breaker which had been criminally abused and taped into temporary function. Replace that, it beeps like 5V and teedles away... bad connection? Ignition switch??

I was at this, off and on, for months.

Jimmy came over and in an hour broke a fusable link much too easily. Had turned to green powder inside. Patched that, and it ran.

YAY!

But some trouble is just built-in.

> I have an interest in seeing how inexpensively products can be designed and manufactured.

There are an ASTOUNDING lot of wires in a 1991 Chevy K2500 truck. At least six wires to work a possibly pointless clutch in the front axle. The main bundle across the firewall is well over an inch diameter and tightly packed. On a 1941 Plymouth the starter is engaged by foot, on a 1966 Chevy by a heavy solenoid/relay, but this truck has an ADDED relay to drive the starter solenoid.

As best I can guess, some junior draftsman in GM was paid for every wire he put on the blueprints, and he cashed-in.

 

[JohnRoberts]

> operational evidence suggests that the switch is not working after the case is closed up (or the processor line is non responsive)

So jump-solder the one mode you want.

I have conflicting evidence, since the switch appears to work properly when box is open.

If it stops working in random mode too, that is an obvious option to pursue.

I've been having such bad luck with electric blankets that I put a 240V 1KW electric heater in my sleep cubby. But powered from 120V... hardly gets warm, but that's warm enough. And it does not creak/clank as lectric baseboard is prone to do. Then for temp and time I use a better-grade baseboard thermostat. This one modulates 25%-50%-75%-100% so when it is just a degree cold it simmers hardly-warm, but will kick-up 100% with more temp difference.

500W is not nothing...  You could stick an incandescent light bulb under your bed, but the light may be a bother.

The downside is an un-lit low-contrast LCD and very clumsy program sequence, also 2+5 day mode. (All days are the same to me.) I have the M-F nights working but weekends need a re-re-programming which is just not worth the squint and press-press-press-press.

> How good are you at troubleshooting?? 

Sometimes not good. My plow-truck stalled-out. I spent a LOT of time replacing all the fuel lines, because they broke when I went to replace the fuel filter. Then drop the tank to replace the fuel pump. Whole sender unit too ugly to re-use. Put it tgether, won't run, and gas gauge interacts with pump cycle. Pull tank out again. (BIG tank.) Works fine on ground. Put it in, and now it won't crank or power the gauges. Spent much time finding a blob of tape hiding a circuit breaker which had been criminally abused and taped into temporary function. Replace that, it beeps like 5V and teedles away... bad connection? Ignition switch??

I was at this, off and on, for months.

Jimmy came over and in an hour broke a fusable link much too easily. Had turned to green powder inside. Patched that, and it ran.

YAY!

But some trouble is just built-in.

Cars were much easier to troubleshoot before solid state electronics and miniaturization. Nowadays too many black box potted lumps to ruin your day.

I have a bad wire connection somewhere in my near 15 YO ride's closed door/window/trunk sensor circuit, that causes my horn to go off randomly if I lock my doors. The unintended consequence of this built in security feature is now I can't lock my car.  Since this is intermittent preferring to go off in the middle of the night, it is not trivial to debug. 

> I have an interest in seeing how inexpensively products can be designed and manufactured.

There are an ASTOUNDING lot of wires in a 1991 Chevy K2500 truck. At least six wires to work a possibly pointless clutch in the front axle. The main bundle across the firewall is well over an inch diameter and tightly packed. On a 1941 Plymouth the starter is engaged by foot, on a 1966 Chevy by a heavy solenoid/relay, but this truck has an ADDED relay to drive the starter solenoid.

Cars were the original sharp pencil exercise (even before TV sets), but they buy wire by the trainload so can be generous.

As best I can guess, some junior draftsman in GM was paid for every wire he put on the blueprints, and he cashed-in.

;D  I inherited one product at my last day job that had $20 worth of amphenol connectors inside it...  So remarkable that the amp sales rep was aware and complained (with tongue in cheek) when I repackaged it to use a small fraction. The not quite engineer who was responsible for the original design quit, right around when that product hit the production floor. I could see why. I carved a ton of cost out of that product and still don't think we ever made a profit on it.

JR

 

[Emperor-TK]

One final amusing tidbit, the product warranty is for the "product lifetime" .... WTF is the product lifetime?  I warranty this product for as long as it is working?

JR

I think that means they will replace it as long as it's an active product.  They aren't going to warehouse parts and replacements to fulfill warranty claims once the manufacturing run has been sold out.  They also let themselves off the hook for equivalent replacements (the new improved version) in the future in case they stop making timers.

 

[JohnRoberts]

One final amusing tidbit, the product warranty is for the "product lifetime" .... WTF is the product lifetime?  I warranty this product for as long as it is working?

JR

I think that means they will replace it as long as it's an active product.  They aren't going to warehouse parts and replacements to fulfill warranty claims once the manufacturing run has been sold out.  They also let themselves off the hook for equivalent replacements (the new improved version) in the future in case they stop making timers.

Yup, and what is the life cycle for such cheap blow through products a year, a season?  My take is they are counting on the typical consumer not bothering to save the warranty or make claims after very long in use. I am only bothering them myself because it malfunctioned when brand new. I suspect the failed one I am replacing had a similar warranty. 

I just bought a fairly expensive coffee roaster and it only has a one year warranty, but I know the company supports the product with repair parts, and the new product, is 90% identical to the old version I bought 10 years ago so I expect similar reliability.  I just bought a replacement gear  ($4) for an Italian coffee grinder that stopped working.  That was a pleasant surprise. Not that it stripped a gear, but that I could get repair parts reasonably. 

i recall back in the late '70s I had a bad bearing on the little plastic widget that centers the floppy disk in my dual 8" floppy drive... The replacement plastic part, cost me something like $150  rather painful as I recall. I originally ordered two of them to have a spare in case the other side broke, but not after seeing the IMO usurious price.

I find their choice of words amusingly vague, not what you want in a contractual agreement, but they know they will probably not have to defend that in court. 

JR

 

[PRR]

> 500W

V^2. 1KW @ 240V is, linearly, 250W at 120V.

OK, the 1KW rating is hot-enough for resistance rise, so it will draw more than 1/4 watts at 1/2 voltage. But 500W is constant-current.....

> too many black box potted lumps

Half or more of those lumps never fail.

But every lump has 1 to 27 wires. Each wire has one to four+ series connectors. They all live outside 24/7. The ones "inside" the car get soaked in morning dew and roasted in afternoon sun.

Not to mention the 2002 Accord's tank-vapor system. It's hidden under the rear floorboard, with solenoid in the splash from the rear tire. As soon as I found it, I thought that was a bad place. 9 years later my thought was reality. Not only had the solenoid rusted-up (and the ECU can tell), its screws went into brass inserts (they expected a problem) and those screws were too corroded to remove. This is normal. Stock advice on Honda forums is to buy a tank with a solenoid because they usually won't come apart.

The minivan had intermittent ABS errors. New sensor not a long-term fix. For a while I was fixing by pulling the connector, glaring into it, and putting it together. Still erred in a few months, same as a new $160 sensor. Pulled the dashboard to look into it....

> they buy wire by the trainload so can be generous.

The Honda uses a lot of THIN wire, wire smaller than I would use in a stomp-box. The plastic is pretty tough, but I don't doubt there is some fatigue-cracked copper held in semi-contact by the plastic.

OTOH there is a 1990 Honda in Maine which just turned one million miles.

> Cars were the original sharp pencil exercise

One could argue "sewing machines" after the patent-trust expired.

Bro's 1941 Plymouth is remarkably simple, mostly. Frame-body is very few pieces. Engine seems to have just one "excess" part, a water-tube, which apparently _was_ needed because the previous version tended to under-cool the back cylinders. Only over-designed system seems to be the brake-clutch pedal assembly... 1964 Falcon did the same function with a dozen fewer parts.

 

[JohnRoberts]

> 500W

V^2. 1KW @ 240V is, linearly, 250W at 120V.

OK, the 1KW rating is hot-enough for resistance rise, so it will draw more than 1/4 watts at 1/2 voltage. But 500W is constant-current.....

:-[ :-[ :-[  oops yes  1/2 V * 1/2 I= 1/4 W.

====
> they buy wire by the trainload so can be generous.

The Honda uses a lot of THIN wire, wire smaller than I would use in a stomp-box. The plastic is pretty tough, but I don't doubt there is some fatigue-cracked copper held in semi-contact by the plastic.

Yup, seen lots of that "all hat- no cattle" wire inside cheap japanese,chinese,korean, whatever made appliances. Copper cost more than vinyl. 

Some audio cables are seriously lacking in actual wire inside a fat jacket.

OTOH there is a 1990 Honda in Maine which just turned one million miles.

> Cars were the original sharp pencil exercise

One could argue "sewing machines" after the patent-trust expired.

I'll concede that, .

Bro's 1941 Plymouth is remarkably simple, mostly. Frame-body is very few pieces. Engine seems to have just one "excess" part, a water-tube, which apparently _was_ needed because the previous version tended to under-cool the back cylinders. Only over-designed system seems to be the brake-clutch pedal assembly... 1964 Falcon did the same function with a dozen fewer parts.

Yes, old cars easier to wrench... small motors with magnetos, even easier.. got spark, got fuel, it should run... (at least back before ethanol).


JR

 

[PRR]

After all the "fun" with the 1991 Chevy P/U's mystery wires (nother half-hour today re-splicing and re-bundling), I am shocked that I got all eight spark-plugs out and replaced in under 2 hours. And no blood!

Had a '79 T-Bird this was a 3-hour 2-day chore. (Had to stop after one bank and let the knuckles stop dripping.)

Some cleverness needed. Lacking magic boot-pliers, I wrapped clothesline around the plug boot, fed it out between fender and frame, past tire, and yanked.

 

[bruce0]

I had a Honda Accord (1992?) Electric everything, AC etc.

It was a great car, but the air conditioning would never go on when my wife was driving.

We could never figure it out, nor could the dealer.  Years later, I traded it in, and as I was cleaning out the odds and ends I had left in the car, the problem became apparent.

Any guesses?

 

[JohnRoberts]

After all the "fun" with the 1991 Chevy P/U's mystery wires (nother half-hour today re-splicing and re-bundling), I am shocked that I got all eight spark-plugs out and replaced in under 2 hours. And no blood!

Had a '79 T-Bird this was a 3-hour 2-day chore. (Had to stop after one bank and let the knuckles stop dripping.)

Some cleverness needed. Lacking magic boot-pliers, I wrapped clothesline around the plug boot, fed it out between fender and frame, past tire, and yanked.

The alpine tiger, Maxwell Smart's ride (in the original TV show), had a 260ci v8 shoe horned into that too small engine room, and IIRC you had to remove the heater core to get at the back spark plug on one side.

My current ride (97 mustang cobra) when the starter motor crapped out, I decided to let the dealer deal with it. As i understand you have to unbolt the motor mount and jack up the motor, may have to drop the header on that side too... too much fun for me to do in my driveway without a lift.

When I drove it up to the dealer and told them my starter motor was bad, they gave me that "your only a dumb-ass customer" look and told me to turn it off so they could measure the battery.  It was parked on a gentle slope so I said OK...  I was able to jump start it again after they confirmed my starter motor was bad.  A few days later when I came to pick it up, the service manger couldn't find my keys on the board, because some kid working in the parts department apparently still had them.. hmmmm  Another reason I don't like leaving my car at the dealer. My old '93 had a special ignition (to tweak timing when the blower was making a lot of boost) where I could plug in a low RPM rev limiter chip, to keep any service dept pukes from having too much fun. Hard to get into too much trouble if it rev limits at 2500 RPM.

JR