4x4x4 Help.......

[blackzakk]

Hi,
I'm no electrical engineer, just a 65 year old taking up kit making. Bought my first 4x4x4 blue and it worked like a dream. So, bought the red and it worked except for 4 LEDs. Took out the LEDs and replaced them, after checking they lit up, and now the 4 white light up but the cube gives me nothing. Nothing gets hot, checked no positives touching negatives. Even took those 4 out to see if I could get back to where I was but nothing.
I'm sure I have done something really stupid and I've looked at the schematic like a child looking at the Hadron Collider.
Can somebody point me in the right direction? I've thoroughly enjoyed putting the cubes together and even sent off for the green one for the set. So, I don't want to abandon this one.
Hope you electronic wizards can help.
Stay safe.

[andrew]

Nice to hear you enjoyed your first kit and I'm sure we can get to the bottom of your current problem. We recently had another customer posting on our blog site with the same symptoms as you, i.e. base LEDs working but nothing else. For him it turned out to be the soldering on the crystal but I'll cut and past my suggestions for what to look for here:

4x4x4_Cube.jpg

>I’m not getting any blinking when testing LEDs. The feet do light up.

As mentioned above the microcontroller comes with the demo sketch pre-installed so if you’re not seeing any blinking LEDs then this would suggest that the microcontroller is not running for some reason. The fact that the base LEDs are on tells me the cube is at least getting power but there could be a number of reasons why the microcontroller may not be running. I would first suggest checking the following things…

1) With the cube powered up is either IC getting hot to the touch?

2) Can you double check the soldering for all the pins on the microcontroller (labelled U1). Especially pins 7, 8, 9, 10, 2 & 22, but also give the other pins a good check. Also make sure that none of the pins of the IC have bent under the chip rather than going through the pad.

3) Check the pins of the crystal X1 and the two 22p capacitors C4, & C5. Also try lifting the crystal slightly off the board by heating one pin at a time and carefully lifting each side slightly

I've highlighted in the above image which pins to check and in your case you can of course ignore step 1

[blackzakk]

Hi Andrew,
Thank you so much for responding. I reflowed the connections you indicated but still no luck. If the four white LEDs are on a separate circuit how would I know if current is flowing to the LEDs? I have put a meter on the cube at random points and getting a reading of 0.25 volts. I don't know if that helps at all.
Of course I could try hitting it with a hammer.
I have wondered, in my ignorance, if a component has failed, like a capacitor or crystal or that the IC chip needs reflashing with the demonstration sequence. If it is the IC chip then I'm done for as dimmer switches still are a mystery to me.

Terry

[andrew]

I re-flowed the connections you indicated but still no luck.

Did you try lifting the crystal (x1) off the PCB a little?

If the four white LEDs are on a separate circuit how would I know if current is flowing to the LEDs?

The LEDs are powered from several pins of IC U1. You can ignore these pins as there would have to be a fault on all these pins for none of the LED's to illuminate. The problem is most likely that the microcontroller U1 isn't running for some reason. However if lifting the crystal didn't work there is one other single point of failure that could cause them not to illuminate which is pin 9 of IC U2. Can you give that a check too. TBH it wouldn't hurt to reflow all the pins on both ICs and double check that none of the pins of these ICs have folded under the chip instead of going through the pad.

I have put a meter on the cube at random points....

If you have a volt meter there are a few places you can check. You should see ~ 5V across the following pins on IC U1:

1) pin 7 and pin 8

2) pin 1 and pin 8

3) pin 20 and pin 22

If you need help identifying any of these pins then let me know

I have wondered, in my ignorance, if a component has failed, like a capacitor or crystal or that the IC chip needs reflashing with the demonstration sequence.

I doubt it's a capacitor issue but again it doesn't hurt to double check the two 22pF capacitors C4 and C5. A fault on the other capacitors wouldn't stop it from working. The microcontroller comes pre-programmed and it can't lose its program without using a separate programmer to erase it so again its safe it ignore that as a possibility.

[blackzakk]

Hi Andrew,
Thanks for your patience.
I reflowed both chips and put a meter across the pins you said. I get 5v from both 7 & 8 and 1 & 8 but only 0.1 across 20 and 22.
I can see the tips of the pins on both chips and the solder has "wicked", and no balling, so I assume the joints are ok. I also lifted the crystal up from the pcb and resoldered it.
What I don't understand is how it was working and then nothing. I know when I first turned the cube on, prior to this buggering about, the LEDs would light but the sequence kept stopping and restarting. If I left it alone for a few minutes it would then go through the cycle.
I don't know if you can help anymore but if you have any ideas I'd be grateful.
Many thanks for your time.

Terry

[andrew]

I get 5v from both 7 & 8 and 1 & 8 but only 0.1 across 20 and 22.

Thanks for checking. You should defiantly see 5V across pins 20 and 22 if everythign is working properly. Could you check to see if you get 5V across pins 7 and 22? That should tell me if it's a bad connection to pin 20 or 22.

What I don't understand is how it was working and then nothing....

Yeah it sounds like something intermittent that was on the edge of working and then stopped. Most likeliest candidate is the crystal but its worth checking the pins above.

[blackzakk]

Hi,
If I connect pins 7 and 20 I get a reading of 3.4. What this means I have no idea but it might mean something to you.
Yesterday I took out each layer, reflowed each each LED and pcb pin connectors and testing them all and they all light up.
I also ordered, and received, the green cube. So, I switched out the crystal with the new one to see if that would perform better but still only the four white LEDs.
Hammer time ?????
As a complete amateur I can only think that the IC1 isn't sending out the commands to the LEDs to go through the sequence somehow. Is there any way I can check this with my meter?
Thanks again for your patience.

Tel ;o)

[andrew]

If I connect pins 7 and 20 I get a reading of 3.4. What this means I have no idea but it might mean something to you.

Odd, I was expecting you to see ether 5V or 0V. The only thing I can think of that would explain why your seeing anything other than that is if there is a bad joint on pin 22. Can you give that pad another check and just doubly make sure the pin of the IC went though the pad? You could give it a continuity check (assuming you meter has that option) between the pin and the actual pad.

So, I switched out the crystal with the new one to see if that would perform better but still only the four white LEDs.

Ok well that can be ruled out at least.

As a complete amateur I can only think that the IC1 isn't sending out the commands to the LEDs to go through the sequence somehow. Is there any way I can check this with my meter?

The reason why i'm inclined to ignore the pins that drive the LEDs is because there are 14 of them and several of those would have to have an issue for the whole cube to not do anything. It's more likely it's one single thing that stopping the whole chip from running (assuming it's that IC that's the problem) like the crystal (which we can rule out now), or power, which it could be as there's something odd going on with pin 22. Unfortunately the only way to check if the microcontroller is actually running would be to measure the crystal with an oscilloscope.

The only other possible thing on the microcontroller is the reset pin. I doubt it's that but you could check it by putting your meter across pins 1 and 8. You should again see ~5V.


One other thing you can check with your meter is pin 9 on IC U2. Again if your meter has a continuity option, with power disconnected from the cube you can check continuity between this pin and the GND pin on the 5 pin serial header to check that it's GND pin is ok.

Hammer time ?????

Hopefully not yet as there's still a few options options open to getting it up and running.

If you need help identifying any of the test point just let me know.

[blackzakk]

Hi,
Ok, so I reflowed the pins on IC2 and pins 20, 21 and 22. Pins 1 & 7 and 7 & 8 give 5v. 1 and 22 I get nothing, 20 and 22 nothing.
I then checked continuity between pin 9 IC2 and ground and we have a beep.. I also checked all other pins to ground on the same chip and nothing.
Pin 22 is definitely in the hole as I sucked out the solder to do it again.
I must be frustrating the crap out of you. I feel your pain as I lecture in maths and I know how I feel when they don't get calculus........
Many thanks, Andrew.

Tel ;o))

[andrew]

1 and 22 I get nothing

20 and 22 nothing.

Yeah you should defiantly be seeing 5V across those pins, In particular pin 1 on U1 is the reset pin and this should defiantly be at 5V. If it's at 0V then the microcontroller will be held in reset and won't even run.

Just to be sure can you check pin one again but with reference to a different ground point? So for example can you test the voltage between pin 1 and the GND pin on the serial header?

When I'm back in the office tomorrow I'll pull up the PCB layout to see if there is any one place where a broken track could cause the above symptoms.

I must be frustrating the crap out of you. I feel your pain as I lecture in maths and I know how I feel when they don't get calculus........

More confusing then frustrating! It's a relatively simple circuit so there's not many places for a problem to hide. But there's defiantly something going on with those pins, it's just figuring out what it is.