Thread: Relay dropping out momentarily

I discovered the other day that output 2 doesn't work. At one point it did. After some testing, I've discovered that outputs 2, 3, 4, 8, 9, and 12 also don't work. At the moment this isn't serious because I'm only using 2 outputs. I'd like to use 2 more in the future, and that's ok too. But I'm limited on my room for expansion now. Of course that won't be a problem if I can't get this clicking relay problem solved!!!

Tim, what are you using to power the brain?

I've decided against trying battery power, because using batteries is not practical for use in my home. I need to power it with an AC/DC adapter. I tried 3 more wall warts today.

The ratings for all of them are 120VAC 60Hz input, and the output as follows, as well as what I measured for voltage with a DMM:
12VDC 1.5A -- measured 12.6 VDC
12VDC 2.0A -- measured 11.5 VDC
12VDC 800mA -- measured 16.4 VDC
12VDC 300mA -- measured 16.2 VDC -- this is the type of wall wart where you select your voltage and polarity. on the 9V setting the DMM gives me 12.5VDC. However, on the 9V setting, using outputs 13-16, I don't have enough power to pull the relays in but they worked when I flipped the switch to 12V. All other outputs work (i.e. the relays pull in) (other than the outputs I've already mentioned don't work) on the 9V setting. I'm only using three relays and have been moving them around from output to output so I can rule out the relays as the problem.

All outputs are centre positive.

The 16+ volts should be ok for the brain right? I see on the website it should be able to handle 17V and I realize that I am approaching it. Could that be why my outputs are all messed up? The LEDs for 9 and 12 are on all the time except when something is plugged into them. The LEDs on the outputs that work light up properly when they're turned on, none of the other outputs light other than 9 and 12 which are on full time.

I'm definitely having quite the array of problems. I'm really disappointed. I'm just amazed that of the 5 different wall warts that i've used, I have the same problem with all of them. Seriously, this thing should be able to work with a 12V adapter.

After checking out the website and reading about it, as well as all the good reviews on this forum, I bought one. I really like the idea, I like the convenience, I like the simplicity, I like the programming, and the support is great when you need help with something.

Has anybody had success with running digital outputs off a brain getting 12V from a wall wart?

All brains (including the one you had originally and the replacement) are tested using a 12v wall wart and 16 relays attached. Every one passes with all ports working and stably before being bagged as a good brain. If any port doesnt work, it is semi-defective. If it is unstable or doesnt flash, it is comatose.

I have about 5 different FBs running around my house for various things right now all being powered off of standard 12v wall warts. Another powered off of the 12v rail of a PSU.

I do not know what to tell you. The one you sent back to us, I plugged in and tested thoroughly. There was absolutely nothing wrong with it. I plugged it in and all was well, I left relays on overnight and the oscilliscope never measured any dropout. (I had it set to capture an image if the voltage dropped accross the relay by 500mV)

The 300mA adapter is useless for the FB unless you are only using it for testing. Also those 12v ones measuring in the 16v, my bet is that they are really really really crappy SMPS with no output filtering or something. So they are using PWM to get a 16v signal into a 12v or 9v signal. So I am guessing if I hooked that up to my scope, I would see a sqaure wave with varying duty cycles to make the different voltages. There needs to be filtering of some sort on that, you should always see a regulated voltage output, not the PWM. Ripple is ok, but that is rediculous. I would NOT use those at all.

The FB uses a standard linear regulator LM7805 (http://www.fairchildsemi.com/ds/LM%2FLM7810.pdf). It can take a voltage of up to 35VDC. This however will make the thing hotter than the sun. All that extra power is wasted as heat. Higher the voltage, more wasted heat. If it gets too hot, it drops power essentially.

My suggestion right now would be to take a spare heatsink or just a large block of metal, and try to fix it to the regulator. Just to see if that is the problem. I really don't know why it wouldn't be working at your two guys' houses, but works perfectly everywhere else in the world... As to the output ports dying, who knows at this point. Usually they die by a whole chip. So outputs 0-7 and then 8 through 15 are each on separate chips and the only way they die is through serious abuse. These chips can handle alot, and the only way I have seen them die is when they are shorted out. Meaning you touch power to the output pin. Such as trying to measure them with a multimeter and the probes crossing both pins. Even then it takes a few seconds of that for the port to explode. If the LEDs cooresponding with that port is dead, then the microcontroller itself has been effected by something because that is on the input side of the output drivers. And if something has gotten to the MCU, that is very very bad and I have never seen it happen. We made the Version 3 have a replacable MCU because we thought this could happen, and it never did. So Version 4 is not replacable. Once the MCU goes the board does too.

Try the heatsink on the regulator, and maybe use a computer power supply to power it. Other than that, I have no idea. We can replace them again, but I am stumped.

When I first discovered the problem I started troubleshooting based on the fact that I believe the brain to be 100% working when I received it.

The two power supplies that read 16V are cheap junk no doubt. The wall warts that I've been using for most of my testing are from external hard drives. For those kind of electronics, I would expect something of decent quality, and they do read approximately 12V. I'm inclined to keep working with those.

I've ruled out my wall wart as the fault. Here's why: If I watch the LEDs with the output on, the LED flickers in sync with the click of the relay... the relay basically just makes the fault audible. If I switch the power jumpers over to USB, and disconnect the 12v power altogether, and turn on the output, I can watch the LED flicker just as frequently as it does with the 12V with the relay connected. So this problem exists without the 12V power connected, the only difference is that I can't power the relay to make it audible.

So then my next inclination is the problem exists with my computer. I am using an old Dell Inspiron laptop as a dedicated computer for the brain. I'm running the bare minimum of WinXP on it, it's only running at 800mHz, and has a small 12Gb hard drive; not much of a computer, but it seems to run MDX just fine. Because of the barely adequate specs of the Dell, I tried my Macbook booted into WinXP. The problem didn't change. The final computer I tried is a Sony Vaio that's approximately a year old and running Vista. Still no change. The LED flickers whether the 12V is connected or not. If I had a scope I'd check it with that to see what's going on. Unfortunately I don't have one.

Based on all of that, my inclination was towards the brain being faulty, so I got it replaced. I'm willing to try one more to see what you can find out about the failed outputs. I think the last thing to try before I send it back is a different USB cable. I really want this thing to work for me.

Different USB cable isn't the solution.

Hi,
I'm having the same buzzing relays like the other guys have. Anyone found a solution yet? it looks to me like the uln chips are semi defective.
Anyone can help?

Originally Posted by stabmegane1

Hi,
I'm having the same buzzing relays like the other guys have. Anyone found a solution yet? it looks to me like the uln chips are semi defective.
Anyone can help?

What model number relay?

I can't find the datasheet now, but i checked the datasheet and they draw 30ma each. i made a pcb with all the relays and i put it just behind the liliput screen i installed. The buzzing makes the picture from my rear camera SO distorted (distortion) i can barely see the details.
I ordered uln chips from ebay i 'll try them and i will tell you.
Something else i have in mind is that i use flat ribbon cables , the ones i use for floppy ( i own a computer shop) to connect the fusion brain with the relay board i made, and i wonder maybe that is the case. parallel cables have a capacitance maybe? and with the coil of the relays we have an oscillation circuit? or something like that. haven't tried other cables there are sooooo many things to take out to reach the relay board.
Maybe the other guys who have the same problem can confirm if they use ribbon cables??
by the way i will tell you how it goes.

P.S. forgive my english i'm from greece.

Every computer I've tried has been a laptop... any reason why my usb output would be any different with a laptop? I can't think of one but maybe I'm missing something.

I test everyone on a laptop with the relays that we supply and I have not been able to make this happen at all. I have hooked up an oscope and set it to trigger, left it overnight and even for a few days, and it never flickers on and off.

So before the brains are shipped out I program each one by hand, then plug in a bundle of relays and have them turn on and off with delay inbetween. No buzzing that I can tell. I am using the wall wart we sell, and the relays we sell.

Another customer that had this problem put a small capacitor (1.0uF) over the output, and it soved his issue. But that sort of solves a problem that should not be there and I cannot reproduce.

As for USB ports varying, they do. But laptop vs. desktop should not make a difference. When powering an output all power comes from the input jack.

The only normal things that can cause this:

a) input voltage too low
b) input current too low, so when it pulls more to flip the relay, the voltage drops and you get in a continuous reset
c) bad physical connection. Either solder on a chip, or wire to a harness. The first is possible but improbable since I test each one individually.
d) It is so hot out that the thermal fuse is tripping at a lower current than it should. I would not think this to be the case if you lived to tell the tale....