Cyclone

What the heck is going on here!

I buit this circuit on one of my breadboards.

http://ourworld.compuserve.com/homep...5.htm#eyes.gif

The circuit asks for a LM324 op-amp but i didn't have one so i used two LM308 op-amps. The circuit worked just great.

but.... when I built the circuit onto a pcb the circuit doesn't work. I checked all my connections and they were fine. I thought maybe there was a wire that wasn't touching so i decided to unsolder everything from the pcb and retest the circuit on my breadboard. Everthing worked and the leds faded properly. So I was like hmm must of been a lose wire or something. So I attempted to re-solder everything back onto the pcb but it still woudn't work. The two leds would just stay on and would not fade in ond out.

I then used my knife and pressed on certain wires to see if there was a lose wire. when i touched the knife on one of the pins of the op-amp the leds started to fade in and out. The circuit seemed to work. Even when i touched that certain pin with my finger the circuit seems to function.... The circuit like finctions ghostly. hmm


Anyone know what the heck is going on here? This is the last circuit I need for my mod and i just don't see why the circuit won't work.


thanks

Eclipsed

Which pin? I would imagine it is a floating problem.

Cyclone

Its pin8 and pin1 on the first op-amp. Whenever I touch those pins for some reason the circuit works? Why would the circuit work on my breadboard and not on my pcb?

oh and the pins aren't even used for anything in my circuit

motion

My guess it's a PCB layout problem. Cascading two of these will produce a high open-loop gain. At high frequency, the phase shift would be enough to produce instability (Nyquist stability criterion).

Unlike the lm324 and 1458, the op-amp you've chosen has no internal capacitor to keep it from oscillating at high frequency. Touching pins 1&8 loads the op-amp just enough to keep it from oscillating. You could put 10-100pf capacitors between pins 1&8 on each of the LM308 you used. See the LM308 datasheet for details.

__________________
"Having to do with Motion Control"

Roff

I agree with Motion, except that you shouldn't need a compensation cap on the Schmitt trigger (the op amp on the right). Leaving it out should allow this op amp, which is being used as a comparator, to operate faster than it would with the cap installed.
The breadboard probably worked because they typically have a lot of capacitance between pins.

Cyclone

YAY thanks Motion and Ron it works now. I connected 100pf cap to pins 1 and 8 to the op-amp on the left and it now works perfectly.

I tried connecting another 100pf cap to the other op-amp but the timing of the leds didn't seem right.


thanks again that was very helpfull. so using the LM308 op-amp always require a cap between those two pins?

Roff

For closed loop gains of 10 or greater, you probably don't need more than 3pF between pins 1 and 8. Using more compensation than necessary reduces the bandwidth of your amplifier.

Cyclone

ok cool.

http://ourworld.compuserve.com/homep...htm#4017-2.gif

now this fading led circuit is attached to this other circuit with a decade counter and timer. I just finished prototyping the timer. When I connect the timer to the counter and turn it on it always skips leds. for example
led#1,3,5,8,10. something like that. The leds that are skiped flicker very breifly but don't stay on but the led next in the sequence does.

When I use a pushbutton to send the input to the counter it works just fine. each led goes on like it should.

oh and I connected my voltmeter to the timer circuit and i noticed right before the voltage increased there was a slight change in voltage before the voltage went to its max, like a slight hick-up in the change of voltage. so it could be sending an extra pulse?

GRRRRR i'm getting sick of all these anoying problems with my circuits.... :evil:

Roff

How is the fader connected to the sequencer?

crust

And did you remember to put bypass capacitors across the VCC/GND pins of your ICs? I have seen some strange things happen when the bypass caps were not in place. I ask only because the circuit in the link does not have any.

Cyclone

The fader is connected directly to one of the output pins of the counter and ground. All i did was replace the led in the schematic with my fader.


hmm no i didn't put any caps across the VCC/GND of the timer. would that really make a difference?

crust

There is no assurance that it will make a difference, but some gates are very fast and can respond to transients in the power which sometimes happen as loads are switched in and out by the gates. In CMOS it happens during the bit transition. In any case, the bypass caps across the VCC and GND filter out any high frequency transients that show up in the power and ground. You want that filtering very close to the IC. If you look at a professional board, there is usually an IC and a tiny capacitor sitting right next to it.


As an aside, I was helping somebody with a project one time that used an MCU to drive a 74XXX (i dont remember which part). The system was mostly breadboarded except the MCU had its own PCB. When the ground from the PCB to the breadboard were connected, the MCU was unable to cause transitions in the 74 parts. When the grounds were disconnected everything seemed to work. As it turned out, the problem was that the power supply rails were noisy when the bus drivers were switching causing lots of noise to show up at the 74 parts. Putting a handful of filter caps across the ICs solved the problem.

Cyclone

hmm I think I fixed the problem. I had all my circuits soldered onto a pcb. But the timer I had on my breadboard. Well I tried soldering it to my pcb and it fixed all the problems. So for some reason my breadboard was
causing me problems.

on a side note a 500mA 9v voltage regulator should be fine to hook my circuit up to my computer's 12volt supply right? It has to power my IR circuit,Timer,fading led circuit and my counter circuit. My voltage regulator should be able to handle that amount of current draw right?

Digi-key sells quite a few voltage regulators all with different features so i think i got the right one. "9v .5amp fixed voltage"


thanks for taking the time to help me out with my problems. I appreciate it