OK, this is driving me nuts and I hope someone here can point me in the right direction. I have a 555 PWM circuit (see attached) I'm using to drive a Mosfet for dimming control. I have breadboarded this and it works perfectly. Probing pin 3 with the scope shows me a square wave at around 120Hz. Turning the pot, I get what is to be expected as a change in duty cycle with the frequency remaining about the same (maybe 2Hz difference between full CW and CCW. I then put this on perfboard and soldered everything in place. When I probe pin 3 the perfboard, I get a change in frequency as the duty cycle changes (from around 128Hz to around 500Hz CW to CCW). The load (halogen lamp) seems to work fine as far as dimming, but I would like know why my frequency is changing. I have checked the perfboard trace by trace to make absolutely sure the wiring is the same. I have also checked the caps and swapped out the 555 with another one to be certain that those things are not the problem. All components are new, unused components. The only difference between the perfboard and the the breadboard is that the 555 is in a socket on the perfboard. Anyone got any suggestions of what I should check?

 

Check the resistance across each of the diodes in circuit (power off). I suspect that one of them is short circuit - either the diode itself is faulty or the layout sorts one of them.

 

OK, this is driving me nuts and I hope someone here can point me in the right direction. I have a 555 PWM circuit (see attached) I'm using to drive a Mosfet for dimming control. I have breadboarded this and it works perfectly. Probing pin 3 with the scope shows me a square wave at around 120Hz. Turning the pot, I get what is to be expected as a change in duty cycle with the frequency remaining about the same (maybe 2Hz difference between full CW and CCW. I then put this on perfboard and soldered everything in place. When I probe pin 3 the perfboard, I get a change in frequency as the duty cycle changes (from around 128Hz to around 500Hz CW to CCW). The load (halogen lamp) seems to work fine as far as dimming, but I would like know why my frequency is changing. I have checked the perfboard trace by trace to make absolutely sure the wiring is the same. I have also checked the caps and swapped out the 555 with another one to be certain that those things are not the problem. All components are new, unused components. The only difference between the perfboard and the the breadboard is that the 555 is in a socket on the perfboard. Anyone got any suggestions of what I should check?View attachment 124011

Look at your supply voltage (with the scope), a 5A bulb may be pulling your supply down a few volts and causing funky output of the 555.

 

Here is almost no change frequency

 

Try 'diode + capacitor' decoupling the supply to the 555. That should stabilise it even if the main load drags the 12V down.

 

Check the resistance across each of the diodes in circuit (power off). I suspect that one of them is short circuit - either the diode itself is faulty or the layout sorts one of them.

An audio taper pot wouldn't balance - it must be linear.

Rarely a problem with pre set trimmers, but a regular potentiometer could easily be a tapered type.

 

An audio taper pot wouldn't balance - it must be linear.

It will affect duty-cycle versus pot position, but that wouldn't affect the frequency.

 

It will affect duty-cycle versus pot position, but that wouldn't affect the frequency.

It won't be balanced - the frequency will change.

It may well be not far out at the extremes of rotation, but there will be a deviation through the swing.

 

It won't be balanced - the frequency will change.

What won't be balanced?
How will the frequency change?
You are making claims without substantiation.

t may well be not far out at the extremes of rotation, but there will be a deviation through the swing.

The PWW will not change linearly with pot position if you use an audio taper pot. but the frequency is determined essentially by the total pot resistance, and that does not change whether the pot is linear or logarithmic.

 

Sorry folks, I have several projects going at once and had to set this one aside. I've got some free time to look at it again.
Albert, I pulled the diodes and checked them, they were both fine. I did run it again and noticed that even when I leave it set a fixed duty cycle (lets say 50% - the point here is that it is not being adjusted), the frequency bounces back and forth as much as 60Hz with no real rhyme or reason. It appears there may be a component failing (my guess is one of the caps), but when I check them with an LCR meter they test fine. Again, all the components were new. One other thing I noticed, I decided to run the circuit at full load and it held fine for more than an hour (I thought anyway), however when I went to check the dimming, it would not dim. Seems it got stuck in full on. I turned it off, let it cool for about 20 mins and checked it again. I once again had dimming control. Seems I may actually have a faulty component, but again they all check out when I remove them from the board and test them.
For the rest of the folks who have responded, my circuit works exactly the way I expect on breadboard. It's only when i solder components to perfboard that I have problems. Could I possibly be getting RF interference due to the large solder blobs that I have used to connect pads together. I have attached pics of my board (please excuse the roughness of this prototype).

 

.........
my circuit works exactly the way I expect on breadboard. It's only when i solder components to perfboard that I have problems. Could I possibly be getting RF interference due to the large solder blobs that I have used to connect pads together. I have attached pics of my board (please excuse the roughness of this prototype).

No. RF interference is highly unlikely at that low frequency.

If the breadboard works fine and the perfboard doesn't, then you likely have a wiring error on the perfboard.
It could be a short from one of the solder blobs.
Continuity check the complete circuit against the schematic.

 

Youcould also have an open crt on the solder blobs.

 

A dry joint could explain the erratic behaviour.
(I would never use dot matrix board. Prefer stripboard, which avoids the need for big solder blobs ).

 

If you have a failing component which appears once the circuit warms up, then I would do a temperature test.

One way is to use a cooling spray. Let it run until the load appears full on. Then, spray/cool each component and see if normal operations occur.

You could also use a heat gun and perform the test in reverse.

 

Youcould also have an open crt on the solder blobs.

Yes, a cold solder joint is probably more likely than a short.

Note that good solder practice means you should not have solder blobs.
The wire you are soldering should always be visible through the solder on the pad.
If not, then you are using too much solder.
Note the bad joint in the center and a correctly soldered joint on the right.

 

Well, I am completely baffled now. I decided to rebuild the circuit on a second board using new components. I noticed that the socket I was using was not taking solder well on the pins, so I fluxed them and tinned them prior to soldering to the perfboard. I also decided to leave off the Mosfet portion and just focus on the oscillator. I made every effort to avoid solder blobs and used wire to connect the circuit where needed. I then decided to power the circuit with a bench supply to eliminate the possibility of that being my problem. I fired up the circuit, probed pin 3 (actually soldered a wire lead to it) and checked the scope. Same exact result as the first board. I then went back to my breadboard circuit, removed the mosfet connections so that it was exactly the same as the soldered circuit, powered it from the bench supply and it is still working perfect! I think I'm done at this point. I guess it's time to pull out the ATTiny85. Thank you to all those who offered up suggestions, I appreciate the help.

 

I started using TLC555 instead of NE555. Its better...i can udjust PWM without freq changes. Some diods and thats it