It's getting better all the time. You might even add a switch to switch out the high frequency filtering so you could do a rudimentary test on new circuits to see if they are relatively impervious to noisy power supplies. I've had new circuit designs work beautifully when I was building and testing them on the bench, but then get intermittent when fired up with a cheap switcher supply like it would be run on when put in service. My Agilent bench supply was great for ensuring my power source was silent, but it didn't tell me if the circuit would work in a real-world application. So I decided that if I ever built my own bench supply I'd have a 'noisy output' switch on it with which to test my circuits' noise immunity.
Great Additions to the Power Supply! You might find using the map function to make the fan variable across the range to work better. Like w1 = map(b0, 64, 125, 0, 1023)or better pwmOut = map(tempC,64,125,0,1023) Just one line to do it all. unless you are in a state of two much power this finds the perfect speed to run and hold the temp steady. If you are keeping the current code you might want to add blank temps between the setpoints to prevent it from oscillating from the temp drop of the fan speeding up and dropping the temp just enough to fall back into the last setpoint. like this If b0 < = 64 then w1 = 0 no setpoint for 65 .... so at 65 it does what is is currently doing. On stays on and off stays off. elseif b0 > 66 and b0
Good idea! The only thing really holding me back from going from going to continuous control is that the phase shift caused by thermal inertia will cause a long period oscillation in the fan speed. It should diminish over time and is probably not going to do any harm. However, it might. I'd need a PID control loop to prevent that. That's the "idea" I talked about in video. Not sure how to do that using positive integer BASIC. I'm sure it can be done .. will just need some thought.
Nice little board for upgrading any project. I recommend you to add some hystereses to prevent fan speed oscillation. Or you could implement a continuous speed variation like this: temp_min = 64; temp_max = 125; pwm_min = 0; pwm_max = 1023; w1 = (b0 - temp_min) * (pwm_max - pwm_min) / (temp_max - temp_min) + pwm_min; if(w1 > 1023){ w1 = 1023; }
Hysteresis might actually cause continuous oscillation due to the relaxation effect + phase shift from thermal inertia. I expect continuous variation will also cause oscillation, but it should diminish over time as long as the cooling is sufficient. I think I'd need a PID control loop to prevent that. That's the "idea" I talked about in video. Not sure how to do that using positive integer BASIC. I'm sure it can be done .. maybe..
@@uni-byte You're right, a system with negative feedback tends to be stable. I implemented myself a PID control loop to regulate the temperature of a solder reflow oven. I'm not familiar with PICAXE but simple arithmatics can do the job
How do you set the current limit to a pre determined value without shorting the leads? It would be a good idea to add a calibrated scale to the current knob somehow. Not particularly accurate but a good first start. Set the fine current knob to the centre. This was an excellent video series btw.
It's getting better all the time. You might even add a switch to switch out the high frequency filtering so you could do a rudimentary test on new circuits to see if they are relatively impervious to noisy power supplies. I've had new circuit designs work beautifully when I was building and testing them on the bench, but then get intermittent when fired up with a cheap switcher supply like it would be run on when put in service. My Agilent bench supply was great for ensuring my power source was silent, but it didn't tell me if the circuit would work in a real-world application. So I decided that if I ever built my own bench supply I'd have a 'noisy output' switch on it with which to test my circuits' noise immunity.
That could be useful. Right now I use some cheap SMPS units for just that purpose.
Great project, thats a very tight squeeze. It's always tempting to use the smallest case, but then later you regret it, but this worked well....
Yeah, it's a tight fit alright. I had actually wanted to put it in a smaller case at first, LOL! Room on my office bench is at a premium..
Nice project 👍
Thank you! Cheers!
Great Additions to the Power Supply!
You might find using the map function to make the fan variable across the range to work better.
Like
w1 = map(b0, 64, 125, 0, 1023)or better
pwmOut = map(tempC,64,125,0,1023)
Just one line to do it all. unless you are in a state of two much power this finds the perfect speed to run and hold the temp steady.
If you are keeping the current code you might want to add blank temps between the setpoints to prevent it from oscillating from the temp drop of the fan speeding up and dropping the temp just enough to fall back into the last setpoint.
like this
If b0 < = 64 then w1 = 0
no setpoint for 65 .... so at 65 it does what is is currently doing. On stays on and off stays off.
elseif b0 > 66 and b0
Good idea! The only thing really holding me back from going from going to continuous control is that the phase shift caused by thermal inertia will cause a long period oscillation in the fan speed. It should diminish over time and is probably not going to do any harm. However, it might. I'd need a PID control loop to prevent that. That's the "idea" I talked about in video. Not sure how to do that using positive integer BASIC. I'm sure it can be done .. will just need some thought.
Nice little board for upgrading any project. I recommend you to add some hystereses to prevent fan speed oscillation. Or you could implement a continuous speed variation like this:
temp_min = 64;
temp_max = 125;
pwm_min = 0;
pwm_max = 1023;
w1 = (b0 - temp_min) * (pwm_max - pwm_min) / (temp_max - temp_min) + pwm_min;
if(w1 > 1023){
w1 = 1023;
}
Hysteresis might actually cause continuous oscillation due to the relaxation effect + phase shift from thermal inertia. I expect continuous variation will also cause oscillation, but it should diminish over time as long as the cooling is sufficient. I think I'd need a PID control loop to prevent that. That's the "idea" I talked about in video. Not sure how to do that using positive integer BASIC. I'm sure it can be done .. maybe..
@@uni-byte You're right, a system with negative feedback tends to be stable. I implemented myself a PID control loop to regulate the temperature of a solder reflow oven. I'm not familiar with PICAXE but simple arithmatics can do the job
How do you set the current limit to a pre determined value without shorting the leads? It would be a good idea to add a calibrated scale to the current knob somehow. Not particularly accurate but a good first start. Set the fine current knob to the centre. This was an excellent video series btw.
I just short the leads, but you are right, I should have included a scale of some sort on the coarse current setting. Thanks on the series!