Well guys, I agree that most of my issues are possibly because of the supply voltage. I started out using a 9v battery, but am currently using a 9v "wall-wart" which obviously cannot handle the current required by the 2 circuit boards. As I mentioned earlier, I am in the process of building a regulated 9v, 1-amp power supply which should give me the proper Vcc. Incidentally, I contacted the manufacturer of the Wind Speed module and he supplied me with the exact sensor. Here it is: http://uk.farnell.com/allegro-micro...fect-sensor-sip-3-1301/dp/1198085?ost=1198085 The "HAL575" reference that I gave you earlier is what is stamped on the sensor, which now may be a misnomer and I may have been leading you guys down the wrong path... my humble apologies if this is the case. Please see if this makes a difference in how we are trying use it's signal. BTW, if you are concerned about the 5v operating voltage for the sensor, the Wind Speed module has a 5v regulator and 2 large capacitors on-board which drives the display, microprocessor and sensor, but module was designed around a 9v supply, hence my desire to design our timer around a 9v supply as well. As far as the magnets are concerned, I am using 1/2" x 1/8" round neodymium magnets (10800 g) on a 3-spoke wheel that rotates with a very slight amount of wind. The sensor is mounted approx. 3/16" from the magnets, which should be close enough to stimulate the sensor, don't you think? I feel confident that I have all 3 magnets oriented the same direction, but they may be backwards... will try to verify the South pole is facing the sensor. I should have the power supply completed within a few days (just waiting for a few components), so please bear with me. I really appreciate your input and expertise with this project. Thank you SO much! Please "stay tuned"

 

So is the sensor you're going to use now an A1301 or an A1302 (i.e. a linear type)? That does make a difference and will require a circuit mod.

 

Yes. According to the manufacturer, he supplied an A1301 linear Hall sensor with the Wind Speed module which is what we need to interface with the 555 timer circuit. Thanks.

 

OK. Here's a suggested mod. There is now a 680k pull-up resistor R1 and a 100 Ohm emitter resistor R4. Simulation says Q1 should switch satisfactorily provided the Hall device can generate peak-to-peak output of at least 1V at >=1Hz.

 

Ok, Thanks Alec. I'll give it a try when I get my power supply built and I'll let you know how it works. Sure do appreciate all your time and effort!!

 

Well, I still have issues... as in it still doesn't work quite right. I got my new power supply built and it's a very stable 9.03 volts DC. I modified the circuit based on your last suggestions and .... well, you get my drift, right? Sorry, we're not quite done yet. The output comes on at the 1st pulse, but it turns off after the 180 second mark and won't come back on unless I cycle power to the timer board. I did some research on 555 timers and found a statement that may be a real stumbling block based on the requirements... "While using the 555 timer in monostable mode, the main disadvantage is that the time span between any two triggering pulses must be greater than the RC time constant.” And... we can obviously expect multiple pluses of varying pulse widths and overall frequency. I have ordered an oscilloscope so I can actually see the pulses from the sensor, but the scope won't be here till next Wednesday. I'm really curious as to what kind of pulse we're dealing with.... sort of anxious to see it!! I have a feeling that the sensor pulses are going to be of varying pulse widths (based on how fast a magnet goes past the sensor) and most likely varying frequency (based on how fast all 3 magnets spin past the sensor). Maybe we should re-address the original requirement with more detail. I have put together a Timer On/Off chart that fairly well depicts what the timer requirements are (see attached). As you can see by the time chart, I have decreased the 180 second "off" requirement to 90 seconds. This should work just fine for my purposes. Besides the RC time constant issue, another problem might be that the timer doesn't start counting until after the last pulse is received and... how long does it wait before it starts counting? 1 second, 2 seconds, 5 seconds...?? I guess perhaps we need to set a "buffer time" say 10 seconds after the last pulse, at which time the timer starts... providing no pulses are received during the buffer time. I know it's rather complicated but I feel with today's technology, there should be a way to accomplish this task and I'm really looking for your help. Please let me know if you need any additional information or if you have questions.

Thanks, Bob

 

OK. If I've understood your requirements correctly this should do the job :-

U1 is a CMOS quad NAND gate. U1a is here being used as an amplifier so as to respond even to weak Hall pulses. The trimpot biases U1a input at slightly above the gate switching threshold, so that in the absence of Hall pulses the gate output is low. When Hall pulses are received the gate toggles and a voltage is built up on cap C3. This drives the outputs of gates U1b and U1c low so that C5 is kept discharged, the 555 output is high and the relay is energised. When Hall pulses cease for less than about 10 secs the C3 voltage drops insufficiently to change the logic states of U1b and U1c, but if they cease for more than about 15 secs the C3 voltage drops sufficiently for the gate outputs to go high. This allows C5 to charge. After about 70 secs its voltage reaches the 555 threshold and the 555 output goes low, de-energising the relay and discharging C5 again via D5 to await another cycle.

 

Oh My!!! A Major overhaul, huh!! I have a very limited supply of parts so I will need to put in an order to DigiKey or Newark before I can build it up. This is sort of ok because I am having some metal work done on the sensor mount and I couldn't test the circuit even if I had the parts. I will let you know what we have in a week or so after I get everything put together. Thanks SO much, Alec.

 

Follow-Up. Alec, I finally got the circuit finished. After looking at the Hall signal with my oscilloscope, I realized another modification to the circuit was required. Attachment #1 shows the Hall signal with and without the timer circuit connected... considerably more complex than I expected!! So, I incorporated a low-pass filter before the 1st capacitor to eliminate the 1.4Khz "fuzz" riding on the signal (see Attachment #2). I re-incorporated the transistor to provide more control of the Hall signal to the 1st gate, and slightly modified the discharge "feed-back" to Gate #3. Bottom line is, it works!! (Attachment #3) The relay turns off somewhere between 85 and 110 seconds, depending on how long the Hall pulses are present, which is perfectly acceptable for my application. All-in-all, I wouldn't have this circuit were it not for your input and expertise. I may call on your help in the future.... Thanks again Alec.

Bob Turner

 

Good to know it's doing its thing ok