|
Homebrew Equipment A place to show, design and discuss the weird and wonderful electronic creations from the hands of individual members. |
|
Thread Tools |
2nd Sep 2020, 3:00 am | #1 |
Nonode
Join Date: May 2016
Location: Cornwall, UK.
Posts: 2,315
|
Bat Detector with Knowles Ultrasonic microphone problems
I've read a few threads on here regarding these detectors, and have discovered a circuit using the Knowles SPU0410LR5H-QB. A catchy title for a microphone with a response high enough that it could make a better detector than the usual electret capsules.
However, I'm getting nothing out of my breadboarded circuit and as I've tried to alter it to fit what I had, I'd be glad if someone can pinpoint why. No major frequency alteration on the 555's output when adjusting P1, a terribly distorted waveform and no sound from the headphones. I've tried separating each bit on the breadboard and scoping them individually, but to no avail. The original circuit's attached, along with the author's write up. The author wanted to run everything from CMOS ICs, which I don't have and thought I could try my hand at replacements from bits I do have. I have a 5V supply (boosted from an 18650 cell with an MT3608). The microphone op-amp is a MC1458P instead of an MCP6002 , with the microphone Vcc dropped from the supply with an 18k resistor (following the datasheet's maximum 3.6V and 120µA). Instead of the MCP6004, I have two LM358s. The switch is a CD74HC4066, and I've got an NE555 instead of an ILC555. I have taken the output of the first half of the second LM358 (unity gain buffer + low pass filter) to an LM368 via the volume control (following the datasheet's component values), planning for it to drive the headphones. So, what have I done wrong? |
2nd Sep 2020, 12:43 pm | #2 |
Octode
Join Date: Jun 2006
Location: St Osyth, Nr Clacton, Essex, UK.
Posts: 1,482
|
Re: Bat Detector with Knowles Ultrasonic microphone problems
I've just looked up ILC555 and it's CMOS. Have you modified the component values so you can use the NE555? I think they need different values.
Graham
__________________
Half my stuff is junk - trouble is, I don't know which half! |
2nd Sep 2020, 12:50 pm | #3 |
Nonode
Join Date: May 2016
Location: Cornwall, UK.
Posts: 2,315
|
Re: Bat Detector with Knowles Ultrasonic microphone problems
Yes, the ILC555 is the author's CMOS version. I've used the same values for the NE555 because after running various possibilities through an online 555 calculator they gave the diagram's suggested oscillator bandwidth. I was under the impression that the Vcc doesn't affect the oscillator, just the combination of R1, R2 and C.
|
2nd Sep 2020, 10:13 pm | #4 |
Nonode
Join Date: May 2016
Location: Cornwall, UK.
Posts: 2,315
|
Re: Bat Detector with Knowles Ultrasonic microphone problems
I have discovered that the circuit's suggested outputs for the high-pass filter and 555 timings don't tally with the calculators I have been using.
The 555 should have approximately 21kHz - 400kHz, but with 10nF and two resistors of 100R and 2k2, I get a bandwidth of about 60kHz - 700kHz. I have to use a 5k pot and 15nF to get it down to what the diagram proposes. The second order high pass filter relies on a 3.3mH inductor. I've never used one before so I thought I'd do a simpler CR filter (100nF and 68R) for 25kHz. However, the circuit's calculated results require a 1.8mH inductor for 25kHz cutoff. I'm going to breadboard the 555 circuit only, and monitor pin 3 with a 10x probe. |
2nd Sep 2020, 10:46 pm | #5 |
Dekatron
Join Date: Sep 2010
Location: Cheltenham, Gloucestershire, UK.
Posts: 3,077
|
Re: Bat Detector with Knowles Ultrasonic microphone problems
I have bats flying about in my garden every summer just as it starts to get dark. They are very acrobatic when they detect an insect and chase it!
I made my own bat detector many years ago to listen to them and I think I did a classic up/down converter. I upconverted to 21.4MHz to a crystal filter with 20kHz BW and then mixed down to AF/baseband. I didn't think the results were that inspiring because I just heard a fairly regular pulsed sound from the bats. However, in recent summers I've watched the bats with a thermal camera at night against the sky. Some of the insects (moths?) look like easy prey because they seem to fly in a slow straight line when viewed with a thermal camera. Without the thermal camera the bats and the insects would be much harder to see at night.
__________________
Regards, Jeremy G0HZU |
3rd Sep 2020, 12:03 am | #6 |
Nonode
Join Date: Mar 2010
Location: Sunderland, Tyne and Wear, UK.
Posts: 2,473
|
Re: Bat Detector with Knowles Ultrasonic microphone problems
Bats are fascinating creatures I love watching them, some people are not too keen on them(when they find out what they are!, for some reason they seem to think they are night time birds!) very acrobatic and accurate flyers too, they look like they are having fun to boot when they first venture out at nightfall.
__________________
I don't suffer from Insanity. I enjoy every minute of it. |
3rd Sep 2020, 12:23 am | #7 |
Dekatron
Join Date: Sep 2010
Location: Cheltenham, Gloucestershire, UK.
Posts: 3,077
|
Re: Bat Detector with Knowles Ultrasonic microphone problems
In my garden there is usually only one bat at a time and it flies in a narrow loop up and down the garden. I think its favourite hunting spot is just in front of the big double patio doors as it flies right at them and then loops away really quickly and often diverts towards an insect. It does this many times in an evening and I think it is looking for insects attracted to the bathroom light just above the patio doors. But that is just a guess.
It is possible to lie on the ground by the patio doors and watch it come within a few feet of me every lap!
__________________
Regards, Jeremy G0HZU |
3rd Sep 2020, 12:46 am | #8 |
Dekatron
Join Date: Aug 2011
Location: Newcastle, Tyne and Wear, UK.
Posts: 11,482
|
Re: Bat Detector with Knowles Ultrasonic microphone problems
Many years ago, driving southwest alongside the shore of Ullswater at night, trees on both sides and over-arching, I was interested to see a bat swing in ahead of me and adopt more or less the same speed and course as me, deviating occasionally but always flying ahead of me along the road. It was shortly joined by several others also flying along in my headlight beams, they stayed with me until the road emerged into the open a few seconds later and then scattered in all directions.
I've never seen this behaviour since, but at the time I wondered if they knew that moths were attracted to the headlights of cars and had actually developed this way of hunting, and if they had, how widespread that was. I've also, on several occasions now, seen Sparrows hopping along under the fronts of cars in car parks and flying up behind the radiator grilles to grab recently deceased insects. After I noticed it the first time, I started to look out for that behaviour. (But that is getting away from the subject of bats). Uncle Bulgaria, where are the bats you are hoping to listen to? (I'm assuming you don't have a belfry). Last edited by SiriusHardware; 3rd Sep 2020 at 12:53 am. |
3rd Sep 2020, 4:07 am | #9 |
Nonode
Join Date: May 2016
Location: Cornwall, UK.
Posts: 2,315
|
Re: Bat Detector with Knowles Ultrasonic microphone problems
Interestingly, I seem to be getting somewhere. The NE555 circuit I've breadboarded has a lovely square output, which changes frequency with rotation of the pot IF:
The 100R is replaced with a 3k. I haven't changed it down to see if there's a lower limit where it stops working. I'm using a 20k pot for the variable as the 2k2 is awkward to adjust (a preset) but would have thought there will be no difference at the lower end of the range. AND I also have to have the USB charging board for the 18650 plugged in to the PC. Monitoring the voltage off the DC boost board into the breadboard (3.7V to adjustable output - set to 5V) shows no change (so the cell is not dead), but as soon as the USB cable is unplugged the waveform disappears from the 'scope. I have verified the 5V is still on pin 8 of the 555 when the cable is unplugged. The detector's a present for my mother, a biologist. There are many and varied bats around the village, and recently I heard a friend's detector when we were in an ancient woodland dining al fresco. It was far better than monotonal cheeps, though it was a very old detector. I thought with this ultrasonic microphone it would only get better! On that occasion I did hear one of the bats myself, or at least some terribly high-pitched sound, which pleased me that my hearing must be up to snuff. |
3rd Sep 2020, 3:53 pm | #10 |
Nonode
Join Date: May 2007
Location: Redruth, Cornwall, UK.
Posts: 2,562
|
Re: Bat Detector with Knowles Ultrasonic microphone problems
When we had some decent summer days (many years ago) several bats could be seen flying low over the garden during dusk. I thought about trying to listen to their ultrasound so decided to build a bat detector from a circuit on the web which worked reasonably well.
I always thought that the limiting factor is the frequency response of the microphone. I had ideas of using a small piezo type tweeter to experiment with but never got around to trying one out. There have been a number of bat detector projects in the various electronics magazines over the years. Magenta Electronic (not sure if they're still in business) did a decent kit as well as fully assembled detectors. I'll try to find my bat detector and the source of the circuit. Regards, Symon |
3rd Sep 2020, 3:59 pm | #11 |
Hexode
Join Date: Mar 2020
Location: Rochdale, Greater Manchester, UK.
Posts: 439
|
Re: Bat Detector with Knowles Ultrasonic microphone problems
Hi, interesting project but changing components is not without its pitfalls. Have you realised that a 100nF capacitor is usually recommended between pin 5 and ground for the bipolar 555, the original cmos version schematic does not show this capacitor? It is required due to the rather large curent spikes that occur during output transitions, and may well be related to your observation that the USB supply was required for proper operation, I'm guessing the battery isn't man enough during these current spikes.
Good luck with the project Chris |
3rd Sep 2020, 4:25 pm | #12 |
Dekatron
Join Date: Apr 2011
Location: Cornwall, UK.
Posts: 13,454
|
Re: Bat Detector with Knowles Ultrasonic microphone problems
Bats are fascinating creatures, got them here where I live, I once had the pleasure of going in one of the premier places for the Greater Horseshoe Bat in one of the old tin mines down here in Cornwall at Prideaux Wood for the baby bat count, Danny Eva was the batman back then, I think he's still around and involved.
We waited outside for a couple of bats to come out for a quick scout around, once they were happy it was then back into the mine to tell the rest of the gang then out they all came, off to get the food, that's when we went in to do the bat count....just above a rather large heap of bat sh*t in an old stope. Lawrence. |
4th Sep 2020, 1:59 am | #13 |
Nonode
Join Date: May 2016
Location: Cornwall, UK.
Posts: 2,315
|
Re: Bat Detector with Knowles Ultrasonic microphone problems
Thank you Chrispy57! That's a good tip. I have a decent square wave now. The 100R does not work at all, but 220R is reliable on the 555 so I'm forced to use that. Probably because it's not a CMOS version. There was a moment this evening where I had a burst of amplified rustling paper in the headphones, but now all is quiet again.
I attach the circuit diagram including my substitutions to see if there's any other error I've made. My assumption was that any opamp will work for boosting the microphone, also for the other stages to the analogue switcher. The author's MCP6004 is apparently able to drive headphones, while my LM358 can't hence the addition of the LM368. I'm not certain I've coupled it to the LM358 properly. I also don't understand the use of VGND. Is the 'V' 'Virtual' and I don't connect it to GND (0V) but leave it floating? I have ordered a couple more microphones, both to check whether mine is broken and because a friend wants a detector...when it's working. See what you think! |
4th Sep 2020, 8:34 am | #14 |
Dekatron
Join Date: May 2003
Location: Edinburgh, UK.
Posts: 3,273
|
Re: Bat Detector with Knowles Ultrasonic microphone problems
The VGND appears to be a half Vcc (2.5v) reference.
Peter |
4th Sep 2020, 12:38 pm | #15 |
Hexode
Join Date: Mar 2020
Location: Rochdale, Greater Manchester, UK.
Posts: 439
|
Re: Bat Detector with Knowles Ultrasonic microphone problems
Hi - one step forward and two steps back here perhaps?
It looks to me as though the output from your 2nd LM358, via the two terminals of the potentiometer, is just being "grounded" since pin 3 of your LM386 is connected to VGND - the "virtual ground" that Peter clarified above. The fact that you are using single supply op-amps means that you have to lift the signal ground to a dc level around half of the supply voltage, otherwise you can only amplify the positive going half of your signal and all the negative going stuff gets lost, squashed by the true ground of the negative terminal of your battery. Feeding the audio signal via a suitable capacitor would also remove the dc level presented from the output of the preceding stage - ac coupling, rather than the direct coupling that you have used. Have a look at some designs using the LM386 that will illustrate this for you. Good luck Chris |
4th Sep 2020, 2:17 pm | #16 |
Nonode
Join Date: May 2016
Location: Cornwall, UK.
Posts: 2,315
|
Re: Bat Detector with Knowles Ultrasonic microphone problems
Thanks, both. I understand now - the VGND reference is set by the two 10k resistors to the opamp in the microphone preamplifier section. This isn't part of the main circuit as published by the author, so I had to find that separately and hadn't twigged the relationship. The signal is then floating on a half VCC DC potential, allowing it to swing both positively and negatively.
All signal aspects should then be referenced to VGND, as otherwise they'll be pulled to 0V. In the original circuit, the wiper of the potentiometer is tied to VGND, as in the attached snippet. Do I really need just a variable resistor to the LM386, or should I use the third leg as a potentiometer as in the LM386 datasheet, where the wiper goes to the input and the resistance track is between the preceding stage and ground? I thought the output coupling capacitor did all the DC blocking I needed, as none was mentioned on the datasheet and the rest of the original circuit doesn't use them, so putting one on the input's another good tip. I'll try a 1µF. |
4th Sep 2020, 4:39 pm | #17 |
Hexode
Join Date: Mar 2020
Location: Rochdale, Greater Manchester, UK.
Posts: 439
|
Re: Bat Detector with Knowles Ultrasonic microphone problems
OK, so C3, 3uF and C2, 100nF will block any dc at the inputs of their following op-amps that may emanate from their preceding op-amps. That's how dc blocking happens, it's a series thing caused by that individual capacitor, which may also have another function eg in your filter, so earning its keep twice over. As you say, the 250uF output coupling capacitor is blocking dc, but only at that point and for the component directly in series with it, ie the speaker.
As far as the pot is concerned, my intuitive guess would be to wire the track from the output of the LM358 to VGND with a capacitor (1uF sounds like a good starting point) connected from the wiper to the input of the LM386 and see how that performs. Good luck. Cheers Chris |
4th Sep 2020, 5:58 pm | #18 | |
Dekatron
Join Date: May 2003
Location: Edinburgh, UK.
Posts: 3,273
|
Re: Bat Detector with Knowles Ultrasonic microphone problems
Quote:
Peter Last edited by peter_scott; 4th Sep 2020 at 6:06 pm. |
|
4th Sep 2020, 6:22 pm | #19 |
Nonode
Join Date: May 2016
Location: Cornwall, UK.
Posts: 2,315
|
Re: Bat Detector with Knowles Ultrasonic microphone problems
Thanks Chris. Really I have to treat opamps like valves, with ac coupling to the grids then? I understand the concept, but have very little experience with ICs though I've made a few valve amps and read quite a lot of theory in my autodidact mode. In my ignorance, I was only considering the DC blocking for the headphones rather than thinking it would cause problems for the opamps.
Thank you for that, Peter. I must admit I'm glad I also found it strange the wiper was connected to VGND in the original - I thought greater minds than mine had drawn this up, and assumed it made some kind of divider to raise the output above VGND without understanding it. At the moment, I've left the potentiometer out of the equation, and connected the low-pass filter to the LM386 via a 1µF capacitor. I have rustling in the headphones, but no response from the microphone, though I've verified it has ~3V on it. I get clicks and buzzes in the headphones if I fiddle with the input to the microphone opamp, so there is amplification happening subsequent to the microphone. I may have bust the microphone. If I apply various kHz signals to the 'IN' point (with the signal generator ground connected to VGND) I get a variety of noises out from the headphones, but no apparent effect from twiddling the frequency control on the 555. I expected hiss until the frequency is 'tuned' but I get whistles and tones of varying pitch, sometimes even changing in pitch as I change the 555 frequency control. There is also some motorboating (well, it sounds more like one of those 'put put boats' that go in the bath) which I put down to inadequate decoupling of the LM386. Apparently it likes VCC decoupling on the pins, and this is on a shambolic solderless breadboard next to a computer so likely to pick up a lot of extraneous noise as well. Definite progress, though! |
4th Sep 2020, 8:22 pm | #20 |
Hexode
Join Date: Mar 2020
Location: Rochdale, Greater Manchester, UK.
Posts: 439
|
Re: Bat Detector with Knowles Ultrasonic microphone problems
Yes, progress as you say, a few thoughts ..... Power Supply:- If you want 3.6V at 120uA from a 5V supply then isn't it going to be (5-3.6)/120x10-6 = 11kOhms resistor that is needed?
You wonder if the mic has been damaged, did you ever apply more than the 5V absolute max voltage to it during setting up the adjustable 28V max boost module? If not then perhaps testing the stages one at a time with your scope might be the way forward? What do you get out of the microphone when speaking into/tapping it? Is the signal about 10 times greater from the output of the MC1458? Sounds like your 555 square wave is running OK, do you get a similar, anti-phase output at pin 7 of the 4066? Drive the input of C2 with a, say 40kHz, sinewave from your sig gen about 1Vpp, do you get similar amplitude twin anti-phase sinewaves at pin 1 and pin 7 of the buffer op-amps? Tweak the 555 potentiometer for about the same frequency squarewave on the scope and look at the output of the balanced mixer - does this vary in frequency and amplitude as you adjust the frequency slightly? Look at the output of the low pass filter, is this a decent audio frequency waveform component of the earlier waveform? Is that now being amplified by the LM386 and audible through the speaker? That would be my fault-finding route through the stages, hope it helps. Cheers Chris |