K8118 ultrasonic bat detector

Okay will try it and let you know what the ears say :zipper_mouth_face:

1 Like

Interesting test. - Waiting for the results.

2 Likes

Hello

Some results.

To my ears the TL974IN doesn’t reduce noise rather it changes the sound character for the worse i.e clipping/duller. will have to compare again though.

One thing unexpected on the breadboard (with TL274IN) is I seem to be getting oscillatory sounds with some higher volume settings and depending on the frequency setting :sweat_smile:

I’m starting to wonder if perhaps the above is caused because of the specs of the electret I have on our breadboard which is a BCM9765P-44 and its interaction with the 15k resistor at freq pot stage.

If i replace the 15k resistor for a 100k I get much less of the oscillation noise but an additional a high frequency whine is introduced. 47k tames the oscillation somewhat + eliminates the wine :laughing: but i think the response is less ideal than the 15k.

An observation: when I eliminate-jumper the 15k resistor, general noise disappears drastically( im referring to the general circuit noise discussed in the thread) but the frequency is fixed.

Otherwise I would say its a partial success although compared with the k8118 the breadboard attempt doesn’t sound as sensitive + oscillation depending on the volume/frequency combination.

So might this be my electrets bandwidth 20-16khz or impedance characteristics misbehaving with the 15 k?

In this case what would be better characteristics for the electret? Presumably the electret used on my K8118 is the 6mm M63 microcel of which I cant see a bandwidth spec?

Bat thanks!

BCM9765P-44 specs

Ok. It seems that changing the amplifier to low noise version doesn’t make a big difference…

1 Like

yeah was fun to try swapamps :exploding_head:

doing some further research the M63 seems to have a bandwidth of 20-16khz and has similar specs to the BCM9765P-44 despite differences in size

just wondering then, shouldn’t that range be higher say up to 30khz for bat calls? I had assumed that the M63 is used as i came across it hunting on the vellman site.

any idea of what might be making the bread-boarded version sound less sensitive, would the M63 help? although if its the same spec i cant see how :thinking:

anyway we have still enjoyed building it so fa,r both the kit and the legwork required for bread-boarding :slight_smile:

thanks!

It seems that maybe the sensitivity of the M63 is slightly higher on high frequencies…
Anyhow, your breadboard version seems to work fine!

1 Like

Yes it is working :partying_face:

Tested in the back garden last night side by side with the k8118 its not as adept at picking up calls though. But as you point out I will try to get an M63, maybe I can scare two bats with one electret, fixing the sensitivity issue as well as the oscillation at the same time.

There are some other elements which spec at 20khz bandwidth, might also be worth a try.

Thanks!

Indeed, maybe even better than M63…

Please improve the documentation on microphone installation. Assembly does not mention making jumpers for the 4 sets of holes nor the 4 jumpers.

The elements are not marked plus or minus so overheating/resoldering can cause damage. By negative you also mean ground as it is called in the company documentation?

A clear image of the back of the microphone with arrows would help and a shot of the bracing prior to soldering.

I had to stop assembly at the last stage to research this. The forum is not as clear on this as it could be. I have built many other kits

Hello @Sealion ,

hopefully this is clearer?

image

Best Regards,
Velleman Support