Bat detector

Hi All,

Edwin and I are slowly wrapping up the development of this winter. I still hope to release the code within 1-2 weeks from now but first we need to test things so you wont end up in trouble ;)

To summarize the most important parts of the development:
- we have added options to extend the hardware with GPS, temperaturesensing and control of the brightness of the TFT-screen. If you want to use these new options you will need to do a bit of hardware work but - aside from the space in the box - this is not very technical. I am sure edwin will explain things in more detail when everything is ready.
- saving of WAV files will be done with extra information in the "GUANO" section of the wave-files. This will allow other software to directly read additional information from the WAV file (eg. time of recording) and also allows processing in for instance the BTO_acoustic_pipeline to get information on the species in a recording.
- both the AUTO_HETERODYNE and TIME_EXPANSION modes will have a new option to change the output on your headphones ... more on that will be shared when testing is done.
- and many smaller changes, often requests from the forum. A more complete list of these changes will be shared later.


kind regards, keep safe,

Cor
 
The forum shows we are already on page 30, but my browser won't open that page. Maybe it will with this new reply.

I did already get some questions about a new revision of the board. I do have to admit I am kind of working in that, but i was not planning to release anything soon.
There are only minor modifications needed to add the teensy-controlled backlight for example. The GPS and thermometer were described already and need a few extra wires but nothing special. For the GPS a metal shield over the Teensy seems important, but is is not to hard to cut away the soldermask if you want to solder that to the board groundplane.

Fot the people wondering about the backlight modification, it requires a simple resistor and capacitor. We did already have this in the code but did not use it because we seemed to but different displays. Today almost every display is sold with Q1 that can control the brightness.

We use pin 17 of the teensy to generate a PWM signal that is driving the display intensity circuit. In order to get a smooth signal and no rapid flickering display with all the noise that comes with it, we add a resistor in series with the PWM line and a capacitor across the display's LED line. I used 1k and 47uF for a first test, but these values are not critical.

The code is not yet released but to make sure people know what to expect form board modifications I am posting the circuit modification here. You could perhaps use the small switch to power the GPS unit, or maybe even switch the detector on and off.

If you do not want to do the modification... you can still use the new (to be released) code. The main advantage is that on autorecord you can choose to put the display in sleep mode. Sleep mode saves about 15mA. You can stil do that and switch the backlight to a dimmed/off state but if you do the display brightness PWM modification the teensy can switch off the display backlight in sleep mode and you can choose different brightness levels, reducing power consumption by working with an even lower backlight setting than just switching between bright and dimm.

Sorry Cor, I fear people will be eager to get the new code soon, but I just had to inform about newer PCB versions which will not be here soon because of only some minor modifications that are needed.

The attached image shows R9 and R12 are not needed any longer, the new parts for backlight control are drawn in blue.



Kind regards,

Edwin

teensybat schematic0.2gpstempbl.jpg

Use the image in this message, i can't get rid of the thumbnail which has a small error. The PWM line in the thumbnail seems attached to the audioboard but it comes for the Teensy of course.
 

Attachments

  • teensybat schematic0.2gpstempbl.jpg
    teensybat schematic0.2gpstempbl.jpg
    149.9 KB · Views: 109
Last edited:
Hi,

As not everybody will be having a TFT with Q1 mounted on the backside (many old models do not have this and are still being sold) there is a simple solution (Edwin shared this with me)
to get PWM controlled backlight also in those cases. I my detector with this type of TFT I have added a small perfboard with this Selection_677.png schematic mounted on it.
That works fine and is not very difficult to build. For those using the same boxes as Edwin has suggested, a thin perfboard (4 lines of connections) can be mounted on the left-side of the box (sideways) below the main PCB. I will share an image of how I have mounted this later.

kind regards
Cor
 
Selection_678.png
This is a (not so sharp) picture of my sideboard inside the housing (green perfboard). The purple and yellow wires are running directly to P17 on the Teensy and the LED pin on the TFT. The sideboard is mounted on plastic distanceholders.

Cor
 
Some good news, after some shipping delays and illness, all the parts have arrived and have been put together and it turns on, can be programmed and even has a nice display. Not so good news is the actual bat detection bit does not work! In either spectrum or waterfall mode nothing is displayed. It looks like there is no signal for some reason. Are there any diagnostics I can try? My guesses would be either I have done something wrong with the microphone or the sound board. I can hear a click on the headphones when I power on or off, but no other sound (in any mode Het/TD/TE). I have turned the gain and volume up to the max! Any thoughts what I have done wrong or how best to test it? (I have tried jangling keys and rubbing fingers in front of the mic, but nothing.)
 
Mein Gerät

Ich habe noch einen zusätzlichen Antennen und Mikrofonanschluss getestet.
Die VLF Empfangseigenschaften sind gut.
Man sieht gut den DCF 60 und 77 Zeitzeichensender.
 

Attachments

  • SDR.jpg
    SDR.jpg
    198.3 KB · Views: 71
Not sure... I can try to find the text online.
The article is by Martin Ossman, who is a forum member. Maybe ask him?
 
Yes it would be, and anybody interested can download/buy a copy of the magazine. But nobody can stop you to evaluate and review :)
 
Some good news, after some shipping delays and illness, all the parts have arrived and have been put together and it turns on, can be programmed and even has a nice display. Not so good news is the actual bat detection bit does not work! In either spectrum or waterfall mode nothing is displayed. It looks like there is no signal for some reason. Are there any diagnostics I can try? My guesses would be either I have done something wrong with the microphone or the sound board. I can hear a click on the headphones when I power on or off, but no other sound (in any mode Het/TD/TE). I have turned the gain and volume up to the max! Any thoughts what I have done wrong or how best to test it? (I have tried jangling keys and rubbing fingers in front of the mic, but nothing.)

Hi,
Ive asked Edwin to help with diagnosing ...

regards
Cor
 
Thanks CorBee. You asked for software improvement suggestions before, perhaps a [self] test option, if you hold down the right push-button during power-on (like the reset option). Not 100% sure what exactly it would do, but could cycle through everything (encoder up/down/press, push-buttons), but importantly for me, play a tone through the headphones. It could also give details of anything coming in from the mic :)
 
Hi,

The problem is that I cannot inject a signal at the microphone itself. Testing most of the functionality can be done by installing an SD_card with a recording of a bat (or anything else) recorded at 192Khz. Those files can be played at the "direct" mode and that will "inject" the audio into the system. But anything before that point (before the audioboard) cannot be tested via software.

regards
Cor
 
Testing most of the functionality can be done by installing an SD_card with a recording of a bat (or anything else) recorded at 192Khz.

That sounds like a good place for me to start! Do you know where I can find a suitable file? (If you have one you don't mind sharing, you could upload it to your GIT repo, with test instructions.)

--
David
 
That sounds like a good place for me to start! Do you know where I can find a suitable file? (If you have one you don't mind sharing, you could upload it to your GIT repo, with test instructions.)

--
David
Hi David,

http://www.batcalls.com/ is a good source, unfortunately I dont know what the samplerate of the files is. But if you load (use original) a file in audacity you can check if it is 192k. Otherwise contact me by pm and I will share a simple file with you.
regards
Cor
 
I'll post some details about the article after takining a nap.. i'm pretty tired and it is friday...
Short: Sound output is via PWM. Input with a ICS40730 microphone, which is, according to the article, difficult to solder.
Martin shows 4 variants of converting the ultrasonic sound to the audible range.
 
For audio output test it it quite easy to play a file and hear it in the headphones.

For the headphones you should use a stereo (3 contacts) jackplug and not a 4 pin headset plug.

The microphone or amplifier can be ruled out by touching the middle if the 3 pins of the microphone connector.
This would give enough noise to show up on the display even if the gain is set to 0.

If the display shows noise the problem is in the microphone.

You can also touch the AF input of the audio board. Set the gain high and toch it with a metal part so you don't touch GND and AF input at the same time.
This should be enough to show something on te waterfall.

If nothing shows there could be something wrong with the audioboard.
If you see something when touching the AF input there might be something wrong with the amplifier.

One of the first things to check is 3.3 volts.


audiotest.jpg
 
Back
Top