To test this feature, visit your live site.

Edited: Oct 23

Same config, same batteries, same SD, vastly different recording times? Audiomoth stopping recording after approx. 60 hours regular issue

I am having a recurring issue where my Audiomoths are stopping recording after only 60 hours at 48kHz. I believe it may be associated with the Enable low-voltage cut-off : Yes setting. Previously you could turn this off in the config, but now it is no longer an option? Below are two deployments with identical batteries, SD cards and configurations. Energiser Pro Max Plus Batteries, Sandisk Extreme 64GB cards. One records for only 58 hours, the other for 166 hours (the amount of time I would expect). This issue happened to me on almost 30 recorders in my last deployment, and I decided it must have been a bad batch of batteries (was using Varta) so I swapped to Energiser Pro Max Plus and the issue still persists. Any ideas why? Summary files and config attached. The early stoppage seems to occur when the battery drops to 3.5v, but this is happening far earlier than it should right? v1.2 Audiomoths. Previously have had no issues getting 160-180 hours out of normal alkaline batteries. Any advice appreciated, thank you!

14 comments

14 Comments

Noëlle Tschirren

Sep 25

Hey Alex and Callan, I currently have 50 audiomoth deviced deployed and just took in the 3 first ones to check. They all have the same settings, 2 of them worked fine, one of them stopped writing files after 60h of good recordings. It still wrote 5 files after, but they are empty. The one with the problem has a SanDisk Extreme 64Gb from Taiwan, the other two are a SanDisk Extreme 64Gb and a SanDisk Pro 64Gb both from China. I will go to the fieldsite tomorrow and check all of them for errors. Also I set them without any sleep duration, so the writing error might also due to that and the card is just a coincidence. Please let me know if you have any updates!

Noëlle Tschirren

Oct 05

Replying to

Hey Alex, had a look at all audiomoths I have now, and it's as expected the Sandisk extreme cards made in Taiwan that cause the issue. I had a look at the summary and the batteries just drain insanely fast with these cards compared to the ones made in China. The only solution for now is to change the batteries frequently and get new ones made in China.

Noëlle Tschirren

Oct 05

Replying to

Hi Callan, we bought them through groupgets in the Netherlands and received a mix of Sandisk extreme made in China, made in Taiwan and also some Sandisk Pro cards. Now I am purchasing new cards from Harvey Norman as this is the only shop close to my fieldsite that I could find that get their cards from China. Did you hear anything from Sandisk on a potential refund? I'll definitely try that as well. As much as this issue sucks I am very glad to have found your post as it would have taken me ages to find out that the cards made in Taiwan are the issue- The print on the back of the card is just so small. So thanks for that!

Alex Rogers

Oct 05

Replying to

Thanks for tracking this down. It's very fustrating as SanDisk Extreme cards have always been very reliable and low-energ in the past. GroupGets are checking their stock. The ones that they've looked at so far seem to be made in Malaysia. We'll try to get examples of all of them so we can measure their energy consumption.

Alex Rogers

Sep 22

I’ll get some SanDisk Extreme PLUS cards to test. SanDisk Extreme Pro cards had very high write currents and generated noise in our tests. We generally only use regular SanDisk Extreme cards.

Alex Rogers

Sep 19

Hi,

I don't think this is a firmware issue. We removed the option to disable the low-voltage cutoff as some people were experiencing damage to the SD card when they ran the battery down very low without it enabled. The low-voltage cutoff is implemented by measuring the supply voltage to the microcontroller and is set to 2.8V. The supply voltage typically drops below this during SD card write operations with heavily depleted batteries.

In this case, the final recordings experienced SD card write errors that prevented them from being closed, indicating that in both cases the supply voltage probably was quite low.

The battery voltage measurement is made before the recording starts with a battery that is quite lightly loaded. The full recording shows a typical profile for an alkaline battery. The early stop recording is very different. The second recording each day shows a much lower battery voltage despite the higher temperature later in the morning. This looks like a depleted high internal resistance battery. Another possibility is the SD card. This is often where variation between expected and actual deployment lifetime comes from. However, if these are all SanDisk Extreme cards I would think this less likely. Double-check the cards to make sure they all look genuine though - there are fake cards in circulation which have poorly printed labels. The fact that there is such a difference in battery voltage between the first and second recording makes me suspect the batteries though. It would be good to try the setup with a more energy-intensive profile - making 55-second recordings with 5-second sleep periods between them at 384kHz - so that you can reproduce this much more quickly on the bench. I would imagine that you should see the difference within a few hours. If you have a good digital multimedia you should be able to measure the current being drawn by the AudioMoth whilst making recordings. If you set a 200mA scale, and switch to DEFAULT, it will make a continuous recording and you will see the average consumption. You can double-check for differences between the AudioMoth and between the SD cards.

The set-up above is supplying 4.0V through a digital multimeter. Average consumption whilst recording at 48kHz is about 14.5mA as expected.

Whilst waiting between recordings the energy consumption will be about 100uA on average with a maximum close to 1mA each time the LED flash.

Or if you can ship one of the devices to andy@openacousticdevices.info, he can check it.

Alex

Edited

Alex Rogers

Sep 21

Replying to

Interesting! All my SanDisk Extreme cards say 'MADE IN CHINA' on the back.

Mike Bennett

Sep 22

Replying to

Alex

FYI. The two Extreme PLUS cards I bought in the US both say "Card made in Malaysia" and "Adapter made in China" I have had no issues so far.

callan alexander

Sep 25

Replying to

Hi Alex,

I measured the current and the 'Made in Taiwan' cards seem to draw more than 3x the other cards. Approx 40mA vs 14-15mA. I also checked the capacity using H2testw and they are actually 64gb. All of the SD cards that were made in Taiwan from this deployment exhibited this issue.

I am not 100% sure which supplier gave me these cards, will have to go back and count the exact number that I have and correlate with my purchasing records. The only places I have sourced my cards though are from large legitimate Australian retailers or the ones that came with my GroupGets order so I have no reason to believe they would be fake. I am going to contact Sandisk and see if I can verify their authenticity.

Thanks very much again for your help, will let you know if i figure out if the cards are legitimate.

*update - sent the card images to Sandisk support who have said that the card is genuine

*another update - have confirmed which supplier I bought them from, one of Australia's biggest tech retailers, should be completely legit. Going to see if I can get a refund through Sandisk

Edited