# Bug: First 1-4 analog samples are 0 and ramping up before data is reliable

**URL:** <https://discuss.saleae.com/t/bug-first-1-4-analog-samples-are-0-and-ramping-up-before-data-is-reliable/3109>\
**Category:** Logic 2 Software\
**Created:** [October 17, 2024, 9:33am UTC](https://discuss.saleae.com/t/bug-first-1-4-analog-samples-are-0-and-ramping-up-before-data-is-reliable/3109 "2024-10-17T09:33:20Z")\
**Posts on this page:** 3\
**Page:** 1

<div class="post-metadata">

**Author:** ![mthu](https://avatars.discourse-cdn.com/v4/letter/m/58f4c7/32.png) [@mthu](https://discuss.saleae.com/u/mthu)\
**Post date:** [October 17, 2024, 9:33am UTC](https://discuss.saleae.com/t/bug-first-1-4-analog-samples-are-0-and-ramping-up-before-data-is-reliable/3109/1 "2024-10-17T09:33:20Z")

</div>

Hi,

I’m seeing an issue where the first few analog samples of a well-known voltage source is “ramping up” before the samples are actually reliable. The amount of unreliable samples depend on the sampling rate. I see this for all sampling rates on the analog sampling except 50 MS/s.

There’s also a general pattern that the first analog samples arrive before the first digital sample, and all samples after the first digital sample are “good”.

Using this for automated testing means we have to manually try and filter out the noise-data, which is frustrating - especially since we don’t know the root cause, and are merely guessing on what the right way to do this is. We see this both in the data exported through the automation API and straight in the GUI, as well as csv dumps.

Example from 50 S/s:

 ![image](https://canada1.discourse-cdn.com/flex030/uploads/saleae/original/2X/1/17df5addf449db1c076367696f824ad2f9994e12.png)

Examples from 12.5 MS/s:

 ![image](https://canada1.discourse-cdn.com/flex030/uploads/saleae/original/2X/8/89066693d9865428c886db13dbebef8885698a50.png)

Software-version: 2.4.14  
HW-revision: 1.2.0, Logic Pro 16.

Can you tell me more about this?

Best regards,  
Martin

---

<div class="post-metadata">

**Author:** ![BitBob](https://yyz2.discourse-cdn.com/flex030/user_avatar/discuss.saleae.com/bitbob/32/2292_2.png) [@BitBob](https://discuss.saleae.com/u/BitBob)\
**Post date:** [October 17, 2024, 2:13pm UTC](https://discuss.saleae.com/t/bug-first-1-4-analog-samples-are-0-and-ramping-up-before-data-is-reliable/3109/2 "2024-10-17T14:13:24Z")

</div>

Do you have a network connection on your PC?

See the warning:

 ![Screenshot_20241017-100018](https://canada1.discourse-cdn.com/flex030/uploads/saleae/original/2X/f/fd0b695708c3b3f2ae66da967dfa475bced829b3.jpeg)

… looks like you don’t have a calibration file installed for your specific Logic hardware.

You can manually download/install the calibration, following:

> **[Device Calibration](https://www.saleae.com/support/specifications-hardware/electrical-characteristics/device-calibration)**
>
> When connected to the internet, the Logic software will automatically handle downloading the calibration file. In rare cases, users may need to download the ...

If you want a deeper dive into the analog filtering & calibration settings, see this post:

> [@Saleae Logic2 Analog filtering technical details?](https://discuss.saleae.com/t/saleae-logic2-analog-filtering-technical-details/2427):
>
> I am trying to understand the technical limitations of Saleae Logic’s analog inputs. I am aware of the support articles: “In addition to the hardware AA filter, there is a digital-analog filter that engages for sample rates lower than the advertised sample rate. This is called the decimation filter. It will filter the data further before down sampling so data sampled at lower rates does not suffer from aliasing from frequency components that made it through the analog AA filter.” “To accu…

(You can optionally ‘hack’ the calibration file to increase the analog bandwidth / reduce the filtering applied)

However, without any calibration file, you can get strange behavior as described above. In particular, you likely don’t have a `mGroupDelay` setting that aligns the analog/digital timing for your specific device/filtering settings, nor do you have a `mImpulseLength` setting that should ‘trim’ the leading analog samples and avoid the behavior you’ve described.

---

<div class="post-metadata">

**Author:** ![mthu](https://avatars.discourse-cdn.com/v4/letter/m/58f4c7/32.png) [@mthu](https://discuss.saleae.com/u/mthu)\
**Post date:** [October 17, 2024, 2:38pm UTC](https://discuss.saleae.com/t/bug-first-1-4-analog-samples-are-0-and-ramping-up-before-data-is-reliable/3109/3 "2024-10-17T14:38:39Z")

</div>

Hi Bitbob,

Thank you - we’ll look into it. We’ve not used calibration so far as the units are not connected to the internet on purpose, and we don’t need absolute accuracy, rather relative accuracy.