[FDS] FDS noise - revisited

[Tom Laue]

Hi all-
Very odd!
We haven't seen periodic noise like this in the past, but I will double 
check our more recent data.
The periodicity is particularly strange- Do all of the data do this, or 
just one cell?
The fact that the periodicity persisted over the entire run is even more 
odd.
You shouldn't have to remove it... it shouldn't be there!
Let me think about what might be causing this.
Best wishes,
Tom

Borries Demeler wrote:
> Greetings, everyone:
>
> Virgil and I had the opportunity to take some successful measurements (this time
> with SYBRgreen, which gave great results when measuring phages.
>
> After fitting the data using SA2D and getting a time invariant noise profile,
> noticed now a very clear periodic frequency in the time invariant noise that I was not
> able before to determine with FFT (panel A, the black line has the low-frequency
> noise removed for easier viewing). But now I get a clear peak at ~33 HZ, which 
> you can see clearly in the ti noise pattern file as a period of ~0.03 cm. (panel B).
>
> The amplitude in my particular experiment is about 1-5 counts in each direction.
> The noise gets larger towards the bottom of the cell.
>
> Any ideas on how to remove that?
>
>
> -Borries
>
>   
>> Hi all-
>> Glad you finally are getting decent data. A voltage of 2222 corresponds 
>> to ~55% PMT setting (2222/4095). Since the signal climbs rapidly with 
>> PMT setting, this signal level is about what I would expect. The 
>> baseline is fine- we have it offset from zero by about 200 on purpose. 
>> The noise is somewhat higher than I would expect. The usual noise is 
>> about 1% of the signal, and you are getting more than 10%. Perhaps the 
>> data acquisition angles are offset or the rotor timing jitter is higher 
>> than normal...
>> While it is possible the noise has a periodic element, I am not sure 
>> where it would come from. At very low signals, we sometimes get a 
>> sawtooth pattern due to the +/- 1 bit digitization "noise." Let me know 
>> if you do the Fourier analysis on your signal.
>> I completed a first-pass at a wizard to help with aligning the dichroic 
>> mirror (misalignment of this is the major cause of low signal), and hope 
>> to test it this week before releasing it. Also, I am looking into noise 
>> sources on the rotor timing pulse to see if the jitter can be reduced- 
>> that work will not start until next week when our machine is less booked up.
>> Best wishes,
>> Tom
>>
>>     
> When you say you would expect a noise of 1% then I really should be seeing
> noise in the level of +/-5, instead of +/-50, right?
>
> I did the FFT of 2 scans from channels A+B, and neither one shows anything
> that I feel is significant. There is a small consistent peak at iabove 0.3 HZ (using
> a sampling of 1, I don't know the speed of your scanner), there may be something
> to this, but not sure if it is significant. I thought there would be something
> with a shorter frequency in the noise, like rotor jitter maybe?  I attached a 
> couple of FFTs.
>
> Still, now that we have respectable signal intensity, I would like to
> know how we could get our noise down to 1% of signal instead of 10%.
> The 1% noise level you see - does it uniformly apply to all signali levels?
> I.e., does the noise scale with the signal? It seems to me that the noise
> should be constant with constant gain setting, regardless of signal strength. 
> It seems to be that way at least on our machine, because our 600 nM
> samples the noise is still about +/-50, with a total signal of ~2700,
> which is more like 4% (same gain and voltage).
>
> So any suggestions on how to reduce the noise we are seeing?
>
> Thanks, -Borries
>   
> ------------------------------------------------------------------------
>
> _______________________________________________
> FDS mailing list
> FDS at biochem.uthscsa.edu
> http://biochem.uthscsa.edu/mailman/listinfo/fds
>   

-- 
Department of Biochemistry and Molecular Biology
University of New Hampshire
Rudman Hall 379
46 College Road
Durham, NH 03824-3544
Phone: 603-862-2459
FAX:   603-862-0031
www.camis.unh.edu
www.bitc.unh.edu


-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://biochem.uthscsa.edu/pipermail/fds/attachments/20061001/56b7d958/attachment.html