All things being equal, and they seldom are, the noise on the TUV is roughly 1/2 of the PDA meaning that a signal to noise value (i.e. sensitivity) is better on the TUV.
The sensitivity difference with the PDA and the TUV is the PDA arises because the PDA must use a bandwidth establish the 1.2 nm resolution across the diodes, to give high quality spectral resolution while a UV can have a larger bandpass (about 4nm). Essentially the UV has more light and hence a lower noise floor. That is the same as in HPLC. However, you can improve the perfromance if collect in addition to 3D data a 2D PDA channel with a 3.6 nm or 4.8 nm bandwidth set and mimic the TUV. You can still collect 3D data at the necessary spectral bandwidth and perform peak purity etc.
This is interesting as we have debated purchasing tuv for our next aquity system (all aquity pda for now) because we observe significant differences in noise/sensitivity with alliance systems (at the same bandwidth).
So what you are saying is that observed sensitivity and noise differences are due to the thoughput of light in the detector due to bandpass?
If so, this must mean that the combined gain and noise contributed from the different detector types (PMT vs PDA) and electronics are on average the same. ill assume that the lamp energy and efficiency of the dispersive optics are similar .
Also with alliance 2996 pda you may choose bandwith on individual channels but with aquity pda's the option is resolution. Do you use bandwidth and resolution in a synonymous manner?
Is my thinking correct on the following?... If you collect 3d with 1.2nm resolution the slits are set to give an effective banpass of 1.2nm per pda diode and if at the same time you also chose a 2D channel with 4.8nm bandpass (or bandwith) then the signal from 4 diodes is used (ie constant slit with)?
Kinda and sorta. Its my fault!
Waters optics engineers will give me heck if I do not say we try not to be shot noise limited. SO long as the lamp is not due for a change the noise of the electronics is not limiting.
The point that I was attempting to make is that a TUV is always better in sensitivity than a PDA. True of ACQUITY and HPLC, the TUV sends all the light through the grating and then into the flow cell and picks out just 4 nm bandpass of light to monitor off a single diode. The bench is optimized for collecting a single bandpass, for Waters about 4 nm. However, the PDA sends all the light through the flow cell and then monitors all the wavelengths at once with a series of smaller surface area diodes that make up the PDA device. Even if every particle of light were transmitted ideally some diodes get less or more light depending on the spectrum of deterium at that wavelength. The Waters' PDA have a strategy that allows for variable sampling across the diode, (longer for those that see less), but still its not as efficient as the optimized TUV. That's why the specs for PDA and TUV differ. However, just as you say, if you bunch the diodes 3.6nm for your 2D data for example instead of looking at just the diode resolution you can get a better result.However, keep the spectral resolution at 1.2 nm so your spectra is good and peak purity it working well.
Now I would like to know what differnces you are seeing between the Alliance and the ACQUITY, I am not sure taht that will be entirely due to the detector!
thanks for the reply
I never really thought about the optical bench of a detector being optimized for a particular bandwith or the relative sizes of diodes in different detector types.
Sorry, i dont have direct Alliance/Aquity comparisons since i am not transferring methods between the different platforms. I have compared alliance with other high dispersion systems and my coworkers have compared Aquity with other low dispersion (UPLC type systems).
PS it has been a long time since i thought about Schottky noise.. "we try not to be shot noise limited" does that mean the systems are capable of sub-shot noise levels or that systems are designed so at the intended signal levels shot noise does not dominate the measurement?