I would interprete that the actual dispense of liquid is completed after ~420 ms, after the curve dropped down to ~600 Pa. But what happens after that? In most pipetting steps, the pressure stays at ~600 Pa, but some ‘start jumping’ or even the pressure drop stops at ~1000 Pa. Some curves finish with > 300 Pa causing an TADM error.
The used dispense type is jet part volume; 100 µl are aspirated and 100 µl are dispensed (no pre and post aliqots)
The method with the mentioned pipetting step runs on several machines, each showing the same effect. Therefore, we don’t think it is a hardware issue.
Do you know this phenomenon? Do the pipetting results of transfers with suspicious TADM curve differ from all other pipetting results? What might be the liquid class (or pipetting) property to reduce or stop it?
@Martin - You can find guidance on interpretation of TADM curves in the TADM programmer’s manual as well as the Liquid handling guide in the Hamilton materials share.
The phenomena you have pointed out at the end of some dispense curves is called fraying, and is generally normal and not cause for concern. If you are using the guard bands for runtime dispense, monitoring, I would recommend raising the final point of the upper band so these don’t trigger errors.
The curves that fray at the end don’t necessarily indicate a quantitative deviation in dispense volume, but it is difficult to be certain without running a gravimetric or photometric protocol to measure volumes.
Given that these are consecutive jet part dispenses, I would check the curve data to see if the transfers with the frayed ends happen to be the last aliquot in the cycle. This may indicate that a minute amount of air is being introduced at the very end of the last dispense after plunger movement stops. This could be completely inconsequential.
If this does indicate a shortage of final aliquot volume, then this could be addressed with a correction curve point adjustment in the liquid class, prewetting the tip upon first aspiration (if not already), or adding a prealiquot volume.
Again, this curve behavior does not mean the transfer is not working as intended. I am just trying to provide the best insight I can without being in front of the system. If your assay is running robustly, this is likely little cause for concern.
Hi @NickHealy_Hamilton and @Martin,
thanks for these fascinating insights! What I would also be interested in, because I have observed it several times: why do a few of the curves remain at a constantly high pressure level after fraying at the end while others remain on a lower pressure level?
@max - I am actually not 100% sure. I will need to inquire internally so see if I can get more information and insight. I will let you know if I hear anything useful.
I’m just starting to learn about using TADM to monitor methods. I have a method that is not being particularly accurate so I enabled monitoring on 2 different Hamilton to compare results. Why are they so dramatically different? The step is aspirating 10uls with 10ul tips.
Can you confirm both pipetting steps are using identical liquid classes, liquids and parameters? Any variance will change the overall results of the TADM curves. Are the tips also fresh for both examples? It looks like some of the tips are either used already or have had a pre-wet step. The ones marked appear to be from fresh tips or what I would expect from fresh tips. Wetted tips will always look slightly different from dry tips on the TADM graph.
Yes the steps have identical liquid classes, liquids and parameters as well as fresh tips. The tips are unfiltered and straight out of the box.
One key difference I forgot to mention in my initial post is that the first graphs are from a CO-RE II hamilton and the second graphs are from a CO-RE I. The tips in both cases are CO-RE II
That’s interesting to hear. We have a new lot of unfiltered tips and the inaccuracies I’m seeing started with the new lot. Unfortunately I don’t currently have another lot to compare with.
but if you use unfiltered tips, it must be a different reason because in our case, the differences in TADM curves were caused by the filter material of different lots.
