Try to use liquid FCA (fixed tip) in our fluent for a charge varients assay. Two questions:
When use water free multi liquid class to aspirate and dispense 450ul water into 2ml micro centrifuge tubes, we noticed it dispensed about 470ul. What and where should we start to optimize?
Also noticed when dispensing water, there are small water drop residues on the side of the liquid FCA fixed tip. We tried to keep submerge as low as possible to be 1mm with liquid level detection but it didn’t help.
I, personally, have never found the included liquid classes to be very useful out of the box. They all need tweaking. It’s been awhile since I’ve used a Fluent but with Evoware you can’t edit the default liquid classes, so you need to copy them and change them for your use.
Again if I recall correctly, one thing to note is that the “multi” methods are for multi dispensing. I.e. you aspirate 500uL and then dispense 100uL five times. There are some considerations when doing that so there’s probably a large offset or even just an overage applied in that liquid class, that’s why you’re getting 470uL instead of 450uL. Again, I would copy that method and edit it for my use until I got what I wanted.
To help with the water drops after dispense, if you are dispensing submerged in liquid you may be able to tweak the speed at which the tips retract from the fluid. A slower retract speed may help limit the droplets formed on the tips.
Pictures, screenshots, videos would help tremendously. There are a lot of experts here that can help and liquid classes have always been the black art of using a liquid handler. There are a lot of considerations and I’ve found that FluentControl has settings that you can tweak that solve pretty much all of them (or the other way around, you can tweak settings to make problems for you).
Tracking variables within microscripts can be very valuable for troubleshooting these sort of things. For example, you can track aspiration and dispense volumes or speeds etc to get a better sense of how variables are handled within the LC.
Most likely, the situation you’re running into is like @jnecr identified: you’re using a multi-dispense class which probably includes conditioning and overage that might make up the volume difference you’re seeing. If this is a used or “hand-me-down” instrument, there may be offsets already defined that you’d need to remove and redefine. There’s usually an accuracy adjustment variable - I usually save the equation and clear this out to first teach LC and reduce droplet formation then add it back in for fine-tuning the final few µl volume difference, if any.
Also, droplets on the SIDE of the tip can be indicative of too fast dispense speeds, or some sort of breakage in the water stream due to liquid compression, gas in system liquid, or particles leading to sputtering. If there’s any droplets at all, its typically best if their on the bottom where modulating pipetting parameters can actually affect and change the droplet formation or breakoff.
Try to identify the source of the problem and that will better form solutions. Take some videos in slow motion if you need to, and collect all the info you have available before troubleshooting.
I will try not to repeat @evwolfson and @jnecr too much, both have great points.
Using Water Free Single if you are not multi-dispensing is a clear advantage.
Fixing droplets can take some time and trial and error @jnecr makes a good point here.
If you still see problems with inaccurate dispensing you could change the “Accuracy Adjustments” under Formulas
(Both under Aspirate and Dispense and remember to do it under the correct Tip Type - I have made the error of changing things in the wrong Tip Type too many times)
The AdjustAccuracy() function can look a bit complicated at first (example below) but its not too bad once you get it:
AdjustAccuracy(volume,5,2.889,10,7.014,50,44.724,100,92.806,500,480.259,1000,967.127)
The first pair of numbers (5, 2.889) represents if I try to transfer 5µl (without correction) I really get 2.889, the next pair (10, 7.014) is the same. If I try to transfer 10µl I will transfer 7.014µl, and so on.
The AdjustAccuracy function will then adjust the actual transfer volume target depending on what volume you are trying to transfer.
You can then either try to adjust the number after 500 up in order to decrease the correction in the area.
There is also an Excel sheet for AdjustAccuracy if you want to get more systematic about it, but make sure to fix any other issues about the liquid class before you do this work.
Don’t hesitate to reach out if you need help or if I did a terrible job explaining myself.
FC 2.5 is quite and old version now so I would look to update your software as the liquid class database is continually being updated to make improvements. With multi-dispensing the first and last dispenses as discarded as these are the least accurate when multi-dispensing.
So you should notice that tips will aspirate a much larger volume than required (base on air gaps and excess required). The tips will actually dispense the first dispense excess back to the source wells before carrying out the multi-dispense in the destination plate. It is important that there is an excess still in the tip so the last dispense is accurate and so it is always recommended that the tips are emptied before repeated any further dispenses. If not you can get a build up of solution and poor airgaps preservation that results in dripping tips, which may look like droplets on the outside of tips.
In terms of optimization the advice above is sound but I always start optimization with speed of dispense and breakoff speed as it could be just a matter of viscosity before I move to accuracy adjustments. But maybe this is a hang up from how I worked on STARs and EVOs. Aspirate slow and dispense fast as a very general rule. However, I think the Fluent’s microscripts and calculated accuracy adjustment alone are sometimes better to fine tune for a specific solution.
Thank you all for the great suggestions. i did manage to find a way to reduce the retract speed. Another question i have is, if i want to have cLLD for dispense into a empty well, is there a way to avoid the error since it is not detecting any liquid from the well initially?
@Jay.liu I think I understand your question. You don’t necessarily need cLLD on for dispensing. You just need dispense at z-max with tracking. This avoids splashing and moves the tips up while dispense to limit solution contact with tips. You can look at the dry contact liquid classes as these will have this preset so might be a better class to optimise from.
Agreed with @MortenSkovsted If you are serious about optimizing the accuracy adjustment I would recommend using the excel file on the installation disk for Fluent Control. The latest on is here: AdjustAccuracy Function 1.4. This will help you build a formula.
I would also recommend pressing F1 when in Fluent Control to access the building manual as it also list out the current Functions and Operators available as well as how each command works.
We also tried turn off cLLD but keep tracking on. Since we set z offset to be 3mm above z max, the tip is not moving with the liquid level. Wonder what setting should we do. Should we do without z offset?
Sorry these extra liquid class can be imported from the installation disk/USB. The dry contact is one I modified myself by just dispensing at z-max and tracking up setting so not difficult to re-create and tune. I forget which one are out the box in Fluent Control.
Just for clarify a Wet contact or contact Wet dispense is normally when you have a well with solution already in the well. Whereas a Dry Contact (normally multi-dispensing) is in to bottom of a dry well.
thank you! Curious to see what is the setting for contact dry dispense? below is our current setting for dispense. We did turn tracking on (“yes”) but i think we didnt see the tip moves with liquid level (instead the system tracks the volume dispensed i think). Are you doing differently with Z position and Z offset? Theoretically i am not sure how the tip can move up with liquid level without cLLD on.
You have to remember there is always a software definition of a block or plate and it’s well or cavity so it’s important to get that definition correct so your tip can track properly according to the well. You can also use FC to calculate how much volume is in a well and even use it to approximate the height it needs to be at if for example you that the well has had 50uL of some liquid dispensed into it. For the last bit you don’t need cLLD at all but the system can behave as if it is using it.
that would make sense, but i have not figured out a way to enable tip retract during dispense when cLLD is off. My understanding is when cLLD is off, the tip is dispensing at fixed location defined by Z position and Z offset (the submerge field is greyed out). the Retract properties only control post-dispensing if i get that right.
There may be something funky happening in your microscript. just curious but if you temporarily change the dispense height to z-dispense, do you see movement?
