[RASMB] Abs optics design question

[Tom Laue]

Hi-
Sorry to be late to the party on this.
The pulsed light source was chosen to allow multi-cell operation. The 
computers and available electronics at the time were not nearly fast 
enough to acquire data from a continuous light source. I have designed 
an absorbance system for the XLA that uses a continuous source and 
computer signal handling based on the AU-FDS electronics (in fact it can 
run simultaneously with the FDS using spare circuits in the FDS 
hardware). Aviv and I went in to the NIH for an STTR to build a 
commercial version of the rapid scan absorbance (RSA) retrofit, but 
unfortunately the proposal was not funded. The reasons for it not being 
funded are major reasons for the formation of Spin Analytical and our 
ongoing development of an entirely new (analytical only) 
ultracentrifuge. Funding for the development is mostly from the sales of 
components... any purchases are deeply appreciated.
In addition to the fact that the XLA was built on the XL preparative 
platform, which dictates many of the design choices, the XLA was also 
designed only for equilibrium sedimentation experiments, not for 
velocity sedimentation.When it first came out, a single scan of a single 
cell took ~6-10 minutes. The 'Continuous scan' option was added which 
sped things up a lot (~1 minute per scan). The design of the XLA is 
remarkable for its time- innovative and fairly robust. It is a shame it 
became too expensive to update and improve it, but the market is just 
not large enough to justify the expense for a company the size of Beckman.
The electronics for the absorbance optics require 8 ms to process the 
signal, so switching to a faster pulsed lamp won't help the scan times.
The interferometer design flaw is entirely my doing. We have built a 
light source that is connected to the rest of the interference optics 
rather than to the absorbance arm. When connected this way, the 'dance 
of the fringes' is reduced by over an order of magnitude, and the 
precision of the data increases concomitantly. Kristian Schilling copied 
our design (literally from scraps of paper) and has built one for his 
machine. He also is putting a large array camera on the interferometer. 
With these two changes, the improvement in the data quality (precision) 
will be between 100 and 1000-fold. Walter Stafford has shown (using the 
E and his own interferometer) that it is possible to do sedimentation on 
~10 ug/ml samples... concentrations below what the absorbance system can 
achieve... and with a precision that is breathtaking.
Just by way of information- The tolerance on the rotor holes is 
+0.0005", -0", and +0" -0.0005" for the cell housings. These are very, 
very tough tolerances to meet. Unfortunately, when a rotor hole is 
drilled larger than this, the cells will slip in and out too readily. 
Our (Spin Analytical) cell housings are all hand lapped to get them to 
this tolerance... cutting a titanium rotor to this tolerance is 
astonishingly difficult. That said, there is no room in the cell 
housings to put a mechanism to fasten the cell tight. I have tried 
repeatedly (and unsuccessfully) to come up with a better cell design. 
The best that can be done is to use a shim (e.g. hair you've pulled out 
in frustration).
The new sapphire windows are great- have you had to reject any because 
of excess absorbance at 230 nm?
Best wishes,
Tom


Hayes, David wrote:
> Hi John,
>
> Just a quick note of explanation:
> My actual question about the sapphire windows is why we have to test
> them rather than the manufacturer testing them.  I am glad to hear that
> all of the windows you have purchased are clear enough to use down at
> 230:  maybe the warnings about absorbing sapphire are out of date now.
> My impression was not to trust sapphire and the buying instructions also
> state that sapphire is not guaranteed to work down to 230.
> About machine design questions, my questions are somewhat rhetorical,
> but I wanted to ask them because the next analytical centrifuge will
> also probably be based on a preparative base and maybe the designers can
> ask themselves if making a few changes might be worth the effort.
> Improved temperature control and measurement is another feature where
> analytical and preparative desired specifications differ quite a bit.
> I think something might be added to the bottom of an AUC cell housing to
> help hold the cell steady in its hole, but only if one were willing to
> make the whole cell bit taller.  The sectors leave room on each side for
> a partial arc:  and though it is certainly easier to manufacture a
> simple round hole in the bottom, it seems possible to put a small
> tightening screw there.
>
> David
> -----Original Message-----
> From: rasmb-bounces at rasmb.bbri.org [mailto:rasmb-bounces at rasmb.bbri.org]
> On Behalf Of John Philo
> Sent: Thursday, June 03, 2010 12:13 PM
> To: rasmb at rasmb.bbri.org
> Subject: Re: [RASMB] Abs optics design question
>
> David is right that it is the pulsed nature of the lamp that precludes
> Andrew's approach, and that the pulse rate of the lamp limits the
> scanning
> rate, but actually the lamp can flash 100 times per second rather than
> 60 as
> David said. My understanding is that there is now a lamp available that
> could go 200 flashes per second, which would be a big help especially
> for
> using 7-hole rotors, but whether Beckman will ever switch to that
> remains to
> be seen.
>
> Regarding David's other questions, the only answer to the first three of
> them is "these are design elements of the underlying preparative XL
> centrifuge and are held fixed in the analytical".
>
> With regard to a cell tightening mechanism, it seems to me there is
> physically no room for such a system in any 12 mm path analytical cell.
> The
> screw ring on the top automatically precludes a tightening adjustment
> screw
> on that end. It might be possible to add such a mechanism in the bottom
> end
> of a special housing for shorter pathlength cells, but there is no room
> with
> 12 mm centerpieces (without raising the center of the cell).
>
> I am puzzled by David's comment about sapphire windows. All of the ones
> I
> have purchased in the last 20 years (directly from Meller Optics, who
> supplies them to Beckman) have had low enough absorbance for use at 230
> nm,
> and more recently many would be usable even at 210 nm.
>
> John
>
> -----Original Message-----
> From: rasmb-bounces at rasmb.bbri.org [mailto:rasmb-bounces at rasmb.bbri.org]
> On
> Behalf Of Hayes, David
> Sent: Thursday, June 03, 2010 8:12 AM
> To: rasmb at rasmb.bbri.org
> Subject: Re: [RASMB] Abs optics design question
>
> Hi Andrew,
> My understanding is that it is not just one cell scanned at a time, but
> one
> sector of one cell scanned at each time.  There is one flash of the lamp
> for
> the reference sector and one flash for the lamp for the sample sector.
> At
> rotor speeds above about 6000 RPM, it is the flash lamp maximum flash
> rate
> (60 flashes per second without overheating) that is the time limiting
> factor
> in collecting absorbance data.  
>
> Besides engineering reasons for choosing the flash lamp back when the
> XLA
> was designed, John Philo did raise the question once if there might be a
> molecular downside to having a continuous lamp shining UV on your sample
> for
> the entire time of the experiment (hours).
> Tom Laue is the expert on this topic of instrument design, but your
> basic
> idea that collecting absorbance data from every cell at once should be
> possible has been confirmed in prototype form by Tom.
>
> Although I know Tom has experienced or heard just about everything, I
> thought it would be fun and useful to have a "reminder" thread about
> inexplicable design decisions for the XLA and add a few questions of my
> own.
>
>
> I would also like to know why the XLA engineers placed all the Peltier
> heating/cooling plates on a circuit board with uninsulated wire traces,
> so
> that when the board heats or cools too fast it warps and short circuits
> (Yes
> this really happened).  Why couldn't the plates be connected
> individually to
> the metal can or why couldn't there be some insulation between the metal
> can
> and the circuit board just in case it warps more than the thickness of
> the
> Peltier plates?  
>
> Why do we have to run the XLA at 0 speed just to get the diffusion pump
> going and get a reliable low vacuum?  Couldn't there just be a switch?
>
> Why does the XLA "wait for diffusion pump cooling" even after we
> replaced
> the diffusion pump with a turbo-molecular pump?
>
> Why can't the interference optics laser be physically connected to the
> interference camera to reduce vibration noise?(sorry Tom, you have
> apologize
> about this before, but maybe some newer AUC people haven't
> heard)
>
> Why can't analytical cell housings have a built in tightening system
> like
> the counterweight? (People have used hair, white-out "paint", and
> kimwipe
> strips to hold the cells a little tighter in the rotor)
>
> Why can't the manufacturer test the sapphire windows for Low UV
> absorbance
> before selling them to us?
>
> David Hayes
>
>
> -----Original Message-----
> From: rasmb-bounces at rasmb.bbri.org [mailto:rasmb-bounces at rasmb.bbri.org]
> On Behalf Of Leech, AP
> Sent: Thursday, June 03, 2010 10:39 AM
> To: RASMB
> Subject: [RASMB] Abs optics design question
>
> Hello all,
>
> This is an idle question about the design of the XL/A (or I) absorbance
> optics. Why are the cells scanned one at once?
> Would it not be possible to scan them all "simultaneously"
> by gating the output from the photomultiplier into a suitable number of
> different channels?
>
> Andrew
>
>   

-- 
Department of Biochemistry and Molecular Biology
University of New Hampshire
Durham, NH 03824-3544
Phone: 603-862-2459
FAX:   603-862-0031
E-mail: Tom.Laue at unh.edu
www.bitc.unh.edu
www.camis.unh.edu