Xuhua Zhou's  Instablog

Xuhua Zhou
Send Message
I am currently an individual investor with focus on event-driven trading and long-short opportunities. I graduated Emory University in 2009 and am also a finance Phd dropout from UCLA Anderson. I could be reached at zhouxuhua@gmail.com
  • Unipixel -- Broken Lines Plus Broken Words Add Up To A Broken Company 11 comments
    Apr 18, 2013 4:16 AM | about stocks: UNXL

    As investors continue to push the stock price of Unipixel (NASDAQ:UNXL) higher and higher, Unipixel's management has finally pulled the trigger on a secondary offering, after months of repeatedly reassuring investors that there will be no dilution. Much has been said about the despicable track record of management credibility. The most recent share offering appears to be simply another data point to be added to a long list of broken words promised by the management to investors. To avoid overloading investors with the same information, this article will focus on the other broken side of the Company -- the broken lines.

    Below is a picture shown in the Company's most recent investor presentation.

    The above picture from Unipixel's own presentation show the lines that make up the metal grid are broken, making this grid not usable as a touch sensor. The fact that Unipixel cannot even find a good image of their grid to present says a great deal about the state of their production capabilities. Visitors to Unipixel's facility have looked at their product under a microscope only to find that these line breaks are everywhere. Unipixel emphatically denies these line breaks are a problem, but followup calls to sensor manufacturers in Asia verify that such breaks make this film useless for a touch sensor. Line breaks will either cause some areas to be electrically disconnected, or cause the conductivity of the grid to vary to such a extent that accurate touch detection is impossible. Both cases lead to a touch module that can't sense touch.

    Why can't they make lines without breaks? Even for good pictures? Yield losses (as high as 20%) during wet patterning processes is a major driver for adoption of laser patterning for ITO, and that is for extremely refined, decades old processes for patterning ITO on PET at a line width of 100 - 200 um, more than 10 times larger than the sub-10 um line widths that Unipixel is targeting. NO ONE has EVER reported sub-10 um patterning using flexographic printing that Unipixel is using. For an overview of different printing methods, see this presentation from Xerox Parc. Flexographic printing is basically using a rubber roll to print ink on a substrate.

    (click to enlarge)

    The following table from the presentation presents a minimum feature size of flexographic printing of 80 um, again more than 10 times larger than the line-width that Unipixel aims to achieve on an extremely reliable basis.

    (click to enlarge)

    Independent university researchers have published very similar results. In the graph below they relate the "nominal grid track width", which is the line on the stamp, or "Plate Cylinder" in the figure below, to the actual printed line width.

    Why does a 20 um stamp line turn into an 80 um printed line? Because the liquid ink spreads out across the substrate, just like the ink from a ballpoint pen. The resolution from this printing process is fundamentally limited by how much ink is transferred to the substrate, and how much is spreads out ("wets"). This line width limitation has simply not been solved for flexographic printing, especially in the sub-10 um range. If it had, why would anyone use photolithography at much larger line sizes? Unipixel may have a stamp with a nominal line width of 5 um, but it is not possible to print sub-10 um lines roll-to-roll (or 3M would already be doing it).

    So what about the Unipixels pictures of sub-10 um lines? Three things:

    1) We don't know how those were printed.

    2) We don't know if the scale is correct.

    3) We don't care because either way, they are broken.

    What if they can't print such small lines? They can't compete and achieve the same performance as ITO, and their film is worthless.

    In conclusion, Unipixel is a complete make-believe story that is literally made up to be believed. When gullible investors throw common economic sense out of the window and lack competency to comprehend the technological complexity, complete loss is the only outcome.

    Disclosure: I am short UNXL. I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it. I have no business relationship with any company whose stock is mentioned in this article.

    Stocks: UNXL
Back To Xuhua Zhou's Instablog HomePage »

Instablogs are blogs which are instantly set up and networked within the Seeking Alpha community. Instablog posts are not selected, edited or screened by Seeking Alpha editors, in contrast to contributors' articles.

Comments (11)
Track new comments
  • Green River Asset
    , contributor
    Comments (309) | Send Message
    Xuhua, the first time I met with the company they went out of their way to show me these gaps in the circuits. They took a lot of pride in the way they could consistently engineer them at precisely the right point along each electrode.


    If you had done more homework you would have learned that these sensors are made up of electrodes with a transmit and a receive end. If you had a solid mesh, you couldn't have a functioning electrode. This is really elementary stuff, Xuhua.


    As for line widths, at one point the technical literature limited line widths at 80mu. Now the technical literature limits the line widths at around 25mu. UniPixel has gone where people said they couldn't, this is true. It's also what makes them unique, and it gives so much value to their IP.


    Keep throwing it against the wall, brother. Something's going to stick!
    19 Apr 2013, 09:40 AM Reply Like
  • Aristides Capital
    , contributor
    Comments (527) | Send Message
    I throw this against the wall:




    Maybe UNXL can do another secondary to raise the funds they'll need to sue LG.
    19 Apr 2013, 11:53 AM Reply Like
  • Green River Asset
    , contributor
    Comments (309) | Send Message
    LG's product has been out there for a while. It's not new. They have some stuff in production but not at 3mu.


    Hopefully LG Chem is successful in starting up its production. The company's production process is rather cumbersome and expensive relative to UniBoss, but it's probably a lot better than ITO. There are several other Asian metal mesh companies who are also at/near production. A couple have products in AIO PCs already. The industry is very excited about the prospects for metal mesh as we head into the back half of the year. This could rapidly grow to a 1Bn ft2 market, which means plenty of opportunity for each of the various players. It's great news and highly validating for UniPixel's technology.


    UniPixel is in a great spot as they will have all the demand they can handle with the Intel and Dell partnerships. Now they just need to scale the operation.


    Try something else SCHFG!
    19 Apr 2013, 01:57 PM Reply Like
  • ouijaman1
    , contributor
    Comments (247) | Send Message
    Zhou, you need to do your homework better. I spent 10 minutes doing a patent search and found a description of metal mesh that explains how it's advisable to have broken gaps in the lines. The gaps then need to be filled with metal for coloring to achieve a visually even look. If I were going to argue that these gaps were "bad", I'd have done a little more research than you appear to have done.
    19 Apr 2013, 04:02 PM Reply Like
  • Ivan Jimenez
    , contributor
    Comments (1025) | Send Message


    A "projected" capacitance touch sensor does has spacing between the capacitors like in pg 33 of this presentation:




    But the spacing in the picture above does not follow a even pattern.


    I agree with Zhou that UniPixel's process is inadequate to produce a consistent quality product. I don't agree Zhou in that the ink spreads because UniPixel is not using ink but a resin like substance per the UniPixel recent patent.




    This resin is what UniPixel calls the "secret sauce", and it is immediately hardened using UV radiation. So I don't think that it will spread but still the challenges of molding the "secret sauce" a 5 microns are overwhelming if not impossible.


    I believe that such process will not effectively place the "secret sauce" at 5 microns creating some serious quality issues that UniPixel has barely started to understand.


    Still UniPixel continues to advertise a product that is grossly flawed.
    23 Apr 2013, 03:59 PM Reply Like
  • ouijaman1
    , contributor
    Comments (247) | Send Message
    Why should anyone listen to what you "believe", without any reasons given for your beliefs? It seems like an arbitrary belief.
    23 Apr 2013, 04:26 PM Reply Like
  • Green River Asset
    , contributor
    Comments (309) | Send Message
    Ivan, you're wrong once again.


    It's comical that you and Xuhua think that you can unearth something from looking at a single slide from a marketing deck that Dell, Intel, Synaptics, Texas Instruments, & NTrig, couldn't find with a team of engineers looking at actual product in an actual laboratory over thousands of hours.


    The reason there's no pattern is that the photo is zoomed in to show only a small section of the grid. Go find a larger picture and you'll see the pattern very clearly.


    You say that the challenge of printing to 5 microns is overwhelming to impossible. Kodak, Dell, and Intel think otherwise. In fact, Kodak is so convinced that they've made UniBoss an important part of their growth plans, and they're investing a lot of money and other resources into a partnership. Intel thought so highly of the technology that they insisted UniPixel scale the operation to many millions of square feet per month to meet demand. Dell spent six months auditing UniBoss and its production process and came away so convinced that they made a financial commitment to build capacity and commanded their supply chain to work with unxl to integrate UniBoss.


    I'm sorry, Ivan, but your view that the product is "grossly flawed" just doesn't carry any weight. The market has decided to trust the real "experts."
    23 Apr 2013, 04:35 PM Reply Like
  • Chris Hofmann
    , contributor
    Comments (755) | Send Message
    I wrote a long, well-researched comment on Ivan's most recent Insta-Blog detailing the numerous errors he claims (some which he KNOWS are errors based on his previous articles). But then he deleted it.


    I'm sensing a pattern... these short detractors of UNXL are switching from real articles to insta-blogs so that they can control the comments which normally blast their derision. Who needs facts when you can control the story.
    5 May 2013, 10:22 AM Reply Like
  • ouijaman1
    , contributor
    Comments (247) | Send Message
    Shorts are going to be buried in a grave they dug.
    6 May 2013, 10:55 AM Reply Like
  • Ivan Jimenez
    , contributor
    Comments (1025) | Send Message


    Carclo's Conductive InkJet Technology also does Flexographic Printing.


    23 Apr 2013, 04:04 PM Reply Like
  • lylemalleniii@gmail.com
    , contributor
    Comments (27) | Send Message
    Every article MUST have a date when published?


    If not included I can only assume that you are not being a journalist.




    Else? don't make me say it..
    4 Mar 2015, 04:22 PM Reply Like
Full index of posts »
Latest Followers


More »

Latest Comments

Instablogs are Seeking Alpha's free blogging platform customized for finance, with instant set up and exposure to millions of readers interested in the financial markets. Publish your own instablog in minutes.