Patent on Stack Focus Rail System

Have questions about the equipment used for macro- or micro- photography? Post those questions in this forum.

Moderators: rjlittlefield, ChrisR, Chris S., Pau

Online
soldevilla
Posts: 684
Joined: Thu Dec 16, 2010 2:49 pm
Location: Barcelona, more or less

Post by soldevilla »

I followed this thread with surprise. I really can not think that such a simple (mechanical) application can be patented. Any mechanical engineer can show a linear motion unit moved by spindle in a good number of catalogs, which we use constantly in the machines we design. I can understand (I always speak of the mechanical part) that the user decides to buy a Stack Shot or the mjkzz version simply because the price compensates for the time to adapt a commercial one, but from there to pretend that it is a patentable innovation, there is a lot of distance . My commercial 3D printer moves on two calibrated rods, pushed by a spindle and moved by a stepper motor. If I remove the extruder and add my camera, I have a conceptually same system as the Stack Shot.

Dave Koerner
Posts: 2
Joined: Wed Jan 17, 2018 12:34 pm

Post by Dave Koerner »

Regardless, a patent was granted. I'm hopeful to hear that the OP was able to resolve his issues with Cognisys and go on selling the product - either via licensing, or simply by ignoring the cease and desist order. I'm considering selling a similar product here in the US, but I don't have deep enough pockets to start the ex-parte re-examination process, which could make that patent null and void.

ray_parkhurst
Posts: 3413
Joined: Sat Nov 20, 2010 10:40 am
Location: Santa Clara, CA, USA
Contact:

Post by ray_parkhurst »

I missed this thread first time around, but after reading through it I am now very glad I am using Peter's system.

anvancy
Posts: 387
Joined: Sat Dec 05, 2009 11:57 pm
Location: India
Contact:

Post by anvancy »

From a layman's perspective:

As a consumer, I have multiple choice to make now. Your rails, Stackshot, We-macro and other customized ones.

Does this look similar to how GoPro thought they are pioneers in action cameras and today are on survival mode and shut down their drone division?

They are now scared as they know their marketshare is in question. I can understand the benefits of a stackshot. But looks like internally they know the $600 cost is way high when a person can directly import a $300 rail from china and do atleast 2/3rds of the job if not at full capacity.
www.anvancy.com

Raynox 150|Raynox 250|Raynox MSN 202|Canon MPE 65mm|Canon 100mm.|Wemacro Rail

mawyatt
Posts: 2497
Joined: Thu Aug 22, 2013 6:54 pm
Location: Clearwater, Florida

Post by mawyatt »

anvancy wrote:From a layman's perspective:

As a consumer, I have multiple choice to make now. Your rails, Stackshot, We-macro and other customized ones.

Does this look similar to how GoPro thought they are pioneers in action cameras and today are on survival mode and shut down their drone division?

They are now scared as they know their marketshare is in question. I can understand the benefits of a stackshot. But looks like internally they know the $600 cost is way high when a person can directly import a $300 rail from china and do atleast 2/3rds of the job if not at full capacity.
The Stackshot, MJKZZ and WeMacro systems are at different ends of the focus stacking capabilities. The Stackshot has just about every feature and is highly parameter programmable allowing different focus rail uses, and has the best user interface IMO when operated from Zerene. The WeMacro is a bargain price system that does a superb job at the fundamentals of automated stacking but lacks some of the flexibility and parameter programing of the Stackshot. If price is a concern then the WeMacro system is a good bargain, but if feature set, programmability and Zerene interface is important then the Stackshot system stands out.

Best,
Research is like a treasure hunt, you don't know where to look or what you'll find!
~Mike

Jan Steinman
Posts: 21
Joined: Mon Jan 22, 2018 5:06 pm
Location: Salt Spring Island, British Columbia, Canada
Contact:

RRS

Post by Jan Steinman »

rayB wrote:I have read that RRS too have recently resorted to such tactics as outlined in your 3rd paragraph.
Unless you're referring to something esoteric, the basic parameters of RRS plates are public domain.

Upon hearing this some months ago, I wrote them a polite email, asking for the specifications. It took RRS about a week, but then I got full engineering drawings in the email!

So it would be interesting to know exactly what it was that RRS was picking patent fights over.
Image
:::: Jan Steinman ::::

mawyatt
Posts: 2497
Joined: Thu Aug 22, 2013 6:54 pm
Location: Clearwater, Florida

Post by mawyatt »

I've been an expert witness on both sides of patent litigation, and currently hold 28US patents, so somewhat knowledgable in this area. BTW we won all those cases :D

Since "prior art" is the 1st attempt at diffusing a patent, and both sides know this, it's unlikely a patent in question will be vigorously enforced if it's very weak and one can easily show "prior art" that's legally binding. Many times patents are applied for and granted for prestige rather than technical merit. The patent office attempts to do a good job of filtering patent applications, and only granting those that are worthy of being protecting by patent, but many times patents based on a weak foundation get through and are granted.

However, sometimes things seem obvious in hindsight, but that doesn't prevent the original idea from being novel and patentable at the time of creation. For example, I was granted a patent on a voltage divider, yes this is true!!

https://www.google.com/patents/US5030848

If you study the patent you will find it was considered novel, and in fact won the Design Idea of the Year in a popular technical magazine. The idea utilized a CMOS Flip-Flop to divide a voltage precisely in half (PPM accuracy), yet the component tolerances weren't important, so 20% components could yield <0.0001% accurate results. At first this seems like a ridiculous patent for a voltage divider, but a more thorough examination of the technical merits revels some novelty in the original idea.

Anyway, always consider that the patent litigation process is a legal process, and thus controlled & processed by specialized law & lawyers, and sometimes rulings and results just don't make any common sense.

Best,
Research is like a treasure hunt, you don't know where to look or what you'll find!
~Mike

mawyatt
Posts: 2497
Joined: Thu Aug 22, 2013 6:54 pm
Location: Clearwater, Florida

Post by mawyatt »

BTW if anyone is interested I'll be happy to explain how the Voltage Divider works, but don't want to take away from this thread.

Best
Research is like a treasure hunt, you don't know where to look or what you'll find!
~Mike

ray_parkhurst
Posts: 3413
Joined: Sat Nov 20, 2010 10:40 am
Location: Santa Clara, CA, USA
Contact:

Post by ray_parkhurst »

In my experience with writing and defending patents, I have found the "patent world" extremely arcane and non-intuitive. New rules changes have further complicated some areas and simplified others, but overall still just as arcane.

For sure you can patent basic core components that are being used in novel ways. So although the wheel has been around as prior art for quite a long time, a wheel that is used as a guiding mechanism for a computer mouse is a novel invention. Mike's "voltage divider" patent probably falls into this category. So it's in the novelty of the application that matters, not the core component. It's actually rather rare these days to see any truly new inventions. Thus most new patents are products of engineering rather than discovery.

Jan Steinman
Posts: 21
Joined: Mon Jan 22, 2018 5:06 pm
Location: Salt Spring Island, British Columbia, Canada
Contact:

Post by Jan Steinman »

mawyatt wrote:BTW if anyone is interested I'll be happy to explain how the Voltage Divider works
Must have to do with transition points of the J a X outputs? Wouldn't that require a comparator, too?
:::: Jan Steinman ::::

mawyatt
Posts: 2497
Joined: Thu Aug 22, 2013 6:54 pm
Location: Clearwater, Florida

Post by mawyatt »

Jan Steinman wrote:
mawyatt wrote:BTW if anyone is interested I'll be happy to explain how the Voltage Divider works
Must have to do with transition points of the J a X outputs? Wouldn't that require a comparator, too?
No comparator required, just CMOS Flip-Flop and 2 resistors and 1 cap. If you look at the patent diagram, it has 2 resistors R1 and R2. These can form a simple voltage divider with a voltage source. The transfer function is R1/(R1+R2) if R1 is the bottom resistor, R2 is the top, and R2/(R1+R2) if R2 is the bottom resistor, and R1 is the top. The average of these is 1/2(R1/(R1+R2) + R2/(R1+R2)), or 1/2(R1+R2)/(R1+R2) or 1/2!!! The capacitor provide the time average, and the Flip-Flop interchanges R1 and R2 at every clock period. The resistors need to be much larger than the "on" resistance of the CMOS FF Q and Qbar outputs. You can build this with a 74AC74, or 74HCT, or 74ACT, or 74HC CMOS logic Flip-Flop, a couple 100K~500K resistors and a 1uF cap. Clock the FF at 1-10KHz. Use a precision DVM (account for the DVM input impedance, reason for the unity gain op-amp in the 2nd schematic) to measure voltages.

Best,

Mike
Research is like a treasure hunt, you don't know where to look or what you'll find!
~Mike

mawyatt
Posts: 2497
Joined: Thu Aug 22, 2013 6:54 pm
Location: Clearwater, Florida

Post by mawyatt »

Excellent wheel analogy Ray!! Better than my Voltage Divider!!

Best,
Research is like a treasure hunt, you don't know where to look or what you'll find!
~Mike

Jan Steinman
Posts: 21
Joined: Mon Jan 22, 2018 5:06 pm
Location: Salt Spring Island, British Columbia, Canada
Contact:

Post by Jan Steinman »

mawyatt wrote:
Jan Steinman wrote:
mawyatt wrote:BTW if anyone is interested I'll be happy to explain how the Voltage Divider works
Must have to do with transition points of the J a X outputs? Wouldn't that require a comparator, too?
No comparator required, just CMOS Flip-Flop and 2 resistors and 1 cap.
Got it.

But you've now moved your source of imprecision from matching resistors to exactly 50% duty cycle on your clock, no?

Not that it isn't orders of magnitude easier to have a fairly symmetrical clock than it is to have matched resistors, but…
:::: Jan Steinman ::::

mawyatt
Posts: 2497
Joined: Thu Aug 22, 2013 6:54 pm
Location: Clearwater, Florida

Post by mawyatt »

Jan,

Nice work, great question!! This would be true in a time average integrator, where voltage is directly proportional to duty cycle, and requires a duty cycle of precisely 50% from the clock since half time on and half time off, average is 1/2.

But it actually doesn't depend on the clock duty cycle, it's completely independent of such. Almost independent of the clock rate if the rate is low, and completely independent of clock jitter :D

It's also almost completely temperature independent as well.

The biggest sources of error is the asymentery in the Flip-Flop output, and the mismatch of P and N Channel ON resistance in the Q & Qbar outputs, both of which with modern CMOS are low. Modern CMOS FF can have asymmetry in the outputs as low as a few picoseconds, even at 100ps error relative to a FF output of 1ms is 1 part in 10^7, and the error doesn't propagate directly, it is directly modulated by the resistor tolerance error. So a 10% resistor error makes the output error 1 part in 10^8 with a 100ps asymmetry error!! The ON resistance is easily handled because it's the differential error between the Q and Qbar outputs that matters, and this is very low in CMOS on the same FF (and can be made arbitrarily low with buffers). The final error associated with the differential resistance is attenuated by the values of the resistors, and can be made to approach 1 part in 10^6 or even better.

I had to make a case to get funding for the patent application way back, this is decided by the Patent Review Board of which I was a senior member. You can imagine the look I got when I asked to file an application for a Voltage Divider :roll:

After a brief analysis on the white board I had a demo prepared. I had setup a precision power supply, and had an 8 1/2 digital voltmeter (DVM) with over 10G ohm input impedance and NBS traceability calibration (both expensive equipment). Had an old function generator for the clock and a 74ACT74 wired as Flip-Flop, a handful of 1% resistors (we didn't have 10%, only 1% or better) from 100K to 1M and couple caps (0.1, 0.22 & 1uF). The power supply had warmed up (oven controlled voltage reference) and set to precisely 5 volts with the DVM. I asked someone to pick a couple resistors and plug them into the test circuit, so 100K was selected. The DVM was at the junction of the resistors and shunt cap and it read something like 2.5135675 volts, so I got a really bad look from everyone, so I said turn on the function generator at 1KHz and the DVM then read 2.4999988 volts!! Then I said change one of the resistors, so a 120K was put in place of one of the 100Ks and the meter read 2.4999988 with a purpose 20% resistor error! Then I changed the cap and DVM didn't change, changed the function generator clock and nothing changed until we started getting above 100KHz where the 74ACT74 asymmetry starting effecting the output. Even change the resistors to all sorts of values with ridiculous errors and the output held pretty much the same.

After the demo I got the board approval to apply for a patent :D

Best,
Research is like a treasure hunt, you don't know where to look or what you'll find!
~Mike

Jan Steinman
Posts: 21
Joined: Mon Jan 22, 2018 5:06 pm
Location: Salt Spring Island, British Columbia, Canada
Contact:

Post by Jan Steinman »

Wow. Never seen an 8.5 digit DVM. And I used to work in metrology!
:::: Jan Steinman ::::

Post Reply Previous topicNext topic