Processor TR Iteration that got scrapped…
If you follow my Instagram feed, you know that last summer I had a fully developed new processor called B’s processor TR ( for Timed Reversal). I had developed this machine very quickly since I was mostly home due to you know what! So I had plenty of time to spend on it. I was happy with the machine and went on buying a few thousands of dollars of parts and started production in parallel of my personal testing plus my two beta testers giving me feedback.
That proved to be a mistake…. the machine worked properly and results were ok but I realized that there is more to it than just turn one way and the other… and that bugged me! After this extensive testing, I had learned many things about what happens inside a tank in rotation. The TR worked fine but I knew I could now make an even better machine.
By the end of the summer, a very hard question came up: what should I do? Build this machine and sell it or scrap it to make it into a better machine?… As you can guess I chose the latter. I bit the bullet and went back to the drawing board… By that time I was back at work full time so between work and family's time, design took a good while add another round of testing and here we are many months later….
Here are the things I’ve learned, hopefully that will explain and make sense of my new design decisions:
Colorful tests units of the new B’s Processor final design! The production units will have the same Carbon fiber and grey color scheme the TR had.
What was good to keep:
-General form factor was good, compact and practical, so it did not change much on the outside…
-The wobble wheels were proven to soften hard flow as I intended them to do.
2. Slow RPM is best!
The “TR” was designed for speed and that was also a mistake! So many people had asked me why my original processor turned so slow, I guess because other machines are fast, that I was convinced it was the way to go. The 10-12rpm of my original machine was actually a much better speed than the 60rpm the “TR” could do. Here is the reasoning behind what I’m saying: All reels have different shapes/geometries; those geometries are most of the time in the way of the liquid and will create hard accelerated flows behind them. (Think of a rock in the middle of the river). Those flows, with the wrong circumstances, will create overdevelopment marks. You can’t avoid it, but at slow rpm, those will have less effects. This is why B’s large format reels are very minimalistic, it’s meant to reduce geometry in the liquid’s way to its minimum. To help get better and softer flows, I had designed a special “wobble wheels” extension for my original machine that you will find now integrated into my new processor. The new B’s processor comes with two sets of interchangeable wheels: the circulars and the wobbles.
Another innovation added to the machine is the new “pulsed Rotation” program. This program makes the tank go thru short acceleration and deceleration cycles while doing its 2 turns one way and the other. Avoiding a constant speed helps smooth out those flow around the geometry. I’m pretty proud of this one, it proved itself well and in combination with the wobble wheels it provides the best evenness you can get with this type of machine :D
3. Variable speed is unnecessary and a problem for results repeatability.
This is why B’s processor now have only three programs: Pulsed, slow (~10rpm) and fast (~38rpm). At first, I was trying to make this machine in a simple way, using off the shelves electronic boards. This was also a mistake and I lost lots of time trying different combinations. Everything pointed to just make it Arduino based so it can move any way I want and that’s what I did. It allowed for a lot of flexibility to add sensors and for the future upgradability of the firmware running the machine! Any time I tune it you will be able to get the benefits of it, just plug the machine on your computer and upload the new firmware! It also allowed to use a very good motor driver witch provides good torque even at slow speed.
4. Equal rotation both ways is important to even out the liquid flows.
The “TR” prototype was hard to set the proper times for equal rotation. So I looked into a system that would insure the tank always turns the same amount both ways. I came up with a new contactless system using a magnetic band on the tank. That band is simply held with a rubber band. Baked into the programing is a routine that also detects the tank and magnet band position so it can re-synchronize itself if its position was lost or if the tank was set on the processor in the wrong position. Like that the tank always change its rotation at the exact same spot. Since the detection is contactless, there is nothing here to break compared to a mechanical switch.
The “Penguin”
5. The Penguin!!
Not something directly linked to the machine but more about the Paterson tanks. During my 4x5 testing, something evaded me for a while, one batch would be perfect and the next one was a disaster! I realized that the problem always happened when my reel was only half loaded, that led me to understand the unbalanced load could freely rotate at the direction change since the Paterson tank column is not linked to the tank and can spin independently. So comes the “Penguin”! This is a simple piece I designed to go at the bottom of the column, it just frictions in and grab onto the three prongs at the bottom of the Paterson Tank. It’s flat and thin so it does not get in the way of the chemistry coming in and the friction is enough to avoid any shift.
Here is the new specs:
Genuine Arduino Nano (no crappy clones here ;)
High efficiency motor driver for torque (a full 8 reel Paterson turns like a charm)
Now runs on 12V for better more stable power
Touchless rotation switch (equal rotation even if the tank slips on the wet wheels)
Light indicator to show if the magnet band is properly triggering the switch
Interchangeable smooth and wobble wheels
3 programs for perfect repeatability
Paterson column linkage, aka the “Penguin”
Tank position automatic detection
Future proof. Front hatch allows access to USB plug to upgrade the programming.
Splash proof
So when can you get one? As you understand now my time is limited so I can’t really put a date on it yet! But I can tell you where I’m at :D Production started of a first small batch, Packaging need to be finalized, Shipping cost needs to be priced and I have to finish writing the manual. The manual is going to be a long one since I want to integrate all that I’ve learnt into it and give you the best processing tips so you have the best possible starting point.
If you made it here thanks for reading me, I know that was a long one! I hope this all made some sense…
I wish you good developments!
Benoît.
Say Hi! to the very first B’s processor production part :D