|20201124-0001-DevTime_Internal.jpg||2 years ago|
|20201124-0002-DevTime_External.jpg||2 years ago|
|README.md||3 months ago|
|TODO||4 years ago|
|devtimer.py||3 months ago|
|time-temp.xlsx||5 years ago|
|tm1637.py||5 years ago|
Developing traditional film and silver paper is very temperature dependent. The warmer the developer solution, the less time the film or paper needed to be developed and vice-versa. Historically, people read the temperature and then manually corrected their development time accordingly, using a correction table provided by the manufacturer.
It turns out that these corrections for temperature are quite similar across different manufacturers of film and paper, although the corrections are different for film versus paper.
Many years ago, a company called "Zone VI" realized this and created an analog timer that corrected for this effect. You placed a temperature probe into the developer and it corrected - via analog adjustments - what a "virtual second" actually had to be. The photographer just looked up the normal development time for 68F developer and the timer ran faster or slower based on the actual temperature. Better still, if the temperature of developer varied during development, it corrected for that in realtime. The Zone VI Compensating Timer had settings for film, paper, and realtime.
The timer was a work of genius engineering and a really nice addition to the serious photographer's wet dakroom. I've depended on one of these for years to make my darkroom work repeatable with minimal thinking or measuring. Mine is getting kind of old now and I began to wonder what I would do if it broke. The timer did come with a "Lifetime Warranty", Sadly Zone VI and its founder, Fred Picker, are both now long gone making warranty claims ... difficult.
While I could design an analog replacement or just figure out the circuit of the Zone VI, it occurred to me that it would be easier to just design a "work alike". Thanks to the explosion of interest in robotics and the Internet Of Things, there is an embarassment of riches of computers, sensors, switches, temperature probes, and so forth. Not only can we build something like this ourselves, doing so has several advantages over the old Zone VI timer:
It's digital, not analog, so we don't have mess with a bunch of precision parts and corrective feedback circuits.
It's software controlled so you can customize how this timer works to suit you. Don't like my compensation factors? Want to adapt this for a different application? Both are easily done with software changes.
It's cheap. You can build one of these for well under $50. (The original Zone VI timer was around $200 if memory serves, and that was when money was still worth something. :)
As of this writing, you'll need to be an accomplished Raspberry Pi hacker or have access to someone who is. The goal is to eventually package this up in a way that a relatively inexperienced person could build it, but I wanted to make the code and design available as early as possible for those who have have already expressed an interest in this project.
The hardware design is really simple because it's based on a Raspberry Pi Zero platform. The whole timer uses a little over a dozen parts in total ... that's including a case and USB power supply.
The code here on the master branch should be considered stable and usable.
The code will tell where to connect things on a Raspberri Pi Zero.
Note that these reference GPIO pin numbers, not device pin numbers.
A few notes:
Where there are pullups or references to Vcc below, I've chosen to use a 5V pin on the Raspberry Pi Zero.
You'll need two TM1637-compatible 4-digit LED displays - one for time and one for temp. Their CLK and DIO assignments are noted in the code. Of course, you'll also have to connect them to Vcc and ground.
Connect a piezo buzzer between GPIO 26 and ground. No resistor needed.
Connect a momentary contact footswitch between GPIO 6 and ground. Pull up that pin with a 4.7K resistor.
Connect a SPDT switch for profile selection to GPIO pins 23 and 24, each pulled up with a 4.7K resistor. The switch common goes to ground.
Connect the data line of a DS18B20 temperature probe to GPIO 4, pulling it up with a 4.7K resistor. Again, you'll have to connect it to Vcc and ground as well.
Take note to read the code concerning the temperature probe. Each probe has a unique serial number and you have to configure a symlink in Linux to point to your specific device.
There's not much to this:
Get your favorite Raspberry Pi Linux distro running.
Setup 1-wire support. In
dtoverlay=w1-gpio (At least, that's how you do it on Raspbian. Other distros may be different.)
Make sure python3 is installed - this project requires it.
Make sure to properly create the symlink to the temperature probe as described in the code.
pip3 install the wiringpi module. You can either do this systemwide or in a pew virtual environment. Either way, you need to be
root because this code - at least for now - requires superuser to run.
tm1637.py files onto your Pi Zero. You can then start the timer with
By default, the code currently writes a lot of debug info to the Pi's tty so you can watch all manner of things fly by. If it bugs you, just change
False at the top of the code.
You can add a reference to
/etc/rc.local to automatically start the timer when the Pi boots. Right now, this takes about 30 seconds but a future effort will work to rip out the stuff not needed and make booting way faster (hopefully).
At program start, the temperature display will briefly show 999F. This is a "sentinel" to let you know the timer is looking for the temperature probe. The display should rapidly update to show actual temperature as reported by the probe. If it does not (i.e., 999F remains displayed), it means that something is wrong with the temperature measurement hardware.
If the timer detects an error reading the probe - for example, if the probe is defective or the wiring is compromised - the temperature will display 998F to let you know. If you are in a film or paper correction profile, the timer will blink the display and also sound an audible alert (see below) to remind you that it cannot correct for temperature. Both the blinking and the audible alerting take place whether the timer is actually running or not. This is intentional to make you take action.
Pressing the momentary contact footswith start/resets the timer.
The SPDT switch selects Film correction, Paper correction, or realtime (no correction). You can change this while the timer runs.
If you are in a film or paper correction profile, and the temperature is beyond the range of the timer to correct, the temperature display will blink. This lets you know the timer is not capable of correcting for temperatures in that range. This does not happen in the realtime profile. There is also an audible alert (see below.) Both the blinking and the audible alerting take place whether the timer is actually running or not. This is intentional to make you take action. If you want to do compensating timing, you have to get the temperature back in range. If not - say, you just want to measure temperature - then you should switch to the realtime profile.
The timer "chirps" at startup to let you know it is initializing.
The timer beeps twice each time you start- or stop the timer.
The timer provides a single beep at each virtual 30 second mark.
In film or paper correction profiles, if the temperature is beyond the range of the timer to correct, you will hear 5 short beeps repeated.
In film or paper correction profiles, if the temperature cannot be read from the probe, you will hear 5 short beeps repeated.
The timer should pretty much be ready to run "out of the box". But, here are a few ideas on how it can be tuned or adjusted:
VERY IMPORTANT: Do NOT run the timer with
DEBUG = True. This option is for the sofware developer only and introduces extra output that makes the timer less accurate. So, unless you are hacking on the hardware or software, leave this set to
CALIBRATION_OFFSET variable is designed to help compensate for the amount of time the master timing loop needs to run on your particular platform. The default value should be about right for a Raspberry Pi Zero. If you use a different platform or overclock your Pi, you may wish to adjust this to calibrate the timer. This is best done by running the timer in realtime mode (no temperature compensation), adjusting this variable and seeing how the timer speed compares to a known good reference timer.
You can adjust the individual compensation entries for film or paper to your liking. The values there are pretty good first approximations based a fair bit of research, but you can "tweak" them to suit your style of photography and film development.
There isn't any. You can send mail to
email@example.com and I'll do what I can to help as I able, but this is very much a part-time activity.
What would be VERY welcome would be pull requests, bug reports, patches, etc.