On the trail of digital camera circuitry: ELV Digital Experiment Board DEB100

[Andreas Thaler]

You don't need to understand the details of digital camera circuits, like those found in the Canon T90 or Minolta X-700. Neither as a user nor as a DIY repairer.

It's enough to say: "The electronics reside in the small black blocks on the circuit board. Conductor tracks and cables connect them to the camera components. The electronics either work or they don't. What happens in detail is irrelevant." It's the same with dishwashers and smartphones.


However, if you want to understand how digital circuits work

you have to familiarize yourself with the basics of digital electronics.

This seems technically and abstractly complicated, but it's surprisingly simple when you immerse yourself in the binary world, which consists only of 1 or 0, high or low, yes or no.

There are good books on the subject.

Practical exercises with digital components such as logic gates, flip-flops, or counters are somewhat confusing, as you have to make many connections on the breadboard.

This is where the ELV Digital Experiment Board DEB 100 where basic digital circuits are already pre-wired comes into play.

You only have to connect the individual components via connectors to build circuits. Numerous switches, display elements such as status LEDs and seven-segment displays are also available. A detailed accompanying manual introduces the topic.

This allows you to work through the fundamentals on which digital camera circuits also function.

The next step is to work with microcontrollers, which can be programmed and contain these basic circuits as virtual units in a single component. This also makes newer SLRS and DSLRs more understandable in their operation

ELV Digital-Experimentierboard DEB100 | Bausätze

ELV ELV Digital-Experimentierboard DEB100 from 99,95 €? - novomind iSHOP

de.elv.com

(German)

 

[koraks]

Hah, that's nice, those kinds of boards/kits used to be quite common once upon a time, especially in the1 1980s-1990s. They went out of fashion with the advent of microcontrollers and ASICs.

This also makes newer SLRS and DSLRs more understandable in their operation

Only to an extent, since these rely on highly tailored/application-specific controllers. Some of the principles used are the same as in generic microcontrollers, but the controllers in cameras (both analog and digital ones) involve a lot of black box behavior that you really can't fathom without company-internal documentation. How metering is handled, how shutter speeds etc. are generated and of course in DSLR's how image data are processed are all proprietary, black-box elements that we often don't even know to what extent they're implemented in hardware or in software (firmware).

 

[Andreas Thaler]

Hah, that's nice, those kinds of boards/kits used to be quite common once upon a time, especially in the1 1980s-1990s. They went out of fashion with the advent of microcontrollers and ASICs.



Only to an extent, since these rely on highly tailored/application-specific controllers. Some of the principles used are the same as in generic microcontrollers, but the controllers in cameras (both analog and digital ones) involve a lot of black box behavior that you really can't fathom without company-internal documentation. How metering is handled, how shutter speeds etc. are generated and of course in DSLR's how image data are processed are all proprietary, black-box elements that we often don't even know to what extent they're implemented in hardware or in software (firmware).

I knew this would appeal to you

But it's true. The fundamentals of digital electronics remain valid, whether as a TTL (transistor transistor logic) component or as a virtual program.

The implementations of camera manufacturers must also be based on these principles

This is not very pleasing to programmers who believe that code is everything (I don't mean you of course.)

 

[Andreas Thaler]

Hah, that's nice, those kinds of boards/kits used to be quite common once upon a time, especially in the1 1980s-1990s. They went out of fashion with the advent of microcontrollers and ASICs.

At least the right practice tool for 80s cameras.

 

[Andreas Thaler]

And at least to get into the basics.

Although computers are already at work there, combining and condensing these basics into computing units („Rechenwerke“ in German).

 

[koraks]

The implementations of camera manufacturers must also be based on these principles

Well, yes, in the end they're also based on fundamental physics, Ohm's Law etc. The question is always what you want/need to understand for what purpose. Having said that, it never hurts to understand the basics of logic ports or how to work with a microcontroller.

This is not very pleasing to programmers who believe that code is everything

There's even an analogy within electrical engineering, where the analog people sometimes argue that all digital circuitry in the end is analog just as well. Which is very difficult to argue against, especially if you get down to the nitty gritty details of signal quality, HF etc.

 

[Andreas Thaler]

Today I put the digital experiment board into operation. I'm using it to review the content of the course I completed some time ago.

The supply voltage can be selected between 3 and 15 VDC, and a power supply can be connected via a socket plug. Alternatively, three AA batteries can be inserted into the holder on the bottom of the board for 4.5 VDC.

The individual digital components are connected to each other via cable bridges.

Here, I've connected the two inputs of a NAND gate (CD4011) to the operating voltage via two push buttons.

I've also connected the NAND gate's output to an inverter (NOT, CD4069), thus creating an AND gate.

Both outputs are connected to LEDs that indicate the respective levels HIGH or LOW.

With this arrangement, the function table for two inputs can be switched with a total of four possible combinations.

Here are the two combinations B = 0 (LOW), A = 0 and B = 1 (HIGH) and A = 1:

Both inputs (B, A) are LOW, the buttons are not pressed.

This results in the NAND output above being HIGH and the AND output being LOW.

Both inputs are HIGH (buttons pressed) so the NAND output is LOW and the AND output is HIGH.


Logical circuits can be constructed using AND, OR, and NOT

This allows you to create the three logical gates AND, OR, and NOT using NAND gates and observe the individual switching combinations.

Since there are only four NAND gates on the board, additional inverters are provided for NOT. However, NOT can, of course, also be generated with NAND.

For example, in an experiment, an LED should light up when ASA = 100, EV = 12, and aperture = 5.6. If all three conditions are met via AND, the LED is activated.

This allows you to get a sense of the camera circuits, even if you don't actually construct the experiment and instead press buttons.

If you like lights, you'll get your money's worth here

---

For me, the experiment board is a perfect complement to circuit simulation on the PC and studying digital circuits in books. You build circuits practically and have a tangible sense of achievement when something works.

You can continue working on the other digital components on the board and deepen your understanding. In any case, you get an insight into how digital circuits fundamentally work.

This should help me when familiarizing myself with the literature on the digital driven Canon A-1 and AE-1.

 

[Andreas Thaler]

Digital display of numbers: BCD-to-Seven-Segment-Decoder

Displays for numbers can be found on many electronic devices, either as LCDs or LEDs.

How they work and how the numbers are formed is usually not something you ask yourself, as long as they display correctly.

The DEB100 experimental board has two such digital displays as LEDs, located at the top center.

If you look at a number more closely, you will see that it consists of seven individual bars that either light up or do not light up.

Here, the 0 is displayed twice, and for each six of the seven bars are lit. The seventh bar in the middle remains dark.

As a graphic it looks like this, with one more dot added.

File:Slanted7SegmentDisplayWithDot.svg - Wikipedia

en.m.wikipedia.org

Here you can see that each bar is assigned a letter (we ignore the dot).

File:7 Segment Display with Labeled Segments.svg - Wikimedia Commons

commons.wikimedia.org

Seven-segment display - Wikipedia

en.wikipedia.org

In order for a number to be displayed, certain bars must light up together.

This way, all the numbers from 0 to 9 can be formed.


A digital circuit

is needed to generate these digits. This circuit works with binary numbers. This means that there are only 0 and 1.

Since we as humans mostly work with the decimal system, i.e. with ten digits instead of two, one system must be coded into the other.

see Jürgen Feldhoff, Digitaltechnik - Grundlagen der Digitalelektronik und Mikrocontrollertechnik, Lehrbrief 4.


This table shows the coding.

On the left, the decimal digits 0 to 9 are arranged vertically.

Next to each decimal digit, in the pink highlight frame, is a binary code consisting of 0 and 1, labeled with D, C, B and A.

For example

• 0 decimal corresponds to 0000 in binary,
• 1 decimal is written as 0001 in binary, or
• 5 decimal is written as 0101 in binary.

We've now converted all the digits from decimal code to BCD (binary coded decimals).

A digital circuit can operate with these binary numbers.


Now the right-hand part of the table comes into play

You can see that for each decimal digit and its BCD code, there is another code: the 7-segment code from a to g.

To use our example from before, the 7-segment-code for decimal 0/BCD 0000 is 1111110. This means that for zero, bars a to f are lit, and g remains dark.

For decimal 1/BCD 0001 the 7-segment-code is 0110000.

And for decimal 5/BCD 0101 it is 1011011.

So the 7-segment code also works in binary with 0 and 1.


So all the information is available to control the digital LED display

In digital terms, this works via logic gates that use AND, OR, and NOT operations to assign the respective BCD codes to the 7-segment codes.

Fortunately, this complex digital circuit doesn't have to be assembled from individual parts; instead, it uses a BCD-to-seven-segment decoder, which simply needs to be connected.

The decoder as IC (integrated circuit) contains the complete digital circuit in miniature format and is included as a component in the DEB100 experimental board.


Display of the number 5 via the 7-segment decoder

The number 5 is displayed.

Below, on the right, you can see the decoder schematic.

Four cables are connected to the four BCD inputs A, B, C, and D.

Above, the seven outputs QA to QG for the LED bars are shown graphically, leading to the right-hand LED.

Using a slide switch and three buttons, the BCD code for decimal 5 is entered in binary form as 1010.

• Switch closed/button pressed = 1,
• switch open/button not pressed = 0.

For visual indication, switches and buttons are connected to four control LEDs that indicate the status of the input:

• 1 = HIGH = GREEN
• 0 = LOW = RED

The decoder now converts the four-digit BCD code into the 7-segment code and lights up the bars corresponding to the 5.

 

[Andreas Thaler]

In practice, you'll no longer work with individual digital ICs like the BCD-to-seven-segment decoder, as the design of such hardware-based circuits quickly becomes confusing.

The circuits can also be programmed on a microcontroller, such as the Arduino. Because they're software-based, the programs can be edited at any time, and there's no need to change any of the electrical connections on the board, which could require completely rebuilding the hardware circuit.

However, for learning purposes, working with digital hardware components offers a lot of visualization.

In camera circuits from the 1970s and 1980s, one can still encounter individual (discrete) digital components. Even though these components might be themselves combined into ICs, the technical documentation refers to such basic digital circuits.

And once you have understood the basics of digital technology, you will understand these descriptions better and learn how these camera circuits generally work.

 

[koraks]

In camera circuits from the 1970s and 1980s, one can still encounter individual (discrete) digital components.

In darkroom timers etc. from the 1970s-1990s it's common to find a decent selection of the kind of IC's you practiced with, or at least variants thereof. They often have some kind of oscillator, followed by some counters and then something to drive a 7-segment display. In cameras, space has always been more on a premium, so manufacturers started bunching stuff together in more complex IC's very early on.

 

[Andreas Thaler]

Flip-flop: a binary one bit memory

A flip-flop can assume two states: 0 or 1.

With one bit this makes it the smallest memory unit in a digital circuit.

One bit corresponds to 2^1 = 2 different states.

As long as the flip-flop is supplied with power, these states are retained until it is set back to 1 or reset to 0.

The DEB100 experimental board features a CD4013B CMOS dual D-type flip-flop. This provides two D-type flip-flops in one IC.

The schematic for one of them is shown here.

This flip-flop can be operated in two ways:

1. statically, by setting the states 0 and 1 via RESET and SET, and
2. dynamically via DATA and CLOCK; here the setting and resetting is done by a clock signal.

RESET/SET

The flip-flop is in state 0, it is reset.

This state is indicated by a red LED at the output /Q.

Using a button I set the state 1 via SET.

This state is indicated by a green LED at the output Q.

Via RESET I reset the flip-flop back to the state 0.


DATA/CLOCK

State 0

I set the flip-flop to state 1 via DATA and then activate CLOCK. Only then does the flip-flop switch to state 1.

For resetting to state 0 I activate CLOCK again. DATA is not activated.


Using a clock signal as a regular square wave, the flip-flop can be switched at different frequencies.

This allows, among other things, the combination of individual flip-flops into counters and storage registers, which can then be controlled automatically.


Several flip-flops form a register

Each flip-flop has a storage capacity of one bit (= 2 states).

Two flip-flops, for example, can store two bits, resulting in 2^2 = 4 states.

With three bits you can store 2^3 = 8 states and with four bits you can store 2^4 = 16 states and so on.

• For 1 bit, these are the states 0 or 1.
• For two bits, these are the states 00, 01, 10 and 11.
• And for three bits 000, 001, 010, 011, 100, 101, 110 and 111.

Assuming we assign a shutter speed to each of these states, we can use 3 bits to represent values binary such as 1/1 second, 1/2 s, 1/4 s, 1/8 s, 1/15 s, 1/30 s, 1/60 s, and 1/125 of a second.

Or various aperture, ASA, or other values.

A digital circuit with a microprocessor can process these digitized values and thus control a camera. Everything is based on 0 and 1.


+++

All information provided without guarantee and use at your own risk.