Digital TTL quench signal generator

The aim of this first experiment is to measure of the amount of light generated by the strobe as a function of the delay between the trigger and TTL quench signal. The delay circuit is required to trigger the strobe and generate a quench signal at a precise time thereafter. The delay should be easy to configure and cover the range from a few 10's of microseconds to a few 10's of milliseconds, i.e. over a range of at least a thousand. It's also advantageous to be able to record the amount of delay easily.

Probably the best way to solve the problem would be to use a microcontroller, which could likely do almost everything required in software with a minimum of external components. I don't have experience with microcontrollers so I went with an old-fashioned discrete IC design, based around a 1MHz crystal oscillator and the 4536 programmable timer, which divides the oscillator signal by any power of two between 1 and 24, giving a period between 2 microseconds and 16.7 seconds, more than enough to cover the range of potentially interesting delay times. The only problem using the output of the 4536 directly is that the delay duration can only be doubled or halved from one setting to the next (i.e. the delays would be logarithmically separated). To give more precision a 74193 (or 40193) decade counter is driven from the output of the 4536 and used to generate the delay before the quench signal. The possible range of delays in then:

• D = 1uS * m * 2^n

where m is between 1 and 9 and n is between 1 and 24. In practice n is limited to a range of ten consecutive values whose lower value is set with jumpers on the circuit (I usually set this to 2, so that n is in the range of 2-11). This design requires only two digits of input (i.e. n and m) and two seven segment LED displays to show the selected values.

The schematics for the circuit, drawn with the fabulous (and free for non commercial use) Eagle layout package are shown below:

The schematics and board layout files are also available in their original Eagle format. The circuit is relatively straight forward. The heart of the interesting functionality was described above. The final component of the primary functionality is a set of S-R latches which drive high voltage transistors (KSP44) that short the strobe trigger and quench signals to ground. These outputs are connected to the strobe using a Nikon AS-18, as described last time. The rest of the circuit (which makes up a lot of the board area) allows m and n to be selected (using up/down buttons) and drives two LED displays to show the values. An input transistor and a switch activates the circuit when triggered by an external device or by the user. I had the board fabricated by Custom PCB in Malaysia (www.custompcb.com) and was very happy with how it turned out. They take the Eagle board layout files directly so it saves you having to do the CAM processing yourself.

Next time the results of using the delay generator on an SB-25 and SB-600.

The magical AS-18

OK.. before I get on to the TTL timing circuit, a quick note on how the old-style TTL system works and some cheap ways to interface an electronic circuit with on of Nikon's TTL strobes. Basically the camera sends two control signals to the strobe, a trigger pulse to start the flash pulse and a quench pulse when the camera determines that enough light has been produced. The strobe charges an internal capacitor up to a high voltage (using a voltage multiplier). When the trigger signal is received the stored charge is dissipated through the flash tube producing a flash of light. The quench signal stops this discharge, stopping the light pulse. In TTL-metering mode the camera monitors the amount of light reaching the focal plane and generates the quench signal when the correct exposure has been reached. The flash calibration board that I will describe in the next post generates the quench signal a fixed time after the trigger signal and a flash meter can then be used to calculate the amount of light produced as a function of the delay before the quench signal is generated.

Modern Nikon speedlights have 4 contacts on their feet to allow communication with the camera. The foot itself (or two prongs below the foot on older speedlights with a plastic foot, such as SB-25) connect the grounds of the camera and strobe. Many speedlights also have additional connectors to allow off-camera flash photography, for example the SB-25 has a standard sync port which has a single contact (and ground) wired up to the trigger circuit of the strobe and a 3-pin TTL connector which allows the trigger and quench signals to be sent from the camera to a remote flash using a 3-wire cable sold by Nikon (SC-18 and SC-19 for example). Nikon make a bunch of converters and cables to help in wired off-camera applications, such as SC-17, AS-10 etc. Most of them are expensive to buy new and can be found on eBay for $20-$30.

For home experimentation the Nikon AS-18 is a great option. It is described as a "TTL Multi-flash Adapter for 950 990 995", and is possibly the cheapest product available from Nikon, selling at $3 on Amazon as of August 2008. Possibly this is because it is technically for use with the Coolpix range of cameras, which have been largely overtaken by the range of DSLRs. The AS-18 has a hot-shoe on the top and a 3-pin TTL socket on the front - it is very similar to the more expensive AS-10 ($40 on Amazon) which has multiple 3-pin sockets to allow multiple flashes to be chained using Nikon's expensive cords. You can see from the picture that the AS-18 has plastic tabs where the AS-10 would have extra 3-pin connectors. The AS-18 can easily be opened and modified to accept any signals you like.

The three contacts of the hot-shoe used by the strobe in TTL mode are shown in the picture to the left. The trigger signal is on the large pad in the center of the shoe, and the quench signal on the smaller pad to the bottom right of the shoe. The rails of the shoe itself are the ground for the camera. The camera fires the strobe by shorting the trigger pin to ground (hence you can fire your strobe by shorting the trigger pin to the foot), and stops the flash by shorting the quench pin to ground. These pads on the AS-18 are simply connected to the 3-pin connector on the front, the idea being you connect the Coolpix to the AS-18 using a cable (such as SC-18) and this allows you to use a standard Nikon speedlight.

It's relatively simple to solder a 3-wire cable onto the pins inside the AS-18 and to route the cable out by removing one of the plastic inserts. On the inside of the AS-18 the:

• white wire is connected to trigger pad, the
  
• red wire to the quench pad, and the
  
• black wire to the ground of the hot shoe.

In conclusion, the AS-18 is a cheap way to interface home-brewed electronics to Nikon's range of TTL flashes. Get 'em while they are still available!

First post

Welcome to the 2nd most boring blog ever! I'm using this blog to record some of the things I am working on at the moment. For the first while it will probably concentrate on some experiments I have been doing with an old Nikon SB-25 speedlight I bought on ebay a few weeks ago. I had the idea that it might be possible to make a controller for Nikon's old speedlights that would make them compatible with the wireless CLS protocol, or at least partly compatible. For example with CLS you can use the camera (e.g. a Nikon D300) to remotely control CLS compatible flashes, such as the SB-600 and SB-800. One of the simplest things you can do is to fire them remotely and set the power of the strobes from the camera or from an external CLS controller (SB-800 or SU-800). At a more advanced level you can do i-TTL metering wirelessly and invoke the model light. It should be possible to do something similar with Nikon's older TTL compatible strobes with an appropriate controller. For example, the controller could monitor the signals (flash or IR pulses) generated by the commander, decode them and fire the connected flash generating an appropriate light pulse.

Producing such a controller will be a long term project that will require learning a lot about strobes and the Nikon CLS system. Also my electronics knowledge is quite rusty,since I have not done anything serious in many years. To start with I decided to produce a circuit that will generate a flash pulse of a fixed duration in time using a TTL compatible flash (my SB-25). Specifically it will trigger the flash then quench it at a fixed time later. The time delay should be generated digitally, so that it is accurate and repeatable. The delay must range from 10's of microseconds to 10's of milliseconds. My next post will describe the circuit to do this.

Thanks for reading the first post on the second most boring blog ever!