Timing diagram trigger quench signals.

[fotoopa]

I like to start measurements on my SB-80_DX and SB800 flashes to
control remote the power. It will be a great help if someone have an
idee how this signals looks. How about the minimal pulswidth of the
signals. Are there some timing diagrams? I know about the functions of
the trigger and quench signals.

I've multiple flashes in use. I look for some information (levels,
function, signals etc) of the other signalpins on the flash hot shoe
contact. As hardware I use also the AVR controllers and have FPGA and
CPLD modules to make exact timings to test this. I've also RF12B
modules to make prototypes.

For experimental photography I use multiple flashes. I like now to
control all the flashes from my central controller. For in-flight
photography of insects I use also multiple flashes. There i have to
adjust very fast all the flashpowers by turning only 1 rotary encoder
buttom. So all this applications are similair to this SPOT project. I
suppose that this tests can be usable to this project to.

Frans.

[fotoopa]

The activity on this forum looks very low! No one give an answer.

So I've make the tests myself to adjust multiple SB-80-DX, SB88
flashes via a controller. I've make also a shoe as interface at nocost
(d'nt like to pay to much for it). Now I can regulate all the timings
to build the correct controller and hardware. In soon I will put all
the results on my Flickr website: http://www.flickr.com/photos/fotoopa_hs/

Frans.

[Ed Piebiak]

Hi Frans, it took me a couple minutes, but I realized that I'm a fan of your stuff on Flickr and Fred Miranda.  Keep up the awesome work dude!

Sadly, I've been too busy with work to do anything regarding this project, aside from researching the Canon ETTL protocol.  I've had an idea, but I was going to experiment with it first, but what the heck I'lll throw it out here now.

Imagine putting a photocell on the remote trigger and programming that strobe to say "show me f4.5" and it adjust the power automagically.  That might be handy.

Another development is that my buddy is getting some of the new PocketWizards.  I'm going to play with them and, uh, study them.  I think it would be great to get HHS and Second Curtain Sync working.

Aside from that, I've for the MSP430 wireless dev kit sitting on the shelf next to a well thumbed copy of World War Z, so that's about the progress I've made with anything Strobit.  The way the farm is, I doubt anything will materialize from me until at least winter.  I apologize for slacking.

--
Virus pwn3d this outgoing message.
Checked by Microsoft Nimda.Free Edition.
Version: 1.1.666 / Virus Database: pwn3d/j00/Ph001Z - Release Date: 1/24/1984

[fotoopa]

Hi Ed,
Thanks for your replay.

My application is really simple. I only have to adjust the flashes all
together with one turn on a button. Adjusting means add a few steps (1
step = 1/3 EV) to all the flashes from the current level. This adjust
is need for flyiing insects if thy are very white. There is no time to
change this power separately. A second buttom can regulate every flash
to a basic value. In the future I will use 4 extra flashes for the
environment light. This flashes need also the setup remote at a manual
level. Ofcourse I need feedback of every flash to see at which power
they are set. This feedback will be diplayed on my central LCD display
on the setup unit.

Now that I understand how this TTL regulation works on the Nikon SB800
or SB80 I can easy implement this into my controller. I 've also the
same RF12B modules to control them wireless and remote. Cost is very
low in case of homemade. Electronics is easy, I've just to put this
information into the controller. But this change is only for the next
version because I need a few extra I/O pins, cables and buttoms.

Now I'm testing the pulswith of the signals needed. I just found that
the minimum startpuls need 10 usec, so I have put him at 15 usec. Next
I wil calibrate all the values to have 1/3 steps EV value. The setup
to measure this is now ready, The logicanalyser is connected to a
photodiod and the FPGA controller can give me the pulses. I just have
to record all this. Also i wil test how fast 2 pulssequence can be
given in succession.

If you need Second Curtain Sync working, this is just a question of
delay. You even can give the flashes at other points into the capture
window of the camera if you like it.
I've only Nikon flashes, so I have no idea for Canon flashes but there
is for sure a same sort of solution for this.

Frans.

[Ed Piebiak]

I wonder if this would help you?

http://www.chem.helsinki.fi/~toomas/photo/flash-discharge/regular.html

[fotoopa]

I've approximately the same lightforms. I know that the flashes need a
time to rise the power. This make time measurements a bit difficult.
Therefore i do it otherwise. I compare the TTL driving with the manual
settings. Example I put the flash first manual at 1/128 power and
record the timing. Then I drive the TTL signals to record the same
time. Thats mean, I've now also 1/128 power. I use this time now to
generated via my controll the powerlevels via TTL mode.
I know also the time between the startpuls and the first light on the
flash. This time is very constant and near 32us. I have also the
stoptime, time between the stoppuls on my controller and the stop of
the light from the flash. This delay change together with the
powerlevel. If low power, this delay is short but if the power rise,
the delay rise also. This only mean that at higher power the drive
signal of the TTL mode need a shorter puls to compensate this extra
delay. But if I compaire all the manual power settings then I can easy
determine the corresponded timing.
I've the first setup on my Flickr site:
http://www.flickr.com/photos/fotoopa_hs/3709619306/

Frans.

[Till Seyfarth]

Hi there,

I've simply measured full power and then repeatedly divided by two
from there - trial and error (approx. 50 recompiles or so). The
approach works fine as long as you have a light meter at hand. I
thought this would be necessary for each flash type anyway - so I kind
of waited for measurements for other flash types. I think I may have
missed something in the discussion but it seems quite complicated the
way you need curves and all that...?

Cheers
Till

[Ed Piebiak]

Yeah, I think you got it right.  We're overcomplicating this, but I think that Frans has to look at this carefully because, if I recall, he can actually shutter sync to 1/1000 or something with his bug rig.  Timing will be an issue there.

However, most of us don't trigger our cameras with lasers.  I find the shark too heavy to carry around strapped to the 70-200 f2.8 IS unless on a tripod.  ;)

I was looking into this trying to figure out how Second Curtain Sync and HSS (which is really getting the flash to go stroboscopic), and uh, this will be tricky.

One thing I'd like to have implemented though is having an adjustable delay between shutter and flash trigger.  This would be, in effect, second curtain sync.  Having groups of flashes fire sequentially would also be neat.  I'm sure there could be some cool uses for that.

I think whipping up a photocell test rig like Frans's would be a good thing to do.  The info from that would be kinda handy.  I'll see if one of my old proto boards could be hacked to be made useful for that.

e

[fotoopa]

I believe the way i measure the timings are very accurate and fast. I
Use the flash himself first into manual mode to see how long the light
output falls on my photodiode, Then I switch to TTL mode and set the
digitale readout delay (5 digits on my controller display at
resolution of 1 us) so that the same light output time is reached in
TTL mode. This way is very accurate and fast. I d'nt need a light
meter to do this job. I have done this measurements at 1/3 step EV
from 1/128 tot full power in less than 15 min. The pulswidth of the
signals are not linaire because more power need longer swith-off time
after the stoppuls on the Q signal. Ofcouse all the timings depend on
the the type of flash. So I have multiple SB80/SB800 but this types
are the same.

The timing table stay now on my Flickr site: http://www.flickr.com/photos/fotoopa_hs/3709381751/

How to drive the flashes with this delay is not the problem. For my
wireless application I like to use the same way as into this spot-
development project. An simple AVR controller together with the RF12B
module for the wireless connection. For the flashes fix on my rig for
flying insects there is no need for wireless, so there i can use a
standard SC-27 nikon TTL cable connected from the flashes to my
controller and using the same delay table to drive the signals.

For Second Curtain Sync there you may use the X-sync connection from
the camera as startpuls to drive camera wireless controller. I see no
problem different to the front Curtain Sync. For HSS there the
problems are different and without a test I can not given advice.
Frans.

[Ed Piebiak]

Yeah, great work on that timing table.  That will cut out a lot of guesswork later.  I suspected that the timing would be non-linear.  I'll have to make a setup like this sometime too.

What photocell did you use? 

[fotoopa]

Oh, thats an very old type. I've him into a short cable, fixed with a
sort resin. I can't see any more the type but any photodiode will
work. I use for the laser detection the BPW34, BPW26 and the BPX43-3
types. Because the light is high at short distance the diode is not
critical to measure the flashes. Just look that the diode is not
saturated. I put a low resistor to have a short rise and fall time.

[fotoopa]

Hi Ed,

I've just replaced the photodiode on the setup with an BPW34 because
I've multiple of this photodiodes in stock. The results are the same.
I 've tested also the newer SB800 flash and the light output for the
same manual settings are a little higher. An other SB-80-DX have
results in the range as the SB800. But global all the values are
useful. They give alway a 1/3 EV higher value if you take the next
value from the table.

An other methode to measure or calibrate all the flashes to the same
level can be done with the camera and a gray card. Take a photo at fix
setup and look for the histogram in an photoprogram like Nikon
caputure or photoshop. You will see a small histogram value. To
calibrate multiple flashes, you have only to adjust the Q-signal to
get the same histogram level. If you adjust now the camera for a next
1/3 step (ISO or aperture) , you can calibrate all the steps and
flashes. This way is more time-consuming but very accurate at no extra
tool cost.

Frans.