X sync socket problem

My memory may not be recalling this 100% correctly, but I seem to remember, from a thread here long ago, that some studio lights have negative trigger voltages - and that caused issues.

(Not sure, but I think reversing the connections on the PC sync lead sorted it out. Personally, I would use a volt-meter to check the trigger voltage on the studio lights and check that the centre pin is positive with respect to the outer shield of the PC sync lead.)

Can anyone remember exactly?

Matt

Sorry if the following is a bit technical and lengthy, but I am pretty sure it is the answer to the problem.

I think that the explanation here lies in the nature of the interface electronics. I experienced the same problem myself recently whilst constructing an interface system for a flash project of my own.

I wished to build a system that would interface with any modern camera so I designed a circuit around a standard opto-isolator IC. The idea was to construct a system giving complete isolation from any high voltages generated by connected flash equipment and have a low interface voltage of my choosing, in this case 4.5 volts.

On my Pentax bodies (K-10D and K-20D) it fired the first flash perfectly but no more unless the system was de-powered or disconnected. Then it would fire once again. And so on…

The reason – well Pentax appear to use a latching power switching device (ie an SCR aka Thyristor) in the camera body. These devices have the property of latching “on” once switched and the current flow has to be reduced to zero by the externally connected circuit (ie the flashgun or interface) to allow the device to unlatch and be ready to switch again.

With a conventional flashgun this is exactly what happens, the flash is fired and once the capacitor is discharged, or more likely in the case of an automatic flashgun, the thyristor in the flashgun has terminated the flash, the voltage at the contacts is momentarily zero thus breaking the current flow.

In the case of a simple LED coupled opto-isolator circuit the camera will fire and conduct illuminating the opto-isolator LED thus firing the flash. But the camera will latch up the LED current as there has been no external way of reducing the current to zero and the LED stays illuminated thus preventing further flashes. It is the change of state from off to on that causes the flash to fire so some means of unlatching needs to be provided.

In my circuit I added a system to remove the voltage to the LED momentarily after firing which allowed unlatching and now it works correctly.

I would surmise that some cameras on the market from other manufacturers have a switching element based on a transistor type switch which does not latch, in fact when researching I found a simple opto-isolator system said to work on Canon’s products. However I would consider the thyristor type switch to be much more robust, it is hard to find these rated under 300 volts or so.

I spoke with Calumet and they are clueless. He had me test the connection by shorting the PC connector on the camera side. It fires the flash just fine (and repeatly) as I open and close the contact on the wire..

I called Pentax and they thought it was an issue with polarity as well. So I bought some parts to build my polarity reversing cable (1/8" mini mono plug). And still no luck with the flash.

I'm going test my friend's K7 next week to see if the problem is isolated to my camera. He also has the pocket wizards I could try.


digikid-tv:
Thanks for your in depth explanation (I'm still not entirely certain I understand it all). It sounds like a design flaw by Pentax and sh!tty for us to have to seek a work around this issue if this is the case. I may just be unlucky and have a buggy k7 I guess.

This Wikipedia entry is useful: http://en.wikipedia.org/wiki/Thyristor
Look under "Switching characteristics" and 'forced commutation'
Matt

I've just been measuring my *ist-D (because it's to hand)...

* The hotshoe & PC socket are separate. There's no connection between the centre pin of either. So that's good

* The centre pin should be the positively biased pin. Measuring with a multimeter, I get about 700 Ohms when the shutter is open (ie flash is going off), and it goes open circuit when the shutter is closed.

* I'm not 100% sure about the thyristor comments above... the flash does need to know both the 'connect' and 'disconnect' moments from the PC cord. This allows a flash to do trailing curtain sync by deciding which 'edge' to trigger off.

Think I'll have to build something up to test these various ideas... and try it all out on the K20D to see if anything's changed.

Will keep you posted
Matt

Matt,

I did my tests with the K10D. I made a test circuit with just an external battery, a LED and a current limiting resistor all connected to the hot shoe. It definitely latches when the shutter fires, I just got it out and tried it again to be sure.

Maybe the earlier cameras are different, I do not have one to test as my first digital SLR was the K10D. I have not tested the K20D as it is not at work with me. (I keep it for best).

I would be interested to know how your tests go. PM if you wish to take this further off forum (it is a bit of a geeky topic!)

please, leave this on forum. Geeky it might be but there are people like me who can follow and would like to do so

Bret

Just tested my K20D which latches on both the hotshoe and X-sync socket just like the K10D.

However I have managed to borrow a Canon EOS7D body this morning to test, this behaves quite differently and does not latch up. It resets itself without removing the applied voltage at the hotshoe. This would be quite suitable for use with a simple opto-isolator interface.

Slappy 76: Perhaps you could try your triggers on a Canon body to see if this is the root of the problem as I hypothesised in my earlier ramblings?

...I'll try and setup a test tomorrow with my K20D.

I'm still puzzled, because AFAIK, the trailing curtain sync is taken from the 'rising edge' of the sync pin - that wouldn't be possible if the sync latched (which it obviously does in your cases) - unless the trailing curtain is done in some other manner... (It's got me wondering!!)

Matt

... or unless the flash is set up so that it drains a cap in some way? the only other way I can think of for the second curtain is that the initial flash would happen - blitz - reload and wait - go.

Second curtain doesn't work with triggers, remember. Is this important?

Bret

Just coming off the subject! Which I know Zero oo about.

Welcome to the forum Jefferson.
regards
albert

Just metered my K20D... looks like the same as the *ist-D.
Now I'm puzzled

Perhaps it's the amount of current that is sunk through the PC/hotshoe connector that's locking it up? Using a multimeter will be taking microamps. Perhaps using an LED (and maybe with the studio strobes) they are drawing a larger amount of current (milliamps) and locking up the electronics.

I can test this tomorrow with the kit at work (assuming I get time)

If this is the case, then a little 'booster' would sort the problem out.

OP: Out of interest, were your studio lights 1/4" mono jack (i.e. like the big old headphone connector)? Or 3.5mm like newer headphones?

Matt

Matt, agreed that there is not enough current from a modern multimeter on the resistance ranges to demonstrate the latching effect. A current of say 5 mA is needed to show this. My test circuit used a 1k ohm resistor in series with a LED powered from 6 volts (4xalkaline AA batteries), so I reckon somewhere about 4.5mA is flowing. A typically quoted hold current for a small thyristor would be around 2 mA.

Perhaps that's what's happening then? Maybe the OP needs a high impedance 'booster' that can sink the current required on the studio lights.

I'll try and measure the trigger current on my AF360, and setup some tests with different currents on the *ist-D and K20D. Hopefully should be able to do this later today...

Matt

Ok, test results in

Setup
* DC Power supply set at 5v (current limited to about 100mA just in case!)
* Various resistors
* Multimeter to measure DC volts
* Tested on *ist-D (probably same for K20D)

Results
* 1k - locks up as described
* 100k - perfect
* 10k - perfect
* 4k7 - perfect
* 2k - locks up
* 3k - perfect

So, it would appear that a current of more than about 2mA will lock up the PC socket/hotshoe - exactly as you suggested John

Suggestion
How about a little transistor booster:

* PNP transistor (e.g. BC327, BC547, etc, etc)
* Collector to ground (the outer of the PC connector)
* 100k between base and emitter (to turn off the transistor)
* 10k between base and the camera's PC centre (to limit current into the camera's thyristor)
* Take the emitter to the Flash's centre

(Basically you just want a high impedance input - greater than about 5k - to pull down a slightly larger current at the input of the studio flash. The Calumet's are 5v trigger voltage, so no big deal.)

You could probably blob that lot into a little adapter lead. If you've got problems, just let me know which jack size it is (and if it's mono), and I'm sure we could work out something

Matt