LED Strobe Trigger For Olympus OM-D E-M1 Mark II

[vsc 5]

An older thread discussed modification of an Olympus flash to reduce cycle time and battery consumption. While the results were mixed, I was motivated to create a functionally equivalent solution for the Olympus OM-D E-M1 Mark II without any limitations. Using an Olympus FL-LM3 flash I substituted LEDs in place of the Xeon tube, enabling triggering an external strobe (e.g. Sea&Sea YS-D2) preserving TTL functionality while reducing cycle time and battery consumption.

 

Micro 4/3's cameras use a serial protocol between the camera and flash which is closed to the public. Looking at the bit stream I quickly decided it was a waste of time to reverse engineer the protocol and turned to simply modifying the Xeon strobe subsystem of the flash. In short remove the Xeon tube, trigger transformer, disable the discharge path for the photoflash, substitute a small (e.g. 1 uF photoflash cap) for good measure and add a LED driver and LEDs. The FL-LM3 is a traditional modern strobe using a IGBT to not only drive Xeon tube, but also the trigger transformer. Stop the strobe as part of TTL metering, the IGBT device is driven into pinch off, thus quenching the Xeon tube, leaving whatever charge is left on in the capacitor. This enables the flash to emit pre-flashes, as well as the main flash.

 

By connecting the LED driver to the IGBT's gate, the LEDs emit exactly the same duration light pulses as would have been emitted by the Xeon tube preserving the pre-flashes, etc.

 

The following is a link to the instructions for this modification,

 

https://www.instructables.com/id/LED-Strobe-Trigger-for-Olympus-OM-D-E-M1-Mark-II/

 

Also a link to a video showing an E-M1 Mark II in an Olympus housing with the modified flash, triggering a Sea&Sea YS-D1 running at frame rate limited by the communication rate between the body and FL-LM3. Note this is somewhat of a trick given if the YS-D2 recycle time increases the frame rate must drop, but there is no feedback to the body. Thus setting the body to run at a lower frame rate is the path forward.

 

 

Edited May 23, 2017 by vsc

[Yako 8]

great!

[Draq 135]

Great idea, If you or someone else were to make some of these and sell them, I would definitely be interested, in the absence of a commercial ttl trigger for Olympus. For me DIY involving electronics and other small delicate things is likely to turn into BIY (break it yourself) so I don't do much of that.

[ProfF 12]

Hi, cool idea!

 

I am wondering whether a similar mod can be done to the FL-LM2 (the smaller flash of Mark I OM-Ds). Anyone with electronic background can comment on whether this mod will work?

 

If it helps, I disassembled my older LM2 and attached photographs. As far as I can tell, the xenon tube assembly is similar to that of the LM3. Any suggestions how to proceed will be appreciated...

 

 

 

Edited July 4, 2017 by ProfF

[Jock 14]

Ha! Yesterday I disassembled an old non-functional FL2 myself and wanted to post photos and the very same question as well. Funny. Hope someone can help.

[Joss 15]

Hello, thanks for putting this together and sharing.

I'm definitely interested.

However, i don't have any electric / electronic skills and the best for me will be to get your bill of materials and instructions to a electronic repair shop.

 

Thanks!

Joss

[vsc 5]

The forum was (is?) not sending me updates on this thread. Perhaps my changes to the settings fixed that, but I'm not confident that what I changed will resolve the problem. And it isn't a spam email filter issue either.

 

With the camera cited in my original posting, with the Olympus housing, I found that the body can be rotated in the housing sufficient to upset the LED / fiber alignment such that the light emitted from the a fiber is insufficient to trigger the strobe. One of the two strobes in use was out of commission for the dive. This only occurred once where I found out in the process of shooting pictures of a pair of frog fish. Switching to an overhead strobe position with the remaining functional strobe yielded good shots... Actually very good shots to the point that I need to use that strobe position more often.

 

Normally I test the rig, shooting a frame or two, but I got lazy that day. It appears I also got hurried installing the camera into the housing. As such as part of setting up the rig look into the fiber mounting holes to ensure you can see the LED's in the center of the holes and/or shoot a frame or two ensuring your strobes fire. A standard strobe with its diffuser and much higher light output would never have this issue.

[vsc 5]

It is hard to say given these pictures given enough of the circuit card is obscured by the black tape. That said I strongly believe something similar could be done given most strobes use more or less similar implementations (e.g. IGBT, etc).

 

 

Hi, cool idea!

 

I am wondering whether a similar mod can be done to the FL-LM2 (the smaller flash of Mark I OM-Ds). Anyone with electronic background can comment on whether this mod will work?

 

If it helps, I disassembled my older LM2 and attached photographs. As far as I can tell, the xenon tube assembly is similar to that of the LM3. Any suggestions how to proceed will be appreciated...

 

 

Edited August 29, 2017 by vsc

[vsc 5]

I just got back from two weeks of diving where the E-M1 mk II rig was in the water for dives averaging 90-120 min of bottom time. With the standard Olympus strobe mounted to the camera, I found that I still had 55-60% of the camera's battery life remaining after two such dives. Thus from a battery savings perspective the LED modification most likely doesn't make sense. Point in fact with the mk II, once it was sealed for the day I kept going and left well enough alone.

The YS-D2 strobes recycle time specified by the manufacture as 1.5 seconds maximum with Enloop batteries. While it is easy to show much quicker recycle times above water, thus permitting a rate of multiple frames per second, underwater at nominal subject ranges the loss through water is sufficiently great that I found it difficult to get the YS-D2 recycle time to under a second. On the other hand the Olympus strobe takes three seconds to recycle, which seems like eternity at times. The from this aspect the benefit is an overall recycle time of ~1 second. Clearly sometime this makes the shot. Note that the YS-D2 recycle times would be shorter in macro photography but there are few instances where this makes a difference and in those cases I normally use a different method to get the subject into view.

[Kraken de Mabini 204]

Hi:

Interesting mod to the flash, I like it. A follow up question: Has somebody designed an LED powered strobe, using say a Cree COB LED? I ask because Inon has withdrawn the Z240 strobe and Sea&Sea has also withdrawn several strobes, plus the YS-D2 has flaky xenon tubes and also has other problems.

[Pavel Kolpakov 82]

Has somebody designed an LED powered strobe, using say a Cree COB LED?

Yes, of course.

There is a great difference between xenon and LED strobes:

 

Photographer needs a very short flash duration to freeze image - about 1...5 ms. This is a main purpose of the flash use.

Xenon tube does this job very effective, discharges several hundred joules energy at 1...5 ms time, because of gas discharge inside the glass tube. Easy to reach most popular GN=24 for our uw photo at 3ms pulse duration..

The another good fact is that xenon tube is very cheap, it costs about 3 $.

But most powered big size COB LEDs (cost 200...600 $) can discharge only about 1...3 joules energy at the such short time. Usually most powered LED strobes have GN=3...4 at ISO 100 standard, not much. Do you need wide-angle uw strobe with GN=3 ? I think, no.

COB LEDs have very small pulse power, but bigger size and much more expensive than xenon tubes.

 

As you see, LED strobes are absolutely inefficient for uw market, until any interesting LED technology solutions appear in the future.

To date, it is more efficient to build xenon strobes.

Edited October 2, 2017 by Pavel Kolpakov

[ProfF 12]

It is hard to say given these pictures given enough of the circuit card is obscured by the black tape. That said I strongly believe something similar could be done given most strobes use more or less similar implementations (e.g. IGBT, etc).

 

 

 

Trying again (was away on a trip so did not have the LM2 under my hands). Hopefully pictures w/o tape would help.

 

Any suggestions on where to start (my electronics knowledge is minimal) would be great

 

Thanks!

 

[vsc 5]

I'm trying to read the part numbers from some of the devices on the boards. The (IGBT) transistor that is used to trigger the strobe tube in the newer product is a RJP4010AGE (nominally marked as R4010 on the package). That device is from Renesas where their catalog shows they also over a 4009 IGBT. Perhaps you can look to see if one of packages with 8-pins is marked with either 4010 or 4009.

 

The idea used in my original posting, it is grab the signal going to the IGBT switch and use that through a transistor suitable to drive an LED. In short the IGBT works fine for the strobe tube with the very high voltages involved but is difficult (impossible?) to make work with an LED operating at a few volts.

 

The data sheet for the 4009 is,

 

https://www.renesas.com/ko-kr/doc/products/transistor/002/r07ds0370ej_rjp4009ans.pdf

[ProfF 12]

I can read R4006/1302 on one of the packages (the one adjacent to the one marked "f2m"). There is another small 8-pin package on the other side but as far as I can see (with magnifying glass) it does not have any marking....

 

The R might be a P and then this might be P4006 which is actually a code for another (older?) IGBT from RENESAS with very similar specs. https://www.renesas.com/en-us/doc/products/transistor/004/rej03g1865_rjp4006ageds.pdf

[vsc 5]

I can read R4006/1302 on one of the packages (the one adjacent to the one marked "f2m"). There is another small 8-pin package on the other side but as far as I can see (with magnifying glass) it does not have any marking....

 

The R might be a P and then this might be P4006 which is actually a code for another (older?) IGBT from RENESAS with very similar specs. https://www.renesas.com/en-us/doc/products/transistor/004/rej03g1865_rjp4006ageds.pdf

 

Somehow when I did a search I didn't come up with the P4006... I'm sure you found the correct data sheet.

 

I believe you can follow a similar modification path using what I have documented. The gate of the IGBT, pin 4, is used to drive the LED driver (e.g. bipolar transistor).

 

Removing the resistor marked 753, disconnects the high voltage from the IGBT as well as prevents charging up the capacitor used to drive the trigger transformer.

 

Finally I swapped out the photoflash capacitor for something smaller though this is more cosmetic than anything else. The photoflash cap is never discharged thus the size of the capacitance is not relevant. Though it did make more physical room in the strobe by putting a physically smaller cap in place of the larger Olympus cap.

 

What remains is to figure out where to pick up +3.3v and ground. You might need to use an ohm meter and trace from the hot shoe to the circuit card to find suitable points to solder wires.

 

The strobe I modified had wires going through a rotary joint, thus I used the existing wires. Your pictures show you can simply disconnect the strobe tube CCA and hook up the LEDs with whatever means makes sense.

 

Overall the hardest part of doing this mode is getting the LEDs positioned correctly behind the fibers. I started out with using Kapton tape with piercings to hold the LEDs with the strobe partially assembled. Once I found the right positioning I drilled the strobe diffuser to match the test LED position. I'm sure other schemes are possible to judge the correct LED position.

Edited October 15, 2017 by vsc