Building A Professional Grade Shutter Tester

[jfk]

I'm having a rough time locating an acrylic sheet in Europe that fits? I'm not sure how important the light transmission percentage is for this or if it being any acrylic sheet that is matte on one side would just work? Any help regarding that would be appreciated

EDIT: I ended up ordering a sheet off of the US amazon because the shipping was only 8 euros

 

[ic-racer]

I think the acrylic sheet is important, if it is too thick or dense, the voltage will have to be set higher than 12V. If it is too transparent, the light distribution may show a hot-spot in the middle.

 

[Bill Burk]

For a similar project I found a frosted white acrylic clipboard at Office Depot.

 

[ic-racer]

Since the calibration curve of the EV light output is fixed in the firmware, the 'correct' acrylic is important, as is the 'correct' LED.

 

[srozum]

Hello there!

I want to let you know that I just published the source code of Film Camera Tester SRT-2 on Github.
I am glad that after so many years there are still interest in building that project.

What you possibly can do with that code?

You may want to modify calibration tables to match your LED. Add different sensors or measuring modes. What else?

Consider that latest version takes almost all memory of Atmega 328 chip. So if you want to add new functionality you have to sacrifice some existing functionality.

I doubt that it can be easily transferred to more powerful architecture. But who knows?

Best,
Serhiy

P.S. Big thanks to @ic-racer for believing in this project and guiding other builders.

 

[ic-racer]

Thank you Serhiy! What an awesome device you created. Can't imaging repairing a camera without it.

Hope so see others complete the project!

 

[aconbere]

Hello there!

I want to let you know that I just published the source code of Film Camera Tester SRT-2 on Github.
I am glad that after so many years there are still interest in building that project.

What you possibly can do with that code?

You may want to modify calibration tables to match your LED. Add different sensors or measuring modes. What else?

Consider that latest version takes almost all memory of Atmega 328 chip. So if you want to add new functionality you have to sacrifice some existing functionality.

I doubt that it can be easily transferred to more powerful architecture. But who knows?

Best,
Serhiy

P.S. Big thanks to @ic-racer for believing in this project and guiding other builders.

Amazing news! This is huge gift to the community. Thank you!

 

[Arborix]

Hello to everyone,

being a long time lurker in this forum, I am planing to build this nice project. Big thanks to srozum, ic-racer and other participants!
First parts are ready, others in the mail.

I will be using a FDM printer. I am wondering what temperature will be reached in the led mount area while running at highest current? Is PLA+ still sufficient or should I use something like ASA for this part? I prefer PLA+ because I like the matte black more than the shiny ASA. And it is easier to print, of course.

Thanks again!

 

[srozum]

Hello to everyone,

being a long time lurker in this forum, I am planing to build this nice project. Big thanks to scrozum, ic-racer and other participants!
First parts are ready, others in the mail.

I will be using a FDM printer. I am wondering what temperature will be reached in the led mount area while running at highest current? Is PLA+ still sufficient or should I use something like ASA for this part? I prefer PLA+ because I like the matte black more than the shiny ASA. And it is easier to print, of course.

Thanks again!

No issue using PLA for the light source. LED runs max at 1A, which is below its maximum rating. Plus heat sink keeps it cool and stable.

 

[nusproizvodjac]

Hi everyone, I'm planning on building the tester, and I have most of the parts already, but I need the 3D printed ones, and the light source.

I'm thinking of ordering everything from JLCPCB, but for the case and some other small parts, I can't get them printed in PLA. Any suggestions what other materials to use that are available from their site?

I also ordered these LEDs, hopefully they will be ok.

Epileds 7070 20W 12V 6V Led replace Cree MKR XHP70 XHP50 LED Emitter 6500K

https://a.aliexpress.com/_EIlqDyM

Thanks!

 

[ic-racer]

Hi everyone, I'm planning on building the tester, and I have most of the parts already, but I need the 3D printed ones, and the light source.

I'm thinking of ordering everything from JLCPCB, but for the case and some other small parts, I can't get them printed in PLA. Any suggestions what other materials to use that are available from their site?

I also ordered these LEDs, hopefully they will be ok.

Epileds 7070 20W 12V 6V Led replace Cree MKR XHP70 XHP50 LED Emitter 6500K

https://a.aliexpress.com/_EIlqDyM

Thanks!

Building A Professional Grade Shutter Tester

When you upload the Gerber files to JLCPCB there is an option to have them put the surface mounted components on the boards; The other 2 files along with the Gerber file have the components locations and values for JLCPCB to put on. I did it that way and got the boards with the SMD components on...

www.photrio.com

 

[nusproizvodjac]

Oh, I started reading the topic backwards, looks like I didn't go far enough, sorry!

One more question regarding the Sensor 6 housing, there seem to be three diagonally positioned rectangle holes on the inner part of the lid STL. They don't correspond to sensor orientation, at first I thought
Sensor 6 and 1 got mixed up, but now I wonder, what's supposed to fit inside those?

 

[nusproizvodjac]

I've assembled the tester, complete with the power supply, and I've tested it a bit. It measures nicely, tested with the simple sensor and a Nikon F5, it's dead on.

I powered everything with a Meanwell LRS 50-12, but the voltage regulator on the Nano gets hot, and I think I'll add a buck converter and power it with 5V, and reserve the 12V for the light unit.

However, I am running into an issue with the LCD or booting. Sometimes the LCD only shows blocks, until I reset the Nano repeatedly 2-3 times by pressing the reset button on the Nano, and then it powers up normally and runs normally. Any ideas what might be the issue?

 

[srozum]

I've assembled the tester, complete with the power supply, and I've tested it a bit. It measures nicely, tested with the simple sensor and a Nikon F5, it's dead on.

I powered everything with a Meanwell LRS 50-12, but the voltage regulator on the Nano gets hot, and I think I'll add a buck converter and power it with 5V, and reserve the 12V for the light unit.

However, I am running into an issue with the LCD or booting. Sometimes the LCD only shows blocks, until I reset the Nano repeatedly 2-3 times by pressing the reset button on the Nano, and then it powers up normally and runs normally. Any ideas what might be the issue?

Hello,

Option 1 it could be corrupted Nano memory.

Option 2 interference with i2c bus. Socket 2 has i2c pins, you might want to remove those wires because socket 2 doesn’t use it.

Option 3 power issue. If regulator gets hot, something drains the power. Check for shorts, bad soldering, etc

 

[nusproizvodjac]

Hi, thanks for the answer!

The B socket isn't connected at all, I haven't soldered it yet, only A, so I don't think that's the issue, since even without the A socket it had the same problem.

With only the board and LCD connected, the LCD goes haywire, probably because the keyboard is missing. As soon as the keyboard is connected, it is stable, not cycling through the menus.

I'll recheck the soldering on the keyboard, but it should be ok, I tested the continuity after soldering. The shield is pretty straightforward, I don't think the issue is there.

I suppose the regulator is getting hot, because I could smell something giving off a burning smell running it on 12V, that's why I think something is getting hot, and my first thought was the regulator. It shouldn't be an issue, it's rated for 12V, but with these knockoff boards, it's a gamble.

Corrupt memory may well be the issue as well.

 

[ic-racer]

Yes, maybe a short. Check for things like stray strands of wire, etc.

 

[Arborix]

but the voltage regulator on the Nano gets hot, and I think I'll add a buck converter and power it with 5V, and reserve the 12V for the light unit.

Just regarding power supply for the Nano. I would not recommend running the Nano from 12 V, because the onboard power regulator is very simple and just generates a lot of heat, perhaps adding to the rest of problems; better add a small voltage regulator und supply it with 5 V, as you already suggested.

 

[nusproizvodjac]

I think I solved it. Jeez, this took a while.

For uploading the firmware I used the USB ASP dongle, but I also used a program called AVRDUDESS. I don't really like the cmd layout, I prefer GUIs.

Basically, I read the fuses and the lock bits, didn't touch anything, loaded the firmware and the settings and went my merry way. I failed to realise that the command line on github for setting fuses and LBs didn't match what I read from my Nano.

After chasing cold solder joints, shorts and whatnot, I said to myself, no way there's an issue with connections, since I always check my connections with a multimeter after soldering.

So, I turned to software issues, retraced my steps and compared the settings and saw the fuses and LBs. I set them according to the github code and now the tester reliably starts at 5V from USB.

However, I will add a buck converter as Arborix suggested, because it runs fine from the MW PSU up to about 11V. After that it kinda starts falling apart. The screen goes nuts, doesn't turn on the backlight, doesn't respond to reset command. No need to stress the board further.

 

[nusproizvodjac]

Finally finished.

Still waiting on the acrylic diffuser to arrive, in the meantime I used a sheet from an LED ceiling light. Seems to be made of polycarbonate, judging by the way it cuts. Even so, the light output is linear, compared to the meter on my F5, dead on. The output voltage from the PSU is around 12.2V.

Aside from the F5, the most consistent and accurate shutters in my collection are Nikon FA and Canon AE1P, and by accurate I mean always within 0.1EV. Funnily enough, Almaz 103 and even Zenit TTL have proven to have pretty accurate shutters.

 

[ic-racer]

Very nice!

 

[nusproizvodjac]

Thanks! And thank you for your help, I've also been talking to Serhiy on FB, he's also been of tremendous help troubleshooting a few things.

It turns out that even though I used metallic film resistors with 1% tolerance, a one of them turned out to be way out of spec, for example 1.8k instead of the declared 2k, which is 10% out, enough to throw off the ID circuit for the light unit.

The shop selling them also mislabeled a few of them, and I ended up using a 4.7k on the shield for the pull-up resistor for the A socket instead of 10k.
I remember thinking, why is one resistor physically larger than the other one, but went with it anyway. The result was Sensor 1 giving me "Not ready" message. I checked and rechecked the joints, shorts and whatnot, until Serhiy in passing mentioned that knockoff Arduino boards can be dicey, and that's why he used his own 10k pull-up resistors for the sockets. I remember thinking, why would he mention that, but then it clicked, and one probing of the board later exposed my mistake.

On the plus side, I've mastered SMD soldering by hand, so there's that.