D-370 Slave Flash circuit

Jon5ja

Well-known member
Joined
Jun 6, 2002
Messages
722
Reaction score
0
Well, I finally got my D-370 camera today and was able to verify that my slave strobe circuit works.  First things first -- here are links to the Preliminary Schematic

http://people.txucom.net/srbarclay/images/Slave4093.gif

Pic of the crude prototype
http://people.txucom.net/srbarclay/images/Slavepic1.jpg

http://people.txucom.net/srbarclay/images/SlaveDiff.jpg

Difference at 33 feet
http://people.txucom.net/srbarclay/images/SlaveDiff.jpg

The flash used was a Vivitar Electronic Flash 2000, the $17 (or thereabouts) one from WalMart.

The pics are severely cropped and shrunken, but you can see a big difference in illumination, even though I had the flash pointed too much at the ground (hard to juggle a dangling battery, the prototype slave trigger, the slave flash, and the camera).  I'd have tried at 40' or more but I filled up the camera's memory (didn't get a SmartMedia card yet).  I'll do that another day.  I didn't put up the 20' pictures, but I think they were as dramatic in difference as the 33' pics.

The circuit in the diagram differs from the prototype only in these respects:

1.)  I didn't have an MOC3023 at the time I built the prototype, so I used an MOC3010.  The MOC3023 should be easier to drive and tolerant of higher voltages, so it should work even better :)

2.)  The prototype doesn't have a regulator on it.  I'm running it off 6 volts (4 AA cells), though the current it pulls is so miniscule, I'd recommend 9V with the regulator (you should be able to run it from the same 9V as the PIR sensor, but I'd use a separate regulator to avoid putting noise onto the PIR's regulated supply).

3.)  The prototype has a red LED that comes on to show when the "window" is open for a synchronous flash.

4.)  I'm actually using a diffused yellow LED as a sensor in the prototype instead of the IR LED specified.

I came up with the circuit Wednesday at lunch, and I ordered parts from Mouser that day with the intent of using a 74VHC132 chip instead of the CD4093.  I couldn't stand it Thursday, though, so I started digging through the junk box and found the CD4093, and prototyped it that evening.  The pinout is different between the two chips, so the pin numbers would change on the schematic if you used the 74VHC132 instead, but otherwise the diagram should work the same.

Note that I'm using an LED for a sensor.  To avoid the current draw of a pull-up resistor if I used a phototransistor for a sensor, I wanted to use a sensor that generated its own power to drive the input transistor.  I searched the world over looking for a photovoltaic sensor and couldn't find anything reasonable and available.  Finally, I remembered that when I was a kid I'd been tinkering with some infrared LEDs and noted that when I pointed them at the light in my room, they'd move the 50uA scale on my meter.  Also, I recalled that Forrest M. Mims III had published a light-beam communications circuit in the 70's using a single LED as both a transmitter and receiver.  Sure enough, I started testing LEDs with my DMM on uA scale, and the output of the LEDs varied from about 0.13 uA for some visible to 26 uA for some high-output IR LEDs when held up to my 40W desk lamp.

I used the IR LED initially, then later started testing other LEDs that were in my junk box to see how they worked.  As it turned out, EVERY LED I tried worked, except the blue ones (which don't seem to be photovoltaic)!  If you're just going to tape the sensor LED near the flash of the D-370 or whatever, I think I'd recommend just using an ordinary diffused LED.  If you want to trigger it from 50' away or something, one of the high-output infrared LEDs might be more suitable (you might want to put a tube around it and ensure it didn't look at the sun directly or that might fire the circuit).  I used the Radio Shack 276-143c with good results -- it was the highest output LED I tested.  I have a handful of different ones ordered from Mouser, I'll report my findings later.

Holler if you have questions.  I may try to do a surface-mount version that would be very small, using a 74VHC00M, a 74VHC14M, and getting rid of a couple of the transistors in doing so.

Jon
 



MCinIL

Inactive
Joined
Feb 2, 2002
Messages
1,236
Reaction score
0
Jon, great work! I can tell you are really getting into this cam stuff. I can't believe that today in 2002 you remembered a circuit from the 70's. The amount of light is definately better. Also I see you have a cute helper and her kittys. Thanks for sharing and all your work on this stuff. MC.
 

coyotebandit

Well-known member
Joined
Jan 23, 2002
Messages
1,892
Reaction score
0
Very impressive again! You do exellent work! I can't wait to try one out, looks like it makes a big difference in the pictures:)
 

Archilochus

Well-known member
Joined
Mar 27, 2001
Messages
1,285
Reaction score
0
Great job Jon.
Just curious what kind of battery life you're getting out of the strobe (Vivitar "Electronic Flash 2000")??

Most strobes I've played with in tyhe past eat up the batts fast if left 'ON'.

Archilochus
 

Jon5ja

Well-known member
Joined
Jun 6, 2002
Messages
722
Reaction score
0
Hi, Arch.

I've only played with the flash on the bench (and in the yard last night), but the Vivitar in question pulls nearly 100 mA just idling after the capacitor is charged.  The inverter circuit obviously isn't too efficient at idle.

I took the flash apart a couple of days ago and looked at the circuit.  Two things I see are constantly draining the capacitor:  

1.) the neon "ready" indicator circuit and
2.) a pull-up (pull-down?) resistor or two in the automatic flash adjustment circuit.

If you were willing to run the flash at full-smoke all the time, you should be able to disconnect the indicator and disable the automatic flash adjustment.  Then, you'd either need a "refresh" timer to periodically kick the inverter on to keep the cap charged, or some type of comparator circuit to watch the voltage at the cap and kick the inverter back on when needed to top off the capacitor.

I'm still kicking around ideas on that one, because I think it'd be cool to be able to remote-mount a couple of slave flashes on trees away from the camera to really light up the area of interest.
 

Brian

Inactive
Joined
Jul 30, 2001
Messages
1,194
Reaction score
0
Great job Jon,


I was trying to get the hot shoe slave to do the same thing but with this you will not need the hot shoe slave.

I will try it out since I use the exact same external flash unit and I have all those parts handy.

Arch,

This unit eats up batteries if left on.  It will only last a short time.  I use the micro chips with addressable switches to turn it on and off automatically when needed.

A circuit with a day/night and a relay will do the same.  Use the day/night to determine if the relay needs to be pulled in when the PIR is triggered. Then give power to the board and external flash using the relay.  Set this relay on time to equal the time for the camera to take a picture and then turn it off and no extra juice will be used to get the job done.

I essential do the same with the microchip.  It will give you the maximum number of flashes the unit can give only when they are needed at night.

My first unit turned on even in the daylight and sometimes it was dead by night time due to the pictures taken during the day powering it up even though it did not flash.

Just a thought.  Keep up the good work Jon.
Thanks
 

Jon5ja

Well-known member
Joined
Jun 6, 2002
Messages
722
Reaction score
0
Thanks, everyone.

MC, I was getting the "Take my picture! Take my picture" treatment from my 3-year-old, so I told her I was going to take a picture of the swingset and she'd have to get in it if she wanted her picture taken, ha.

As a quick follow-up, I did test several more LEDs last night as sensors, and the Radio Shack 276-143c still had the highest output of all of them.  The QED234 was also good.

I replaced the original MOC3010 with an MOC3023M, and it actually does work better -- it takes much less LED current to drive, and it has considerably less leakage than the 1987-vintage MOC3010's I was using.

Jon
 

Brian

Inactive
Joined
Jul 30, 2001
Messages
1,194
Reaction score
0
Jon,

I laid the circuit out with my software in a surface mount style.  The board was 1.25 in. X 7/8 in. when I was finished.  I did not include the regulator since I want to power it from the four AA's in the external unit already.  I am hoping it will not effect the circuit.  If it does I will add the regulator later and power it from another source.  I will etch it with my next batch of stuff.  The parts should be here Friday or Saturday I hope.

Thanks again
 

Jon5ja

Well-known member
Joined
Jun 6, 2002
Messages
722
Reaction score
0
Great, Brian!  I'll be looking forward to seeing your results.  With it that small, you could probably put a spot for 3 watch batteries on the back and make it nearly self contained if you didn't want to incorporate it into the flash directly.

Given the noise margin of the CMOS stuff and a pretty stiff supply (besides the fact that we're using Schmitt-trigger input gates), you probably don't terribly need the regulator.

BTW, using nominal CD4093 specs with the 0.1uF timing cap and 47K charging / 2.2M discharging resistors, I calculate 6.5 mS "inhibit" time and 133 mS "window" time.  So by these numbers we should be OK as long as 6.5 mS Time-Between-Flashes 139.5 mS.  Do you think a different time range would be more appropriate for a wider range of cameras?  My D-370 looks (on the scope) like about 70 mS between flashes but I don't know what other cameras are using.

Also, are you using 74VHC logic or something else on your SMT version?
 

Brian

Inactive
Joined
Jul 30, 2001
Messages
1,194
Reaction score
0
Jon,

I think the time you have is great.  The EZ is about 80ms best I can tell and most all of them fall into that range.  There might be one outside it but I doubt it.  I haven't tested very many of them.

I actually plan on putting it inside one of the hotshoe slave cases that I have.  I have one that I messed up and I think it will fit.  Otherwise, I will have to mount it on the flash unit itself.  Should be no problem though.

I am using the HCF4093B style in an SOP-14 style package.
I am using the MOC3023M style in an SO-6 style package.
I am using the MMBT2907A style in an SOT-23 style package.
The resistors and Caps are 1206 style and I ordered an infrared LED for the sensor.

I will probably etch it this week but the parts will probably get here Saturday or Monday. I am looking forward to trying it out of my EZ camera.

I will post some pictures of it when I get it done and let you know how it works.

Thanks
 

coyotebandit

Well-known member
Joined
Jan 23, 2002
Messages
1,892
Reaction score
0
Sounds like this project is coming along pretty good.

Jon,

What changes in the circuit need to be made to get rid of the transistors? Do you have a schematic for the version that uses the 74VHC00M, and 74VHC14M?

Brian,

Where did you order all your parts from? I was looking around, and it looks like mouser stocks all the parts, I was thinking of ordering them, but I might wait to see how yours works since you already have a slave flash and an EZ digital. Getting rid of the voltage regulator circuit should extend the battery life also. It would be nice to have a circuit that would work with both the EZ Digital and the D-370.

Thanks

CB
 

Brian

Inactive
Joined
Jul 30, 2001
Messages
1,194
Reaction score
0
coyotebandit,

I got my parts from Mouser.

The voltage regulator I would use only draws 1.2uA to regulate but you have to make sure the supply to the transistors comes from the raw voltage so you don't overload the regualtor.

I am hoping that it will work with both. I will find out soon enough.  I hope.
 

Jon5ja

Well-known member
Joined
Jun 6, 2002
Messages
722
Reaction score
0
CB, I had a diagram using the 74VHC00M and a 74VHC14M but I didn't like the layout so I started moving stuff around... now I've got it all haggled up and I didn't save the original :-(  When I get a few minutes I'll clean it up and post a link to it.

Doing it with the two chips only gets rid of the two transistors driving the MOC -- other than wiring 3 LEDs in series, I haven't figured out a way to get rid of the input transistor.  All this is in the vein, of course, of trying to keep it a zero-power (at idle) circuit.  I used the two transistors at the MOC because I really needed a NOR gate but didn't want to add another IC.  As it turns out, the pair of transistors may not take up any more space than another IC anyway.  Hmmm... now I'm thinking for SMT maybe an MC74VHC1G02 would fill the bill instead of the transistors...

Brian, the Seiko regulator should be happy to 40 mA, so it shouldn't be a problem to run the whole circuit from it.  With the MOC3023M, you should be able to increase that 390-ohm resistor to 680 ohms or maybe 1K since the MOC3023M should only require a maximum of 5 mA to make it work (and will probably work with less).

The way the circuit is configured, you really can't pull the emitter of the top transistor away from the supply voltage to the chip -- if you pull it higher, the transistor will be turned on all the time or else it'll try to pull the output of the gate above Vdd. (I have to look at my monitor upside down to work with PNP transistors, but I'm convinced that's right)... but, hopefully you won't need the regulator anyway.

This circuit will only be pulling the 5 mA or whatever for the 2 mS or so each flash;  the rest of the time current draw on mine is so low that it doesn't move my uA meter off 0.01 uA that it sits at unconnected  :)
 

Brian

Inactive
Joined
Jul 30, 2001
Messages
1,194
Reaction score
0
Jon,

Thanks for the clarification on the current draw I assumed it drew more.  I am hoping to get away with not having to have it either.

I will do some testing with it when all the parts get here and I can put it together.

Thanks again
 

Archilochus

Well-known member
Joined
Mar 27, 2001
Messages
1,285
Reaction score
0
Hi Jon,
Curious how well the sensor portion of the circuit works in distrubed ambient??  Any falsing??  The cheap slave triggers I've used are almost impossible to make go off without an actual strobe 'flashing' them.

Archilochus
 

Jon5ja

Well-known member
Joined
Jun 6, 2002
Messages
722
Reaction score
0
Arch,

With a fairly insensitive LED (most diffused ones) it won't false even pointed at the sun.  If you were using a sensitive IR LED outside and wanted to slave a flash from some distance, you'd want to be careful that it did not point directly at the sun, or else you'd have to add something to the circuit that would make it dV/dT (or dI/dT) sensitive.  Right off the top of my head, maybe put a resistor across the sensor LED (low enough in value to keep the voltage across it pretty low at ambient), then a small series capacitor, then a shunt resistor at the B-E of the transistor.  I'll play with that when I get a chance.  I just calculated a value for the capacitor, but it was too small for me to believe it will work :)

My intent was to mount the LED sensor near the flash of the camera, probably inside the box, so I wasn't really worried about ambient since the flash was the only source that was really going to be bright enough to fire the slave.  Obviously, this isn't intended as a general-purpose replacement for a slave hot shoe.  Even with a very sensitive LED you have to get it within a few inches of a 40W bulb to trigger the circuit (which gives me a clue about just HOW MUCH brighter that flash is than a bulb).

To make the circuit sensitive enough for general slave use, I expect you'd have to add another active amplifer stage, and then of course you'd have to have an R-C circuit to make it risetime sensitive -- and you'd likely be drawing some power all the time, so you'd probably just as well have used a phototransistor.

(Edited by Jon5ja at 5:52 pm on July 11, 2002)


(Edited by Jon5ja at 5:54 pm on July 11, 2002)
 

Steiner

Well-known member
Joined
Mar 18, 2001
Messages
62
Reaction score
0
I can't say enough about the progress you all are making on the digital unit.  Very impressive....  There is one item that I am not clear on.  It seems as though most are using Brian's chip to control the camera.  I gather that the D370 needs a refresh pulse about every 2.5 hours to keep the camera from shutting down.  My question is if we keep the camera awake all of the time, doesn't that allow the camera to eat the batteries?  I originally thought that Brian's chip turned the camera on once the PIR detected motion, fired the camera to take a picture, and then shut the camera off all in an attempt to save battery power. I may be confusing this function with a different digital that was being tested earlier. I read where someone got about 1 week out of rechargeable batteries on the D370.  Is the battery life more dependant on time or on the number of pictures taken?  I realize both of them consume power but which one dicatates when the batteries will be replaced the most.  Sorry that this is slightly off topic, but you guys seem to be the ones pushing the digital capability.  Thanks.....
 

Latest Posts



Top Bottom