PDA

View Full Version : stepper motor / PC control of emulsion making



Pages : 1 [2] 3 4 5 6 7 8 9

Photo Engineer
04-04-2007, 10:06 AM
Ben;

Just FYI, I have a complete PIC manual here and have all kinds of development tools. I gave up as it is just as hard as you point out. Our advantage at EK was being able to buy a full controller board with AD/DA controllers, and then buy the pumps already assembled with motor and input jacks.

I can't afford that route and I figured no-one else here could and that does not include the time of development. Simple syringe injection or dropwise addition using burettes will work just fine with a few mods to technique and using it I have good, reproducable emulsions for at least an Azo type and a Kodabromide type.

The film emulsion is more of a problem, and I'm working on that!

And that brings me to the final problem... Expense of Silver Nitrate. You don't just whip one of these systems up and make an emulsion. The first few will be scrap but will teach you something, so finally you will get a good one and keep on making it. But, until that point, you have to remember that each tiny step forward may be costing you from 10 - 50 grams of Silver Nitrate and that is not inexpensive.

There are people out there making emulsions this way, and I applaud their efforts. I believe that they will make good emuslions. However, the time and expense that will be involved is not for the average photographer. It may come to this at some future time however, so lets not forget it.

PE

Jon King
04-04-2007, 10:10 AM
What I was suggesting was to make all the individual systems "inteligent" so there is less load on the controling PC. The PC simply tells each chip what to do, and they get on with the job until the next instruction.
I would ramp flow rates in steps of 1/256th of the total flow. Although there would be much clunking from the valves!


If you are going to make each valve or set of valves intelligent, consider a serial bus to hook them up.. think I2C, with bus buffers. I don't know about PICs, but many microcontrollers have the i2c control logic built in. If you need to use a feedback system to control the rates accurately, instead of the open loop approach you are considering, I think you'll appreciate the local controllers even more.

rmazzullo
04-04-2007, 10:17 AM
There are 12 others referenced in this patent, but this might shed some light on vAg electrodes. Warning: wading through this patent is not for the faint of heart.

go to www.uspto.gov, and do a search for pat number 5102528.

Thank you to everyone for contributing to this discussion so far. I am very encouraged by the ideas and information presented in this thread.

I would like to know if multiple electrodes positioned in the mixing vessel and polled repeatedly would give a better picture of the mix progress and how well the ingredients were incorporated throughout. Would they have to be placed in areas that are affected last by the mixing (in the corners of the container) as well as the areas affected first? Should the mixing vessel be rounded on the bottom, instead of having a flat bottom, to mix the emulsion more completely?

Another issue is how to place the electrode(s), and still be able to mix the emulsion.

Thanks,

Bob Mazzullo

rmazzullo
04-04-2007, 10:27 AM
P.S. I know that I probably started something in this thread that is out of reach for most folks, and I apologize for this. The ability to make and coat emulsions by hand is, I believe, the only way to go to get a good understanding of what is going on. All the electro-mechanical assistance in the world won't mean didley if you don't know what is happening to begin with.

I am not backing down from the electromechanical control idea at all, and will still ask plenty of questions. But solder and breadboard time is still down the road a bit.

Thanks,

Bob

Photo Engineer
04-04-2007, 10:28 AM
Bob;

Geometry of the electrodes is critical as is mixing. Kettle shape is also critical, as different shapes give different results, especially when scaling. Mixer type is also critical as is method of delivery of solutions.

These change as an emulsion is scaled up or down.

PE

ben-s
04-04-2007, 10:29 AM
Ben;

Just FYI, I have a complete PIC manual here and have all kinds of development tools. I gave up as it is just as hard as you point out. Our advantage at EK was being able to buy a full controller board with AD/DA controllers, and then buy the pumps already assembled with motor and input jacks.

I can't afford that route and I figured no-one else here could and that does not include the time of development. Simple syringe injection or dropwise addition using burettes will work just fine with a few mods to technique and using it I have good, reproducable emulsions for at least an Azo type and a Kodabromide type.

The film emulsion is more of a problem, and I'm working on that!

And that brings me to the final problem... Expense of Silver Nitrate. You don't just whip one of these systems up and make an emulsion. The first few will be scrap but will teach you something, so finally you will get a good one and keep on making it. But, until that point, you have to remember that each tiny step forward may be costing you from 10 - 50 grams of Silver Nitrate and that is not inexpensive.

There are people out there making emulsions this way, and I applaud their efforts. I believe that they will make good emuslions. However, the time and expense that will be involved is not for the average photographer. It may come to this at some future time however, so lets not forget it.

PE


Aha! a fellow sufferer from the MOVLW and MOVWF bafflement!
I know a couple of people who can sit down at the PC and have a working PIC app in a few minutes, serial lines and all, but I have a certain amount of difficulty getting my head around the small command set. :rolleyes:

Old friends from BASIC like for-next and if-then and do-loop are either just not there, or very cut down. :confused:

I must admit I was a bit surprised by the cost of silver nitrate.
This is one reason I'm interested in a computer driven system - it might be possible to scale a make right back to just a few grams of silver nitrate and still get reasonable results. Would this be possible?

Just out of curiosity, are bigger makes more resistant to anomalies in injection rates and such? I would have thought that the bigger the make, the more of a buffer there is for mistakes? Is this correct, or twisted logic?

ben-s
04-04-2007, 10:33 AM
If you are going to make each valve or set of valves intelligent, consider a serial bus to hook them up.. think I2C, with bus buffers. I don't know about PICs, but many microcontrollers have the i2c control logic built in. If you need to use a feedback system to control the rates accurately, instead of the open loop approach you are considering, I think you'll appreciate the local controllers even more.

It's not open-loop. There are a pH and vAg probes in the vessel.
The valve pack design I suggested was just for flow control.
To be quite honest, I've never used i2c, but I was suggesting a bus driven system: (see post#11)

These (PICs) would all be connected to one serial line, and given individual ID's.
i2c may well be the way to go

Photo Engineer
04-04-2007, 10:42 AM
Ben;

At EK, good emulsions can be made at all scales, but at home here with my hand methods the larger a batch I make, the better it is. However, I'm still limited to about 600 grams unless I change my mixer. I would need more energy to go beyond 1 Kg.

Ask yourself how repeatable is adding 10 ml in one minute vs adding 100 ml in one minute.... Or adding 1 ml in one minute evenly vs adding 10 ml in one minute evenly using a dropper or a pump. You see the problem.

But, scaling this to 1000 Kg becomes a major pumping and delivery problem. Oscillations or changes in delivery rate are quite harmful to an emulsion unless it is used in a smooth ramp upwards or downwards. We were always concerned about developing pulsations when using a peristaltic pump at low flow rates.

No matter how you design the computer system, expect expensive failures at startup. Be prepared for the cost of the equipment and chemistry, but once established it is the best way to go, I'm sure. But, as I said above, not everyone is prepared to pay that price.

So, your logic is not twisted, just expensive. :D

PE

ben-s
04-04-2007, 10:45 AM
...
So, your logic is not twisted, just expensive. :D

PE


Which I guess sums emulsion making up to a tee;)

Photo Engineer
04-04-2007, 10:46 AM
Let me go beyond this question and bring up a case where you are running 3 systems. One is AgNO3, the second is NaBr, and the third is KI. You are using the 3 solutions due to the fact that vAg is being controlled, and if the KI were in the NaBr as a mix, then as the system performed its control function you would get banding of iodide which is bad.

Now consider that the peristaltic pumps oscillate (they all do AAMOF), then the sync of the KI and AgNO3 are out for another reason and introduce regular oscillations or banding as opposed to random bands in the mixed salt case.

There are a number of solutions to this problem, but before you get a good emulsion by whatever means, you MUST solve this problem.

PE

Photo Engineer
04-04-2007, 10:49 AM
One of the great emulsion makers once wrote "There is nothing more expensive than a hobby of emulsion making unless it is dabbling in the stock market" (that is an approximate quote) and I think it was Wall in the introduction to one of his books in the 20s.

PE

ben-s
04-04-2007, 10:54 AM
I follow you to an extent, and then my understanding falls over.
I can understand oscillations and pulsation in the pump systems, but what is banding?
Does it mean that the grains of silver will build up as a series of "shells", each of a different compound?
That is the only meaning I can think of...

rmazzullo
04-04-2007, 11:06 AM
One of the great emulsion makers once wrote "There is nothing more expensive than a hobby of emulsion making unless it is dabbling in the stock market" (that is an approximate quote) and I think it was Wall in the introduction to one of his books in the 20s.

PE

Well, I work in the Market, and now I'm getting into emulsion making. At tax time.

Damn the torpedoes, full speed ahead!

Bob Mazzullo

rmazzullo
04-04-2007, 11:07 AM
I follow you to an extent, and then my understanding falls over.
I can understand oscillations and pulsation in the pump systems, but what is banding?
Does it mean that the grains of silver will build up as a series of "shells", each of a different compound?
That is the only meaning I can think of...

Does "banding" refer to layers of emulsion which have higher (or lower) concentrations of ingredients than in other areas of the mixing vessel?

Photo Engineer
04-04-2007, 11:53 AM
Banding refers to bands or shells of Silver Halide in a crystal with unintentionally different compositions. If Silver Nitrate and Sodium Bromide were running constantly, and Potassium Iodide was pulsating, you would get bands of Silver Iodide within the crystal which can be quite ungood.

As one very nervous Navy person said "Damn the torpedoes and lead me to the head". That can also be used to express the nervous gut feeling one gets after spending so much on emulsion making.

Now, on to another topic....

Making an emulsion that is 'perfect' means nothing if you have no suitable way of coating it with reasonably high quality. So, you might want to apply some thought to that problem as well.

PE

rmazzullo
04-04-2007, 12:48 PM
This PDF is online at Jim Browning's Dye Transfer site, and includes construction details for his sheet film coater. Perhaps this type of coater can be modified or scaled down to coat a smaller sheet size?

http://www.dyetransfer.org/images/DyeTran.pdf

Thanks,

Bob Mazzullo

Photo Engineer
04-04-2007, 01:11 PM
Both Jim's emulsion and coating equipment are very good. He has made some very good emulsions and coatings with this setup. I have posted references to his work before.

My coating blades are a hand version of his large coater.

The only objection I have after reading through everything time and again is th fact that his emulsion uses a decelrating flow in the 2 growth portions of the make. It shows 4' and 5' for these. If I were to make it, I think I would use 5' and 4' thereby increasing flow rate as the crystals grow. I have not tried this and I don't know if Jim has either. It just seems intuitively better to me based on prior experience.

If you do build something, avoid alumiinum and copper containing materials as well as lead and iron. Stainless Steel and Titanium are best, as long as the stainless is non-magnetic.

PE

wirehead
04-04-2007, 02:00 PM
Well, I tend to prefer AVRs over PICs, on account of it being easy to get C compilers for the AVR.

Dono, I think I have more interest in building useful bits for others to use to experiment with emulsions than trying to do it myself in my apartment. vAg meters and computer controlled syringe pumps are easier to cook up in a small space than a prototype robotic coating machine. :)

ben-s
04-04-2007, 02:24 PM
Well, I tend to prefer AVRs over PICs, on account of it being easy to get C compilers for the AVR.

Dono, I think I have more interest in building useful bits for others to use to experiment with emulsions than trying to do it myself in my apartment. vAg meters and computer controlled syringe pumps are easier to cook up in a small space than a prototype robotic coating machine. :)

All very well if you know C ;) I don't.
I have just been playing with a few bits for a syringe pump. I think a leadscrew attached to a stepper, plus a plate with a threaded hole in it to pull the plunger down should work.

Photo Engineer
04-04-2007, 03:15 PM
Ben;

I believe we used a screw to push the plunger down, rather than a mechanism to pull it. I guess it makes no difference in the long run, but there may be commercial screw driven push equipment. I've even seen a design for an auger driven feed device, but never heard of it being used.

PE