Printalyzer Densitometer - A compact budget-friendly densitometer project

[BHuij]

I mostly just don't want to pay a premium for someone else to 3D print something for me when I could do it myself for pennies.

 

[tih]

I mostly just don't want to pay a premium for someone else to 3D print something for me when I could do it myself for pennies.

On the other hand, we want Derek to make money off this product, after all the work he's put into it, right? Partly because it's deserved, but also because that ought to increase the likelihood that he'll want to invent more useful stuff for us to buy.

 

[koraks]

@tih indeed.
Moreover, I'd gladly pay a 'premium' for a design that's properly engineered and verified to work. Knowing my own skills in 3D engineering and printing, I'd need at least one trial version and then a final one to correct the issues. Drawing the thing in CAD would take me a couple of hours when all is said and done with all the little details that go into a design like this.

 

[dkonigs]

Okay, so there's a lot to unpack here. I'll attempt to paraphrase the various questions/comments and answer them. So here goes...

Will you be selling this as a DIY kit?

I have no plans for doing that. Making such a kit would be a tremendous amount of additional work, and a support headache, and likely wouldn't end up saving you that much money (depending on where the line is drawn between "provided" and "DIY" sub-assemblies).

The circuit board is indeed almost entirely surface-mount components, and you absolutely do not want to be hand-soldering those.
The basic process for building up one involves:

1. Use a stencil to apply solder paste in all the right spots on the PCB
2. Working under a microscope, using tweezers and/or a vacuum tool, carefully place all the components
3. Bake the PCB in a special oven to reflow the solder

I've personally done this many times for building all of the prototypes, however the production devices will have component placement done by machine.

What about 3D printing myself, with a supplied PCB and other ancillary components?

While technically possible, this would also be setting yourself up for an immense amount of frustration for an inferior result. Getting all of the dimensions and tolerances of the process dialed in can be a lot of work, and even the slightest misalignment or offset can lead to assembly difficulties and/or inconsistent measurements (especially with transmission).

That being said, there are basically two ways to 3D print something like this:

1. An FDM-style 3D printer, like many of us have at home. This is cheap, and easy to fiddle with, but the whole problem of support material is not to be understated. If you don't carefully manage where supports go, or make supports out of a second material, or slice/redesign the model to minimize them, you may end up with horrible surface quality where you least desire it. I used this method for many of the early prototypes, and it can be made to work, but takes some effort and doesn't always look the best.
2. A commercial 3D printer such as a Multi-Jet-Fusion machine. This is what all of the production devices use, and the print quality is very good. The problem is that you're not likely to own such a machine yourself, so you have to outsource to someone who does. And they're absolutely going to charge you for the privilege of using it. And if they're a faceless company on the Internet (e.g. Shapeways, 3D Hubs, Craftcloud, etc), you'll pull your hair out getting parts made to acceptable tolerances. Often they'll hide behind official process specs, which when multiplied across a multi-part assembly, let them get away with crooked parts that don't fit right and say its still "within spec."

Can I flash firmware over USB?

Yes, you absolutely can! In fact, I explicitly chose a microcontroller with built-in USB DFU support to make this possible even on a completely blank device.

In practice, there are three ways to install firmware onto the device:

1. Using a debugger (e.g. STLink) attached to a special connector on the PCB. This is what I primarily do for firmware development, and some of the PCB checkout and testing.
2. Over USB, by pressing a hidden button that starts the device in USB DFU mode and then loading with special software (e.g. dfu-util)
3. Over USB, by holding down more obvious buttons on startup which run a custom bootloader that implements UF2. This approach means the device will show up as a USB mass storage device on your computer, and you simply drag-and-drop the firmware file to update it.

For normal users, the plan is to basically always use UF2 for updating device firmware. The only time I expect DFU to be used, is to update the custom bootloader itself.

But could I still build it myself, if I really wanted to?

Yes, you absolutely could.

In fact, I'm keeping the entire design as open-source with all the necessary schematics and models available here:

GitHub - dektronics/printalyzer-densitometer: Reflection and Transmission Densitometer for Photographic Darkroom Use

Reflection and Transmission Densitometer for Photographic Darkroom Use - GitHub - dektronics/printalyzer-densitometer: Reflection and Transmission Densitometer for Photographic Darkroom Use

github.com

Just be aware that its going to cost you a lot more money to DIY than it would to buy one from me once I start selling them. There are a lot of parts to order, many aren't cheaply bought as a one-off, there's associated shipping costs, you may or may not have all the necessary tools, etc.

Also, don't forget about the global electronics component shortage going on right now. There may be parts that are difficult to buy, even if you want to buy them, unless you're willing to pay some shady scalper/broker. I made sure to order a lot of things up-front as a hedge against this.

And if you have any more questions, please feel free to ask!

 

[albada]

And if you have any more questions, please feel free to ask!

Seeing that your desktop app supports MacOS/Linux/Windows, please add support for iPhone and Android.
JUST KIDDING!
Seriously, I am amazed at how much work you've put into this project. I'm surprised that a desktop app exists at all; I expected users to read numbers off the display. I looked over some of your software and firmware, and they alone must have consumed a huge amount of time. Plus designing electronics and machinery! And this is just the densitometer; you're also working on the Printalyzer. How do you do it all...

 

[dkonigs]

Seeing that your desktop app supports MacOS/Linux/Windows, please add support for iPhone and Android.
JUST KIDDING!
Seriously, I am amazed at how much work you've put into this project. I'm surprised that a desktop app exists at all; I expected users to read numbers off the display. I looked over some of your software and firmware, and they alone must have consumed a huge amount of time. Plus designing electronics and machinery! And this is just the densitometer; you're also working on the Printalyzer. How do you do it all...

I know you asked that a a joke, but...



As nice as the desktop app is, I actually made it irrelevant for most users. The Printalyzer Densitometer can be configured to act as a USB HID device (a.k.a. simulated keyboard). This means that it can simply "type" readings into whatever program you have open on your computer. And if you have a smartphone with USB-OTG support and the right cable, well, you can see the video

 

[albada]

That's a great idea! I would not have thought of that.

 

[adeyoujo]

I know you asked that a a joke, but...



As nice as the desktop app is, I actually made it irrelevant for most users. The Printalyzer Densitometer can be configured to act as a USB HID device (a.k.a. simulated keyboard). This means that it can simply "type" readings into whatever program you have open on your computer. And if you have a smartphone with USB-OTG support and the right cable, well, you can see the video

That is great! Talk about a time saver!

 

[dkonigs]

As of now, the Printalyzer Densitometer is finally available for sale!

Printalyzer Densitometer — Dektronics

Compact budget-friendly densitometer for anyone who wants to take a more analytical approach to testing black and white photographic film or paper.

www.dektronics.com

(Initial batch at an introductory price. As it sells out, I'll build more inventory.)

 

[Michel Hardy-Vallée]

As of now, the Printalyzer Densitometer is finally available for sale!

Printalyzer Densitometer — Dektronics

Compact budget-friendly densitometer for anyone who wants to take a more analytical approach to testing black and white photographic film or paper.

www.dektronics.com

(Initial batch at an introductory price. As it sells out, I'll build more inventory.)

YES!

I'm order #5, so I guess that makes me a late adopter...

 

[tih]

#6 here - early Christmas present for myself!

 

[albada]

#10 - I didn't quite make the single digits.

 

[dkonigs]

If it makes you feel any better, order #1 got device serial number 18

(I started counting with the initial prototypes, and the numbering is based on circuit board testing. So the pre-release numbers include prototypes, test units, evaluation units, etc.)

 

[MattKing]

I shared the information with those of my friends who have old densitometers - some of which have old, in need of replacement bits and pieces like bulbs.

 

[Michel Hardy-Vallée]

Well, here is user-provided family of curves #1, then. This took me maybe 10 minutes using the Dektronics and Excel, measuring and graphing the data from a stack of Kentmere RC VC test prints made with a Stouffer projection strip. I will redo them when I have the time to sit down and characterize my materials properly, but this is a thing of beauty:

 

[Steven Lee]

@CatLABS would be a perfect distributor too

 

[munz6869]

Order #23: super excited!

Marc!

 

[bernard_L]

Well, here is user-provided family of curves #1, then. This took me maybe 10 minutes using the Dektronics and Excel, measuring and graphing the data from a stack of Kentmere RC VC test prints made with a Stouffer projection strip. I will redo them when I have the time to sit down and characterize my materials properly, but this is a thing of beauty:

This paper seems to have a problem. Hardly a slope difference between 2 and 3, contrasted with big slope difference between 3 and 4. And Dmax in the range 1.5-1.6. Old paper? Matte surface?

 

[koraks]

Kentmere VC RC. I bet it's old alright. Maybe fogged as well.
Regardless, it looks like the densitometer worked nicely on that one!

 

[albada]

Mine arrived today (that was fast!).
And one paper-graph deserves another. Here is Ilford RC (> 1 year old) compared to Foma RC (2 months old). Both stored refrigerated. This shows both papers at maximum contrast (blue light). On the X-axis, three steps is one stop.

Foma has higher max contrast than Ilford, but Ilford has whiter white and blacker black.

EDIT: I'm going to love this densitometer. Thank you Derek!

 

[albada]

@dkonigs : In your manual, under "Theory of Operation", you mentioned that "These tests consisted of a repeatable temperature ramp from 0°C through 45°C."
Question: How did you ramp temperature? Do you have an environmental chamber?

EDIT: By the way, I'm impressed that you performed temperature tests on the densitometer. Many engineers would have overlooked that.

 

[dkonigs]

@dkonigs : In your manual, under "Theory of Operation", you mentioned that "These tests consisted of a repeatable temperature ramp from 0°C through 45°C."
Question: How did you ramp temperature? Do you have an environmental chamber?

EDIT: By the way, I'm impressed that you performed temperature tests on the densitometer. Many engineers would have overlooked that.

I covered this a bit in my video on product testing. However, here's basically what I did...
I put the device in the freezer for about a half-hour or so, which cooled it down to a bit below -5C. Then I took it out, put it into a heated chamber (originally designed as a 3D printer filament dryer) and hooked it up to my instrumentation. Then I monitored both internal temperature and sensor readings while waiting for it to warm up to room temperature, and then as I heated the chamber up past 50C.

I then repeated this process across a wide variety of measurement targets, normalized the data relative to the same point in the ramp, and produced that graph you see in the docs.

I will admit that the thought of renting an environmental chamber did occur to me, and that would have been necessary if I wanted to characterize consistent or steady state temperatures. However, the improved method I used turned out to be good enough to collect the data I wanted.

 

[Michel Hardy-Vallée]

This paper seems to have a problem. Hardly a slope difference between 2 and 3, contrasted with big slope difference between 3 and 4. And Dmax in the range 1.5-1.6. Old paper? Matte surface?

It's a quick and dirty graph, so no pretense at exactitude intended.

 

[Zathras]

I just ordered one of these yesterday and it showed up today. Turns out that Dektronics is not very far from where I live.
Now I can start doing those BTZS film and paper tests I've been wanting to do! It looks like a well made unit that will be
very useful for my needs.

 

[ProfessorC1983]

Hey @dkonigs I just wanted to let you know I'm excited to purchase one of these units, however it seems like your website is not loading for me, either the link posted above or the one on your Twitter profile. Hope this can get corrected soon as I'm sure others are looking forward to it!

Um, nevermind! Seems this was either a temporary glitch in the 'tubes or user error. Order #34 is in and can't wait to play with it!