Re: [PLOTS] Re: Want to build a "lab" spectrometer with improved camera, slits, alignment methods, etc.

81 views
Skip to first unread message

Jeffrey Warren

unread,
Feb 5, 2016, 9:22:28 AM2/5/16
to plots-spe...@googlegroups.com
Hi, Jacques - I'm going to bump this over to the spectrometry list, and I do think that a number of things you're asking about have been explored: 



Hope this helps as a starting point! Both @cfastie and @stoft are on this list :-)

Jeff


On Fri, Feb 5, 2016 at 6:37 AM, Jacques Blanchart <jacques....@gmail.com> wrote:
Forgot to add that I'm an "hasbeen" (70 y old) and a "newbee" in optics and spectrometry.
I'm interested by reflectance spectrometry (measurement of pigment influence in soaps and cosmeteics).


Le vendredi 5 février 2016 12:33:49 UTC+1, Jacques Blanchart a écrit :
I purchased the v3 spectrometer kit in order to refresh my knowledge in optics and spectrometry.
My goal is to build a "lab" spectrometer with enhanced capabilities : sensitivity, spectrum range, etc.
"Lab" spectrometer means that it does not need to be portable (anyhow, less than 1 cubic meter !!! ;-)
I'm looking for help in :
- spectrometer design calculations : distance from camera to slit, grating choice...
- adjustment possibilities : ajustment of slit position, camera, focus...
- camera choie
and so on.

Anyone interested by thos project ?
Or does this already exist ?

ps : please apologize for my french-english dialect...

--
Post to this group at publicla...@googlegroups.com
 
Public Lab mailing lists (http://publiclab.org/lists) are great for discussion, but to get attribution, open source your work, and make it easy for others to find and cite your contributions, please publish your work at http://publiclab.org
---
You received this message because you are subscribed to the Google Groups "The Public Laboratory for Open Technology and Science" group.
To unsubscribe from this group and stop receiving emails from it, send an email to publiclaborato...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

Jacques Blanchart

unread,
Feb 5, 2016, 12:30:49 PM2/5/16
to plots-spectrometry, je...@publiclab.org
Many thanks, Jeff
I already tried to find out some information but I missed 3 of the 5 links.
Thanks again
Jacques

Chris Fastie

unread,
Feb 7, 2016, 11:11:22 AM2/7/16
to The Public Laboratory for Open Technology and Science, plots-spe...@googlegroups.com
Moving over to plots-spectrometry...

Wow, Those CCD linear arrays are now under US$7.00 delivered from China. An Arduino Nano clone is under $4.00. The quality of the spectral data from such a system could be an order of magnitude greater than a camera based spectrometer. I wonder if a lens is required to collimate the light after it has been spread by the grating and before it hits the CCD. In other words, how perpendicular to the CCD do the incoming rays have to be?

It would be great to see links to the best software, drivers, and libraries to make this work.

Chris

On Sunday, February 7, 2016 at 5:37:15 AM UTC-5, Dennis Meulensteen wrote:
I just ordered a CCD and an el-cheapo arduino clone to do just that. Basic driver software is already available on the 'net.

I am using a TCD1201D CCD device couples with an Arduino nano 3.0 unit. It communicates over a USB Serial connection.
The device has 2048 photosites and can function slow enough for the Arduino to keep up with it.

I will start off with a DVD grating and see where that gets me.

I can document what I am doing as I go along but can't submit to design by commission. That said, I am always open to suggestions!

My ulterior motive is to calibrate my monitor and prints. Don't know if that is too much of a stretch target but if it is it will be a fantastic learning experience.

Dennis

Jeffrey Warren

unread,
Feb 8, 2016, 9:55:19 AM2/8/16
to plots-spe...@googlegroups.com
On the software front, we're starting to use an Arduino-compatible Trinket as a device to control the light on a spectrometer with a dial. I believe the Trinket (and other arduinos) can act as a USB keyboard device, and "type" --

One way I'm thinking of to transfer the spectrum would be to have it type out the values it detects, as a comma-delimited string, into a web form on Spectral Workbench. This sounds a little primitive, but the advantages are that:

a) we have already started to integrate a compatible device into a spectrometer setup
b) no drivers, any platform
c) outputting CSV format would make it easy to work with any program
d) possible advantages in having the device connected to the input light be the same as connected to the sensor

I'd be very very happy to build a CSV input form for this purpose into Spectral Workbench if anyone builds one of these!

Jeff


William Macfarlane

unread,
Feb 8, 2016, 10:42:29 AM2/8/16
to publicla...@googlegroups.com, plots-spe...@googlegroups.com
Jeff, if I recall correctly you were using the 3V version of the Trinket -- is there a reason why it needs to be the 3V and not the 5V?  I ask because this tutorial/code here -- https://learn.adafruit.com/trinket-usb-keyboard/overview -- suggests that one of the more straightforward ways of having it be a USB keyboard might not work with the 3V version.

I haven't actually investigated why this may or may not be the case, or what might be done to get around it, but it might make sense just to use the 5V version?

Chris Fastie

unread,
Feb 8, 2016, 10:56:54 AM2/8/16
to plots-spectrometry
Public Lab author bhickman had some success making a spectrometer from a linear CCD array: https://publiclab.org/notes/bhickman/08-27-2014/3d-printed-ccd-spectrometer-wheetrometer-3-0.  He included a lens and a curved mirror in the optical path. These focusing and collimating elements are probably required but can add distortion unless you use good elements and know how to use them. He used a LaunchPad instead of Arduino.

Jeffrey Warren

unread,
Feb 8, 2016, 10:57:31 AM2/8/16
to publicla...@googlegroups.com, plots-spe...@googlegroups.com
The super-bright ultraviolet LED we're using is a 3.3v LED, and I just thought that getting the 3.3v trinket was easier than doing math. 

But I seem to recall you know how to turn 5v into 3.3 volts? 

William Macfarlane

unread,
Feb 8, 2016, 11:07:52 AM2/8/16
to publicla...@googlegroups.com, plots-spe...@googlegroups.com
How much current does the LED want?

Jeffrey Warren

unread,
Feb 8, 2016, 11:09:56 AM2/8/16
to publicla...@googlegroups.com, plots-spe...@googlegroups.com
I posted links here: https://publiclab.org/notes/warren/01-05-2016/testing-high-brightness-405nm-leds-in-fluorescence-spectroscopy-of-oils


wattage:
3w
max drive current:
700mA
forward voltage (@ 700 ma):
Min 3.0v Max 3.6v
light intensity:
640mW - 680mW
wave length:
405nm
viewing angle:
120°
color:
Hyper Violet
star size:
20x30mm Solderless Board

Jeffrey Warren

unread,
Feb 8, 2016, 11:10:53 AM2/8/16
to publicla...@googlegroups.com, plots-spe...@googlegroups.com

William Macfarlane

unread,
Feb 8, 2016, 11:14:59 AM2/8/16
to publicla...@googlegroups.com, plots-spe...@googlegroups.com
By my math you want like 2.5 ohms 

(5V-3.3V)/.7A = 2.43 ohms.

My guess is that if you're pulsing the LED anyway to dim it, you can get away with not having a resistor since the resistance value is so low and not having it on all the time will help it dissipate heat, but that's "definitely-not-an-engineer"-me talking.

Jeffrey Warren

unread,
Feb 8, 2016, 11:24:22 AM2/8/16
to plots-spe...@googlegroups.com, publicla...@googlegroups.com

Hi Peter, did you see my earlier post with links a few minutes ago?

Thanks, will! I burned one LED out almost instantly connecting it directly to 5v (empiricism ftw), so I'm inclined to actually be careful on overloading it in this case.

I think we can test this out and if it works, note that either trinket cariant works but that the 5v requires the extra resistor.

Thanks, all!

Post to this group at plots-sp...@googlegroups.com

 
Public Lab mailing lists (http://publiclab.org/lists) are great for discussion, but to get attribution, open source your work, and make it easy for others to find and cite your contributions, please publish your work at http://publiclab.org
---
You received this message because you are subscribed to the Google Groups "plots-spectrometry" group.
To unsubscribe from this group and stop receiving emails from it, send an email to plots-spectrome...@googlegroups.com.

William Macfarlane

unread,
Feb 8, 2016, 11:27:59 AM2/8/16
to publicla...@googlegroups.com, plots-spe...@googlegroups.com
If we're being sensible, there's probably no need to drive the LED at its _max_ current, so a 3.9ohm resistor might be a good value that is commonly produced and is what I'd recommend trying next.

Kevino

unread,
Feb 8, 2016, 1:04:10 PM2/8/16
to plots-spectrometry, publicla...@googlegroups.com
Wanted to add a few things here:
1. The trinket tutorial suggests using 5V trinket, not because of the voltage, but the clock rate. You can't reliably do the USB 1.1 HID timing at the recommended 8-12Mhz clock rate for 3.3v, you need 16Mhz.
That said, the atmega usually runs fine at 3.3v and 16Mhz (I haven't checked the design but you might have to change the crystal) Running an atmega 328 (trinket pro) at 16Mhz and 3.3V is technically "overclocking"
2. The arduino nanos, are based on 32U4 which has excellent support for HID emulation
3. I found this for driving the CCD: http://mauromombelli.com/2015/02/20/TCD1201D.html. I looked at some other things that suggest to get optimal performance you need something faster like a CPLD, but depending on what you are trying to do, an arduino might be ok. Once again, though you need to go as fast as you can.

Note there is a breed of new arduinos based on an ARM Core M0 (see adafruit feather, arduino zero) that has a ton of memory and higher clockrate.
To unsubscribe from this group and stop receiving emails from it, send an email to plots-spectrometry+unsub...@googlegroups.com.

For more options, visit https://groups.google.com/d/optout.

--
Post to this group at publicla...@googlegroups.com
 
Public Lab mailing lists (http://publiclab.org/lists) are great for discussion, but to get attribution, open source your work, and make it easy for others to find and cite your contributions, please publish your work at http://publiclab.org
---
You received this message because you are subscribed to the Google Groups "The Public Laboratory for Open Technology and Science" group.
To unsubscribe from this group and stop receiving emails from it, send an email to publiclaborato...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

Dave Stoft

unread,
Feb 8, 2016, 1:31:36 PM2/8/16
to plots-spectrometry, publicla...@googlegroups.com
The listed voltage range on the LED is the typical forward voltage drop the diode will have dependent on the amount of current that is made available; it is not a voltage that you "set". The max forward current appears to be 700mA. Therefore, yes, if you supply a 5V source directly, the current will immediately rise to the maximum available. USB is usually limited to a max of 500mA (although that is really pushing the limits of the typical USB source so is actually being rather hard on that electronics. The other issue is that running the LED at high DC currents can easily exceed the thermal properties of the device -- unless the mounting has sufficient cooling to dissipate the excess heat.

Controlling the current is made more difficult by having a very low voltage drop from source to diode; i.e. 5V source and current sufficient to create a 3+V drop. This is why a higher voltage source is used such that either a voltage dropping resistor (simplest) can be used (think the typical low-power red LED running a <10mA). Under some conditions, one or more series diodes (DC regulator type, not LED) can drop the source voltage by 0.7V per diode -- but the method is only helpful when regulation or precise voltage is not critical.

Yes, it is common at high powers to pulse LEDs so as to achieve high peak power levels but still protect the LED. This generally means synchronization between the LED drive and the detector. Otherwise, it becomes more like pulse-width modulation and again issues of thermal heating, source current, control of max LED current arise.

A quick look at the spec sheet of the root chip of the "Trinket" series appears to accommodate a range of supply voltages from 3-5. However, that is only the uC, and does not include the other parts on their board. They do not have sufficient specs that I could see which provide accurate supply details for their module. Since they appear to make two versions, a 3.3V and a 5V, it would be prudent to assume that some of the hardware on board does have a 3.3V limit for that version -- which means it could be over-stressing the 3.3V module to attempt to run it on 5V from a USB port supply.

If you need to run a high power LED, it would probably be best to plan on a separate "wall wart" power supply just for the LED (at some higher voltage and then add a regulator, current limit, etc for protection) but you could still run the module off USB -- just taking some added steps to protect the separate supply from the USB power/signals (ground loops, etc.). This would free you from the restrictions of USB when designing for the LED.

All these specs and limitations are there for a reason.

Dave Stoft

unread,
Feb 8, 2016, 3:32:39 PM2/8/16
to plots-spectrometry, publicla...@googlegroups.com, je...@publiclab.org

Be aware that these LEDs, like most LEDs (vs the 405nm Laser) have wide bandwidths. Since the narrow-band 405nm laser produces a broad spectrum of fluorescence energy, using a broad spectrum source might? change that fluorescence spectrum which could be more difficult for extracting meaningful information. This is just a conjecture, but it is probably worth at least a few well-designed comparison tests to find out.

Jeffrey Warren

unread,
Feb 9, 2016, 9:04:13 PM2/9/16
to publicla...@googlegroups.com, plots-spe...@googlegroups.com
Well, I've been running this LED directly on this Arduino Trinket for a few days, plugged into an analog pin and the ground, but it seems like (makes sense) it can't power it to full brightness. Great to know about the MOSFET; which of these? http://www.digikey.com/product-search/en?keywords=IRFZ44N

I'm not myself working on the HID thing, just the LED dimming. 

Thanks!
Jeff

On Tue, Feb 9, 2016 at 4:38 AM, Dennis Meulensteen <den...@meulensteen.nl> wrote:
I took a look at the LED schematics but those will kill the Digital port on your Arduino; guaranteed! Its only rated for something like 40mA peak and just 20mA typical! Great for a signal LED in a control panel, not so much for lighting anything.

If you really want to keep the component count low I would use a MOSFET to switch the LED. It doesn't need a base resistor like a normal transistor would because it is voltage driven, not current driven. So for those of you who hate Ohm's law, that's got to be a plus. 

A cheap one that packs a lot of punch for LED projects is an IRFZ44N. I have one PWMming my aquarium LEDS as I type this; that's a lot of LEDs. It doesn't even get warm to the touch.

I even found a beautiful (yes; really) description of this component on the net. 

If you want to combine the LED project with the CCD project you will need to go to an Arduino Micro board I think. It has excellent USB-HID (keyboard and so on) support. It has 2.5K RAM and runs at 16 MHz, small form factor like the trinket pro.

Should be a breeze to implement the output as typing. I don't have the CCD at home yet but I could mock something up for testing to see if RAM usage will allow a fully buffered frame on that device with all the required drivers loaded....

Combining the LED with the CCD is trivial as we can just switch it during CCD integration ("exposure") and use a timer to switch off (both) as needed.

The site that was posted convinces me that this can be done without resorting to overkill hardware and the complexities of OSes and such. 

Op vrijdag 5 februari 2016 12:33:49 UTC+1 schreef Jacques Blanchart:
I purchased the v3 spectrometer kit in order to refresh my knowledge in optics and spectrometry.
My goal is to build a "lab" spectrometer with enhanced capabilities : sensitivity, spectrum range, etc.
"Lab" spectrometer means that it does not need to be portable (anyhow, less than 1 cubic meter !!! ;-)
I'm looking for help in :
- spectrometer design calculations : distance from camera to slit, grating choice...
- adjustment possibilities : ajustment of slit position, camera, focus...
- camera choie
and so on.

Anyone interested by thos project ?
Or does this already exist ?

ps : please apologize for my french-english dialect...

--

Dave Stoft

unread,
Feb 9, 2016, 9:48:35 PM2/9/16
to plots-spectrometry, publicla...@googlegroups.com, je...@publiclab.org
Right, some internal current limiting by the Arduino has probably prevented damage but you'll only get ~40 mils. Yes, the Z44 hexfet is a nice method as they provide some internal compensating/protection at the gate. There are 4 package types on Digikey which show minimum buy quantities of 1. Any one of those should be fine -- I prefer the leaded devices for prototype work. At 700mA and 0.1V drain-source drop, it's only about 50mW so the load will be easy on the part. A TO-220 pkg would just warm a little under continuous duty.

You'll need a 100K (no real current) "pull down" from gate to ground to "bleed off" charge when the gate is off. You could also add a 1K in series with the gate if you like but the current through the 100k will be very low. You will still need to add a series R for the diode to set the max current. The nice thing is that the fet's drain-source voltage drop is down in the milliohms and the device curves suggest it will remain below 0.2V under most conditions up to 1A.

This leaves the voltage drop of the diode to consider. Assume it is 3.6V at 700mA, then you'd probably have ~3.5V UV diode drop for the max current available from USB -- or 1.5V drop to handle with a 5V supply (1.5V at 250mA is 6 ohms with 375mW dissipation). So you might want a 5-10 ohm 1W resistor to limit the current -- unless you use an external supply where values might be different with a different voltage.

PWM would keep the average current low but the peak would remain a consideration. Remember, pulsing a few hundred mils of current out at a module on the end of a USB line with USB 5V power is going to generate substantial transient spikes -- on the supply line and on the data lines. You'll need a significant cap (47uF+) at the module/LED to dissipate that energy. (Two cap types: electrolytic for the DC buffering of the high current spike demand from the diode/fet and a ceramic type (smaller) (0.1uF) to shunt the RF energy transients from the switching. Probably also add a back-biased diode (1N4148 etc) across the diode/fet to help clip the inverse transients from the capacitors.

So a few additional parts, but not too bad.
Reply all
Reply to author
Forward
0 new messages