I'm designing a device for blood glucose measurement. The wavelength of the IR led is chosen as 1550nm as it is only at this frequency, the glucose molecules are absorbed.
i've attached the circuit below.
this circuit gives the output only for 950nm but i require a circuit that can detect a wavelength of 1550nm.
kindly suggest the changes to be made in the circuit for proper output.

 

I'm thinking all you need to change is the LED transmitter/emitter pair. If those require different circuitry, you may need to adjust this.

 

I don't see anything in the circuit that differentiates between 950nm and 1550nm.
Change the IR emitter and detector to match the desire wavelength.

For your application, I would use two identical circuits, one at the desired wavelength paired with another at a wavelength just outside of your desired wavelength. Take the difference between the two readings. This provides a reference or control to cancel out all other effects (e.g. temperature) not related to the parameter of interest.

 

Is the 39K @ pin 18 of PIC a little large ? Base drive of around 65 uA; might be better @ 3 mA ?. R of 1K ?

 

And by the way I would use a dedicated pair that employ internal filtering of the carrier, which you might want to be higher than just 1kHz. There's no reason to reinvent the wheel and make your own filter. Very cheap components can do an excellent job of this for you.

 

Is the 39K @ pin 18 of PIC a little large ? Base drive of around 65 uA; might be better @ 3 mA ?. R of 1K ?

Thanks Bernard..
I changed the resistance to 1K..
but the problem is that the output comes properly with just the transmitter and detector and no significant output voltage occurs when a finger is placed between them.

 

Thanks Bernard..
I changed the resistance to 1K..
but the problem is that the output comes properly with just the transmitter and detector and no significant output voltage occurs when a finger is placed between them.

It could be that the amplification is too high and the output is saturating in both cases.

 

I have the same problem
have you solved the problem if there is a finger between transmitter and detector?

 

Hello,

As said in the previous posts, the IR diode and detector should match to the same wavelenght.
Also the sensitivity comes to play.
Perhaps saturation is reached with the 1K resistor.
Making the sensitivity adjustable by putting a potentiometer of 25K in series with the 1K resistor in the base of 2N2222 would help.

Bertus

 

Hello,

As said in the previous posts, the IR diode and detector should match to the same wavelenght.
Also the sensitivity comes to play.
Perhaps saturation is reached with the 1K resistor.
Making the sensitivity adjustable by putting a potentiometer of 25K in series with the 1K resistor in the base of 2N2222 would help.

Bertus

I thinks he/she can detect signal from transmitter directly but still has a problem if there is a finger between transmitter and detector. it is from his/her last post

 

Hello,

I have located the original circuit.
It can be found in the attached PDF.

Bertus

 

Hello,

As said in the previous posts, the IR diode and detector should match to the same wavelenght.

While there are other problems as you point out, and this is an old thread with very few details, e.g., part numbers for the IR components, it is unclear whether you (and others) intend "match to the same wavelength" to mean both source and detector need to be "monochromatic" or effectively so (e.g., filtered). Some very successful research spectrophotometers have only one monochromator. That is, if the incident light is monochromatic, the detector can be broad band or visa versa. The latter being quite common.

In the present instance, it might be more practical to have a monochromatic source and broad spectrum detector. The only requirement is that the wavelength of the incident light be within the detector's range of detection.

Edit: @bertus, your post appeared while I was writing.

 

Hello,

I have located the original circuit.
It can be found in the attached PDF.

Bertus

Thank you for your reply. I have read the PDF.
but I want to ask something. what photodiode type(part) you used to detect 1550nm?
In my experiment, I use FGA015 from thorlabs and the photodiode just have 0.15um active area. and when the finger placed between transmitter and detector it can't read the signal.

 

Hello,

It is not my project.
What kind of IR led did you use?
The detector is able to detect the 1550 nm signal.

Bertus

 

Hello,

It is not my project.
What kind of IR led did you use?
The detector is able to detect the 1550 nm signal.

Bertus

ye the photodiode can detect 1550nm signal
i have proved it by transmit the signal directly to the detector.

 

Hello,

Are you using the exact schematic as in the openings post?
Then you must also have the program to use it.
The IR diode is pulsed and the detector signal is filtered.

Bertus

 

1) You may choose wide-spectra sensor and emitter where apply the narrow-band optical phase-filter between.
2)You may choose the narrow band emitter like specific MIR LEDs, Hamamatsu and some other have em but not cheap
3)The same but cotrary, wide band emitter and narrow band receiver. Althese must be analysed from signal to noise standpoint
4) I would hardly influence to use a "synchronous detector" aka PLL circuit for data capturing. You may add a transimpedance OP before it, but anyway the keywords ought be the 4046 or 567 depending on case. For example, in early 80ies was such russian 140UD13 PLL what I had realized 1 km data link at ordinary LED in bright sunlight. And You have the very similar case where room light brings in the significant dirsturbance.

 

Do all transistors and Op amps work in high frequency? Cause 1550nm wavelength is equal to about 193thz and if infra red diode and photodetector work well at this frequency, other elements must be work well at this frequency.

 

Are you measuring pulse rate, hemoglobin, oxygen saturation, glycated hemoglobin, or glucose? (To me, plethysmography = pressure/volume. )
Can you provide a link for the method you propose?

I am familiar with other non-invasive glucose monitoring methods (e.g., https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5772583/ ) but not the one you describe. This article describes using IR in that range (1400 nm to 15xx nm), but sensitivity is quite low: https://iopscience.iop.org/article/10.7567/JJAP.55.106601/pdf

 

Exists a theoretical fiction working at THz, it is even approved in lab environment (nano-scale diamond semiconductors), costs millions and stay in few exemplar over whole world. Let forget it and for long.

No-one and never are using phase modulation of light-beam by impact of carrier frequency or phase. Instead are worked on all possible modulation kinds, yet at the frequency where exist a plenty of normal transistors.

However Your task was to detect, is there light or not, and how bright, wasn’t so?? Then there is no any need to have a fast technique at all. Any transistor will give You some Megahertzes, and specialized IC like HEF15xx/25xx series will give a sure 10 nsec or even shorter response time.

Do You are sure is worth to pay maybe thousand-fold more for to have a few-fold better response time?? Yes, exists such techniques, astronomers measuring distances of 65 000 km to the satellites (named SLR tech) call it miraculous "Riga Event Timer" giving an accuracy of sub-femtosec thus that distance is measured by 1...3 milimeters (!!) accuracy. But You will not want even to hear the price, it is mind-blowing (may check the www.eventech.lv).