Post Go back to editing

ADPD4001  | ADPD1081 | ADPD188GG | Best Configuration for PPG (2 Green vs 3 Green, Red LED, and IR LED)

We are working on a real-time wearable PPG solution. We understand that either ADPD4001 or ADPD1081 can be used.

Q1) I am wondering if you could comment on the accuracy and power consumption of those two ICs. Which one is recommended by you for a wearable solution?

Q2) The evaluation board of ADPD1081 only uses Green LEDs, while ADPD4001 has 3 green LEDs, 1 red and 1 IR LED! What is the main reason behind this? ADPD1081  cannot support 1 red and 1 IR LED? Which one is better in terms of accuracy? Do we need Green, Red, and IR at the same time?

Q3) To be able to work with the evaluation boards, it seems that we also need an EVAL-ADPDUCZ. I am wondering if there is any solution to directly connect ADPD4001/ADPD1081 evaluation boards directly to PC without an extra EVAL-ADPDUCZ or we need to purchase one EVAL-ADPDUCZ?

Q4) I studied the user guide of "ADPD1081". On page 7, I can see the schematic of the evaluation board. But something is wrong with that. In the schematic, the Analog used "ADPD107-WLCSP"! which is an old chip, and according to the website, ADPD107 is Not Recommended for New Designs... I also noticed that the Green LEDs from OSRAM are not also recommended for new designs anymore. Indeed, they are not even available in the market anymore. Could someone please comment on that? Is ADPD1081 an old solution/IC? Should we use ADPD4001 in our solution?

Q5) The evaluation board of ADPD4001 and ADPD1081 are not available in the market. It seems that you are going to release them on April 30. Is there any reason behind that? 

Q6) Analog also provides a module entitled ADPD188GG. This module already has 2 green LEDs, 2 PDs with IR cut filter. What is its main difference between this module and ADPD4001/ADPD1081 ICs. 

Q7) What is the best (accuracy vs power consumption) for PPG monitoring? How many LEDs? 2 Green, or 3 Green? why three not 2? How many RED and how many IR (do we need RED and IR at the same time or just IR is fine)? and how many PDs? ADPD188GG has 2 PDs rather than just 1 PD? In summary, what is the best configuration?

[edited by: ffirouzi at 1:55 PM (GMT 0) on 3 Feb 2020]
  • Hello,

    Thanks for contacting us. 

    1) We currently recommend ADPD4000 for the wearable solutions, as it has less power consumption and noise. Could you explain what you mean by accuracy? It can mean differently in different context. If that means better SNR-so better quality of the output, ADPD4000 is better. If that means linearity of the system, they are very similar. 

    2) Green, red and IR would give info about different physiological properties. Red and IR are mainly used for SpO2 measurements. Green gives a high quality PPG signal, but you may not be able to extract some physical properties with that wavelength. ADPD1080 can also support red and IR. In fact, they all can support any light if the photodiode you choose is responsive to the LED you use. It's independent of the IC. Accuracy is probably not a too valid question in this case. Green, due to physiology, gives a better PPG signal but you cannot extract some info. The need for green, red, IR and even more colors depends on what you want to extract out of the output signal. 

    3) If you would like to evaluate the eval boards, you would need EVAL-ADPDUCZ. If you want to use your own microcontroller to run ADPD4000/1080, you can do it. The information on SPI and I2C is in the datasheets. EVAL-ADPDUCZ is for easy evaluation. 

    4) ADPD1080 eval board is older. That's why some components may be/are obsolete. APDPD4000 is a newer IC and we recommend ADPD4000 for the designs at this point. 

    5) I would recommend you to contact one of our sales offices. They could be able to help you.

    6) ADPD188GG is a complete system level design for PPG, whereas ADPD4000 and ADPD1080 are only ICs. They have their own eval boards, but they are not full modules. If you would like to use your own LEDs, PDs, etc., we recommend you to use ADPD4000. If you would like to use ADI solution for everything, ADPD188GG makes the designs simpler. 

    7) ADPD4000 is the best in terms of SNR and power performance. The signal quality you would get depends on the optical and mechanical design too. Even if you have a great IC, if you do not have the right optical and mechanical design, the performance would get affected. Number of LEDs for each color depends on what you would like to achieve and where you would like to use. For example, if you have low CTR(current transfer ratio), you may need a few of the same LED to increase the signal level. If you need to measure different physiological properties, you may need red, IR, blue, etc along with typical green LED. If you use more selective, narrower band PD, you may need to use multiple PDs to capture different colors of light. In summary, the best configuration highly depends on the system level requirements and there is no single answer that applies to all cases

  • Hello,

    Fantastic. I appreciate your detailed answer. It is really helpful. Thank you. Only two more questions:

    1- Is ADPD188GG module based on ADPD4000 or 1080 or something else? For how long will you support this module? 

    2- I understand the main difference between the Green LED and red. Could you please also comment on the major difference between RED vs IR? 

    Thank you

  • Happy to be able to help you! 

    1) It's based on 1080. There is no certain deadline to stop the support for the module, so we're going to be supporting it.

    2) A typical algorithm to calculate blood oxygenation level(% SpO2) is ratio of ratios, 

    % SpO2 = (RMS_AC of Red / DC of Red) / (RMS_AC of IR / DC of IR)

    Physically, since red and IR has different wavelengths, oxygenated and non-oxygenated hemoglobin have different sensitivities(absorption and reflection) to red and IR. So, you can find oxygenation level by using both.

    Plus, since red and IR have different penetration levels in skin, the info you get out of PPG signal itself may slightly differ.