Talk:Photodiode

History section
Needs a History section. See "Active pixel sensor" as a typical example.

Kenyoni (talk • contribs) 11:00, 26 March 2017 (UTC)

"Photoconductive mode" vs. "photodiode mode"
Hello, I'm totally new.

The text references the reverse bias mode as "Photoconductive mode", while the picture on the right calls this "photodiode mode". I think either the text should mention the alternative name, so the picture is easier to understand, or the alternative name should be removed from the picture. I could do either, but don't dare.

Dneise (talk) 06:57, 21 November 2016 (UTC)

"Other modes of operation"
More and more photodiodes these days are used in avalanche (APD) and even geiger mode (SPAD and SiPM).

I will come back and write about this in the main page.

I propose to list these modes explicitly, not just as "other modes".

In the photodetector page, plan to write about SiPM which is an ensemble of photodiodes.

Construction, principle of operation, and saturation
Hi there. Can someone please put some stuff up for the construction and how the photo diodes work please?

What are some possible reasons photodiodes do not reach saturation?

Semiconductor physics needed
I think that this article needs to mention semiconductor physics or link to a discussion therein. Therobotbuilder 19:57, 12 November 2006 (UTC)

Source of energy
I have some conservation of energy angst. The diode's built-in field is what accelerates the electrons to create the drift velocity (hence current) in the diode. So the energy that is generated from a photovoltaic cell would seem to come from the p-n junction (and its' charge carrier gradient), not the light source which generated the electron-hole pair in the first place. I know the light must be the source of energy behind the current that is generated, but I can't quite see how. can anyone explain? Andybuckle 16:48, 15 January 2007 (UTC)
 * Perhaps I can explain, belatedly. When a photon is absorbed in a semiconductor, it donates ~1eV to separate an electron and hole by perhaps a few nanometers. It only takes a fraction of a percent more, donated by thermal motion, to finish separating them farther than the Debye Length, beyond which local free carriers shield the pair from each other, and they wander around idly, influenced by much smaller energy gradients or thermal diffusion. Much like a hydrogen atom which has been ionized in a plasma. Whichever carrier is in the minority eventually diffuses to the depletion layer and is rolls "downhill" to the other side, but that downhill run is draggy, not ballistic, so it's lost energy, roughly 15% of the pair energy in an optimally biased solar cell, 100% in a zero-biased photodiode.  It's maybe ~1000% in a reverse-biased photodiode, but in that case the energy above 100% comes from the external power supply.  jimswen (talk) 08:00, 10 February 2016 (UTC)

Injected into the base
"electrons that are generated by photons in the base-collector junction _are injected into the base_" in the description of phototransistor work in the first para. Is that true? Aren't the electrons swept into the collector, increasing its current and the holes gather in the base region increasing its voltage? --Janislaw 10:35, 16 January 2007 (UTC)

Picture of photodiode
Can someone confirm that the pictures is actually a photodiode. It looks like an LED. 194.200.145.5 13:24, 2 February 2007 (UTC)


 * An LED can serve as a photodiode. In fact, every diode is photosensitive. --Phil Roan, 71.116.81.130 (talk) 00:15, 24 December 2007 (UTC)

A number of different devices come in certain common packages. The clear bullet-shape, called T-1&3/4 or 5mm, can contain an LED chip, a photodiode, a phototransistor, or a few other things. The main thing in common is the device uses light. jimswen (talk) 08:18, 10 February 2016 (UTC)

Reverse Bias
The article states that reverse-biasing "strengthens the photocurrent". Does this refer to some minor effect, or is it just false? If you're getting only one electron per photon, the current will be set by the photon rate and be independent of applied voltage. Also, I'm not familiar with the effects of "expanding the reaction volume". Is that going to create better efficiency? I thought the only reason to reverse-bias a photodiode was to reduce the capacitance, thereby increasing the response speed. Spiel496 15:54, 24 May 2007 (UTC)


 * Spiel is correct. Reverse bias does not strengthen the photocurrent (For a visual see the graph at ). In fact, the article has it backwards. Reverse bias causes photocurrent, decreasing the sensitivity of the device. However, it significantly lowers the capacitance, which causes improvements in the response time because the time constant is a product of the load resistance with the junction capacitance.

--Phil Roan, 71.116.81.130 (talk) 00:15, 24 December 2007 (UTC)

Like a solar battery
Picture of square device looks exactly like a solar battery I have from an LCD power supply. Much different in that it creates electron flow versus controls electron flow. Now about that gain in a phototransistor.

Graphs or schematics
It might help to add a graph with I-V curves under different levels of illumination or a schematic for an equivalent circuit, e.g. current source (dependent on light level) in parallel with a regular diode and possibly some resistors or capacitors Nrnkpeukdzr (talk) 04:20, 13 January 2008 (UTC)

A graph similar to Figure 1 in the Encyclopedia of Laser Physics article on Photodiodes would be helpful. http://www.rp-photonics.com/photodiodes.html — Preceding unsigned comment added by 129.215.139.65 (talk) 18:29, 21 June 2011 (UTC)

Photocurrent polarity
The photocurrent flows from cathode to anode, against the diode's arrow. This can be confirmed by the graphs at : the photocurrent is negative but the current from "turning the diode on" is positive. The polarity paragraph is wrong, the first paragraph of Principle of Operation is correct. 128.146.36.20 (talk) 02:21, 2 October 2008 (UTC)
 * I think it used to say that photo-electrons flowed with the diode's arrow -- that would have been correct. It looks right now, though. Spiel496 (talk) 17:29, 2 October 2008 (UTC)

Phototransistor: Faster or slower?
Under "Other_modes_of_operation" it says "Phototransistors also have slower response times." but the article on "Photoresistor" says "Phototransistors react much quicker to light change. As such, they're preferred for receiving data." and links to this section. Which is it? —Preceding unsigned comment added by 72.37.171.52 (talk) 14:16, 21 November 2008 (UTC)


 * A Photodiode is faster than a Phototransistor, at least if it is the sensor that sets the speed. The Phototranistor is often faster than a typical Photoresitor. However there are also some special very fast Photoresistors, even comparable to fast photodiodes. --Ulrich67 (talk) 22:18, 12 January 2011 (UTC)

Incorrect redirection
Obviously photo transistor have nothing to do with photo diodes, yet somebody seemed to have mislinked the two together.

(To the post above) As for their speed, a photo diode should be faster than a photo transistor, which in turn is faster than a photo resistor. Both the photo diode and the photo transistor can be used for receiving data. Interarticle (talk) 12:58, 19 September 2009 (UTC)

How the devices works
please explain how the photo diode works —Preceding unsigned comment added by 117.254.157.237 (talk) 12:20, 30 September 2009 (UTC)

"at least one of these sentences is wrong"
Someone put a superscript saying that at least one of these is wrong, but made no mention of what was wrong on this talk page (that I can find):


 * 1. Due to the intrinsic layer, a PIN photodiode must be reverse biased (Vr). The Vr increases the depletion region allowing a larger volume for electron-hole pair production, and reduces the capacitance thereby increasing the bandwidth.
 * 2. The Vr also introduces noise current, which reduces the S/N ratio. Therefore, a reverse bias is recommended for higher bandwidth applications and/or applications where a wide dynamic range is required.
 * 3. A PN photodiode is more suitable for lower light applications because it allows for unbiased operation. (*at least one of these sentences is wrong!)

I suspect that they have misread number three to be referring to the PIN photodiode rather than a PN. I can't think of a way to rewrite the sentence to rely less on the PN vs PIN wording, but the content (when properly read) seems sound to me.

Because I can't see a problem with any of the sentences, I have removed the superscript.

118.208.163.172 (talk) 07:01, 30 April 2010 (UTC)


 * There is a slight Problem with Sentence no 1. It's not really wrong, but misleading. It's true that the depletion region gets larger in a PIN diode, but the increase in sensitivity is way smaller than the increase in volume. In a PN-diode most electron hole pairs created close the junction also contribute to the current. The sensitive range is about one diffusion length (e.g. 200 µm) beyond the junction. It's only close to the long wavelength limit (e.g. more than about 950 nm in silicon) that the electron - hole pairs are created deep inside the material so that the sensitivity goes down. Here a PIN diode can have a slight advantage. --Ulrich67 (talk) 22:48, 12 January 2011 (UTC)


 * The "at least one" was a bit of an understatement, since all three are deeply flawed. In 2, the Vr does not introduce noise current unless the junction is leaky.  In 3, I have no idea what's behind it. Dicklyon (talk) 02:49, 13 January 2011 (UTC)


 * That section came from this hit-and-run editor: Special:Contributions/24.20.100.227. I think I'll just take it out.  Dicklyon (talk) 04:06, 13 January 2011 (UTC)

Comparison with photomultipliers
Somebody flagged this comparison section as needing references and threatens to remove the section. I strongly urge that it be kept. It is very difficult to provide references on twenty different statements or more. Readers can certainly check the validity of any of the statements if needed. I really like sections like this in Wikipedia articles and would urge critics to lighten up a little on them. The authors who post them should be commended for clarifying the differences between similar topics that may otherwise be confusing to people. I say keep it. 3dimen (talk) 13:53, 7 August 2011 (UTC)

Photovoltaic mode vs. photodiode mode
Apparently in both vases the cathode shld be positive. That is wrong as these are different modes. In Voltaic mode it should be in conductivity mode with the cathode on the minuw — Preceding unsigned comment added by 88.159.150.221 (talk) 16:01, 29 September 2011 (UTC)

Base unconnected instead of emitter?
Regarding (subsection Other modes of operation):


 *  If the base and collector leads are used and the emitter is left unconnected, the phototransistor becomes a photodiode.

This is inconsistent with the symbol of the phototransistor. Isn't it the base that is left unconnected? --Mortense (talk) 12:29, 8 March 2018 (UTC)
 * To be a phototransistor, yes. But if you use the base and collector junction alone, it's a simple photodiode.  Electrically they behave differently; phototransistors have gain and were (at the time) more sensitive. Andy Dingley (talk) 12:49, 8 March 2018 (UTC)


 * But it is in the paragraph starting with "A phototransistor is a light-sensitive transistor."


 * From Phototransistor Symbol and Circuit Configurations:


 * It can be seen that the phototransistor symbol shown does not give a base connection. Often the base is left disconnected as the light is used to enable the current flow through the phototransistor. 
 * --Mortense (talk) 13:01, 8 March 2018 (UTC)
 * I don't recall ever encountering a phototransistor where the base was physically disconnected: they've always had the third connection emerging from the case and left unused - probably because they were being made on a standard packaging line.
 * Electrically, phototransistors have a base, and this is used (they wouldn't be transistors otherwise). However the current to it arises from an internal photoelectric effect, so they don't need a connection to this base, although that's not the same as not having a base. Andy Dingley (talk) 13:08, 8 March 2018 (UTC)
 * Also, for the early alloy-junction transistors, the base is the mechanical basis of the transistor and needs to be mounted, unlike the later diffusion transistors, which are grown from the collector upwards. Andy Dingley (talk) 13:16, 8 March 2018 (UTC)

Comments on the article
I did not enjoy reading this article. I don't find it well organized. The text is not easy to interpret.

1. "The common, traditional solar cell used to generate electric solar power is a large area photodiode." and "This mode exploits the photovoltaic effect, which is the basis for solar cells – a traditional solar cell is just a large area photodiode." are redundant. One of the two should be removed, preferably the second.

2. The "Other modes of operation" subsection was listing avalanche photodiodes which are a type of photodiode and phototransistors in bipolar and field-effect forms which are related but not diodes. Solaristors were also listed. I promoted the subsection to a section and renamed it "Related devices".

3. I think the "pinned photodiode" should move under "Related devices".

ICE77 (talk) 07:47, 23 May 2020 (UTC)

Pinned photodiode history
I reverted your attempt. Please say here what the issues and relevant sources are, and we can discuss and help massage it into the right form for the article. Dicklyon (talk) 15:56, 3 February 2022 (UTC)

About the invention and historical development efforts of Pinned Buried Photodiode.

Not to be confused with PIN photodiode. The pinned photodiode (PPD) has a shallow implant (P+ or N+) in N-type or P-type diffusion layer, respectively, over a P-type or N-type (respectively) substrate layer, such that the intermediate diffusion layer, like the base region of a bipolar junction transistor, can be fully depleted of majority carries, which gives the compete signal charge transfer action, realizing the completely image lag feature. Early charge-coupled device image sensors with the floating surface N+P junction type photo sensor suffered from shutter lag. This was largely resolved by replacing the N+P single junction type photo sensor with the invention of the double and triple junction type pinned photodiode (PPD). The PPD (usually PNP) is used in CMOS active-pixel sensors; The original NPN variants with the MOS buffer memory for Global Shutter Function and also the PNP type PPD with a anti-blooming structure both were invented at Sony in 1975 for use in CCD image sensors. Sony also developed the PNP type PPD in FT CCD image sensor in 1978. This led to their invention and the first development of the pinned photodiode, a photodetector structure with low lag, low noise, high quantum efficiency and low dark current.[21]

Buried Photodiode with the serious image lag was proposed by NEC in 1980.[22][23] They also recognized that lag can be eliminated if the signal carriers could be transferred completely from the photodiode to the CCD. It was first publicly reported by NEC in 1982, with the addition of an anti-blooming structure.[22][24] The new photodetector structure invented first at Sony in 1975 and developed first at Sony in 1978 were given the name "pinned photodiode" (PPD) by KODAK. In 1987, the PPD began to be incorporated into most CCD sensors, becoming a fixture in consumer electronic video cameras and then digital still cameras.[22]

NASA's Jet Propulsion Laboratory (JPL) proposed an improvement to the CMOS sensor: the integration of the pinned photodiode. A CMOS sensor with PPD technology was first fabricated in 1995 by a joint JPL and Kodak team. Since then, the PPD has been used in nearly all CMOS sensors. The CMOS sensor with PPD technology was further advanced and refined by R.M. Guidash in 1997, K. Yonemoto and H. Sumi in 2000, and I. Inoue in 2003. This led to CMOS sensors achieve imaging performance on par with CCD sensors, and later exceeding CCD sensors.[22]

Reference to be added additional references by proper numberings:

[1] Yoshiaki Hagiwara, “Invention and Historical Development Efforts of Pinned Buried Photodiode”, Proc. of the International Conference on Electrical, Computer and Energy Technologies (ICECET) 9-10 December 2021, Cape Town-South Africa Should be linked to the following site for details. https://202011282002569657330.onamaeweb.jp/AIPS_Library/P2021_ICECET2021Paper75_PWD_897_992_647_542_870_423_776_till_Dec_10_2021/ICECET2021_Paper75.pdf

[2] Yoshiaki Hagiwara, Japanese Patent Application JPA 1975-127646 on N+NP+NP-P+ Triple Junction Type Pinned Photodiode with Back Light Illumination with the CCD/MOS Buffer Memory for Global Shutter Function. Should be linked to the following site for details. https://202011282002569657330.onamaeweb.jp/AIPS_Library/3_JP1975-127646_NPNP_triple_junction_Pinned_Photodiode_Patent_32_pages.pdf

[3] Yoshiaki Hagiwara, Japanese Patent Application JPA 1975-127647 on N+NP+N Double Junction Type Pinned Photodiode with Back Light Illumination with the CCD/MOS Buffer Memory for Global Shutter Function. Should be linked to the following site for details. https://202011282002569657330.onamaeweb.jp/AIPS_Library/4_JP1975-127647_NPN_double_junction_Pinned_Photodiode_Patent_22_pages.pdf

[4] Yoshiaki Hagiwara, Japanese Patent No. 1215101 (Japanese Patent Application JPA 1975-134985) on the Pinned surface P+NP double junction type Pinned Photodiode on N-type substrate wafer (Nsub), forming a P+NPNsub triple junction dynamic photo thyristor type PPD with the VOD function. Should be linked to the following site for details. https://202011282002569657330.onamaeweb.jp/AIPS_Library/5_JP1975-134985_PNP_double_junction_Pinned_Photodiode_on_Nsub_Patent_7_pages.pdf

[5] Yoshiaki Hagiwara, Shigeyuki Ochi and Takeo Hashimoto, Japanese Patent Application JPA 1977-126885 on Electrical Shutter Clocking Scheme with OFD Punch Thru Action. Should be linked to the following site for details. https://202011282002569657330.onamaeweb.jp/AIPS_Library/6_JP1977-126885_Elecric_Shutter_Clocking_Scheme_by_OFD_Punch_Thru_Action_13_pages.pdf

[6] Yoshiaki Hagiwara, Motoaki Abe and Chikara Okada, "A 380H X 488V CCD Imager with Narrow Channel Transfer Gates", Proceeding of the 10th Conference on Solid State Devices, Tokyo 1978, Japanese Journal of Applied Physics, Volume 18 Sup 18-1, pp. 335-340 November 1979. Should be linked to the following site for details. https://202011282002569657330.onamaeweb.jp/AIPS_Library/9_P1978_Pinned_Photodiode_1978_Paper_by_Hagiwara_7_Pages.pdf

[7] Yoshiaki Daimon-Hagiwara,” Advances in CCD imagers”, an invited paper at CCD’79 international conference at Edinburgh, Scotland, UK September 1979. Should be linked to the following site for details.

[8] https://harvestimaging.com/pubdocs/089_2005_dec_IEDM_hole_role.pdf

[9] https://www.sony.com/en/SonyInfo/News/notice/20200626/

And also I would like to add the following references.

[A1] https://202011282002569657330.onamaeweb.jp/AIPS_Library/3_JP1975-127646_NPNP_triple_junction_Pinned_Photodiode_Patent_32_pages.pdf

[A2] https://202011282002569657330.onamaeweb.jp/AIPS_Library/4_JP1975-127647_NPN_double_junction_Pinned_Photodiode_Patent_22_pages.pdf

[A3] https://202011282002569657330.onamaeweb.jp/AIPS_Library/5_JP1975-134985_PNP_double_junction_Pinned_Photodiode_on_Nsub_Patent_7_pages.pdf

[A4] https://202011282002569657330.onamaeweb.jp/AIPS_Library/6_JP1977-126885_Elecric_Shutter_Clocking_Scheme_by_OFD_Punch_Thru_Action_13_pages.pdf

[A5] https://202011282002569657330.onamaeweb.jp/AIPS_Library/P1978_Pinned_Photodiode_1978_Paper_by_Hagiwara.pdf

I would like to add also my recent published papers.

[A6] https://202011282002569657330.onamaeweb.jp/AIPS_Library/P2019_3DIC2019Paper_on_3D_Pinned_Photodiode.pdf

[A7] https://202011282002569657330.onamaeweb.jp/AIPS_Library/P2020_EDTM2020_PaperID_3C4_by_Hagiwara.pdf

[A8] P2021_IJSSA2021_Paper_20210616https://202011282002569657330.onamaeweb.jp/AIPS_Library/P2021_IJSSA2021_Paper_20210616_on_Electrostatic_and_Dynamic_Analysis_of_Pinned_Photodiodes.pdf

[A9] https://202011282002569657330.onamaeweb.jp/AIPS_Library/ICECET2021/001A_ICECET2021_Paper061_Pinned%20Buried%20PIN%20Photodiode%20type%20Solar%20Cell.pdf

[A10] https://202011282002569657330.onamaeweb.jp/AIPS_Library/ICECET2021/002A_ICECET2021_Paper075_Invention%20and%20Historical%20Development%20Efforts%20of%20Pinned%20Buried%20Photodiode.pdf

— Preceding unsigned comment added by YoshiakiHagiwara19480704 (talk • contribs) 16:49, 12 February 2022 (UTC)


 * First revision, addressing conflict of interest issues:
 * I suspect the material has been rejected in the past because of your user name, and the appearance that you are inserting self-interested details about the history of development. The importance of those details to an already verbose article may be exaggerated by personal judgement. (I don't see where you've addressed the advice on your talk page at Managing a conflict of interest.) I've made a first revision removing all of that detail:


 * About the invention and historical development efforts of Pinned Buried Photodiode.


 * The first pinned photodiode (PPD) and the Buried Photodiode (BP) both have a shallow P+ implant in N type diffusion layer over a P-type substrate layer. PPD is always BP. PPD does not have the image lag. But BP may have the serious image lag problem. PPD is also not to be confused with the PIN photodiode.[21].
 * The first PPD with the no-image lag, the low-surface dark current and the high short-wave blue light sensitivity features was invented in 1975 at Sony.[1]
 * Sony developed the first double junction type PPD and used it in FT CCD in 1978 [6]. This was the first invention of the pinned photodiode (PPD), a photodetector structure with low lag, low noise, high quantum efficiency and low dark current.
 * Early charge-coupled device image sensors with the single N+P photodiode with the N+ floating surface suffered from shutter lag. The serious image lag was largely resolved in 1978. The PPD was used in CCD sensors in the past analog TV era and is now still used in the present digital TV era in CMOS active-pixel sensors.
 * In 1980, BP was used in ILT CCD, eliminating if the signal carriers could be transferred from the photodiode to the CCD. [24]
 * The BP with the serious image lag problem was further developed and used in ILT CCD in 1982, with the addition of an anti-blooming structure.[22][24] The BP proposed in 1980 and developed in 1982 by NEC [22][23] reported the serious image lag problem, which was by definition not PPD. BP is not always PPD. [21].
 * The photodetector structure was given the name "pinned photodiode". In 1987, the PPD began to be incorporated into most CCD sensors, becoming a fixture in consumer electronic video cameras and then digital still cameras.[22]
 * In 1994, Eric Fossum, while working at NASA's Jet Propulsion Laboratory (JPL), proposed an improvement to the CMOS sensor: the integration of the pinned photodiode. A CMOS sensor with PPD technology was first fabricated in 1995 by a joint JPL and Kodak team. Since then, the PPD has been used in nearly all CMOS sensors.
 * The CMOS sensor with PPD technology was further advanced in 2000 and 2003. This led to CMOS sensors achieve imaging performance on par with CCD sensors, and later exceeding CCD sensors.[22]


 * (references as above, not repeated here)  signed, Willondon (talk)  17:17, 12 February 2022 (UTC)

Reverting to last stable
For the moment I am reverting to the most stable version in my opinion. There are lots of changes in the article that are not explained and unsourced, which I don't think will be constructive to the article. I don't see any sources changed, but the content are changed in the article, that is why per WP:BRD I am reverting and starting this discussion. I think that massive changes to the article as I see should be discussed here before being placed in the article, as it is clear that there are concerns to the content being added.

I also noticed that you have some WP:COI concerns on your talk page, can you resolve that? Thank you and have a nice day everyone. SunDawn talk  02:53, 21 February 2022 (UTC)


 * Yes, we definitely need better sourcing if we're to say much more about Hagiwara's contributions. I will try to formulate a way to acknowledge his contribution, based on sources, without going overboard in claiming priority for the PPD invention, which is usually attributed to Teranishi of NEC based on the definition that Kodak authors wrote, in which the PPD has an n-type collection layer and p-type pinning layer to accumulate photoelectrons; Hagiwara used a p-type collection layer and n-type pinning layer to collect holes (later known as Sony's "hole accumulation diode", HAD) to make a similar or equivalent structure; was it as effective and important?  Hard to tell.  There are important points to be made on both sides of this priority dispute.  Hagiwara wants his early contributions to be acknowledged in the history; Fossom makes the point that Teranishi's PNP structure is what's widely used in today's image sensors.  Neither is wrong, but what we say about "first" and "invention" needs to be based on reliable refereed sources.  This looks to me like a case where more analysis and publication outside of Wikipedia is needed before we can say much more about it. Dicklyon (talk) 04:51, 21 February 2022 (UTC)


 * One issue I have is with the inclusion of a large number of individual names, and sometimes companies. Secondary sources not only verify facts but, often forgotten, indicate notability. Some names (Teranishi and Fossum) have their own WP articles and secondary coverage. Others do not, and while you hate to diminish the valuable contributions of others, in those cases where they are listed alongside Teranishi or Fossum, could they be referred to as "et al."? — Preceding unsigned comment added by Willondon (talk • contribs)

This is Hagiwara writing with apology for confusions since I am not familiar with Wikipedia editing rules and knowhows.

I just wanted to show the evidence that I am the true inventor of Pinned Buried Photodiode ( SONY HAD ). I wanted to upload the evidence that my 1975 inventions defined the NPNP triple junction type PPD [A1] the NPN double junction type PPD [A2] and the PNP double junction type PPD [A3]with anti-blooming structure.

The problem was these patents [A1][A2][A3] were all written only in Japanese. I am now 73 years old and I forgot completely the details of my 1975 inventions. I stopped working for image sensors in 1980. Since 1980, I was busy working at Sony for AIBO robot and PS3 Cell Processors and AI robots for future. I retired from Sony in 2008 at age 60.

In 1977 and in 1978 as a young engineer I published two SSDM conferences [B1][B2] on the first PPD structures. Since then I invited to the eight international conference[C1]thru[C8]. Recently I found out Fossum wrote a misleading paper attacking SONY(Hagiwara) inventions on Sony HAD which is identical to PPD. Sony was late in producing PPD on the market till 1987. Sony invented PPD in 1975 [A1][A2][A3]and reported the first PNP junction type PPD delay line at SSDM1977 [B1] and then the first PNP junction type  PPD image sensor array at SSDM1978[B2]. I was invited at eight international conferences[C1] thru [C8]. Recently I found out my invention was stolen somehow. I was really absent-minded and busy in other business. Recently I challenged to publish eight papers and was happy to have five papers [D1] thru [D5] accepted. These five papers explain the details why I believe that I invented the NPN variant and the original PNP type PPD in 1975 and that I developed PNP type PPD used as a delay line in SSDM1977[B1] paper and as a area image sensor in SSDM1978[B2]. The details PDF copies are liked to the dedicated site shown below.

http://www.aiplab.com/Evidence_that_Hagiwara_at_Sony_is_the_inventor_of_Pinned_Photodiode.html

I wish editors of Wikipedia help me to claim my belief on the invention of the first PPD and the development efforts.

I am not happy at all at Fossum misleading paper. Fossum did not quote my inventions and development efforts properly. Fossum did not quote the important in-pixel active source-follower 3T1C amp circuit invented by Peter Noble in 1968. The modern CMOS image sensors use PPD invented in 1975 by Hagiwara at Sony, Japan and also the in-pixel 3T1C active amp circuit invented in 1968 by Peter at Plessy, UK. We now longer used CCD. CCD was used as the charge transfer device (CTD) for a long time with PPD as the light detecting dev. Now with the modern advanced CMOS scaling technology we now have enough space in one pixel to include the 3T1C amp circuit area for the modern CMOS type CTD image sensors.

Yoshi  http://www.aiplab.com    hagiwara-yoshiaki@aiplab.com


 * Wikipedia cannot help with that, I'm afraid. The encyclopedia doesn't do research to judge information; it relies on secondary sources to report on things, and relies on them to indicate notability. Because of that, there are very important concerns when it comes to primary sources and editors who have a connection to the subject. What you would have to do is find reliable secondary sources that have published information more reflective of the facts. And unfortunately, it may be the case that Wikipedia doesn't perfectly reflect history or reality (what does?). If the secondary sources have got it wrong, Wikipedia will have it wrong. That's a fact we all have to live with.  signed, Willondon (talk)  16:07, 22 February 2022 (UTC)


 * There are some small steps we can take, I think. Not in deciding who invented what, but it clarifying the objective history of who said what in publications.  The Fossum paper that Yoshi refers to is not to his liking, and possibly not unbiased, but is a secondary source that analyzes his claims and his patents at least a little bit.  Thieuwissen also has a paper that touches on the subject.  So there are things that can be said without taking sides in disputes, quoting these peer-reviewed articles instead of leaving Hagiwara's contribution completely out of the story.  It's clear that he patented in 1975 and published on in 1978 and later some photodiode structures that have a resemblance (with opposite polarity) to the one that Kodak later dubbed "pinned".  Their published definition of "pinned" was not so flexible as to apply to the analogous structure with opposite polarity.  Some of these things can be mentioned while avoiding priority and inventorship disputes.  Dicklyon (talk) 18:48, 22 February 2022 (UTC)

Hi, this is Hagiwara writing again. Please examine the two publications [A] and [B] shown below by neutral people. I wish these two publications be quoted in Wikipedia.

[A] Please see  https://www.shmj.or.jp/english/pdf/dis/exhibi1005E.pdf

which quotes the three original publications on the first PPD development efforts by Sony.:

[3] Y. Hagiwara, Japanese Patent JP1975-134985 on PNP jucntion type PPD with anti-blooming structure. [5] Y. Hagiwara, M. Abe, and C. Okada,“A 380H x 488V CCD imager with narrow channel transfe [6] I. Kajino, M. Shimada, Y. Nakada, Y. Hirata and Y. Hagiwara, ”Single Chip Color Camera Using

I really wish you would kindly include these publications [3][4][5] and [A] to Wikipedia "Pinned Photodiode"

As shown n Fig. in the publication [5], Sony developed the PNP junction type PPD directly pinned by the the adjacent P+ channel stops formed by ion-implantation technology while KODAK and NEC used the conventional LOCOS technology which induced the undesired oxidation-induced thermal stress, degrading the chip yields.

[B] Please see also https://www.ijert.org/research/cmos-image-sensors-recent-innovations-in-imaging-technology-IJERTCONV7IS12001.pdf

which quotes the old forgotten IEEE paper [4] and explains the importance of the in-pixel amp 3T1C circuit invented by Peter Noble in 1968.

[4] Peter J. W. Noble, IEEE Trans. Electron Dev. 15-4 (1968)202–209

I really wish you would kindly include these publications [4] and [B] to Wikipedia "Pinned Photodiode".

The modern CMOS image sensor is made of two parts. One is the light-detecting photo-charge storage which is now widely called as PPD. The other is the digital low-power CMOS type charge transfer device (CTD).

We used CCD type CTD with PPD sensor in the analog TV era. But we no longer used CCD type CTD in our modern digital high definition TV era. Now we use the digital low-power CMOS type CTD with PPD and also the in-pixel active 3T1C source-follower AMP circuit invented by Peter Noble in 1968.

Yoshi   http://www.aiplab.com      hagiwara-yoshiaki@aiplab.com

P.S.

Please note that the publication [A] by Japan Semiconductor History Museum, which is shown above, commented as "In 1975, Sony proposed using a PNP transistor as the photodetector [3]. By providing a P+ layer (emitter) for the light incident section, the sensor electrode that covers the entire light receiving surface of the photodiode can be eliminated to improve the light sensitivity greatly. It was a basic proposal for a pinned photodiode with a P+ layer on the surface of the light receiving part"

That is, in short, Japan Semiconductor History Museum said in public, "In 1975 Sony proposed PNP transistor type dynamic photo transistor" which is identical to the PNP type pinned photodiode.

In conclusion, Sony invented not only the NPN variant in Oct 23, 1975 but also the first PNP type Pinned Photodiode in Nov. 10, 1975.

The evidence is given by the Fig. 6 of the JPA1975-134985.

Fig. 6 shows the photo transistor with the completely depleted empty potential well in the base photo-charge storage region, implying the complete charge transfer action and with no image lag feature.

In the example shown in Fig. 6 of the JPA1975-134985 patent, the emitter region is used for the in-pixel vertical overflow drain(VOD) for anti-blooming function. The structure is now widely called as Pinned Photodiode.

Yoshi   http://www.aiplab.com      hagiwara-yoshiaki@aiplab.com

+++++++++++++++++++++++++++++++++++++++++++++

This is Yoshi, speaking again and my apology for making my talks on the story very very long. I would like to summarize my requests to kind editors:

Can you and other kind editors help adding these references shown below in the talks to Wikipedia on “Pinned Photodiode” and “Pinned Photodiode History” ?

[1][2][3] and [4] to the Wikipedia on "Pinned Photodiode"

[1]   https://www.shmj.or.jp/english/pdf/dis/exhibi1005E.pdf

This is a public notice by Japan Semiconductor History Museum which quotes the three original publications on the first PPD development efforts by Sony.:

[2] Y. Hagiwara, Japanese Patent JP1975-134985 on PNP junction type PPD with anti-blooming structure.

[3] Y. Hagiwara, M. Abe, and C. Okada,“A 380H x 488V CCD imager with narrow channel transfer gates”, Proc. The 10th Conference on Solid State Devices, Tokyo, (1978): Japanese Journal of Applied Physics, vol. 18, Supplements 18-1, pp. 335-340, (1979)

[4] I. Kajino, M. Shimada, Y. Nakada, Y. Hirata and Y. Hagiwara, ”Single Chip Color Camera Using Narrow channel CCD Imager with Over Flow Drain”、Technical Report of The Institute of Image Information and Television Engineers、vol. 5, no. 29, pp. 32-3S, (1981)

For further understandings, specially on Modern CMOS imagers which are composed of (1) PPD type photon detecting device (PDD) and (2) CMOS type charge transfer device (CTD).

Historically confusingly we had (A) N+P junction type PPD (B) MOS capacitor type PPD and (C) PPD type PDD. and also confusingly we had (A) MOS passive type CTD (B) CCD type CTD and (C) CMOS digital Active type CTD. please add the paper [5] shown below which was published on International Journal of Engineering Research & Technology (IJERT), ISSN: 2278-0181, Published by, www.ijert.org NCRIETS – 2019 Conference Proceeding

which quotes the old forgotten IEEE paper [6] and explains the importance of the in-pixel amp 3T1C circuit invented by Peter Noble in 1968.

[5] https://www.ijert.org/research/cmos-image-sensors-recent-innovations-in-imaging-technology-IJERTCONV7IS12001.pdf

[6] Peter J. W. Noble, IEEE Trans. Electron Dev. 15-4 (1968)202–209 kind regards, Yoshi   http://www.aiplab.com      hagiwara-yoshiaki@aiplab.com

A proposal
I think a bit of rewrite is in order. First off, the pinned photodiode section doesn't make a lot of sense under Applications. There should be perhaps a section on Variants instead. Second, we should start with a subsection on buried photodiodes, and describe Hagiwara's 1975 structures, among others, there. After that, a section on pinned photodiodes, noting that Kodak picked that name and definition to describe Teranishi's device. We can note that the question of whether Hagiwara's device should be considered a pinned photodiode is disputed, and give reference to papers by him and Sony and by Fossum and by Theuwissen, without taking a position on the question. I believe there is good reliable secondary sourcing for enough stuff here, and we're obviously not going to make everybody (or anybody) happy either by taking sides or by ignoring Hagiwara's contributions. I can try to draft something in a while, if nobody else does. Dicklyon (talk) 05:01, 25 February 2022 (UTC)

There is an old feud between Sony and NEC on this invention. When writing my review article I spent some time on Hagiwara's patents and comparing them to Teranishi et al's patents at NEC. I don't think there was any initial bias. But as stated in the IEEE Trans. Electron Devices peer-reviewed PPD review paper I co-authored, in the end, modern PPDs differ significantly from Hagiwara's similar and earlier invention. If only he did not draw a metal contact to the top layer (which causes several drawbacks) and "clock" that contact, he could claim he was the inventor of the modern PPD. We all stand on the shoulders of those that came before us, and I do believe Hagiwara's device should be included as an independent precursor to the NEC work, even if ultimately it was the NEC invention that was adopted nearly universally, not the Sony precursor embodiment. Hagiwara definitely deserves credit for his early and advanced ideas. But, it is more than the NPN v. PNP structure that is different. Best not to overly simplify this. FossumER (talk) 22:52, 5 March 2022 (UTC)


 * Hagiwara proposed the pinned photodiode with the surface metal contact as an option in his three 1975 original Jjapanese Patent Applications. However, Hagiwara also invented in 1978 to used the adjacent P+ channel stops to pin the surface P region od the PNP double junction Pinned Photodiode.Hagiwara reprted the data in the SSDM1977 and the SSDM1978 conferences. For a very high frequency global shutter with a very short exposure time, any RC delay time is not desired. In modern CMOS image sensor applications for digital HD TVs, the metal direct contact is the best option. It has zero RC delay time. The transistors and their metal contacts now have become much smaller then the pixel size, thanks to the CMOS scaling advancemnts. 111.108.209.27 (talk) 19:18, 27 February 2023 (UTC)