Different Types of Transistors

[EEVblog]

I'm working on a video about the pros/cons/applications of each type of transistor, and the table is looking rather, err, fun
And I'm researching more info to potentially add to the list, or spark anything I've missed.
Such a comparison table is going to be inherently messy in any case 

Anyway, I recall a thread on the forum somewhere that IIRC had a good comparison breakdown of the different types, but in extensive searching I can't find it. Does anyone recall this thread/post?
Thanks.

[floobydust]

What did you mean by "types"?

The die design like epitaxial, perforated emitter, that kind of thing? JFET, MOSFET, BJT, that kind of thing?

[coromonadalix]

sillicium,  germanium ?

[EEVblog]

What did you mean by "types"?
The die design like epitaxial, perforated emitter, that kind of thing? JFET, MOSFET, BJT, that kind of thing?

Sorry, yes, the major functional types, JFET, MOSFET, BJT, IGBT.
Don't really want to go deeper down the rabbit hole than that I think, but maybe a mention of some sub categories?

Basically the video was going to be titled "Which Transistor" and would be a basic breakdown of which type of transistor is best/better/worse for which application etc.
Basically just a big whiteboard table and discussion, no practical experiments in this video anyway.

[Alex Eisenhut]

How about historical oddities like the UJT?

[Doctorandus_P]

I remember reading a story about a fet designed by Philips specifically for shorting the antenna input of a video recorder to it's antenna output, so the TV could still be used even when the whole video recorder was turned off or even unplugged. It must have been a depletion mode MOS fet or Jfet working up to GHz frequencies.

I'd think you could also do a whole series about opamps. You've made some video's of some circuit full with SMT and a few THT opamps and stated there must have been a good and special reason why exactly that particular opamp was selected in that place. Some more in depth circuit analyses and what tradeoffs are made to select components could be interesting.

[T3sl4co1l]

Uhhh, that probably sounds like something I listed off, but damned if I remember what the thread was called.

Do ya... just want to review the script instead?  Probably faster than searching through my posts...

Tim

[EEVblog]

Uhhh, that probably sounds like something I listed off, but damned if I remember what the thread was called.

Yeah, I had a vauge recollection it was you and search for your posts but also came up with nothing.

Do ya... just want to review the script instead?  Probably faster than searching through my posts...

Script? 
At best my detailed videos comprise a coule of post-it notes.
I might post something here once I get something decent together and everyone can battle it out 

[bsfeechannel]

Anyway, I recall a thread on the forum somewhere that IIRC had a good comparison breakdown of the different types, but in extensive searching I can't find it. Does anyone recall this thread/post?
Thanks.

Maybe this one:

https://www.eevblog.com/forum/chat/bjts-being-phased-out/

[EEVblog]

Anyway, I recall a thread on the forum somewhere that IIRC had a good comparison breakdown of the different types, but in extensive searching I can't find it. Does anyone recall this thread/post?
Thanks.

Maybe this one:

https://www.eevblog.com/forum/chat/bjts-being-phased-out/

Nope, but that one didn't come up in searches.
IIRC there was almost an ASCII character box like format to it, like:

------------|----------------|-----------
BJT                 PROS              CONS
MOSFET
etc

I could be completely mis-remembering though.

[PartialDischarge]

Maybe not this but interesting info to cover:

https://www.eevblog.com/forum/beginners/transistor-vs-mosfet-for-linear-power-supply/

[Berni]

I would also include Triacs and SCRs since they have particular qualities for high peak power or AC switching that are not as easy to do with regular transistors.

Also mention "Delpletion Mode MOSFETs" because they are modern parts that are commonly available, but very few people know about. I suppose they act a lot like JFETs but these can be used to actually switch a serious amount of power. I'm sure a lot of people wished for a transistor that worked like the normally closed contact on a relay.

[floobydust]

How about historical oddities like the UJT?

I died of boredom in university EE class as the prof plowed through math on the intrinsic stand-off ratio for UJT's. 40 years, used a UJT three times. I don't think they are worth covering.

But definitely cover SiC MOSFETs, they are driving green energy products and newer tech.

These have a gate but are transistors, kinda confusing: https://www.genesicsemi.com/sic-junction-transistor/

[EEVblog]

I would also include Triacs and SCRs since they have particular qualities for high peak power or AC switching that are not as easy to do with regular transistors.

Considered that. Will decide once I have somethign fleshed out and see if the video is already possibly too long.

Also mention "Delpletion Mode MOSFETs" because they are modern parts that are commonly available, but very few people know about. I suppose they act a lot like JFETs but these can be used to actually switch a serious amount of power. I'm sure a lot of people wished for a transistor that worked like the normally closed contact on a relay.

Probably worthy of a separate mention.

[Alex Eisenhut]

I would also include Triacs and SCRs since they have particular qualities for high peak power or AC switching that are not as easy to do with regular transistors.

Well, in that case, we should cover the modern ones.

https://en.wikipedia.org/wiki/Integrated_gate-commutated_thyristor

And maybe talk about the crazy "Neutron-Transmutation-Doped Silicon used as the IGCT base substrate". Great way to make uniform doping is to transmute silicon into phosphorus in situ...

[sandalcandal]

HEMT and eHEMT are some relatively new transistor types which are of importance in new GaN based technology.

In order to achieve the absolute highest efficiencies (>99%) by using the absolute fastest switching speeds (>150V/ns rise times) without causing excessive ringing and EMI, TI (and maybe some others?) are building co-packaged driver-transistor ICs. http://www.semiconductor-today.com/news_items/2020/nov/ti-091120.shtml

I also see "cascode" being used in some SiC implementations, this isn't discrete semiconductors connected to form a cascode circuit within a package but a monolithic doping/epitaxy structure which creates an effective cascode behaviour. UnitedSiC in particular likes to tout their cascode SiC.

Overall these cascode SiC FETs seem to tend towards having worse parasitic capacitance but slightly better Rds_on so whether or not they are better than more conventional MOSFETs such as those from Wolfspeed/Cree is dependent on the exact application circumstances. My personal design efforts and review of reference designs and papers for the most efficient and cost effective designs tends away from cascode SiC towards "conventional" SiC MOSFETs (or GaN eHEMT when voltage is low enough) when given complete freedom over input/output voltage and current levels.

I had some notes on some good reverse engineering videos of GaN structures being used in some very modern SMPS in another thread. There are also some other light notes from other posters in that thread too.

I recently saw a webinar from a reverse engineering firm that did a  brief segment on GaN-on-Si AC adapters. https://techinsights.wistia.com/medias/7ma357gv90 @16:50 to 21:20 Also https://www.techinsights.com/blog/navitas-found-inside-ravpower-rp-pc104-w-gallium-nitride-45-w-usb-c-power-delivery-charger and https://www.eetimes.com/ac-adapters-gan-sic-or-si/ (from the same author) They looked at a few including RAV Power, Made in Mind Mu One and mentioned Anker and FINsix. The implementations appear to be just based on reference designs of GaN ICs from Navitas and Popwer Integrations (which @Weston has also mentioned).

There was also some discussion on GaN in my resonant converter project thread with JohnG providing some insider insight on GaN tech which opened the way for my own initial understanding before I did further research. Look through the thread for the continued discussion. Some good notes on the very different reverse behaviour of HEMT.

I've been out of the converter design loop for a few years now, but at one point I had spent some time with resonant and high frequency designs. But, I can comment on the devices, having spent the last decade looking at SiC and GaN power devices.

All the low voltage, commercially available GaN devices intended for power conversion are enhancement-mode HEMTs on a silicon substrate. There are some depletion-mode devices available, but the latter are for RF amps where they tolerate such nonsense  .

They do not have a PN junction, so they do not have a body diode in the usual sense, but they all act as a body diode when the gate is connected to the source (via gate drive or some other means). They are essentially turning themselves on in the reverse direction, so the voltage drop is the threshold plus whatever extra voltage is needed to conduct the current, usually about 2.5-3V. It is higher than a SI PN junction. However, there are a couple mitigating factors. First, there is no reverse recovery. This may not matter for a resonant converter operating at nominal load, but if you have to run at light load, or during startup, it may become very difficult to completely avoid hard switching, and you can end up spending a lot of time getting this to work as desired. Since reverse recovery losses can dominate total switching loss, this can be an important advantage.

Note that the lack of a PN junction is also the reason that GaN power FETs do not have an avalanche rating. The breakdown voltage is often much higher than the voltage rating, but it is wise to follow the spec and not exceed it, because it affects the lifetime of the part (exceeding the rating causes a gradual increase in on-state resistance).

While the low voltage GaN FETs tend to have lower output charge Qoss, we are talking 10s of percent lower vs Si. The gate charge might be 5-10x lower, and if you are running at a MHz, this has an impact on total losses. Housekeeping power is often an afterthought, but it can be a real headache if you at looking at efficiencies in the upper 90s.

The GaN Systems part has a package, and the gate can tolerate a higher voltage. This certainly makes the life of the layout engineer easier, and the higher voltage rating on the gate is a little more forgiving. The package adds size and inductance, though the added inductance very small. The chip-scale package of the EPC parts requires more attention to layout of the power loop and handling, but you can get a 7 mohm, 100V part for $0.90 in 1k quantities (Digikey pricing).

Disclaimer: I work for a GaN semiconductor company.

Cheers,
John

I've also sat in on a bunch of industry talks on the latest and greatest in transistor technology but I'm not sure those talks or their archives are available to the general public... you might be able to find some if you have a look but I'll post them if I can find them on youtube or somewhere else that doesn't require login access.

There does seem to be significant and fairly rapid movement and development in power semiconductors thanks to the impetus of renewable power systems and electric vehicles.

Edit: Even within MOSFETs there are "differences" with different semiconductor structures which produce different characteristics for the overall device, e.g. HEXFET and super-junction trench. Most of the newest and best performance silicon MOSFETs (for the applications I look at) seem to be some flavour of super-junction, deep-trench.

[sandalcandal]

Anyway, I recall a thread on the forum somewhere that IIRC had a good comparison breakdown of the different types, but in extensive searching I can't find it. Does anyone recall this thread/post?
Thanks.

Maybe this one:

https://www.eevblog.com/forum/chat/bjts-being-phased-out/

Nope, but that one didn't come up in searches.
IIRC there was almost an ASCII character box like format to it, like:

------------|----------------|-----------
BJT                 PROS              CONS
MOSFET
etc

I could be completely mis-remembering though.

I don't recall seeing something like that but there's a good chance that and a bunch of other relevant power electronics notes have all been deleted along with a certain user... even that thread is missing some posts from that user.

[David Hess]

Also mention "Delpletion Mode MOSFETs" because they are modern parts that are commonly available, but very few people know about. I suppose they act a lot like JFETs but these can be used to actually switch a serious amount of power. I'm sure a lot of people wished for a transistor that worked like the normally closed contact on a relay.

Probably worthy of a separate mention.

I consider them the 4th type after bipolar, JFET, and enhancement mode MOSFET.

[magic]

But definitely cover SiC MOSFETs, they are driving green energy products and newer tech.

These have a gate but are transistors, kinda confusing: https://www.genesicsemi.com/sic-junction-transistor/

That link is not SiC MOSFETs. Not sure what they are, ordinary BJTs made of SiC? What's the point of that terminal naming, then?

[David Hess]

That link is not SiC MOSFETs. Not sure what they are, ordinary BJTs made of SiC? What's the point of that terminal naming, then?

My guess is that an intern made the datasheets with copy/paste.

[Alex Eisenhut]

How about the dual gate MOSFET?

https://www.hamanuals.com/S1/CX7/Articles/Mosfet.pdf

And all the weird IC transistors like the Schottky transistor and multiple emitter ones?

[T3sl4co1l]

SiC BJTs, which they call SJTs and label the terminals like FETs for some bizarre reason, but yeah, they're BJTs as near as I can tell:
https://www.genesicsemi.com/sic-junction-transistor/
Astonishingly, they show full SOA, at least the one I clicked on.

JFETs, MOSFETs and IGBTs are also available in SiC, I think the JFETs were more of a transitional thing? but they have alright specs too.

Supposedly they make em up to 10kV, but I've never seen one for sale.  Wonder if that's an ITAR thing?

GaN FETs are available from low voltages (EPC eGaN in CSPs being the most prominent family) to over 20A, 600V (in large DFNs).  Very fast switching, no power dissipation or avalanche capacity (unsurprising given the tiny die area).

Si MOSFETs are still going strong, higher voltages especially benefitting from SuperJunction technology (since it was widely introduced in the 2010s ish?).  Amazingly, despite the higher than ever power density, they often sport DC SOA -- they can be a much cheaper option than "linear mode" MOSFETs.

What else... BJTs have always been there, not sure there's been much improvement in that domain over the last decades; though SiGe HBTs are pretty stonking fantastic.  RF BJTs have been slowly dwindling, with PNPs over 1GHz now being completely extinct AFAIK.

SiGe is pretty standard too, you can get op-amps made with them (e.g. AD's XFCB process) -- with fairly ordinary ratings and modest (if not jellybean) prices, and all's the wiser is the quiescent current is a sizable fraction lower than previous generations could offer.

Power (Si) MOSFETs have been getting incrementally faster, more efficient and so on, though still slow compared to RF types (mainly just because of aspect ratio: low Rds(on) is the higher priority, at expense to speed/capacitance and power dissipation).  RF MOSFETs (typically/all? LDMOS) place comfortably in the low GHz, and are reasonably affordable (ca. $1/W).

Not much improvement on jellybeans like MMBT3904/6, or BC847/857, or 2N7002/BSS84, or etc.  Tinier jellybean I'd recommend, RUM001L02T2CL and RZM001P02T2L.

Tim

[sandalcandal]

GaN FETs are available from low voltages (EPC eGaN in CSPs being the most prominent family) to over 20A, 600V (in large DFNs).  Very fast switching, no power dissipation or avalanche capacity (unsurprising given the tiny die area).

Certainly the EPC eGaN devices are relatively low power but there are plenty of quite high power HEMT GaN devices available. eHEMT devices by GaN Systems start at 4A 650V (tested at 850V) and go up to 150A https://gansystems.com/gan-transistors/
The HEMT/eHEMT structure is incapable of avalanching.

Edit:I haven't seen much of BJT for WBG semiconductors but everything else Tim has said lines up with what I've seen. I think I've also seen somewhere that high voltage semiconductors are regulated by ITAR
Edit2: Looking at the SiC BJT technology GeneSiC stuff seems relatively old and less "sophisticated". Looking at datasheets specs, the SiC MOSFETs look much better both in terms of cost and loss (parasitics and R_dson).
Comparing 1200V 100A switches
$98 https://www.genesicsemi.com/sic-junction-transistor/GA50JT12-247/GA50JT12-247.pdf
$33.70 https://www.wolfspeed.com/downloads/dl/file/id/1511/product/0/c3m0021120k.pdf

[magic]

SiC BJTs, which they call SJTs and label the terminals like FETs for some bizarre reason, but yeah, they're BJTs as near as I can tell:
https://www.genesicsemi.com/sic-junction-transistor/

Well, my first thought was that somebody from marketing didn't want the association with "obsolete" technology

[Momchilo]

Maybe the differences between Si and SiC technology. And a short explanation why PNP is not used as often as NPN.
If the video is not long enough and you still need content, maybe mention thyristors and why they can sometimes be the better alternative to transistors.