diode temperature model

in eq'ns (10.23), (10.24) and (10.27) in the "temperature model" section of

formatting link

Does what do B and ni mean?

Reply to
Michael Robinson
Loading thread data ...

P.S. I mean B in (10.23), not kB. And what physical meaning if any ni might have, because it doesn't appear in the list of parameters.

Reply to
Michael Robinson

Intrinsic carrier concentration of silicon?

Reply to
Andrew Holme

Yup, and the B is a term that is device specific. Get a text on semiconductor physics. I think in contains terms like the junction area and doping profiles.... but don't quote me.

George H.

Reply to
George Herold

Puzzle me this. Let's say you have a current mirror and two current sources, I1 and I2, with I1 = 10*I2. The currents are low enough to avoid self heating effects, and Q1 and Q2 have good thermal coupling. Q1's collector will sit 60 mV higher than Q2's collector (correct me if I'm wrong). You raise the circuit's temperature while holding I1 and I2 constant. What happens to the C1-C2 differential voltage?

I1 I2 | | | | | ,------+ | | | c\\ | /c Q1 |----+----| Q2 e/ \\e | | | | '------+------' | gnd

Reply to
Michael Robinson

Erm, I think the voltage differential would have to be witnessed by the emitters instead, which means redrawing the circuit with a voltage source between the emitters, or resistor(s) in one or both emitters. Anyway, the question I'm really asking is whether such a circuit would have the negative temperature coefficient of a diode junction, or is it a PTAT circuit like figure one in this link:

formatting link

Reply to
Michael Robinson

the circuit I drew is wrong, it needs the voltage differential to appear between the emitter, like this

supply---+-----Rs-----+---load | | e\\ BCV62 /e |----+---| c/ | \\ | | | out---+ '-----+ | | R1 R2 | | | ,-----+ | | | c\\ | /c |----+---| e/ BCV61 \\ | | '------+-----' | gnd

Reply to
Michael Robinson

In eqn. 10.23, B*T^3 is the product of the effective densities of state in the conduction band and valence band. Eqn. 10.23 is usually written as N(sub)c * N(sub)v * exp(-E(sub)g/(k(sub)B*T)). I would assume the reason Eqn. 10.23 is written as it is, is to make explicit the temperature dependence in the two effective density of states terms.

n(sub)i, as others have mentioned, is the intrinsic carrier concentration of silicon.

See, for example, Streetman & Bannerjee, "Solid State Electronic Devices", 5th ed., pg. 88, Eqn. 3-26. (That's n(sub)i, but you can easily square it.)

Hope this helps.

Bob Pownall

Reply to
Bob Pownall

DQoiQm9iIFBvd25hbGwiIDxyZXBvd25hbGxAbmV0c2NhcGUubmV0PiB3cm90ZSBpbiBtZXNzYWdl IG5ld3M6aDZzdnBsJGQ1YyQxQG5ld3MuZXRlcm5hbC1zZXB0ZW1iZXIub3JnLi4uDQo+IE1pY2hh ZWwgUm9iaW5zb24gd3JvdGU6DQo+PiBpbiBlcSducyAoMTAuMjMpLCAoMTAuMjQpIGFuZCAoMTAu MjcpIGluIHRoZSAidGVtcGVyYXR1cmUgbW9kZWwiIHNlY3Rpb24gb2YNCj4+IGh0dHA6Ly9xdWNz LnNvdXJjZWZvcmdlLm5ldC90ZWNoL25vZGU2OC5odG1sDQo+PiANCj4+IERvZXMgd2hhdCBkbyBC IGFuZCBuaSBtZWFuPyANCj4+IA0KPj4gDQo+IA0KPiBJbiBlcW4uIDEwLjIzLCAgQipUXjMgaXMg dGhlIHByb2R1Y3Qgb2YgdGhlIGVmZmVjdGl2ZSBkZW5zaXRpZXMgb2Ygc3RhdGUgDQo+IGluIHRo ZSBjb25kdWN0aW9uIGJhbmQgYW5kIHZhbGVuY2UgYmFuZC4gIEVxbi4gMTAuMjMgaXMgdXN1YWxs eSB3cml0dGVuIA0KPiBhcyBOKHN1YiljICogTihzdWIpdiAqIGV4cCgtRShzdWIpZy8oayhzdWIp QipUKSkuICBJIHdvdWxkIGFzc3VtZSB0aGUgDQo+IHJlYXNvbiBFcW4uIDEwLjIzIGlzIHdyaXR0 ZW4gYXMgaXQgaXMsIGlzIHRvIG1ha2UgZXhwbGljaXQgdGhlIA0KPiB0ZW1wZXJhdHVyZSBkZXBl bmRlbmNlIGluIHRoZSB0d28gZWZmZWN0aXZlIGRlbnNpdHkgb2Ygc3RhdGVzIHRlcm1zLg0KPiAN Cj4gbihzdWIpaSwgYXMgb3RoZXJzIGhhdmUgbWVudGlvbmVkLCBpcyB0aGUgaW50cmluc2ljIGNh cnJpZXIgDQo+IGNvbmNlbnRyYXRpb24gb2Ygc2lsaWNvbi4NCj4gDQo+IFNlZSwgZm9yIGV4YW1w bGUsIFN0cmVldG1hbiAmIEJhbm5lcmplZSwgIlNvbGlkIFN0YXRlIEVsZWN0cm9uaWMgDQo+IERl dmljZXMiLCA1dGggZWQuLCBwZy4gODgsIEVxbi4gMy0yNi4gIChUaGF0J3MgbihzdWIpaSwgYnV0 IHlvdSBjYW4gDQo+IGVhc2lseSBzcXVhcmUgaXQuKQ0KPiANCj4gSG9wZSB0aGlzIGhlbHBzLg0K PiANCj4gQm9iIFBvd25hbGwNCg0KDQoNCllvdSBzb3VuZCB2ZXJ5IGtub3dsZWRnZWFibGUgQm9i LCBob3cgZG8geW91IGluY3JlYXNlIHRoZSBmbG93IG9mIGNvbmR1Y3Rpb24gYmFuZD8NCg0KSSBo ZWFyZCBzb21lIHBlb3BsZSBjYW4gZG8gdGhpcy4NCg0KDQo=

Reply to
Richardson

section=20

I2,=20

Q2=20

if=20

source=20

question=20

link:

IIRC the 60 mV is a tradeoff between "emitter area" and "BE offset".

Reply to
JosephKK

Yes.

...Jim Thompson

--
|  James E.Thompson, P.E.                           |    mens     |
|  Analog Innovations, Inc.                         |     et      |
 Click to see the full signature
Reply to
Jim Thompson

section=20

I2,=20

and Q2=20

if=20

constant.

the=20

source=20

question=20

link:

May i make that: the 60 mV is the BE junction area and current density versus BE offset? For the same junction area a 10:1 current density provides a 60mV difference in junction voltage. Likewise, a 10:1 area difference will give a 10:1 current difference with no offset.

And finally, it should not have a easily detectable tempco without having thermal gradients across the device.

Reply to
JosephKK

60mV is for 10X, area OR density. 10X is usually picked for avoidance of offset issues, but I've done as little as 4X and as much as 100X (on a chip design).

Easy in the ASIC world.

...Jim Thompson

--
|  James E.Thompson, P.E.                           |    mens     |
|  Analog Innovations, Inc.                         |     et      |
 Click to see the full signature
Reply to
Jim Thompson

May i make that: the 60 mV is the BE junction area and current density versus BE offset? For the same junction area a 10:1 current density provides a 60mV difference in junction voltage. Likewise, a 10:1 area difference will give a 10:1 current difference with no offset.

And finally, it should not have a easily detectable tempco without having thermal gradients across the device.

But for equal area and a 10:1 current ratio you get a positive temperature coefficient, as in equations 1 and 2 of the Toronto U paper

formatting link

Reply to
Michael Robinson

temperature=20

OK. But does not the 10:1 current density cause a thermal gradient?

Reply to
JosephKK

Not enough to write home about ;-)

...Jim Thompson

--
|  James E.Thompson, P.E.                           |    mens     |
|  Analog Innovations, Inc.                         |     et      |
 Click to see the full signature
Reply to
Jim Thompson

Since we're on voltage references, I notice that LTC has a really nice CMOS bandgap--the LTC6655--with

Reply to
Phil Hobbs

I'll take a look after Christmas. Add an emergency repair event to the shopping scramble and I'm behind :-(

Skylight failed open... lever arms sliding in grooves failed. I managed to disassemble it 16' up on a ladder and close it just before the thunderstorm hit ;-)

.... 66ºF and pouring the rain ;-) ...Jim Thompson

--
| James E.Thompson, CTO                            |    mens     |
| Analog Innovations, Inc.                         |     et      |
 Click to see the full signature
Reply to
Jim Thompson

800 nV that is--still very good for a voltage reference. Merry Christmas, and may God bless you and your family even though you don't believe He exists. ;)

There's an old story told about Enrico Fermi--he had a horseshoe nailed above his front door. One of his colleagues said to him one day, "Enrico, surely you don't believe that ridiculous object is going to bring you good luck!" Fermi replied, "Of course not! But I am reliably informed that it will whether I believe in it or not."

Cheers

Phil Hobbs

Cheers

Phil Hobbs

Reply to
Phil Hobbs

But I'm a great traditionalist. Celebrate everyone's holidays... they all drink don't they ?:-)

Holidays are for children and grandchildren anyway!

Merry Christmas to all and to all a good night!

And we have a Hanukkah bush as well ;-)

Plus we've added a new event this year... "Meet the families" ;-)

Absolutely!

...Jim Thompson

--
| James E.Thompson, CTO                            |    mens     |
| Analog Innovations, Inc.                         |     et      |
 Click to see the full signature
Reply to
Jim Thompson

ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.