Christian,
Thank you for this data. We should get together in San Diego to discuss the data I took and this data, and figure out a way forward for these critical components.
The data does show something we noticed as well. As the voltage/current goes up, the curves bend due to heating which causes the threshold to go down. Given that the current simulations show very large current imbalance, this will cause the diodes with the most current to heat up, lower their threshold, and create even more imbalance.
I may be able to show simulations on how this will affect current imbalance in San Deigo
Regards,
Jeff Heath |
|
paper: | 402 East Carrillo Street, Suite D |
Santa Barbara, California 93101 | |
voice: | 805.965.6400 |
fax: | 805.965.1701 |
computer: | |
From: Christian BEIA [mailto:christian.beia@xxxxxx]
Sent: Wednesday, July 02, 2014 5:44 AM
To: Jeff Heath; STDS-802-3-4PPOE@xxxxxxxxxxxxxxxxx
Subject: RE: [802.3_4PPOE] Diode mismatch values
Hi Jeff, all,
The diode data were provided by me.
They come from production data, so they consider the spread among different lots of production.
I also derived the linearized model, which is not coming from a measurement, but the linearization of the worst case operating points at 0.5A
This is the relevant slide of beia_1_0514.pdf.
Please let me know if you have any question
Best regards
Christian
From: Jeff Heath [mailto:jheath@xxxxxxxxxx]
Sent: Wednesday, July 02, 2014 1:05 AM
To: STDS-802-3-4PPOE@xxxxxxxxxxxxxxxxx
Subject: Re: [802.3_4PPOE] Diode mismatch values
Sorry for my typo. My lab data showed only a 0.030V Vt mismatch, not a 0.30V mismatch…
Regards,
Jeff Heath |
|
paper: | 402 East Carrillo Street, Suite D |
Santa Barbara, California 93101 | |
voice: | 805.965.6400 |
fax: | 805.965.1701 |
computer: | |
From: Jeff Heath [mailto:jheath@xxxxxxxxxx]
Sent: Tuesday, July 01, 2014 4:00 PM
To: 'STDS-802-3-4PPOE@xxxxxxxxxxxxxxxxx'
Subject: Diode mismatch values
Yair,
The diode mismatch parameters I your presentation show a wide variation of threshold (.39V vs .53V) and a zero difference in resistance (0.25 ohms vs 0.25 ohms).
I have included my previous lab data on just six diodes which shows just the opposite: a threshold that is very close (Varying by only 0.30V) and a resistance that is wide (0.28 ohms vs 0.40 ohms).
Given that in the short channel case, the threshold offset is the major contributor to offset currents, I think we should change the values for the diode mismatch.
Regards,
Jeff Heath |
|
paper: | 402 East Carrillo Street, Suite D |
Santa Barbara, California 93101 | |
voice: | 805.965.6400 |
fax: | 805.965.1701 |
computer: | |