expert opinion needed on Raspberry Pi power issue

I've seen the other threads discussing this subject and I think that what you need is not so much an expert opinion as an informed opinion - from somebody at SMSC who really understands the chip.

My generalised expert opinion is that (to paraphrase form a 1970's Nat Semi data book) if you operate outside specifications then all bets are cancelled.

SMSC is in the process of being acquired by Microchip which might well eventually improve the availability of detailed information but the takeover is being held up by legal nonsense (IMHO it's nonsense) so right now I expect they are all worrying about their futures.

I can think of ways to measure the power dissipated by the SMSC chip but you would need a supply of hot and cold Pi's.

Why not try selsinork's mod on a hot Pi and see if it gets better.

Michael Kellett

Michael,

   Yes, an informed opinion is what's needed, and I have provided

a link to SMSC's tech support for just that reason.  But nobody has

taken that step yet, perhaps because the form asks for company

information, and maybe people don't want to involve their companies.

   The SMSC datasheet doesn't give any information on maximum

power draw from the VDD18CORE pins, so it's not so much a matter

of exceeding specifications as operating in a mode where no

specifications are provided.

   It is possible there are experts in this forum who have experience

using multiple LDOs in parallel.  Perhaps if the voltage thresholds

of both LDOs are similar, minor manufacturing differences might

account for which one has the lower threshold.

  Selsinork's mod removed RG1, (after measuring its current to be

extremely low), proving that the board would function using only the

1.8v regulator in the LAN chip.  This mod wouldn't be expected to

solve the problem of a hot LAN chip.  User Mackay did the opposite.

He bypassed the connection between the LAN chip and the 1.8v

power plane, forcing RG1 to do all the work (and measured the

resulting current at RG1 to be up to 80mA during startup).  This

mod would potentially solve the problem of a hot LAN chip, but

Mackay's LAN chip unfortunately wasn't hot to begin with.

Has anyone tried contacting Pete Lomas personally, at his company?  He probably wouldn't reply on behalf of the Foundation as the RPF non-engineers might consider this a political issue, but he might well reply personally just out of professional courtesy and decorum, in answer to a simple technical question.  (Especially if the question is phrased in a way designed to avoid any silly politics.)

During the short period when Pete was active on the forum, he was quite open and matter of fact about technical issues, very much an engineer, so I don't think it's a lost cause.  Yet, anyway.

Morgaine.

User Mackay did the opposite.

He bypassed the connection between the LAN chip and the 1.8v

power plane, forcing RG1 to do all the work (and measured the

resulting current at RG1 to be up to 80mA during startup).  This

mod would potentially solve the problem of a hot LAN chip, but

Mackay's LAN chip unfortunately wasn't hot to begin with.

Do you have a link to that ?  I couldn't find a way of easily seperating the lan9512's 1.8v from the rest of the board without removing the chip, so it would be interesting to see what was done.

Has anyone tried contacting Pete Lomas personally, at his company? 

No, but a personal approach like that is the only way I think can succeed.  The whole silly politics thing is also why I'm unlikely to bring it up on the other forum - as an engineer I'm crap at playing that game.

   Yes, an informed opinion is what's needed, and I have provided

a link to SMSC's tech support for just that reason.  But nobody has

taken that step yet, perhaps because the form asks for company

information, and maybe people don't want to involve their companies.

I've submitted a generic query to the smsc tech support web page. All I've asked for is clarification on whether the VDD18CORE can be used to power external devices and if so what the current capability is.

Lets see what reply I get, if any..

Mackay's link is here:

http://www.element14.com/community/thread/19236?start=14&tstart=14

I would swear that wasn't there before, don't know how I missed it when I already replied later in the same thread.. Thanks!

selsinork wrote:

 

I would swear that wasn't there before, don't know how I missed it when I already replied later in the same thread.. Thanks!

Some users have noticed threads playing hide-and-seek.  Maybe individual posts do so as well :-)

Reply from SMSC:

We have received your request for Technical Support.  The following reviews.

SMSC Technical Support Request # 534992-DD1256

 

31/07/2012 10:06 ISaturle: 

Hi Iain,

 

VDD18 is not intended to power external nets.  Equally, it's not intended to ever be driven by external sources. 

 

Seems clear enough.

Reply from SMSC:

We have received your request for Technical Support.  The following reviews.

SMSC Technical Support Request # 534992-DD1256

 

31/07/2012 10:06 ISaturle: 

Hi Iain,

 

VDD18 is not intended to power external nets.  Equally, it's not intended to ever be driven by external sources. 

 

Seems clear enough.

Selsinork,

>Seems clear enough.

Yes, perfectly clear! 

I hope SMSC adds that clarification to their datasheet.

Maybe someone from element14 can comment.

Otherwise, I think now you have a good basis for

contacting element14's tech support.