Raspberry Pi - Hardware Flaws and Fixups?

"On the 'pi we can reduce the 1.7V to about a milivolt by adding a 1 Ohm resistor between 5V and ground. 25W of extra power during normal operation, but not good enough to satisfy the microchip datasheet spec."

I asume you are joking right ?

The voltages are no "parasitic", just product of not good design IMHO.

Like asuming that cellphone chargers are a good option as a power supply for the Rpi.

-J

What I'm trying to say is that if you TRY to design a raspberry pi-with an MCP23017 and would like the MCP23017 to power-on-reset, you HAVE to reduce the powerline to ZERO to get any type of guarantee from microchip that the chip will perform the reset.

So even if I were to install a 1 ohm "bleeder" resistor to ground, I'm guessing the voltage on the powerlines with no powersupply but with an HDMI screen connected will be about 1 mV: There is a 1800:1 resistordividor between a 5V and the raspberry pi 5V. It can be 2 or 4 mV for all I care. But that won't make the board achieve the microchip specs and it WOULD increase power consumption during normal operation tenfold.

In short, the microchip specification for this chip is unworkable. Bad specifications.

The problem probably lies in the SPECIFICATIONS. Not in the chip.

And then we'd be using a "typical" value from the datasheet, so in practice, microchip tells you these chips will perform a "power on reset" the first time from the factory, and all bets are off after that....

It's not easy to write good specifications. It's not all that hard either. Broadcom are better at it than Microchip. The broadcom datasheet for the BCM2835 nicely states for unimplemented bits: write zero to allow for upgrades to the chip from our side. Those are the things that make a datasheet worthwhile.

I remember from a long time ago, datasheets that used to mention: 'minimum hold time' for an output. They specified that the chip wouldn't change the output for XXX ns after the clock. This means that they have themselves in trouble: Suppose they can make these chips 2x faster. Alas, they cannot sell them under that datasheet: they won't meet the minimum hold time. And their customers are in trouble. Suppose a design trusts the minimum hold time. But 10 years later the chip needs to be replaced and is replaced by a faster chip, possibly from another manufacturer. Big trouble.

Bad specifications cause trouble. This part of the microchip MCP23017 datasheet can be considered such a bad specification.

This is a bit of a non-sequitur I know, but for some reason your mention of MCP23017 made me think of how much more massively expandible Pi would be if the CSI/DSI board area were replaced by the cheapest Cortex-M0 and some more headers.

Morgaine.

That wouldn't pay for EU bonus at Broadcom

Knowing how to tell your company to sod off without losing your bonus is an essential engineering skill.

About the temperature issues, I can't remember now if there was a different thread for it.

There is a great article (in spanish, google translate is your friend) with very cool thermal images of the board operating in different modes.

It confirms what I observed with the IR gun that the SMSC chip gets hotter than the BCM+Mem.

http://www.geektopia.es/es/technology/2012/06/22/articulos/se-calienta-el-ordenador-raspberry-pi-estudio-de-sus-temperaturas-en-funcionamiento.html

-J

I'm not so sure that the power line has to be reduced to zero to be assured of a POR.  Whilst they state "VDD Start Voltage to Ensure Powe-on Reset " = Vss they also state this hasn't been 100% tested, the Vdd supply has to be above 1v8 for the device to work, and as Vil = 0.2 * Vdd = 1v1 @ 5v5 an assumption could be made that anything below 1v1 is considered as 0 < Vss <1v1.

True they don't say this but so long as Vdd rises no faster than 90ms from zero to 1v8 a reset ought to occur or 35ms. from 1v1 -> 1v8.

Oh, sure, you can make assumptions on how YOU think the device works  and then draw your conclusions from that. But why do you have a datasheet then when you're going to assume specifications anyway?

The datasheet is there to

* Provide workable design guidelines for the designers.

* provide testable parameters for production testing.

* form a sort of a "contract" between a chip producer and a PCB designer.

Suppose a datasheet says:

V_in_low           max 0V

So this is a bit better than in the microchip datasheet: we have a testable limit. But it is unworkable for a designer.

Suppose the line is connected to a pullup. Say a 1k8 pullup to 5V.

Suppose I have an AVR (which has one of the biefiest output drivers around!) that drives the pin low. The output resistance is about 25 ohms. The voltage will end up at 68mV.

You can start guessing that most chips will accept a voltage below about 0.5 VCC as low, with the absolute lowest value that might be recognized as  a 1 being 0.3VCC. Sure. You're probably right. But WHY is there such a weird spec in the datasheet? Why don't they just say:

Maybe they made a mistake, and by documenting it in the datasheet: "its a feature, not a bug". Now if someone gets "bitten" by this, they get to say: Did you read the datasheet?. And they get to fix it at their own leasure, while producing and selling the chips.

Of course, this is not desirable if the "bug" is big. But if it is something that might effect less than 5% of the customers? They will surely try to sell the chips.

Different field. Same thing. I fly paragliders for a hobby.

They come in different sizes. For example "small", "medium" and "large", for different weights of people. So each has a weight-range. These things are tested by an independent body for their safety. But only within the weightrange given by the manufacturer.  So a small will be for 65-85kg, the medium 80-100 and the large 95-120 or something like that. The safety test has been performed for the medium at 100kg, and it COULD in theory be a dangerous mean *** at 101 kg. Not likely, but the tests don't give any guarantees outside the official weightrange.  All the weightrange numbers are round numbers, 5kg overlap between sizes.

Now suddenly a new model is announced. The weightranges end up having just 1kg or even 0kg overlap, and the ranges are suddenly values like 82 - 97 kg. That makes you wonder why they didn't just make that 80-100? Did the glider fail the tests at 81 and 98 kg? Nowadays I know enough about these things that I'm willing to be a "test pilot" for these things. But that's just me 99% of the people should stick with the specifications.

So, back to the chips. Given the specifications you cannot rely on the power-on-reset. Maybe it will do the power-on-reset, maybe not. The device is probably usable, but the values mentioned as "power on values" should not be trusted. Initialize everything you want in a certain way.

So I guess it is official that Pete has been abducted by some obscure organization to steal all the secrets of the Raspberry Pi ....

I expect he's just busy.  Norcott Technologies does a lot more than just Raspberry Pi after all.

It sure was refreshing to have him here briefly for rational technical discussions though, free of Liz's personal dictatorship and vindictive bans.  I hope he returns when he has time.

Morgaine.