Circuit insight anyone?

[s3c]

Was bored today so I opened a dongle I had laying around, I've attached a pic and a circuit diagram I drew up for it. It only contains an I2C eeprom and a couple of passives but it's pretty smart, it uses a diode and capacitor to power the eeprom from the eeprom clock line.

Anyway, anyone have an idea what the resistors R1 and R4 are for? I'm assuming PIN1 and PIN2 are connected to Vcc through external resistors and are connected as SDA and SCL for normal I2C operation.

[ian]

Parasitically powered I2C EEPROM. Interesting. Have you dumped it yet :)

It seems like 10K (R4) would be counter productive and fight the I2C pull-up.

R1 is probably to drain the cap to avoid accidental writes/etc after it's disconnected?

I couldn't find anything like this via search, have you made the circuit on a bread board to test it? I might have to test it myself :)

[s3c]

Well I know the circuit works since I use the dongle every day, figured the same after thinking about R1 for a bit. I suspect R4 is to overcome the zener capacitance and prevent it from holding the clock line high longer than desired, agree/disagree?

Haven't dumped it yet but that's why I put it in the Bus Pirate forum :)

[Sjaak]

hehe. I'm just curious where do you use it for?

I would expect that r4 was in series to protect (together with the zener) the input voltage. I still think it is for protecting the input.

[s3c]

Don't actually know for sure that it's a zener but what would a normal reverse biased diode be doing in that part of the circuit? As for the function, I'll tell you when I've done my writeup :)

[EDIT]

The zener isn't in series with R4 so it's not helping with the regulation but that could be done by the external pullups so R4 remains a mystery. If you keep PIN2 in mind it's connected to the circuit without a voltage clamp so obviously the voltage required is readily available, why the zener on PIN1 then?

[ian]

Here's a png of PDF schematic. I'm going to post this to the blog and offer a PCB for additional insight.

[s3c]

Thanks Ian, mind if I use the PIC for the blog post?

[ian]

Sure, no problem.

[Sjaak]

Don't actually know for sure that it's a zener but what would a normal reverse biased diode be doing in that part of the circuit? As for the function, I'll tell you when I've done my writeup :)

[EDIT]

The zener isn't in series with R4 so it's not helping with the regulation but that could be done by the external pullups so R4 remains a mystery. If you keep PIN2 in mind it's connected to the circuit without a voltage clamp so obviously the voltage required is readily available, why the zener on PIN1 then?

I could be just to assure the supply voltage is stable (kinda protection). the other input could be protected with normal (internal) clamping diodes. For writing you want to have a stabilized voltage. I have a datasheet  which mentions 2mA when writing an 0.2mA when reading. I suspect this will give voltage drops.

[LeissKG]

I have seen similar circuits with a variable input voltage. At lower voltages you have a resistive voltage divider. If the voltage becomes to high for the device you limit the voltage with a zener diode. What is missing here is an external resistor for this voltage divider.

Klaus Leiss

[DTR2]

After simulation with LTSpice i found that the circuit sinks about 600ma assuming a 5v source and 4.7v zener in the inverse clock cycle.

Im not very clued up with the analog side of digital circuits but it can be any of the following.

• Stability / Filter issue with eeprom clock due to power circuit sinking current in opposite cycle -Excess current resulting in timing lag
• Some kind of security check/feature on host side to see if someone is "leeching" on line

[DTR2]

I forgot to add current and voltage annotations
Clock voltage is green
Diode current is blue

[s3c]

Dumping the EEPROM is giving me some problems, I'm assuming it's because my clock speed is too low and the circuit drains the stored power and resets before the communication is finished.

[ian]

Solder a fly-wire onto the power pin and power it with the bus pirate :)

[Sjaak]

You could try to desolder the protection resistors and connecting there or use the bp as sniffer

(or what ian says.. I'm not fast with replying today ;))