Earlier today, I saw that Jeff Keyzer (aka MightyOhm) had posted on his blog about power supply oscillations on the Bus Pirate. The Bus Pirate is one of my go-to tools, and 99% of the time it works as I’d hope, so I thought I’d check mine for issues.
Jeff’s SparkFun built Bus Pirate was showing a 15kHz, 150mVpp signal on the 3.3V power supply rail. Surely I would have noticed if mine was doing this?
Ian at Dangerous Prototypes replied to Jeff on twitter:
@mightyohm we’ve had reports about that on the sfe bus pirate. We use a differentand have not seen it on ours. Will pass your post to sfe
— Dangerous Prototypes (@dangerousproto) February 26, 2013
So I took a quick look at mine, hoping it wouldn’t be the same. I have a v3 Bus Pirate ordered from Seeed Studio many years ago.
It wasn’t the same, but it also exhibited an issue.
Turning on the 3.3V power supply gave me the following trace:
So I have fairly strong bursts of noise, sometimes up to 400mVpp! These seem to repeat at about 150kHz.
Zooming in on one of these bursts:
It looks like something is ringing at about 20MHz.
It was a very similar story on the 5V rail.
So the quick fix is to whack a small electrolytic capacitor in:
I did this for the 3.3V regulator. As soon as I had, the severe noise and ringing on the 3.3V rail stopped.
Interestingly, the noise on the 5V rail stopped as well…
I thought I was going loopy. Yet desoldering the cap causes the noise on both to return.
So what is going on here? Are two regulators oscillating and interacting in a bizarre manner? Who knows. I’m just going to put a second cap on the 5V regulator as well.
Does this make the Bus Pirate worthless? No. I doubt the noise would cause problems for most people. I doubt the noise is as serious when load is placed on the regulators. The fix is incredibly quick and easy. It’s still an incredible tool.
Ian
March 6, 2013 at 1:02pmThanks for the great investigation and writeup. I pulled all my Bus Pirate production models and tested them. Take a look here:
http://dangerousprototypes.com/2013/03/06/possible-issue-with-voltage-regulation-on-some-bus-pirates/
I’m really curious about this case because this version does have the 2.2uF tantalum capacitors recommended in the datasheet of the MIC5205 (though as I recall Seeed and I worked to find an alternative even for this version that is ok with ceramics).
cybergibbons
March 6, 2013 at 1:11pmHi Ian – thanks for your investigation. I’m really intrigued by my situation here – the oscillations seem entirely related to serial communications. If I am not connected to the board, I don’t see the issue.
I’m suspecting the protocol to the FTDI chip is causing it but need to look into it further – work has been rather busy for the last week but I should have time at the weekend.