Skip to main content ↑ ↓
Topic: Possible issue with voltage regulation on some Bus Pirates (Read 8658 times) previous topic - next topic

Possible issue with voltage regulation on some Bus Pirates

Just a quick heads up... after a post about power supply noise on a SparkFun Bus Pirate, I thought I'd check my SeeedStudio one as well. It exhibits a similar issue.

I have never had a problem with this noise before, but it might be worth bearing in mind if it matters to you, or if you are getting strange results.

http://cybergibbons.wordpress.com/2013/ ... illations/

Re: Possible issue with voltage regulation on some Bus Pirat

Reply #1 –
[quote author="cybergibbons"]Just a quick heads up... after a post about power supply noise on a SparkFun Bus Pirate, I thought I'd check my SeeedStudio one as well. It exhibits a similar issue.

I have never had a problem with this noise before, but it might be worth bearing in mind if it matters to you, or if you are getting strange results.

http://cybergibbons.wordpress.com/2013/ ... illations/[/quote]

Thanks for the investigation, we'll look into it more closely, and try to fix the problem ASAP.
best regards FIlip.

Re: Possible issue with voltage regulation on some Bus Pirat

Reply #2 –
For what it's worth...

A comment from Ned Konz on June 4, 2010 at 3:18 pm at http://dangerousprototypes.com/2010/04/ ... irate-v3b/
Quote
I think that ceramic output caps may be a bad choice for the MIC5205 family because of their low ESR. [...] Ultra-low-ESR capacitors can cause a low amplitude oscillation on the output and/or underdamped transient response.

This February 24th, 2013 post shows that the Sparkfun.com version of bus pirate is indeed oscillating.
http://mightyohm.com/blog/2013/02/simpl ... illations/

Does the "official" BP as currently shipped have a fix for this ?

Re: Possible issue with voltage regulation on some Bus Pirat

Reply #3 –
AFAIK there is no DP version of the buspirate with ceramic capacitors in the wild. We made some prototypes with it but exactly that comment made the prototype never leave the workshop. Perhaps Ian knows for sure.

Sparkfun did design their own version of the buspirate and the productpictures suggest they are using ceramic capacitors.

@cybergibbons: did you also check the input voltage (Vusb?) I only see the output traces on your blog. Could you also post the markings of the vregs here?

I'll also check my two buspirates tonight.

Re: Possible issue with voltage regulation on some Bus Pirat

Reply #4 –
Sjaak,

Yes - I checked the Vusb as soon as I saw the issue, suspecting either the laptop or noise being picked up from something else on my desk. It's very clean - I'll expand the post tonight.

I'll also look at the regulator markings.

Re: Possible issue with voltage regulation on some Bus Pirat

Reply #5 –
I have been looking into this using my V3a pirate, there is no doubt in my mind this can be improved with better filter caps. i think i have isolated the source of the noise too, time is short right now but ill update after more testing.

Re: Possible issue with voltage regulation on some Bus Pirat

Reply #6 –
VR2 is marked LB33
VR3 is marked LB33
VR4 is marked KB50

I was double checking the USB power (on an additional laptop and a couple of USB power supplies) just now and I have noticed something quite curious...

VR2, the regulator that powers the entire board, exhibits little to no noise when initially plugged into the board.

Start minicom and connect, and the noise starts.

Close minicom and the noise stops.

I've just double checked using a couple of other devices and I can't observe the same effect.

Re: Possible issue with voltage regulation on some Bus Pirat

Reply #7 –
the usb lines are very close to the main regulator and uneven length and spacing. the noise on the power is time correlated to usb packets, between usb packets the noise is magnitudes better. a layout revision is needed to minimize the capacitive coupling between power and usb.

the usb signals are still far from clean, I recommend using only a usb cable with ferite cores at each end.

I have added a 100nf ceramic cap parallel to C21 and c23, this seems to drop the noise down quite a bit.

C22 and C24 are still tantalum, I have not changed them yet but noise is down on the respective rails.

Re: Possible issue with voltage regulation on some Bus Pirat

Reply #8 –
I would have thought that power rail noise correlated to USB packets would be mostly because of current drawn from the power supply to drive the USB signaling voltage by the USB transceiver block, rather than just capacitive coupling? But I have not looked at the circuit myself.

Re: Possible issue with voltage regulation on some Bus Pirat

Reply #9 –
[quote author="jbeale"]I would have thought that power rail noise correlated to USB packets would be mostly because of current drawn from the power supply to drive the USB signaling voltage by the USB transceiver block, rather than just capacitive coupling? But I have not looked at the circuit myself.[/quote]

probably contributing to some of the noise.

the ft232 d+ and d- pins are backwards imho. the usb traces on the board are horrible, it is such a small area to get a good differential pair layout out. one is 2x the length of the other on V3a, it is better on V3b and V3.5 but one trace is still on the bottom.

V4 could use some clean up around the usb lines too, but much better than previous versions.

Re: Possible issue with voltage regulation on some Bus Pirat

Reply #10 –
Hi, up to v3.5 we used Vias with the D+ D- crosing each others' paths (uunderneathe) ..
In V3,6 we fixed this with having one trace go around the SSOP pins... This makes one longer then the other, but there are no Vias on the Data traces, and no path crossing...

Do you thing we should revert to 3.5 layeout (Via's and path crossing), or keep 3.6 (longer lenght), which will produce less usb noise in the power supply.
best regards FIlip.

Re: Possible issue with voltage regulation on some Bus Pirat

Reply #11 –
[quote author="arakis"]Hi, up to v3.5 we used Vias with the D+ D- crosing each others' paths (uunderneathe) ..
In V3,6 we fixed this with having one trace go around the SSOP pins... This makes one longer then the other, but there are no Vias on the Data traces, and no path crossing...

Do you thing we should revert to 3.5 layeout (Via's and path crossing), or keep 3.6 (longer lenght), which will produce less usb noise in the power supply.[/quote]

The V3.6 has a little more board room from the switch to the SOB board size. keeping same component count but isolating the ground around the connector routing bolth usb lines under the ic, and using the internal regulator to power the ftdi chip would also help isolate noise from the rest of the circuit.

I have a few free hours today, ill download the board files and see what can be done without major changes.

Re: Possible issue with voltage regulation on some Bus Pirat

Reply #12 –
should i start with V3.6 or V3.8 from svn?

Re: Possible issue with voltage regulation on some Bus Pirat

Reply #13 –
Hi sqkybeaver, please use v3.6a-SSOP, its the current production and v3.8 is not yet final

Re: Possible issue with voltage regulation on some Bus Pirat

Reply #14 –
[quote author="vimark"]Hi sqkybeaver, please use v3.6a-SSOP, its the current production and v3.8 is not yet final[/quote]

OK