Showing posts with label Keithley. Show all posts
Showing posts with label Keithley. Show all posts

Sunday, December 18, 2016

Low Cost Bi-Directional Level Shift Module Characterization (TE291)

-----
From time to time the level shift module pictured above has come in handy.  They do a great job shifting 3.3VDC to 5VDC logic or 5VDC logic to 3.3VDC logic.  The cost is about $1USD and hookup is simple.  Of course they are designed for low speed digital signals but, we wondered how the module would handle higher speeds.
----
The Bench Setup:
Keithley 2230G-30-1 power supply for the 3.3VDC and 5VDC power.









Tektronix AFG3252C function generator for the square wave (3.3V and 5V logic) stimulus.










Tektronix MDO4104C oscilloscope to capture the input/output signals.

-----
The Result:
Performance was very good, especially considering these module are often used in the 100KHz and below range.  Leveling from 5V to 3.3V had better results.  The signals start looking ridiculous over 1MHz. Take a look at the scope shots below.
-----
3.3V Level Shifted to 5V at 10KHz, 100KHz, 500KHz, and 1MHz
(Yellow = Input; Blue = Output)




-----
 5V Level Shifted to 3.3V at 10KHz, 100KHz, 500KHz, and 1MHz
(Yellow = Input; Blue = Output)




 -----

Sunday, April 10, 2016

Graphing Current Drain of the Sparkfun BadgerHack

-----
Sparkfun had a nice booth at SXSW in Austin that let you build and walk away with their BadgerHack project.  Very generous and Thanks!  Also at the booth were some of the 'moviestars' often seen on their YouTube channel.  It was great to meet a few of them.

I was curious about the current drain of all those LEDs dancing on the display.  I had access to a Keithley 2461 SourceMeter/SMU which makes it easy to run the experiment.  The video below tells it all. 

-----
Thanks again, Sparkfun!

Wednesday, January 13, 2016

5VDC Boost Converter Characterization

 -----
Boost Converters are a type of switch power supply that magically turns a small DC voltage into larger DC voltage.  Learn about how that magic is preformed here.   Used properly they are great, but the magic of a boost converter does have it's limits.
----- 
For the measurement setup a Keithley 2230 power supply is set to 1.5VDC (think AA battery voltage) to act as V(in) for the boost converter.  A Keithley DMM7510 HiRes multimeter is used to measure the ripple on V(out).  A Keithley 2450 SourceMeter SMU is used to simulate a load on the boost converter in static and swept mode. The Keithley 2450 also produces a V(out) graph under the swept load.

-----
The video shows how the 5VDC boost converter starts performing poorly at loads close to 100mA.  The output voltage begins to drop and ripple frequency increases as the boost converter works harder under the higher loads.  The Keithley 2450 is set up to graph the boost converter voltage output as it simulates the swept load.
-----
[EDIT: February 20, 2016]: We got a comment that the boost performance could have been effected by not having a constant 1.5VDC input into the converter during the entire load sweep due to lead resistance.  Turns out the comment was correct.  We retested the same boost converter with 4-wire measurement connections at both the INPUT and OUTPUT.  This 4-wire setup reduces any errors due to lead resistance.  The video below shows the improved result.  Measurement setup is everything!!!
 -----

Saturday, August 22, 2015

Keithley DMM7510 monitors Charger Doctor Output



The Charger Doctor is a sub $2 device that displays voltage and current levels while charging a USB device.  There have been several teardown write ups on the Charger Doctor.  It's useful, but does have some limitations.  For example, it doesn't charge iOS devices and it seems to effect the reliability of the WiFi connection on our Raspberry PIs.  This made us want to take a closer look at what the output was doing.

Here we use a Keithley DMM7510 to monitor the Charger Doctor output.  The DMM7510 is a precision instrument with 7.5 digits of resolution and 14 PPM accuracy.  It has a touch screen interface with graphical output, advanced triggering options, cursors, and a whole lot more.  Plus, it is fast (very very fast); capable of sampling at 1M/s.  To sum it up, it's pretty darn cool!
-----


In the video the DMM7510 shows a decrease in voltage when the Charger Doctor displays current.  Also, some instability when displaying voltage.  But remember, the Keithley DMM7510 is a very sensitive instrument.  The change in voltage looks big on the screen, but it is really very small.
-----
The overall voltage swing is only 7.536mV or 0.00753 Volts.  That's not a lot.
-----
And taking a closer look at that instability when the Charger Doctor displays voltage we see that it is only 1.256mv peak to peak; that's just 0.001256 Volts.  Really, that seems exceptional for a sub $2USD device.
-----
There is a lot of discussion on the web as to why iOS devices don't charge via the Charger Doctor.  With the stability shown, I'm not sure why the Charger Doctor effects the WiFi reliability of the Raspberry PI, but I can say it does.  We set the Keithley DMM7510 to camp on the Charger Doctor output signal and trigger on a larger glitch or voltage spike to give us a clue.  After about 15 minutes the Charger Doctor remained stable within the trigger limits and we moved on with our life.  We do have to take into account that the Charger Doctor was effectively unloaded while making these measurements and that could make a big difference.
-----
Thanks for stopping by.
-----

Monday, September 15, 2014

Testing the CREE 5000 Lumen XML U2 LED HeadLight

The CREE 5000 Lumen XML U2 LED HeadLight was a topic of discussion at Ride Dual Sport.  The product seemed well built from the pics and at less than $18 shipped it seemed like a must try.

The unit comes complete with a the LED light, rechargeable battery (with a nice nylon case), handle bar mount, and charger.


The light feels solid and well built.


----
So how does it perform?  First off; it's bright.  Really bright!!!  There are three brightness setting, but honestly, the dimmest seems plenty bright.  I ran the light from full charge to dead on LOW and HIGH.  On LOW a 4.5 hour run time was observed.  A Keithley 2110 5 1/2 digit precision digital multimeter was used to record current draw.


-----
The battery pack puts out 8VDC, so don't connect the headlight straight to a 12VDC motorcycle or car battery without spending 60 cents for 7808 regulation.  


-----
Current shown by the Keithley 2110 with the unit in STANDBY mode; ~50 hours estimated.



-----
Current shown by the Keithley 2110 with the unit in LOW mode; 4.5 hours observed.


 -----
Current shown by the Keithley 2110 with the unit in MEDIUM mode; ~2 hours estimated.  



-----
Current shown by the Keithley 2110 with the unit in HIGH mode; 1.5 hours observed.

 -----
The unit is built well, works great, and is a fantastic value.