A couple of months ago I joined the element14 community and decided to sign up for a Road Test which I did, for a Keithley 2110 DMM, I was not really expecting to be selected but I was which was great – so thank you element14 for the meter and the opportunity to do this review. Now while I am not planning to turn my blog into a review site, I figure that the odd review on new kit might be nice, especially if there is the opportunity to tear it down after the review – in fact I will not do any review that does not involve tearing down the kit to have a look inside.
I have specifically made this a video review so I wont write too much about the review here, you can watch the video for the detail (and the puppies too).
In summary though, I hated the display on first site and that made me want to dislike the meter too. But apart from the display, everything else about the meter, its performance, its build quality and apparent accuracy (so far as I can tell) is absolutely first class. If I had a wish list – or more to the point, if I were in charge of sales at Keithley this is what I would have done differently…
- Design in a decent display, anything other than what is in there – a nice VFD, or an OLED perhaps, or even a nicer LCD if you have to
- Ship a thermocouple
- Make the continuity beeper useful – its way too slow response-wise
- Write the software so it can be used on Windows, Linux and OSX – it really is not that hard to do
- Did I say a better display?
Lui Gough also reviewed the meter and has done a really nice write-up for his review, he also tested the software and data collection aspects of the meter, you can see his review right here: Lui Gough Keithley 2110 Review
Other Links
Coming up next, a teardown of the Keithley 2110….
This content is published under the Attribution-Noncommercial-Share Alike 3.0 Unported license.
With your Mac simply download and install Oracle’s Virtualbox, then you can install a version of Windows ( or any other OS you want, even the same as the outer machine) and then you can connect the USB devices and have them handled by the guest OS in the VM with pretty much no problems. That way you can just have a copy to install the software in, and if it gives problems just wipe and use another copy and it will be new again.
Hi Sean, yeah I had a look and it seems straight forward, I will give it a try. Gerry
Hi Gerry, wondered if you might be able to throw some light on a problem I have with my 2110?
I wanted to test how well it works with high speed measurements. I decided to measure a 1KHz signal going in and take 1000 measurements – all being written to the internal store of the device.
The manual says to set the NPLC to 0.001 for a really low integration time – 20uS integration. It also states in the manual it is possible to take 50,000 readings per second so I think it should work.
Well I assumed I would end up with a set of results that would represent a square wave. But it didn’t come out that way, it came out as a stable average voltage of 50% the peak square wave voltage. I do not know if the ADC is holding charge but can anyone help with setting the DMM up to make this type of measurement work? Maybe my understanding of the meter’s performance is wrong?
My end goal is to measure uA current of a processor and measure changes of current over a few 100ms so I need to know the measurements are good and working. The 1KHz voltage signal is my way of checking I can set the DMM up properly and calibrate how fast the ADC converter tracks changes in input.
Really appreciate any help.
Thanks
Trev
Hi Trevor, I am not sure to be honest but at a guess I would be pretty certain that the frequency measurement is not being done via the ADC. I suspect there is some simple waveform shaping to get a square wave and thats being fed into a pin on the micro controller. If you think about it, the meter is capable of measuring up to 300Khz, when I checked it, it seemed to be good up to 550Khz so there is no way an ADC, even at 50,000 samples per second could do that. So if the micro controller is counting the frequency its almost certainly using some kind of gate and count scheme and will not be capable of anything like 50,000 reads per second. Anyways, I suspect that your problem is the input bandwidth on the meter, its probably averaging out the square wave to get the average DC voltage, think PWM into low-pass filter. If the voltage/current you want to measure is smaller fluctuations then you will get better results, but I dont think you are likely to get the meter to output a series of numbers that look like a sine wave, This is all guesswork of course, I am happy to be corrected.
Try the same thing at say 50Hz and see what you get, then double the frequency and try again, you should start to see where its rolling off.
Gerry
Hi. I wasn’t very clear. I am measuring the voltage of a 1khz signal. Using the signal to see how many readings I can get but I don’t get anything but an average.
Hi Trevor, yeah I understood what you were trying to achieve but I suspect that the meter will not have that fast a response time. As I suggested try a low frequency first, it will probably do what you expect at 50Hz, then you can increase the frequency to see where it rolls off. Does that make sense?
Gerry
Hi again. Well the manual states integration times of 20us and the ability to take up to 50000 readings per second. I don’t think it’s pushing it to measure that fast. I can’t see where it states it can not be done. I just can’t figure out to configure it properly.
The one feature I liked about this meter was super accurate measurements with an internal 2000 buffer store.
Not convinced by the display either. No idea why they didn’t add a contrast at least.
Hi Gerry
As a follow up to this I did put a message on Keithley’s forum. I have linked this in below. They do say it is possible but at present the results are not great. As you suggested I could wind down the frequency and find a place where it starts to look good but it seems this is an interesting point about this meter. It can do 50,000 readings/s into internal memory. This is a speed figure only as it can only store 2000 readings. But also there must be some other filters in there because I am not currently sure of what range it can work to. Maybe it can only measure small changes in signal at this higher rate.
Anyway, here is the link to the forum posts in case it is of interest to you or anyone else that reads this.
Thanks for your time with this so far
Trev
http://forum.keithley.com/phpBB3/viewtopic.php?f=32&t=89040&p=142067&sid=7aebc9c416aca0cdac6260a1e32c80fd#p142067
Thanks for posting that link, could well be useful to others. Gerry
About Linux/MacOS and the Keithley 2110. The USB interface supports USB-TMC which is a standard way to talk to test and measurement devices. The commands are documented in the manual and there is a module to drive it that ships in the linux kernel. It is probably also in libusb so if you know how to get libusb working on a mac then you can use it from there.
That will let you command and read from the meter via usb. Sadly I have yet to find a nice UI to use for it. There is a python script (but i also hate python).
1. They seem to have killed off the teared pricing that made the USB only version about $150 cheaper.
2. I prefer the LCD they have to VFDs’ which tend to age badly.
3. I wish the math function let you have std deviation when you do average readings.
Hi Evan, thanks for the info and comments. Gerry