
OK folks, behind the scenes I have been compiling accuracy data for various measuring devices have in my possession. It's my nerdy indulgence!
Most are cheap-o meters, inspired by the KOT channel content!
Recently in my reviews you have seen logarithmic "error charts" which are an effort to best communicate how well certain measures are performed by a meter. I recently noticed others presenting data the same way. Not that I invented them or anything - they are just really effective and make a lot of sense.
Voltages: "The Collection"
This first graph of voltage is definitely overwhelming, but it gives the big picture. I measured voltages from 15 mV to 48V in even logarithmic steps. Each measurement is about 3X the previous one.
The values below 0.5 V are demonstrating lack of precision. Usually a meter's ability to measure a low value is constricted by it's number of available digits for it's lowest range. This is compounded by the number of counts it is able to resolve.
The values above 0.5 V vary mostly because of accuracy. Many of these meters do not offer user calibration.

Voltages: "The Meters"
Here's the least cheap-O of the devices, and it shows. It doesn't have a mV range, but it makes the most use of it's digits in the voltage range:

This meter looks like a phone and has a nice display. As you can see, it's all flash and no cash:

Here's a vintage pocket meter which can be user calibrated. Fortunately the adjustment I did around 5V is good for most of the range. I was surprised by it's linearity despite being an older meter:

Now I'll start clumping some meters together. The following graph is of super-cheap meters which I paid under $5 for, or I forget what I paid and they fall into the class. Be aware I did adjust the voltage calibration potentiometer at 5V not too long ago, this is an improvement to how I received the meters. You can see the accuracy tends to vary due to bargain basement circuitry. The best of the bunch is the Harbor Freight Cen-Tech:

I had some fun with the meters above, changing the LED backlight to match the boot and also extending the backlight from a few seconds to a minute or two:

The following are a couple of those black meters that you can get at AliExpress for $1.79 from time to time. They don't come with an adjusting potentiometer, but it's easy to add one (how easy depends on the PCB variant you get). I converted one of them to a 4-wire milli-ohmmeter, but the voltage ranges are still in-tact. More about that in another episode:

Probably the most frustrating thing about these cheap multimeters with a range selector is how wonky the selector works, despite careful cleaning. On all but the Cen-Tech, you can wiggle the switch back and forth on it's centered position and the reading will change a few counts or go completely out of bounds. Doesn't inspire confidence, thus I don't recommend them. On better meters, this never happens.
The AN870 is an outstanding value at just over $20. The range, accuracy, and precision is hard to match at this price point. I had to get a second one:

Another very good value is the AN8009, under $20 when on special. I consider it to be the AN870's little brother:

As you can see, for some meters I purchased copies of some meters with outstanding value for the purpose of these kind of comparisons.
The next meter has been around for a while, and goes for $30 when specials roll around. It's a chunky meter, 25,000 counts. It has a few tricks up it's sleeve, including simultaneous AC and DC voltage on the display. However, be aware that these measures are not as accurate in the dual-display mode. This graph shows the discrete DCV range setting.

Here's a bonus graph. It demonstrates the additional error that is present the AC-DC mode:

Here's a couple close-cousins. These cost just over $5. Sometimes meters look very different but perform similarly due to their similar architecture:

When it comes to auto-ranging meters like the SZ02, I put it into the manual range. This meter is very nice considering I paid under $10 for it. Just wished the buttons weren't so tall:

I recently reviewed the Tesmen supplied to the channel. It did fine at voltage:

These little guys were about $4 each when sold on the "pick 3" during this past year or so. They vary quite a bit across their different functions. Voltage is no exception. I tend to keep them around for continuity measurements or kicking around in the garage. You might think unit #3 is really good, but that meter is not as good at other functions:

Here's some really good pocket meters, starting with the cheapest offering from ANENG. This is the newer variant of this meter, let's call it V2.0 or 2024:

The rest of the pocket meters also did well. I believe all three of these offer a user calibration potentiometer, once again I adjusted it at 5V:

These two are essentially identically constructed pen meters which are just over $5 and don't offer a manual mode. Therefore, they report anything under 0.8V as a resistance:

Here are two tweezer-style multimeters. Their voltage range is intended to measure batteries or cells with a small load. Anything over 36V is forbidden. When reverse-biased, they do not place a load on the input. This is the results when reverse biased:

Here's four clamp meters putting their input jacks to good use.

Back to a standard multimeters. This is a full-sized unit with a delightful to operate range selector:

Here's a bench meter from Zotek and a handheld version that has most of the same functions and ranges. It became obvious to me as put the meters through their paces, that they share the same architecture, which is a very good thing:

These two meters are not anything alike, but I did notice suspicious similarities in results during usage. One is a two-in-one clock/multimeter, the other is a thermal camera/multimeter.

And here's a guy that wants to be an oscilloscope plus multimeter:

These two identical internal resistance measurement devices are not considered multimeters, but do measure voltage of cells, batteries, and low impedance power sources across a restricted range. The HRM-10 offers user calibration across positive and negative voltages. I performed this calibration before measurements - looks quite good in the end:

And the last non-multimeter I have to show you is a USB volt-ammeter device, which is able to measure up to 30V or so. Myself and others are impressed with the outstanding accuracy and precision of this device. My only complaint is the last digit of the display wanders up and down 2 counts when a stable voltage is supplied.

Voltages: "The Finale"
Probably the most important benefit to all this work is determining which device is best at the voltage range I am measuring. If I break it up into three ranges (transition at 50 mV and 5V), I can determine that I have three multimeters which will provide perfect & dependable measurements.

Most meters offer their best accuracy specification in the voltage function. Current, resistance, capacitance have a larger percent variation. I think it's because voltage is a direct measurement, and the other functions rely on other circuitry at the front end which compound the total error.
Hopefully you like this information and want to see more. I'll post similar results for resistance, capacitance, current and additional functions. Keep On Testing!
Comments