Cheap Ph Meter Vs. $100 Ph Meter Test

[passedpawn]

It's a free country and I'm not asking moderators to censor you, but I'd also like the freedom to call out a-hole behavior when I see it more than once.

No name-calling here. It's in the HBT Constitution's Bill of Wrongs.

To everyone involved here: I don't think it's useful to post if you don't intend on being effective. To be effective, you can't condescend to your audience. Also, explicit namecalling and veiled attempts at it are both counterproductive and result in the subject changing from technical to emotional. That's not an effective way to communicate.

 

[jekeane]

I want to chime in on the $10 meter. I had 3 crap out on me in a year 1 DOA, 1 that was wildly inaccurate, and 1 that worked well enough for 3 months. They looked like the same yellow one pictured.

Bought a Hach pocket pro+ and it has worked very well I have only recalibrated in once in the time I have owned it as it is always true in buffer solutions. (I brew 2-4 times a month)

 

[ajdelange]

Again - if we were evaluating these meters in a NIST metrology lab, I agree. We're not. I'm trying to determine if a couple of cheap meters can demonstrate accuracy within 0.1pH of a so-called "HBT Approved" expensive meter. So far I only have a couple of meters and a few measurements. We'll see where this thread goes as I make more measurements.

I don't think there is really any question about that. A meter that has just been calibrated in 4 buffer and 7 buffers that reads 4.0 and 7.0 in the stability test for a period of an hour is going to read within 0.1 pH of a stable higher quality meter that has also been calibrated. If 0.1 is sufficient all you really need to do is the stability test and see if the UUT stays at 4.0 for as long as you need it to on your brew day. It is, of course, of some value to verify that they do this as if they only stay within 0.1 for 15 minutes then they aren't worth much (though they could still be used if one were willing to calibrate them before each reading). Thus I applaud an effort to check the stability of these items but as I have strongly encouraged buyers of any meter to do the stability check I don't think you are adding much to the body of knowledge as if the brewer does check his new meter (and common sense says that he would on a $10 item) what you experienced isn't as important to him as what he saw. Of course your 2 data points are 2 more data points and there is some value in that.

If one goes to the user reviews on these meters one doesn't see many complaints about their failure to hold calibration (as the function of calibration is to remove error any meter is error free at calibration if it has the resolution to display the actual buffer pH and inaccuracy in a meter is thus failure to hold calibration). What one does see is frequent reports of failure after a month or two such as in the post just before this one. Thus in assaying accuracy I think you are addressing a problem the community is not interested in.

I'm glad you quoted my posts as a Moderator has apparently removed much of my last post. This is not my area of expertise, and I don't mind being proven wrong. I'm also an engineer and work daily with the top decile in intelligence - the big difference is that very few people I've come across talk to people the way you do. The immediate assault on others' intelligence if they disagree is an undesirable trait in the scientific community. You do this every single time someone does not immediately fall in line with your view on a particular topic.

Then you should have no trouble finding a couple of examples of this in my posts. As I said in an earlier post I am not aware that I have ever called anyone stupid or said that a remark they made was stupid but I don't remember everything I have written so if you could point me to a couple of places where I did that I'd appreciate it. I would also hope that the administrators would have call it to my attention if I did what you say I did but then they can't be on top of everything all the time.

It's a free country and I'm not asking moderators to censor you, but I'd also like the freedom to call out a-hole behavior when I see it more than once.

By all means. Let's see the examples. As far as I am concerned you are welcome to call me on such behaviour every time you see it but you had better have chapter and verse when you do.

Regarding 10/10ths - I'm implying that you are an extreme example of someone that takes into account the smallest minutia when homebrewing. Your effort and rigor far exceed that of any homebrewer I've ever met, and likely a huge percentage of homebrewers at large. This is not a bad thing, just your thing.

Well I readily admit that I am after rigor and again I'll remind you that we are in the Brew Science forum.

I consider myself something like a 7/10ths brewer - someone who wants to take as much as possible into account when brewing, but only so far as is required to produce great beer. I simply don't have the time required to delve into the details and fret about the small stuff.

Better is the enemy of good enough.

This thread was intended to START the process of peeling back the veil on cheap pH meters. I hear a lot about how bad they are so I wanted to see that for myself. So far I haven't been able to replicate the horror stories, but maybe this will change. I don't know yet.

As an engineer you should know that you aren't going to be able to draw conclusions about reliability based on a sample size of 2. When you buy component's from a supplier don't you look at the 'ility' data?

 

[brewprint]

I have the Milawaukee 102 and am curious about the actual temp of the probe.

I had mine in a high 50s basement and then was checking some pHs. Should I store that actual probe at about the same temperature as my samples?

 

[ajdelange]

I want to chime in on the $10 meter. I had 3 crap out on me in a year 1 DOA, 1 that was wildly inaccurate, and 1 that worked well enough for 3 months. They looked like the same yellow one pictured.

From the data in this thread we have 3 meters that worked for a year and 3 that didn't. The best we can say at this point is that the probability of getting a meter that lasts a year from this source is 50%. But we have no (or little) confidence in that number because n = 6 is too small a sample. Supposing the reliability is really 50%. The probability of 3 or more failures is only 34%. Suppose the probability of failure is 90%. Then the probability of 3 or more failures is 98.4%. Thus we can be 98.4% confident that the probability of failure is less that 90% but only 34% confident that it is less that 50%. I wouldn't buy one of these based on those numbers but then I wouldn't want to make the decision based on n = 6.

Bought a Hach pocket pro+ and it has worked very well I have only recalibrated in once in the time I have owned it as it is always true in buffer solutions. (I brew 2-4 times a month)

I really do recommend a full cal every time you brew (with new buffers). The cal isn't really more difficult at all than a cal check unless it flunks the cal check and you wind up having to do the full cal anyway.

 

[ajdelange]

I had mine in a high 50s basement and then was checking some pHs. Should I store that actual probe at about the same temperature as my samples?

The less temperature stress to which the electrode is subjected the better. Changing from 50 °F to 70°F for a measurement isn't much of a stress but if you store closer to measurement temperature the stress will be even less and your probe will come to thermal equilibrium with the sample faster.

 

[brewprint]

The less temperature stress to which the electrode is subjected the better. Changing from 50 °F to 70°F for a measurement isn't much of a stress but if you store closer to measurement temperature the stress will be even less and your probe will come to thermal equilibrium with the sample faster.

Perfect.

I see that there's a liquid inside of the probe and it made me wonder.

 

[Pappers_]

I appreciate the restraint that has been shown here and rules are not being violated. But, sometimes its not just about following the rules. Other members are reporting this thread because it is becoming unreadable - AJ and Eduardo, please refrain from replying to each other for a while. Lets get the thread back on track here.

 

[ajdelange]

I see that there's a liquid inside of the probe and it made me wonder.

The liquid in the pH sensing bulb is sealed in so as the temperature in the bulb rises so does the pressure but I have been led to believe that that is not the concern. Rather it is the stress in the glass itself that is. What happens if you pour boiling water into a cheap glass tumbler? That sort of thing.

The other fluid containing chamber (the reference cell) is open to the outside and so increase in pressure there isn't a concern either.

 

[mabrungard]

No one deserves protected status here and I use my real name in forums to keep me honest and civil. I wish more participants would, but I know of cases where participants had to remain more anonymous to protect their job. I do try to think before typing, try to add REAL value to a thread, stand by my words, and be nice!

AJ's response of Que? was appropriate since I did miss that the original question referred to the water before the grain was incorporated. In a way, I was therefore wrong. AJ does keep me honest and I try to do the same...when necessary. This was not one of those cases.

I do caution the premise of this thread. It is difficult to assess the reliability and performance of an electronic instrument based on this informal forum. But I do believe that it would be difficult to contend that a $10 instrument that has some moderately complicated elements in it, could be as reliable or accurate as a more costly unit. The good thing is that the accuracy and reliability of a cheap pH meter doesn't matter much, since lives or livelihoods aren't on the line in homebrewing. But for a pro-brewer to rely on such an instrument is foolish. I was brought in to consult at a brewery several years ago that was using cheap instruments. They had 100's of barrels of beer that were destroyed due to their inability to measure and control their brewing. Unfortunately, I was brought in too late to fix their problems and they went out of business a couple of weeks later.

If your brewing matters, buy the best equipment that you can afford. That means different things to each of us, but do the best you can.

 

[PGEduardo]

Wow, can't believe I haven't brewed for 7 months! I've been drinking less beer lately, so I guess that's a good thing.

Back at it again with a Maris Otter SMASH-ish. Original batteries on everything so far.

BLUF: These cheapo meters still seem to be holding their own! MW201 - 5.43, Cheapo1 - 5.5, Cheapo2 - 5.5!

 

[ajdelange]

So what does this tell us?

Fiirst, that one person bought two cheepies and they both worked for more than a year. Assuming the population probability of failure in that time to be 0.3 (30% - no claim that that is the number - just that user reports seem to indicate a number like that) the probability of 0 failures in a sample of 2 is (1 - 0.3)^2 = 0.49 or 49%. Even with a pretty high (IMO) failure rate about half of people who buy two will find neither fails in over a year. And the probability of failure of at least one is 2*(0.3)*(0.7) = 0.42 (43%) and of two is (0.3)^3 = 0.09 or 9%. Thus, in statistical terms, we have very little confidence that the failure rate is less than 30%. By comparison, if we tested 10 meters and had no failures the probability of that event would be (1 - 0.3)^10 = 0.04 (4%) and we would be able to say, at less than the 5% confidence level, that the meters have better reliability than 30%. Five percent is often the minimum acceptable confidence level. Thus the survival of two meters for more than a year doesn't tell us anything useful with respect to the hypothesis that these are really reliable. One would need to do this test on at least 10. And then, of course, we'd complain that 5% isn't a very comfortable confidence level but that's the nature of statistics folks!

Second, that the apparent accuracy of these meters would be unacceptable to many brewers. While one can no more draw any statistically significant information as to whether the population of these meters meet their specified accuracy from a sample of two than he can about reliability, it may be helpful to some to look into what these readings tell us about the meters assuming they do meet their specified accuracy of 0.1 pH rms. The meter reads a voltage and temperature. Both of these are corrupted by noise and bias and one of the differences between cheap and more expensive meters is in processing done to minimize the effects of the noise with the biases taken out by calibration (that's what the buffer stuff is all about). The meter takes the measured voltage and temperature and sticks them into an equation which produces a number. This number is then rounded (which introduces 'quantizing noise') and the rounded number displayed. The quantizing noise is, in a good system, about 10 or more dB below the other noises so it doesn't distort the reading. The quantizing noise for an 0.1 precision display is 0.029 pH and this is 10.7dB below the stated accuracy of these cheapie meters which is 0.1 pH. For the Milwaukee meter the quantizing noise is around 0.0029 pH which is 17 db below the specified accuracy of 0.02 pH. Both meters are well designed in this regard. So we accept that the meters both read and display, unencumbered by quatizing noise, the actual pH with the readings corrupted only by the random noises as accurately described by their accuracy specifications (0.1 and 0.02). This means that what the display says isn't the actual, true, pH but something 'close' to it with what close means depending on the accuracy. The meter reads the most likely pH but other pH's are possible.

The curves below show the distribution of probable actual pH's from the two meters given their readings. The most likely pH's are at the 0.5 (median, 50%) level but the distributions of the probable actual pH's about that point are quite different. Looking first at the Milwaukee curves we see that the probable actual pH will be above between 5.44 in 10% of measurements where the meter reads 5.43 and below 5.46 in 90% of measurements so we conclude 5.44 < pH < 5.46 80% of the time. The numbers from the cheapie meters are 5.37 < pH < 5.62 in 80% of measurements.

Now whether that spread is acceptable or not depends entirely on the user's perspective which, in turn, depends on what he is trying to do. If all he cares about is getting his mash into the 5.4 - 5.6 range then likely it is. If he's trying to see if a difference in mash pH between 5.4 and 5.5 makes a perceptible difference in his beer then clearly it isn't.

 

[z-bob]

@PGEduardo Re: "n = 2"
[I know I should stay out of this] I usually have to read everything AJ says about 3 or 4 times before I understand a little of it. I think I'm a pretty smart guy, but chemistry and statistics are way outside my areas of expertise and it's been over 35 years since I studied them. Anyway,

The Milwaukee MW102 meter is sufficiently-higher in accuracy to use it as a standard by which to judge the two cheap meters, so long as you keep it calibrated. AJ could tell you the exact margin of error, and I think he has already in a previous post.

The biggest problem with your test is the sample size. You have shown that your two cheap meters have held up for a year and still provide useful readings. But nobody knows whether your two meters are good representatives or if you just got lucky (or unlucky) in the draw. Taking more meter reading with the same 2 meters does not increase the sample size (it does tell you something, I'm not sure exactly what, perhaps how much those 2 meters drift) you'd have to buy a bunch more meters (100 more? 300 more?) and test them too to have statistically valid results. And the information gleaned from that testing would probably not be worth the cost.

I appreciate you putting your test results out there for everyone to see. A little bit of data is better than no data as long as one doesn't try to draw sweeping conclusions from it.

 

[PGEduardo]

If anyone wants to ship me some cheapos I'll gladly throw them in the mix.

I think I actually agree with AJ here - my goal is to shed some light on whether these cheap meters are useful at all for the casual hobbyist. IE - is your mash in the right range.

If you are perfecting a recipe and really dialing in your chemistry these are not the meters for you.

 

[Tehnicni]

Greetings to everyone
I'm new to the forum and a new owner of a yellow pH meter. Whether one of you measured the pH of the garden soil with it (in 3 x more water, dissolved soil). The results were 8, which is completely false. The litmus strips gave me a pH of 5.
Tap water have 7.2, baking soda 8.1.

 

[ScrewyBrewer]

If you are happy with the results you get using an economy pH meter, then that's great. If you are happy with the results you get using a higher end pH meter, then that's great too. Whether you love or hate sour beers is a personal preference. The same thing applies to pH meters and many other things in life. It seems pointless, and exhausting, trying to convert someone to your way of thinking.

 

[ajdelange]

Greetings to everyone
Whether one of you measured the pH of the garden soil with it (in 3 x more water, dissolved soil). The results were 8, which is completely false. The litmus strips gave me a pH of 5.

I'm not 100% certain how to interpret this (perhaps English is not the first language here) but if it says that this meter read 8 in a soil sample when pH strips said 5 and that, therefore, the meter is wrong I'll point out that you can't draw that conclusion. Yes, the inexpensive meters are often off by quite a bit but litmus paper is pretty useless for pH measurement. Telling us what litmus paper says is about as valuable as telling us your hydrangeas are blue.

Tap water have 7.2,

As tap water pH can range from below 6 to 10 or more that's not very useful.

baking soda 8.1.

Now here we're getting somewhere. A sodium bicarbonate solution has a pH of about 8.3 so at first blush it appears this meter is off by 0.2.

To meaningfully characterize the performance of this meter carry out the stability test in the sticky at the top of this forum.

 

[TANSTAAFB]

Just meticulously tried to calibrate a cheapo meter. Not stable at all, even during the calibration process!

 

[TANSTAAFB]

Came back to it and retested after the meter was off for a while and it's on the money! I'll retest this evening and tomorrow to check stability over a little time. How long will the calibration solution last?

 

[dmtaylor]

My el cheapo Chinese junk from the Amazon still works fine after 2 years. If it ever fails, which maybe it won't, I'll buy the same model again. It's the same kind as the purple one above.

 

[imasickboy]

A sodium bicarbonate solution has a pH of about 8.3

This may be a dumb question, but wouldn't the pH of that solution depend upon the water it was mixed with, and the concentration of the solution itself? Please note that I'm not trying to stir up any *****, but just questioning something that seemed like a illogical generalization to me. Educate me, please?

 

[z-bob]

This may be a dumb question, but wouldn't the pH of that solution depend upon the water it was mixed with, and the concentration of the solution itself? Please note that I'm not trying to stir up any poopye, but just questioning something that seemed like a illogical generalization to me. Educate me, please?

Sodium bicarbonate (actually, any bicarbonate ) is a buffer. The concentration of the solution and the pH of water you started with will affect the pH of the solution a little, but not much. The bicarbonate buffer should overwhelm whatever else is in the water; for the most part, anyway.

 

[imasickboy]

Good enough. Much appreciated.

 

[brewing_clown]

My experience with what I consider a more expensive meter (given my brewing means) seems more akin to what some folks are recounting for the cheapies. I bought a Milwaukee pH56 and treated it per the manufacturer's guidance. The electrode wasn't even stable enough to calibrate.

I'm currently using a cheapie to measure mash, sparge and wort pH, and it seems to be stable enough when I check it with buffer solutions. I'm not comfortable with good enough, but am wary of putting money into something that doesn't result in much better.

That said, and having read this thread, I'm going to give a better meter another shot. I'm willing to believe that my experience is unusual, even though the particular meter I initially chose was probably not a good buy.

Thanks all for the info.

 

[ajdelange]

This may be a dumb question,

Not a dumb question at all.

but wouldn't the pH of that solution depend upon the water it was mixed with, and the concentration of the solution itself?

Yes, it would. I assumed that if one were going to do a check on a pH meter with a sodium bicarbonate solution (which he would only do in desperation because it's not a good buffer but it is in everyone's kitchen) that he would want the solution to have pH 8.3 and that he would, therefore, mix the sodium bicarbonate with DI or at least RO purified water or, assuming that he had neither of those, make a strong enough solution that the bicarbonate would swamp whatever was in the water. For example, a couple of teaspoons full in a quart or less of water. But yes, if the water has high alkalinity and a low pH and only a little bicarbonate is added to it it could easily pull the pH away from 8.3 by at least as much as you observed.

 

[ajdelange]

Sodium bicarbonate (actually, any bicarbonate ) is a buffer. The concentration of the solution and the pH of water you started with will affect the pH of the solution a little, but not much. The bicarbonate buffer should overwhelm whatever else is in the water; for the most part, anyway.

There seems to be a lot of confusion about what buffering is and what a buffer is. Buffering is the amount of acid or base required to induce a small pH change. A solution that requires a lot of acid for a small pH change is said to have high buffering capacity. A solution in which the pH changes quite a bit when only a small amount of acid or base is added is said to have weak, little or poor buffering. The nuance come in the fact that solutions with mixtures of the ions of an acid have pH regions where buffering is strong and others where it is weak. Though such a solution has buffering at any pH if we want to use it as a buffer, i.e. something that holds pH, we must add acid or base to get it to the pH (or one of the pH's) where buffering is strong. Thus we can call a mix of carbonic acid and bicarbonate ions a buffer when the mix is such that the pH is 6.38 and we can call a mix of bicarbonate ions and carbonate ions a buffer when the pH is near 10.38 (and do - this is the blue buffer furnished with many pH outfits) but we cannot call a solution of straight bicarbonate ions in water a buffer because it exhibits near 0 buffering capacity at its native pH of 8.3.

 

[Shwagger]

I use a Ysi at work every day that measures ph, temp, do, orp, and conductivity. Calibrate 2 times a day. This is a top of the line professional piece of equipment. And it drifts pretty regularl. If you are within 0.05 of your 5.2, you're doing great. Look up your states tolerances for ph in drinking water ,for example .Can be a loose tolerance. Fyi

Edit: should have said I check calibration twice a day

 

[ajdelange]

There is no reason to accept drift as bad as ±0.05 as you can buy several meters that will stay within ±0.02 for hours for around $125.

 

[Shwagger]

5 hundredths is pretty dang good in my book. I'll take that all day and 4 times in Tuesday. I am a simple brewer, though.

 

[z-bob]

Here's a cheap pH meter that has my eye. https://hannainst.com/hi98100-checker-plus-ph-tester.html I like the form factor. But the accuracy v. resolution ratings look fishy. Like maybe the extra digit of precision is useless.