Home Brew Forums

Home Brew Forums (http://www.homebrewtalk.com/forum.php)
-   Brew Science (http://www.homebrewtalk.com/f128/)
-   -   Effect of reversing pH meter calibration order (http://www.homebrewtalk.com/f128/effect-reversing-ph-meter-calibration-order-348178/)

afr0byte 08-17-2012 12:32 PM

Effect of reversing pH meter calibration order
 
So, I realized that the last time I calibrated my pH meter I calibrated with the 4.01 solution first, instead of the 7.01 solution (The meter manual says to do 7.01 first.). How much of an effect, and what kind of an effect might this have? Does someone know? I plan to do a DI water pH test of an all pilsner test mash tonight, to see if my probe is in the ball park at least.

DarthMalt 08-17-2012 12:56 PM

I think the reason for having a specified calibration order is that the neutral buffer solution (7.0) is less likely to have a drastic change on the acidic (4) or basic (10) solutions. However, I highly doubt that a small amount of residual buffer solution left on the meter with have any effect on the other solutions for the simple fact that they are all buffered ie. PH stabilized to a certain degree. It would take a fair amount of contamination to change the PH of the solutions.

As far as the probe itself, the order you calibrate should have little effect on the final outcome as long as the probe was kept in the solution long enough to register at the specified PH. I use PH meters at work that are probably far more sensitive than what a brewer might use (depending on the meter) and have made the mistake of calibrating "out of order" before without any marked difference in error.

mabrungard 08-17-2012 01:02 PM

Unfortunately, its an issue with slope and offset. If the offset was adjusted with during the 7 calibration, it affects the calibration setting (offset) at 4. If you then went back and checked the calibration at 4, you may have found that the trimmer needed another tweak to properly set the slope. After that tweak, the instrument should be in calibration. But performing the calibration order in reverse and not revisiting the 4 calibration may mean that the instrument is not calibrated very well in the lower range as pH moves further from 7.

Calibrate at 7 first and then at 4.

afr0byte 08-17-2012 01:08 PM

Rog, thanks for the responses. I plan to do a properly ordered calibration this time. We'll see how the DI water pH ends up tonight. Btw, Martin. I'm brewing an IIPA with 14 lbs pilsner, .5 pounds 9L munich, and .5 pounds C40 in a 6 gallon mash with a Ca level of about 100 and a alkalinity of about 53 (lime softened water). In your opinion, would I need any acid malt? I'll probably do a sample mash tonight anyways, assuming my di mash is reasonable, but I thought I'd ask. EZ Water estimates a no-acid-malt-mash as ~5.7. Bru'n water estimates 5.5, hence the question.

ajdelange 08-17-2012 04:03 PM

Quote:

Originally Posted by afr0byte (Post 4338245)
So, I realized that the last time I calibrated my pH meter I calibrated with the 4.01 solution first, instead of the 7.01 solution (The meter manual says to do 7.01 first.). How much of an effect, and what kind of an effect might this have? Does someone know? I plan to do a DI water pH test of an all pilsner test mash tonight, to see if my probe is in the ball park at least.

If your instructions say do 7.01 first then you should do 7.01 first i.e. you should always do what the instructions say. I can't predict what the effects might be because I don't know how your meter works. In a modern digital meter, and this includes inexpensive things like the Hanna pHep, it doesn't matter one bit which buffer you use first because all the meter wants to know is the millivolt response of the electrode to the buffer and the temperature of the buffer (which it uses to determine the buffer pH but is also needed to compute the calibration coefficients).

If the meter is an old analogue style where you adjust amplifier gain and offset then the amount of error you could expect would depend on the actual slope and offset of the electrode. If, for example, the actual electrode offset is close to 0 and the meter was set (from a previous calibration) to close to 0 offset it wouldn't make much difference if you adjusted the slope first, assuming that is what you did (because the offset doesn't need to be adjusted. If the difference between the offset of the electrode and offset of the meter differ quite a bit then you can expect more error.

Of course the obvious thing to do is recalibrate before you do the tests you are planning and do it according to the directions this time. You should be recalibrating within an hour or so of use anyway.

afr0byte 08-17-2012 04:11 PM

Quote:

Originally Posted by ajdelange (Post 4338851)
Of course the obvious thing to do is recalibrate before you do the tests you are planning and do it according to the directions this time. You should be recalibrating within an hour or so of use anyway.

Thanks for the response. I actually calibrate right before I take my reading on brew day, so I should be good in that respect.

ajdelange 08-19-2012 06:16 PM

Quote:

Originally Posted by mabrungard (Post 4338314)
Unfortunately, its an issue with slope and offset. If the offset was adjusted with during the 7 calibration, it affects the calibration setting (offset) at 4. If you then went back and checked the calibration at 4, you may have found that the trimmer needed another tweak to properly set the slope. After that tweak, the instrument should be in calibration. But performing the calibration order in reverse and not revisiting the 4 calibration may mean that the instrument is not calibrated very well in the lower range as pH moves further from 7.

Calibrate at 7 first and then at 4.

This assumes an analogue style meter. If he went back to check the 4 setting then that's tantamount to doing the calibration properly as the offset adjustment is independent of slope but not conversely.
With a digital meter voltages and temperatures are measured and recorded and inserted into formulas which produce calibration constants. The constants are not calculated until both temperatures and voltages are recorded so it doesn't matter what order the measurements are taken in.


Here's a simplified explanation of how it works with an analog meter.

An ideal (perfect,brand new... electrode produces voltage
E = slope*(pH - 7)
with slope dependent on temperature

A real electrode produces a voltage

E = E0 + slope1*(pH - pHi) where slope1 is not the same as the ideal slope. It tends to become smaller in magnitude as the electrode ages. E0 is an offset voltage. pHi is the isoelectric pH of the meter. It should be close to 7. Let's suppose it is for simplicity.

The electrode is connected to an amplifier which has gain and offset settings. The latter controls a voltage which is added to the electrode voltage and the sum is then amplified (scaled) by the gain factor. The output of the meter is

E' = gain*(E0+ offset) + slope*gain*(pH - 7) (assuming pHi = 7)
Clearly we want to adjust the controls so that offset = -E0 and gain = 1/slope and it's clear that the way to do that is put the electrode in 7 buffer and adjust the offset control until the output of the amplifier is 0 (center scale on an analogue meter and that point is marked 7) and then put the electrode in pH 4 buffer and adjusts the gain until it is equal to 1/slope at which point the voltage is -3. This point on the meter face is marked 4. The meter is calibrated.

Now if one goes into 4 buffer first he'd adjust the gain until the voltage is -3 (meter reads 4) which would result in

gain = -3/(E0 - offset -3*slope)

which would give the correct gain (1/slope) only in the case where the offset happened to have been set at -E0 (from a previous calibration). If one now goes into the 7 buffer he can set the offset correctly to 0 the voltage (put the pin or A/D reading at 7) and this would required simply that the offset be equal (in magnitude) to E0. Thus going into 4 buffer first gives and erroneous slope setting but a correct offset. It should be clear from this that the error one encounters in doing this depends on how far away from pH 7 the sample falls. Close to 7 the readings will be quite accurate. Farther away, less so.


All times are GMT. The time now is 08:06 PM.

Copyright ©2000 - 2014, Jelsoft Enterprises Ltd.