The problem with using meters on solutions with limited solubility. [ESPECIALLY "PPM" meters set up to measure salt water which is very highly soluble]
The ionic portion of CS has very limited solubility and some portion will convert into an un-measureable colloid depending on variable environmental conditions.
 The Silver is still in the water, but a meter won't detect it.
The instant power is off, conductivity starts dropping.
The stronger you made the solution, the more it will drop...by around 40%
Each time you bump it back up, it will drop by about 1/2 as much as the previous time, but anything over 30 uS will drop back to 30 uS over time.

Bottom line:
Using conductivity to measure something where conductivity changes is not the best way to do anything.
 It's like using an ox to pull a freight train.

 A better way is by Faraday prediction, [which is what the programming button does] but that only says how much silver was emitted, not what happens to it [doesn't account for waste]  But the SWAP function limits waste to a great degree over using DC current.

More:

All hand held PWT/TDS/PPM meters work the same way. They actually measure conductivity not PPM. They are simply not the same thing.

TDS (PPM) meters such as the TDS3 also measure conductivity but then convert that measurement to an estimated PPM using water industry standard tables for dissolved salts. [NaCl in this case] (also known as Total Dissolved Solids or TDS)

Since Ionic/Colloidal Silver is not a mineral salt, it behaves differently and requires a different technique for reading the TDS meter.

The reading can generally be taken ‘as is’ when checking pure distilled water or when checking any other water source such as tap or well water. (ie 200 on the meter = 200 PPM)
However, when checking Colloidal Silver made with pure distilled water the reading should be doubled. [If it reads 10, it’s actually 20 PPM of colloidal silver]
 
Due to range and resolution limitations,  when measuring colloidal silver with this meter  you should also allow for an error factor of at least +/- 10% . So for example 10 on the TDS meter could be as high as 22 PPM or as low as 18 PPM  (when doubled).

Meters such as the Hanna PWT, the HM Digital EC3 and COM-100 read out directly in Microsiemens of Conductivity [uS].
The COM100 also reads out in 3 different scenarios for PPM depending on the suspected dominant mineral salt content.
Silver water is NOT salt water.

To get an idea of what the PPM is in CS, use 1uS = ~1 PPM.
 Since the conductivity will start dropping as soon as power is off, Use Meter immediately after shutdown.
Concentrations over the saturation point of Ionic Silver in water [~ 10 PPM ] will progressively and variably form more and more non-conductive “particles” accounting for some of the difference between Faraday Equation Prediction vs Conductivity Monitoring.
 Beyond around 25 PPM in silver water a meter reading will have considerable slew towards the low side and a batch may never register more than 30 uS the day after it’s done even though it may really be 50+ PPM

This method has been checked against samples tested by an Atomic Absorption Spectrophotometer at  a range average of  12 PPM

Getting anything but an eddicated guess is beyond our capabilities. [ "eddicated"  is a step or two below 'educated' ]
To get that eddicated guess, use the meter the instant that power goes off.
The generator does it that way, then adds a time related prediction after current has become a constant in the equations.
But even then, it's probably going to be the low end of actual as it doesn't include what it took to make current into a constant.

The only REAL way is to use an Atomic Absorption Spectrophotometer, which none of us can afford [50 to 250 thousand dollars] and requires high training to operate...and is as big as a desk or a closet and uses a great deal of heat by electric furnace or acetylene flame to evaporate the water and turn the metal itself into a  light emitting gas plasma.
..or pay a "good" lab ~$200 per sample to do that for us. [Even most labs have problems and they ALL average at least 3 tests of the same sample ]

 Better bet:
Don't worry about it.
In the "real world"....USING CS is entirely "seat of the pants" and nearly impossible to over do with the least application of common sense.
 No use recommendations I've ever seen make any scientific sense at all.
When making your own CS,  it won't break the bank to go 'full tilt' in the short run...where any danger is in the long run coupled with massive over indulgence.
..it doesn't really matter how strong it is, so long as the job gets done, which makes *intuition* as good as anything else. [when, frankly, nothing else is reasonably available]

Personally, I use the Pitui test.
 Can't taste it?..make it stronger
Doesn't taste very strong?...use  more.
Pitui!!  [Making faces]  YUP, a little bit should be enough.

In the past, store bought CS was only 3 PPM, only fairly recently going as high as 20 PPM as processes have improved...more generally at 10 PPM.
..and commercial producers aren't all equal [not all of them have an AA Spectrophotometer]  as some samples have tested out at only 3% of the stated content.

 A meter can't tell the difference between Silver and anything else that might be in there...and as you may have just experienced, "when" the meter is used makes a huge difference.

The 'earlier generators'  [Silver 5-7] worked the same way as the Silver 9, but turned itself off at well beyond saturation point where sometimes environmental conditions delayed turn off or it failed to turn off all-together.... and made no automatic predictions...leaving it up to "ignorance is bliss". [which is better than ABSOLUTELY NO CLUE between 1 and 500 PPM *mud* made with uncontrolled exponential *run-away* DC, quite often in salted water..or.. IOW...TOTAL ignerance is totally blissful ]

  An airplane is a long way from being a bird, but a whole lot better than diving off a cliff and calling that flight.
An "Air speed indicator" is not the same thing as a odometer/speedometer that measures distance directly...it won't tell you exactly how long it will take to arrive or exactly where you are.
Flying with a wheel on the ground isn't an option.

Ode

This is the best that I’ve ever heard it stated.

    Dosing is not a science. The key word is "enough."
 If you've used CS before, then you already know that you're not allergic to it, therefore you don't have to start with the few drops on the tongue that I suggest to new users, "just in case."
 Beyond that, it would be good if we knew what kind of CS you've got. If store-bought, which brand and what supposed strength?  If home made,what kind of generator and how brewed.
  The usual advice is to start small and ramp up the dose until you have the effect you want. This is measured by either the symptoms improving or by the symptoms worsening (temporarily) in a Herxheimer reaction.

  In general, flu-like symptoms and a general yucky feeling mark the Herxheimer reaction. In severe cases Herx reactions can be life threatening, particularly in people who are already weakened. If there is a bad infection that is destroyed very rapidly by the treatment it can overwhelm the body, leading to a shock reaction that could require emergency treatment.
 
 Thus the caution to start small and ramp up until some effect is noted, then hold at that level until effects subside. When that happens, ramp up the dose further until effects resume, but remain manageable, and so on. Keep at it a few days longer if all seems to be well, just to be sure...

 So what sort of dose might you want to work up to?

  First, I will *assume* an "average" CS brew of some 5 to 20 ppm in plain distilled water made with a simple generator.

 For a very aggressive treatment, some of our members have reported good `results taking a good mouthful (very precise measurement unit there!), swished around in your mouth for at least a couple of minutes before swallowing, taken every waking half hour to an hour.

 To start slowly, just to be cautious of any severe Herx, I would probably wait a couple of hours between "swallows" at first.

 One other thing is to consider adding Gatorade to the protocol. If you search the archives on that subject, you'll see that it is used to *drastically* increase the rate at which the CS enters the bloodstream.[note:  Mix CS with Gatoraid and use immediately…do not store it ]

 Lastly, I want to say that, if push comes to shove, going to the doctor for antibiotics is not out of the question. I doubt it will be needed, but if your self-treatment is not effective within a reasonable time, the conventional alternative will serve as a safety net. Not without side effects, true, but safer than no treatment.

 You must judge the risks and only you are responsible for the consequences of your health-care decisions. Don't take my suggestions  as a substitute for your own judgement or that of your professional  advisors.      [Authors name with-held by request]