Mizuno's bucket of water

Quote

You said, "input power subjects the experiment to a poor signal to noise ratio based on noise in the output measurements."

I probably should have said that low output power is more subject to errors in measurement than higher power outputs. You will say, sometimes high power is hard to measure. I am saying that output power in the milliwatt range is much more difficult to measure accurately in real world situations by calorimetry, than is power in the dozens or hundreds of Watts.

THHuxleynew wrote:

Quote

Input power matters because it translates to output power. Although (usually) input power can be measured very accurately, that is not so true of output power. Small fractional errors in output power measurement become problematic when input power >> signal.

That is exactly what I was addressing. Thanks for making it clear.

Quote

"What you are missing is that small increases in output power compared to blanks indeed can subject the experiment to a poor signal to noise ratio based on noise in the output measurement. If the out/in ratio is very poor, then the input power contributes to the output noise because more power is needed to run the experiment."

I probably should have simplified this to something like: low level measurements of heat by calorimetry at some point become more error prone than higher level measurements. By noise, I didn't mean noise in the sense the word is used in data transmission or communication. I meant and should have said errors instead of noise. I never said input power is noise nor does your quote say that.

Quote from seven_of_twenty

I probably should have simplified this to something like: low level measurements of heat by calorimetry at some point become more error prone than higher level measurements.

That is a completely different issue. It has nothing to do with input power. The difficulty of measuring low levels of heat is the same whether there is high input power, low input power, or no input power. Input power has nothing to do with it.

You are not "simplifying" by saying this. You are saying something completely different. Previously, you said: "input power subjects the experiment to a poor signal to noise ratio based on noise in the output measurements." That has nothing to do with measuring low power heat.

Quote from seven_of_twenty

By noise, I didn't mean noise in the sense the word is used in data transmission or communication.

Yes, you did mean that. Again, your sentence is: "input power subjects the experiment to a poor signal to noise ratio based on noise in the output measurements." That means input power causes noise in the output measurements. That's what you said, repeatedly. That is incorrect. Input power does not do that.

Quote from seven_of_twenty

THH: Input power matters because it translates to output power. Although (usually) input power can be measured very accurately, that is not so true of output power. Small fractional errors in output power measurement become problematic when input power >> signal.

That is exactly what I was addressing. Thanks for making it clear.

No, this does not "make it clear." THH is wrong, and so are you, for two reasons.

First: "Although (usually) input power can be measured very accurately . . ." is wrong. It can always be measured very accurately, in nearly any cold fusion experiment. It is direct current electricity, which can be measured more accurately than any other force. There are a few cold fusion experiments with complex waveform input electricity which are somewhat more challenging to measure, but none of the ones discussed here, such as Mizuno's, fall in that category.

Second, although output power cannot be measured as accurately, input power does not affects that measurement, and it does not increase the inaccuracy, because it is always subtracted. 99.999% of it is subtracted. It makes no difference whether input is 1000 W, 100 W, or 10 mW. It is all removed, so it does not make the output measurement less accurate.

As usual, you seem to me to be making no effective effort to understand the arguments. OK then, an experiment that provides 1 mW excess power with 100 mW input power (i.e 100 mW in, 101 mW out) is just as good for proving that an experiment has LENR as one which provides 1 W excess power with 1 mW in. (i.e. 1 mW in, 1W out). In your view, the two would be equally convincing and impressive. Or is that not what you seem to be saying?

Quote from seven_of_twenty

As usual, you seem to me to be making no effective effort to understand the arguments. OK then, an experiment that provides 1 mW excess power with 100 mW input power (i.e 100 mW in, 101 mW out) is just as good for proving that an experiment has LENR as one which provides 1 W excess power with 1 mW in.

Nope. The more excess power you get, the easier it is to measure. Excess power, that is. Input power is subtracted, so it does not matter.

1 mW excess is very difficult to measure. No ordinary calorimeter can measure it. Whether the input power is 10 mW or 1 W makes no difference to how difficult it is to measure the 1 mW. (Practically no difference. At such low power, the noise from input power begins to play a role. It is 1 W/10,000.)

". . . (i.e 100 mW in, 101 mW out) is just as good for proving that an experiment has LENR as one which provides 1 W excess power with 1 mW in."

Again, you misunderstand. 1 W excess power is far easier to measure than 1 mW. It would be far easier no matter what the input power is: 1 mW, 10 W, 100 W . . . or zero input power, for that matter. The input power does not affect the measurement. The more excess heat you have, the better. Total output power including the input is irrelevant. Only the excess heat portion matters.

In other words, 101 W total (100 W input, 1 W excess) is much easier to measure than 1 mW total (0 input, 1 mW total), or 2 mW total (1 mW input, 1 mW excess). The size of the excess portion is the only thing that makes it easy or hard to measure.

This is wrong: "an experiment that provides 1 mW excess power with 100 mW input power."

The 100 mW does not "provide" the 1 mW of excess in any sense. It does not cause it. There is no direct connection between them. Input only maintains the temperature, which can also be done by improving insulation. Or by moving to Venus. With electrochemistry, input can be turned off altogether.

Quote from seven_of_twenty

In your view, the two would be equally convincing and impressive. Or is that not what you seem to be saying?

No. You fail to understand simple arithmetic. 1 W of excess heat is easier to measure than 1 mW. Much easier. The input power plays no role and does not make it any harder or any easier. Whether it is high, low or zero, it does not affect the measurement of excess heat, because it is subtracted out.

You are the only one who thinks that is what I am saying. I suggest you read this message carefully, and think about measuring an object that is sitting on a platform, by measuring from the floor to the top of the object. If you know the exact height of the platform, and you subtract it out, why does it matter how high the platform is? Answer: it doesn't matter.

The platform is the input power. Anything higher than the platform is excess heat. Whether the platform is 2 cm high or 2 meters high makes no difference. You can measure the excess height just as accurately either way.

Jed: Nope. The more excess power you get, the easier it is to measure. Excess power, that is. Input power is subtracted, so it does not matter.

Jed, you have made throughout this topic the same - clearly wrong - mistake.

I have not corrected you because we have hashed this out many times, but occasionally, as here, your black and white thinking leads you so far astray I need to point it out.

If you have 1W input power, and 1W excess, it is normally very easy to measure. We would both I think agree that. You can find scenarios, e.g. when the equipment generating the power is very massive and large, where it is still challenging. But still, normally easy.

If you have 10,000W input power, and 1W excess, it is normally very difficult to measure. That is because although you can subtract the input power, and calibrate, any error in the input or output measurements (it is usually output that is most difficult) gets amplified by the difference between the (subtracted) 10kW and the (signal) 1W.

Specifically if your power measurement has 0.01% noise or error that 1W signal will be lost.

Now, the cases we actually consider are usually in between these extremes. In good experiments, well recorded and calibrated, with clear attention to ensuring that systems are the same between calibration and recording results, you can argue that the positive results obtained are well beyond possible errors. You can properly argue that subtraction allows results << input to be reliably deduced. My problem with the LENR examples quoted here is that such results do not get reproduced. What can be reproduced ends up different from the apparently solid positive result. Whereas results that are not solid (like F&P' famous open cell experiments with bubbles) get fully reproduced by people who makes the same assumptions F&P do and do not question or further instrument them.

Anyway, that no doubt you disagree with, as do many here, and that is fine.

What is not fine however is when you make statements that are (a) wrong (b) genuinely misleading, as the one above.

It does matter what is the input power when trying to measure excess. How much it matters depends on the accuracy of the measurements and the degree to which differences between calibration and control systems can be reduced. Both these factors vary enormously over the range of experiments considered as possible evidence of LENR.

Therefore your generalisation here will lead to bad judgement of LENR experiment success. It should be avoided by everyone interested in this.

Regards, THH

Jed: You are the only one who thinks that is what I am saying. I suggest you read this message carefully, and think about measuring an object that is sitting on a platform, by measuring from the floor to the top of the object. If you know the exact height of the platform, and you subtract it out, why does it matter how high the platform is? Answer: it doesn't matter.

To further examine this analogy. You can easily measure the object height from the top of the platform. True. That is not possible in the LENR case.

Otherwise, we are measuring pillars where at one time only one pillar is available. One is 1m high. The other is 0.1mm higher. You need a ruler accurate to significantly more than 1 part in 10,000 to do this.

Actually, for length, this is no problem. We have such equipment.

Quote from THHuxleynew

think about measuring an object that is sitting on a platform,

Such as a bucket....an ice bucket?

I just hope the ice kills coronavirus,

Quote from THHuxleynew

Jed: Nope. The more excess power you get, the easier it is to measure. Excess power, that is. Input power is subtracted, so it does not matter.

Jed, you have made throughout this topic the same - clearly wrong - mistake.

I have not corrected you because we have hashed this out many times, but occasionally, as here, your black and white thinking leads you so far astray I need to point it out.

If you have 1W input power, and 1W excess, it is normally very easy to measure. We would both I think agree that. You can find scenarios, e.g. when the equipment generating the power is very massive and large, where it is still challenging. But still, normally easy.

If you have 10,000W input power, and 1W excess, it is normally very difficult to measure. That is because although you can subtract the input power, and calibrate, any error in the input or output measurements (it is usually output that is most difficult) gets amplified by the difference between the (subtracted) 10kW and the (signal) 1W.

This is pure bullshit. YES, OBVIOUSLY if you input 10,000 W, that will cause a lot of noise. Do you think I did not know that? Is that not obvious from what I wrote? I said any power meter is good to about 1/10,000, so that would be a ~1 W of noise, which is a lot. However, here in the real world, in actual cold fusion experiments, no one has 10,000 W of input power. So that is a contrived example.

"Jed, you have made throughout this topic the same - clearly wrong - mistake." No, I did not make a mistake. You twisted my words to make it look as if I made a mistake.

Quote from THHuxleynew

To further examine this analogy. You can easily measure the object height from the top of the platform. True. That is not possible in the LENR case.

You are missing the point. In my example, you measure from the floor to the top of the object. As you say, in calorimetry, you have to do it that way. You can only measure input power and excess power together. HOWEVER, just as you know the exact height of the platform, you know the exact extent of input power, so both can be subtracted.

I expect you understood that, but you deliberately pointed out what I already made clear (that the two can only be measured together), to make it look as if I made a mistake. You probably understand this, but Seven_of_Twenty does not understand it, and never will. Do not add to the confusion by pretending I did not realize the two can only be measured as one value. That has nothing to do with how easily the input power can be subtracted.

What THHuxleynew wrote.

One problem here seems to be (IMO) that JedRothwell rarely makes an effort to understand what others are saying before charging off into the sunset. As to what I do and don't understand ("and never will":-- you sound like a particular politician!) I may once again try to straighten you out when I have more time, though it is probably only for an audience of one and therefore not worth the bother.

Quote from seven_of_twenty

I may once again try to straighten you out when I have more time, though it is probably only for an audience of one and therefore not worth the bother.

There is no need for you to do that. Your previous statements are quite clear, and wrong. You have not retracted them. Clearly, you still believe them, especially these two recent ones:

"input power subjects the experiment to a poor signal to noise ratio based on noise in the output measurements." That is completely wrong.

You think that I believe ". . . (i.e 100 mW in, 101 mW out) is just as good for proving that an experiment has LENR as one which provides 1 W excess power with 1 mW in. In your view, the two would be equally convincing and impressive. Or is that not what you seem to be saying?"

No, I never said anything remotely like that. You do not understand that power level of the excess heat makes the only difference. Higher excess heat is easier to measure. Higher or lower input power has no effect.