Confirmation of OU devices and claims

[Jeg]

Looks to me that battery presents some resistance for this very brief period, which eventually breaks down. 
Dirt at the conducts or an internal battery's characteristic?

And it is also this stair like decay if you zoom at the picture. Except if Void changed V/div after waveform's capture.

[web000x]

I think the point I've been trying to make is that virtually all replications will still allow the excuse of "not exact enough" when/if they are found not to meet the OU claims of the originator.
But if someone can get their hands on an original unit, that once was claimed to be OU by the Builder Himself, even that might not remove all possibility of the "not close enough" excuse -- as local conditions may vary, or any of a number of other factors not directly associated with the build itself.


Void, that little flat spot has me thinking. Could it be an artefact of the probe you used? Out of range, signal truncation? I wonder if a good noninductive CSR would show the same flat top. It might be interesting to try both your current probe and a CSR on the same pulse.

I have only loosely been following this thread but haven't seen much focus on the actual Jim Watson device as blueprints for building.  It would seem that if that device was the one that caught the attention of so many people, that we should look at its clues.  Has anyone got their sights set on an energizer which uses an unequal number of magnets to coils?  You can see in the photos of the Watson machine that the magnets don't line up perfectly with the coils and that there is an unequal offset between the two..


Just thinking out loud and am curious if anyone is replicating in this direction..


Thanks,


Dave

[Void]

Void, that little flat spot has me thinking. Could it be an artefact of the probe you used? Out of range, signal truncation? I wonder if a good noninductive CSR would show the same flat top. It might be interesting to try both your current probe and a CSR on the same pulse.

Hi TK. Yes, I had thought that the squared off peak might just be due to my not so
very good quality current probe.   I was avoiding using a CSR because even with a
CSR value of 0.22 ohms and with the current pulse peaking at around 18A, that would
be a voltage drop across the CSR of about 4V. Also, the increased resistance will increase
the cap discharge time constant, making the cap discharge time take a fair bit longer. 

I repeated the test with a 0.22 Ohm nominal resistor (lowest ohm value I have on hand) for the CSR,
and it doesn't square off the peak of the current pulse, so it looks like my cheapo current probe was doing that.
See the attached scope shot. It shows a peak of 2.66V across the 0.22 Ohm CSR, so 2.66V / 0.22 Ohms
equals 12.1A peak. I believe the current peak is lower than when using the current probe because of the
added resistance of 0.22 Ohms. Also the current pulse duration was quite a bit longer at a little over 17.5 ms,
which again would be due to the addition of the 0.22 Ohms increasing the cap discharge time constant.
It just goes to show that in a setup like this you want to use appropriate gauge wires and make sure all
wire connections are clean and solid to reduce resistive losses as much as possible.

I repeated the Cap discharge test using the same 5300 uF cap bank charged to about 24V, and
discharged into a much larger lawn tractor type 12V lead acid battery, and the current pulse waveform
using my current probe looks almost identical to the waveform when discharging into the smaller
5 AH SLA battery, so the capacity and size of the battery doesn't seem to affect the current discharge
pulse waveform too much in any noticeable way.

Looks to me that battery presents some resistance for this very brief period, which eventually breaks down. 
Dirt at the conducts or an internal battery's characteristic?
And it is also this stair like decay if you zoom at the picture. Except if Void changed V/div after waveform's capture.

Hi Jeg. It looks like the squared off peak and maybe some of that jagged look to the waveform is
due to my cheapo scope current probe. I didn't change the V/div on the scope after capturing
the waveform. That was how it came out. Part of the jagged appearance may also be due in part
to image rendering aliasing. The scope display is pretty low resolution.

All the best...

[itsu]

I used my current probe (green) in a similar way as described by Void together with a measurement with a 0.1 Ohm csr (yellow).

I have a 4000uF cap loaded to 25.8V and a 12V 7Ah battery reading 12.8V.
The cap at 25.8C contains a charge of 1.33J
After dump into the battery, it contains at 12.8V still 0.327J, so we dumped about 1J into the battery, like Void did.

My current probe (green) set at it max setting (5A/Div.) also showed the flat topping like in Voids screenshot, with some spikes before due to mr Hand making the connection see screenshot 1.
With the 0.1 Ohm non inductive csr (yellow), no flat topping was seen and a whopping 87.6A (8.76 / 0.1) was measured being dumped into the battery.

Using a 1 Ohm csr makes a lot of difference as it seems to limit the current being dumped, see screenshot 2
Now we have 12.6A being dumped via the csr (12.6 / 1) yellow trace and confirmed by the current probe green.


So i think the current probes are being overdriven or saturated.


Itsu

[Void]

Hi Itsu. Thanks for doing that test. It is interesting that your current probe also
showed a squared off top on the current peak.  Strange that you measured such
a higher current peak than me using your 0.1 Ohm CSR, as your setup was similar
to my test setup. Maybe the test leads I was using have too much resistance and
that reduced the current peak quite a bit.

At any rate, it shows that even with a relatively smaller sized cap bank and charged only
to around 25V you should still expect quite large cap discharge peak currents, so if you
are using a relay to discharge the cap bank you may want one that has quite a high current
rating, like Tinman said he was going to try. If you are planning on using a commutator, then
keep in mind that the cap current discharge pulse could take up to 25 ms or possibly even longer,
depending on the resistance in commutator contacts and wiring connections.

All the best...