ECW Orbo O-Cube Testing Week 2 (Feb 15 — Post Questions for Steorn Here)

The Week 1 O-Cube Testing thread has become very long, so here’s a new thread to start out the new week.

For reference:

Week 1 thread

The Google document started by Ged which summarizes key information from testing so far .

The Spreadsheet with the terminal readings from the ocube

The video posted by Steorn yesterday showing some of the inner workings of the Ocube

I will hopefully be talking with Steorn via Skype this afternoon, so if there are questions that you would like me to ask please post them in this thread and I will compile them.

Here’s a quick video of the wiring setup using stuff I have collected. As you can see, it’s quite a patch-up job. I would welcome any suggestions.

Meter readings on either side of the resistor in the unattached circuit

Feb 16 2016

It’s been a long and involved day and there is a lot of information that has been shared in the comments section below. But I thought I would give a quick recap of what has gone one.

According to the video Steorn put on Facebook on Sunday, for the Orbo cells to charge effectively, they need to be exposed to what Shaun McCarthy calls a “reference voltage”, and this is why the ocube has two 9 Volt batteries inside with a 1M Ohm resistor across them. This means that despite the relatively high voltage, the current is tiny. Shaun determined from information I sent him that the ECW Ocube was for some reason not getting this voltage, so he recommended that we apply it externally. So today, I built the circuit with two 9 volt batteries and a 1M Ohm resistor in it. A photo is below:


New photo by Frank Acland / Google Photos

After applying this circuit to the ocube, the charging time of the ocube has appreciably increased. See the spreadsheet linked to above to see data on charging times. You’ll also see that the test was interrupted when while I was out I think one of my cats dislodged one of the screws used in the terminals and a short was created.

I have been in touch with Shaun McCarthy today. I asked him how long he thought it should take to charge with this circuit applied:

Not sure to be honest, shorting the unit will have fecked the domains up, the purpose of the applied voltage is to keep the domains pointing in the correct direction (or rather encourage them to return after a discharge)
I think the first step here is to see if there is any rise in voltage, that will tell us if the circuit is ok

Its tough to say if the applied voltage is really across the power cells if its still rising itself – the rise is very non linear, it goes through a BH curve type of shape, I think leave it for 12 hours and lets see.

Feb 17, 2016

Some more comments from Shaun from a Skype text conversation between him, myself and Rob King:

Q: What sort of voltage climb will we see at this low starting voltage?

A: I don’t know the units them selves are never below 2.9 volts, and the short out will have effected the domain structure. The applied voltage will reinstate the domain structure, but the time frame I really don’t know to be honest.

Q: At what voltage will the charge controller kick in and start charging the cap?

A: the cap is on our side of the controller board, so it will be at the same voltage as you are currently measuring, with the obvious note that something in the circuit is wrong and its unknown what is wrong at this time

Q: so is the cap and orbo in parallel?

A: yep, pre controller board

Q: so there is no controller board and the cap and orbo are connecter to the USB board

A: no, there is a controller board in this non li version between the powercells and the USB board
In fact there are a bunch of them in parallel to allow a strong current draw

Q: so the controller board takes a wide range of input voltages and outputs 3.7 or 5V?

A: it outputs 5v from 1.6v up

Q: the 5V then feeds the usb board

A: yep

Q: what triggers the blue light on the USB board to light up?

A: when a cable is plugged in, if it stops charging then the cable needs to be plugged in again

Q: so it detects a load being attached, and when a load is detached

A: yep

Q: but does not step up any voltage?

A: no

Q: I can see what looked like and inductor but might be a shunt resistor

A: its a standard board we buy in

Q: so the USB is 5v in and 5v out

A: no, sorry, in fact its 3.2 to 5 in, and 5 out

Q: ok, so it must have a boost converter on board

A: yep

Q: the new controllers are true li-ion chargers?

A: there is no LI in the revised cube

Q: what is in the revised in place of it then?

A: with the appropriate number of cells we can store 20k joules, hence no battery

Here’s a link to nice chart that has been put together by Sanjeev providing data at a glance from the readings I’ve been logging:

https://docs.google.com/spreadsheets/d/1SmmDLw9-g_Mamtk7qloWEeWNacHwEBpXxdwdTXnAZpc/pubchart?oid=1920631363&format=interactive

Feb 18, 2016

Here are some comments from Shaun McCarthy today regarding the current issue:

If you short the cells you destroy the domain alignment – we normally just throw them away, never tried to realign the domains before

In a way, however they have not been shorted – that huge spike of voltage is going to feck the domains around – not saying that we cant get them back alaigned, just saying that we have never tried

February 19, 2016

The day started off with something of a surprise. For some reason, there was a sudden spike in the voltage readings across the terminals U1-U2, which according to Steorn measures voltage across the Orbo powerpacks. See the spreadsheet above for the readings. Since the spike to 2.96 V, the reading has gradually decreased. Below are photos of the most recent readings


New photo by Frank Acland / Google Photos

New photo by Frank Acland / Google Photos

Here are some more Q&As with Shaun McCarthy about the current situation:

I showed him the chart of the voltage rise this morning:

SM: Its very strange behaviour, not sure what to make of it, maybe a couple of the cells are destroyed, I think leave it and see, the rise and then decay is strange

Me: Here’s a question about shorting. In the video when you showed the ability for the orbo material to bounce back from a short — you had it shorted out for over 1/2 an hour without any apparent degradation. Is it different in these packs?

SM: yes, but that was without an applied field, as you can see in that video its low voltage. We apply the field to ‘upgrade’ the output – so a short with a lot of energy can really feck around with the domains, so in a way it may seem the same but its like comparing falling off a bar stool to falling out of a plane. The strips in the video have no capacitance so there is no huge amount of energy stored, in the production cells there could be 1500J in storage, you short that and you are going to have a minimum of a couple of hundred watts spiking the domain structure

RK: So this is a common problem you have seen, that shorting the production cells does them no good or kills them

SM: Well it can hurt, that’s for sure – we try not to do it, the heat generated can be very high, a simple measure of the amount of energy stored is the amount that frank took out on his first charge, the only storage in the system is a 5f cap and the cells, so it will give you some idea of how much energy is released in a short, the difference being that in a short the energy is dumped in an instant, that simply causes issues.

Feb 20th 2016

Below is a video that I made this morning of a quick test using an LED that Rob King has sent me for testing purposes. We tested it out using the new multimeter and its rating is 0.22 W on the lowest setting (there are three settings) which I used in this video.

Here’s an explanation of the steps I took.

1. Measured voltage across points U1-U2 (apparently the voltage of the Orbo packs)
2. Measured voltage across the USB port.
3. Plugged in USB light until it was depleted
4. Measured voltage across points U1-U2
5. Measured voltage across the USB port.

The video was made at about 9:25 am, and you will see from the spreadsheet that the voltage acorss U1-U2 has climbed since then — no circuit has been applied to the ocube.

I repeated the test at 11:57 — video below:

Here’s another test. This time we have a 100 Ohm resistor across terminals U1-U2 for one minute, with voltage being measured.

Feb 21, 2016

I did another test this morning. Overnight I put on a reference voltage circuit off 18 V with a 100k Ohm resistor which brought up the voltage measurement from 2.85-3.18 V. When I took off the circuit, the voltage across U1-U2 f dropped from 3.18 – 3.169 Volts over a period of about 16 minutes. I am interested to see whether putting a load on the ocube can actually stimulate the “self-charging” behavior of the orbo cells.

Feb 23, 2016

Below is the first video of the charging test started yesterday. More to follow. In this video there is a 9V battery with a 220 Ohm resistor charging the ocube via the U1-U2 ports.

This is the second part of the test done this morning continued from yesterday. This time I used a 100 Ohm resistor with the 9V battery.

Part 3 — still with the 100 Ohm resistor. You will see the battery is losing power.

  • OM

    4.7v

    In short, the device acts as 4×1.2v accumulators.

    http://img1.114pifa.com/2059/ZLatSBgHC_1398257698.jpg

  • http://www.overunity.com/ Stefan Hartmann

    Frank have you been disconnected the internal 2x 9 Volts batteries ? Otherwise all these tests make no sense as you don’t know if these charge controllers are blown….
    When do you expect the replacement shipment from Steorn ?

  • Tony norris

    Can’t you just boost the output of the usb from the cube to 9v and run it on itself? This would elminate having to use 9v battries

  • http://www.russgeorge.net/ russ george

    What about adding some rechargable Li batteries to the testing… see if the Orbo will charge them repeatedly being sure to show the batts fully discharged before each Orbo feeding.

  • SG

    4.68V -> 4.66V
    ~ 239 seconds

    At this level of charge, the Ocube is behaving like it has approximately 558 F of capacitance. It is a rough estimate because the start/stop times require a bit of guesswork given that the meter shows fluctuations around each 100th of a voltage change. In any case, the amount of capacitance demonstrated, as has been said before, is quite remarkable. It is also interesting that the Ocube quickly recharged itself to the 4.68V level after the discharge test.

    Here is what I suggest next: let’s add the rated bias circuit (2 fresh 9V batteries + 1Mohm resistor) and let it sit for a day or so to see if it will self-charge (in bias assisted mode) any further. Then we can follow up with some longer-term discharge tests.