Fusionchip's Bedini Feedback to Source!!!

[plengo]

@ Fausto

A possible problem I can see using two batteries in series, is that I have found that when there is a difference in voltage between the batteries, one can rise in potential during discharge. I'm not saying that this is happening in your case but it is possible.

I have now reduced my fan speed as much as possible whilst remaining stable and I will monitor this over the next few hours. If I detect a drop in voltage I will then reposition my magnet. I'm using a flat hard disk magnet as per Gadget's suggestion. Once we can get our batteries charging it will be interesting to see how high the voltage will climb and whether it will stabilise at a certain level or eventually fall. Gadget reports that he can 'fry' his batteries so it will be exciting to see if we can get as advanced as him.

Hoppy

I agree with you on the two battery case. I had that. My test is very simple. IF the voltage in one of the batteries, usually the smaller one, goes up and the other stays, it is a good sign but not necessarily the evidence YET. I wait a long time, usually about 4 hours, to see if they either stabilize or go up. IF go UP and the other stays, yeaaaaa, otherwise wait longer until you see that by the time running would be impossible not to have the voltage going down at 7ma. I roughly calculated that in 4 hours at 7ma at least the SLA should loose about 0.01v while the big battery does not change.

To both change I only need about 24 hours to loose 0.01v. That's is my way of knowing that guaranteed it is loosing power.

So, when I have my setup with 24v and a small SLA I will know very fast, in minutes, the direction of the setup.

It would be good if we could come up with a proof of concept that demonstrates what is happening here and be able to test faster.

Fausto.

[Hoppy]

I agree with you on the two battery case. I had that. My test is very simple. IF the voltage in one of the batteries, usually the smaller one, goes up and the other stays, it is a good sign but not necessarily the evidence YET. I wait a long time, usually about 4 hours, to see if they either stabilize or go up. IF go UP and the other stays, yeaaaaa, otherwise wait longer until you see that by the time running would be impossible not to have the voltage going down at 7ma. I roughly calculated that in 4 hours at 7ma at least the SLA should loose about 0.01v while the big battery does not change.

To both change I only need about 24 hours to loose 0.01v. That's is my way of knowing that guaranteed it is loosing power.

So, when I have my setup with 24v and a small SLA I will know very fast, in minutes, the direction of the setup.

It would be good if we could come up with a proof of concept that demonstrates what is happening here and be able to test faster.

Fausto.

Curiosity has got the better of me and I had to answer a question that had been bugging me for a while. I started the test at 22:30hrs on 23 Feb at an on-load voltage of 12.68V with the fan pulling 100mA. This current was stepped down in increments to 20mA where I terminated the test at 16:50hrs today. A total of 3.12A/hrs was extracted from my 20A/hr battery over the duration of the test. The final on-load voltage was 12.65V, so I was naturally curious about why the terminal voltage was not a lot lower than this. I therefore decided to terminate the test and run the motor at 300mA to see the effect on the battery. Within two minutes the battery voltage fell to 12.52V which is 0.13V lower than the on-load voltage at the start of the test. This of course does not in any way suggest that a gain had not been made over and above battery capacity but it is nevertheless interesting in that it shows that terminal voltage is a very bad indicator of battery state of charge at very low discharge currents, which makes it very difficult indeed to detect and quantify any genuine capacity gains that might be achieved using this particular 'feedback to source' test setup.

Hoppy

[plengo]

Curiosity has got the better of me and I had to answer a question that had been bugging me for a while. I started the test at 22:30hrs on 23 Feb at an on-load voltage of 12.68V with the fan pulling 100mA. This current was stepped down in increments to 20mA where I terminated the test at 16:50hrs today. A total of 3.12A/hrs was extracted from my 20A/hr battery over the duration of the test. The final on-load voltage was 12.65V, so I was naturally curious about why the terminal voltage was not a lot lower than this. I therefore decided to terminate the test and run the motor at 300mA to see the effect on the battery. Within two minutes the battery voltage fell to 12.52V which is 0.13V lower than the on-load voltage at the start of the test. This of course does not in any way suggest that a gain had not been made over and above battery capacity but it is nevertheless interesting in that it shows that terminal voltage is a very bad indicator of battery state of charge at very low discharge currents, which makes it very difficult indeed to detect and quantify any genuine capacity gains that might be achieved using this particular 'feedback to source' test setup.

Hoppy

Excellent point Hoppy. This is a very good insight of yours. When I did my over 200 load tests on my SSG i noticed that the battery will have a "curve" that is very peculiar to its load rate (what we call C10, C20, C??). Higher load rates the slope of the curve is higher, lower the load, smaller the slope. Also the curve is in the shape of stairs going down (see the picture). In any case the slope will be there no matter how light the load may be, even on as little as 7ma, but still be there. What will change will be the stair dimension and how long you will be in a certain "step". I even notice sometimes that the voltage can go up but not much while under load, but always in the end of a long run it will go down.

So voltage will not be a very good demonstration of the state of the charge but will definetely tell you where it will not be. In my case this battery of mine is a "baby" because I already know how she reacts to different loads and times because of the shear amount of tests I did on her. So for me she tells me a lot when the voltage changes at a certain rate at the voltage level she is sitting now, around 12 volts.

In the picture you will see different load rates and curves to demonstrate my point. Those curves are not real, real but they are very close to some of my load tests that I still have the data. They resemble a lot my real physical data.

You CAN know where you are on the battery by how it is behaving in voltage fluctuations while you are running the system for a while. You CAN know if you are on the "plataform" or on the "curve of the step" and even know the rate she will fall if you know your battery.

Fausto.

[Hoppy]

@ Fausto

Yes, I call this layering and I too have seen this whilst spending too many hours load testing SG's. Its not so apparent when conventional DC charging. I tend to think its bad cell balance as a result of uneven desulfation during repeated shallow load test cycling. Anyway whetever its real cause its good that it provides you with a reference to which you can measure and interpret voltage variations more accurately. I'm continually learning more about batteries and they are complicated beasts, almost living entities with their own personalities!

Hoppy

[plengo]

@ Fausto

Yes, I call this layering and I too have seen this whilst spending too many hours load testing SG's. Its not so apparent when conventional DC charging. I tend to think its bad cell balance as a result of uneven desulfation during repeated shallow load test cycling. Anyway whetever its real cause its good that it provides you with a reference to which you can measure and interpret voltage variations more accurately. I'm continually learning more about batteries and they are complicated beasts, almost living entities with their own personalities!

Hoppy

In the end I am totally with you Hoppy. They are indeed a living thing with their behavior.

Off course I will only really say that this Fan circuit is really working "after" I have this tested under many different batteries and not ghost voltage as a result, if this really charges the batteries to voltages way higher than when it started, more like from 12v to 13.65v or 14v than I will really believe it myself, backed up by evidence that would be undeniable.

For now, all I can do is test this little mysterious design over and over and over and over again with the best tools that I have at hand.

Fausto.