Bedini SSG - self sustaining

[Groundloop]

Fausto,

The new data sampler and relay board is almost ready soldered. Now I have to
make some software and start testing. I will be using the board in the image
to test the software, and when I'm done, then I will ship this board to you.
You can reuse the three IC's from the old board and just load the new SW
in the pic mcu. I'm still waiting for parts so it will take some time. (A week or two.)

Groundloop.

[mscoffman]

@mscoffman

Concerning a more advanced design for the future, I am on it. Thanks to Groundloop original design, I am expanding on it and creating a 12 switches circuit where I can not only do more but also I will be using a different PIC processor that I can real-time control via the USB of the computer.

I know you guys know how to do that already but for me it has been very difficult trying to program and test. There is lots to learn. I somehow think that a different PIC more inline with USB and ready for MICROCHIP products such as "PicKit 2 Debugger express" is easier for me and others to get ready boards and downloaders and other things to try this out.

>I always like to build an almost infinite expansion capability into these things
>because Like you already know, at first you want to build the u'p into
>the circuit to run it, and then you want to build some instrumentation
>into it, and then you want build some data archival, each step wants to
>control more and more digital lines. The downside to using a big version
>of the PIC chips is that is going to use more power when you eventually
>want to use limited power! I have used the 16F59 a 40pin chip and I
>programmed it with the standard PIC programmer.

> Also the key to software programming is getting some subroutines that
> you both completely understand and you know are completely reliable.
>This certainly takes time to do.

> I often find the easiest thing is to just patch some additional
>goodies into what you already have, then wait till the end to
>redesign it to the best possible way.

@mscoffman

The voltage raizing right after the short is simply phenomenal and not even logical in my opinion. I am suspicious that bigger the battery better it will be the raizing. I can barely wait for the weekend to try many things out. So little time.

BTW, since my board is out of order since last night, I noticed that the voltage on my battery is resting at a much higher voltage than even the previous achievements. Very weird. Soon I will also do a load test to see if it is indeed a real charge or ghost charge.

>You know when a battery becomes unloaded the voltage will rise
>with time. Like when you crank a car just a little to avoid using
>up the battery charge all at one time. The battery recovers some.

> Shorting a battery is definitely problematic, It will probably require a big
> relay (0ohms) whose contacts won't weld with the current.
> You may need to be satisfied with switching a low resistance high wattage
> resistor.

>Believe me, I have not said that "one size fits all" ie. lift-truck batteries
>are going to need to be handled the same way, but with
>vastly different switches.


[/quote]

@mscoffman

Oh, another important thing is, if you dont have at least a 4 digits voltmeter it will be difficult to see this effect at first, but if you somehow do that many times with intervals lets say of one hour in between the "shortings" you might see 1/10 of a voltage increase in a 7amp/h battery.

>By the way I was thinking if you need a very inaccurate voltage
>reading it may be possible to use the "control voltage pin" on
>an Ne555 timer to get timed pulses into the u'p then use a
>frequency to voltage conversion scheme to get an approximate
>but very inexpensive reading.

:MarkSCoffman

[Groundloop]

Fausto,

The SW for the new data sampler and relay unit is done and tested.
(Attached is the HW design files and the SW source for the mcu.)
I selected to put my unit into a metal box. Looks nicer that way.

The unit has 4 unipolar differential inputs for 0 to 40,95 Volt. The
resolution for each input is 0,01 Volt. The accuracy for each input
is +/- 0,01 Volt. (This is by design of the MAX1202 a/d converter.)
(It is possible to use bipolar differential inputs by changing the SW.)
The sampling rate can be set from fast (29 single channel samplings pr. second
and 7 four channels samplings pr. second.) to slow.(1 sec steps up to 99 seconds.)
Unit has a built in real time clock.
The unit can be controlled by using a terminal program. Best result will be
with a application program run on a host PC.

Your card is waiting for some parts. Will ship as soon as I get them.

Design files and newest SW can be downloaded here:
http://home.no/ufoufoufoufo/DS_RL_V_1.rar

Groundloop.

[mscoffman]

@plengo

Associated with your battery shorting problem. How about using
a relay follower like this;

http://shop.advanceautoparts.com/webapp/wcs/stores/servlet/product_Relay-BWD_5972280-P_409_R

It's cheap and could handle the high current. It may need to be fired
from an electrolytic capacitor, if you short it's only power source.
Just an idea.


---

@groundloop

This is a nice looking unit, with nice specs too.

---

@plengo

I am just throwing this out for your future consideration to
think about.

If you are running your PIC processor from a storage battery,
you may want to consider building a stand alone optoisolator
circuit (9pin to 9pin) to optoisolate your PIC processor from
your host PC for data transfer. This would segregate the
system ground from the PC ground for DC conductance
too. I'll try to find a circuit for this on the internet, it's
somewhat of a trick circuit but it will do the task.

:S:MarkSCoffman

[plengo]

Fausto,

The SW for the new data sampler and relay unit is done and tested.
(Attached is the HW design files and the SW source for the mcu.)
I selected to put my unit into a metal box. Looks nicer that way.

The unit has 4 unipolar differential inputs for 0 to 40,95 Volt. The
resolution for each input is 0,01 Volt. The accuracy for each input
is +/- 0,01 Volt. (This is by design of the MAX1202 a/d converter.)
(It is possible to use bipolar differential inputs by changing the SW.)
The sampling rate can be set from fast (29 single channel samplings pr. second
and 7 four channels samplings pr. second.) to slow.(1 sec steps up to 99 seconds.)
Unit has a built in real time clock.
The unit can be controlled by using a terminal program. Best result will be
with a application program run on a host PC.

Your card is waiting for some parts. Will ship as soon as I get them.

Design files and newest SW can be downloaded here:
http://home.no/ufoufoufoufo/DS_RL_V_1.rar

Groundloop.

Oh my, this is AWSOME man. Beautifull design!!!

Hey Groundloop, what kind of relay is that in use? It is a one pole being either connected to either of the other 2 poles? If it is, it is EXCELLENT, no more concerns with positive and negative and much more flexible concerning current throughput, BUT if they are relays or some form of mechanical switching would not that limit the frequency? I am using now 50khz! (in Brazil they say: not trying to throw sand into your soup, BUT).

Fausto.