Author Topic: Ghurd controller with 7805 or similar  (Read 8119 times)

0 Members and 1 Guest are viewing this topic.

SpmP

  • Newbie
  • *
  • Posts: 23
Ghurd controller with 7805 or similar
« on: April 22, 2012, 08:06:15 PM »
Good day All,
 I am looking at the Ghurd shunt controller and wondering if there is a reason why the BDO chip (ZM33064) could not be replaced by a linear voltage regulator such as an LM7805 (or LM317 etc) and a comparator other than component count?

... and... if I was to build this circuit using a 7805, could I overcome the issues, if there are any?

Cheers

Jasper


vtpeaknik

  • Full Member
  • ***
  • Posts: 136
  • Country: us
Re: Ghurd controller with 7805 or similar
« Reply #1 on: April 23, 2012, 03:06:09 PM »
Hopefully ghurd himself will explain why he chose the approach he did.  I know it works!   The 7805 approach would use more quiescent power for sure.  But how about using the MAX8212 chip, which is designed to do exactly what we're talking about here, with minimal power draw, and "programmable" threshold voltage and hysteresis?  Here's an example use of that chip, by a person who posts here too:
http://www.earth.org.uk/low-voltage-drop-out-circuit-design.html

DamonHD

  • Administrator
  • Super Hero Member Plus
  • *****
  • Posts: 4125
  • Country: gb
    • Earth Notes
Re: Ghurd controller with 7805 or similar
« Reply #2 on: April 23, 2012, 03:55:54 PM »
I thought that was starting to look familiar!  B^>

1) I don't want ghurd-megacorp-inc sending out its thuggish tentacles to destroy my feeble competition: I'm told that with the white cat and the laser beams Mr G is quite scary (though the fish in the moat are just goldfish).

2) The ghurd design is much more thoroughly tested than mine, so bear that in mind.  Mine has a user base of, well, 2 now if one of you guys starts using it.  10 more and I could be a dozenaire like him.

Rgds

Damon
« Last Edit: April 23, 2012, 03:58:17 PM by DamonHD »
Podcast: https://www.earth.org.uk/SECTION_podcast.html

@DamonHD@mastodon.social

SpmP

  • Newbie
  • *
  • Posts: 23
Re: Ghurd controller with 7805 or similar
« Reply #3 on: April 23, 2012, 10:22:13 PM »
Thanks guys,

I am needing to use the LM7804/LM317 approach as my friends who will also use this circuit want a circuit that they can build from scavangeable and common parts. Both the bod chips are neat, but not scavanageable.

So it looks like the problem here is the hysteresis! I had a play with hysteresis on coparator/opamps in LTSpice and it is a giant mess! From http://www.ecircuitcenter.com/circuits/op_comp/op_comp.htm . I found I could not control what I wanted very well.
 Is there another way to get hysteresis off a single linear reg and comparator/opamp, or have I mucked up my spice sims 8)

I tried with two levels, one from linear reg, and another from voltage divider (pot) off that level, but as Vreg seemed to depend on the pot position it seems like an unstable solution.

Cheers.

oztules

  • Hero Member
  • *****
  • Posts: 1477
  • Country: aq
  • Village idiot
Re: Ghurd controller with 7805 or similar
« Reply #4 on: April 24, 2012, 04:21:44 AM »
Have a good look at the circuit from Larny. His version of hysteresis is worth a second look. It is completely controllable and .... well thats the thing.

It uses a cunning system to conceive the hysteresis.



.............oztules
Flinders Island Australia

SpmP

  • Newbie
  • *
  • Posts: 23
Re: Ghurd controller with 7805 or similar
« Reply #5 on: April 24, 2012, 04:52:04 AM »
Thanks Oztools,
 are you talking about http://ronald48.lnk.telstra.net/forum/forum_posts.asp?TID=1762&PN=0&TPN=1 from our favorite shed?

If so, I have pretty much just finished the schematic, and now doing the pcb layout for a very very similar circuit that turns a relay for a battery charger on and off... Will post when its done, and _if_ it works. It way more simple tho, and got the component count right down. But same principal, LM317 and voltage divider + 2 comparators.  Damn cunning b*Strd! putting the 10K pots in parallel! nice.

If this is not the circuit you mean, what is pls.

XOKE

  • Jr. Member
  • **
  • Posts: 90
Re: Ghurd controller with 7805 or similar
« Reply #6 on: April 24, 2012, 06:12:02 AM »
Hi,
I use Motorola MC33064 and works fine until now

regards

xoke

ghurd

  • Moderator
  • Super Hero Member Plus
  • *****
  • Posts: 8059
Re: Ghurd controller with 7805 or similar
« Reply #7 on: April 24, 2012, 08:43:17 AM »
"replaced by a linear voltage regulator such as an LM7805 (or LM317 etc) and a comparator other than component count?"

I am not sure I get the reason behind doing that since all "the magic parts" would be different.

In my circuit, the right hand side, from the 2N7000 to the 'output to fet gates', is sort of to get sharp switching and high voltage (above logic level voltages) to the power fet gates.
It is sort of ghurd's version of Darlington-izing mosfets to get the power fet gates up to >12V instead of <5.1V.

Might look for a simpler opamp / comparator circuit with feedback... something that uses a Zener instead of 7805.

Many DIY circuits I tried years ago had stablity issues related to duty cycle or temperature swings.  Close enough to be good enough for something made from salvaged parts.
Not sure what parts caused it, I just know it happened.  Pots, Zeners and ICs often are effected.
G-
www.ghurd.info<<<-----Information on my Controller

oztules

  • Hero Member
  • *****
  • Posts: 1477
  • Country: aq
  • Village idiot
Re: Ghurd controller with 7805 or similar
« Reply #8 on: April 24, 2012, 06:17:23 PM »
Sorry, I screwed up, and forgot the link... it is this one... from the same article/thread
http://ronald48.lnk.telstra.net/forum/forum_posts.asp?TID=1762&PN=0&TPN=4

If this does not contravene any rules it is this one..... else admin...... kill it.

5456-0

Have a good look at how the hysteresis works... it is dead cunning, and may take a bit to figure out first off, but worth the look I feel. As the circuit changes state, it references different reference points (switching op amps on and off as the reference points) They cant cross interfere.

He gives a very good run through in the post referenced above. I intend to use this hysteresis technique in the future.... when time permits... sigh...

...............oztules
« Last Edit: April 24, 2012, 06:26:45 PM by oztules »
Flinders Island Australia

SpmP

  • Newbie
  • *
  • Posts: 23
Re: Ghurd controller with 7805 or similar
« Reply #9 on: April 24, 2012, 07:37:51 PM »
I am not sure I get the reason behind doing that since all "the magic parts" would be different.
I was realising this, so it couldn't even be a corruption of the _Ghurd-Controller_ 8), in this instance it is no longer a personal pronoun, it is a product and idea all to itself.... Feel sorry for the guy whos name is Xantrex.

In my circuit, the right hand side, from the 2N7000 to the 'output to fet gates', is sort of to get sharp switching and high voltage (above logic level voltages) to the power fet gates.
It is sort of ghurd's version of Darlington-izing mosfets to get the power fet gates up to >12V instead of <5.1V.

Might look for a simpler opamp / comparator circuit with feedback... something that uses a Zener instead of 7805.
If I understand a little of the operation of your circuit, it was that due to the small hysteresis and sharp cutoff, the switiching effect was akin to PWM, in that it is of a high frequency, is this correct?

 When I looked at using a zener, all the info and expert-opinions were that zeners have way too much temperature drift, are unreliable off the shelf (voltage wise), and especially bad down the low end, 4.7/5/8V.
 Personally it is what I wanted to use, so if the temperature stability is in reality a non issue, then the inter item voltage variance becomes a non issue, as the o/p needs to be calibrated anyway. Is there any reason why two 100k pots in parallel with the zener would not work, and for example with a 14V supply and say a 20K series resistor?

I am not sure I fully understand how feedback on opamps work to create hysteresis, something to learn and play with I guess... but sounds like the promise of one vRef, and one comparator/opamp.

 The other issue with zeners is: Are they easily scavangeable? A LM7805 etc. is probably more common in day to day stuff... its a funny design spec, scavengability.

Many DIY circuits I tried years ago had stablity issues related to duty cycle or temperature swings.  Close enough to be good enough for something made from salvaged parts.
Not sure what parts caused it, I just know it happened.  Pots, Zeners and ICs often are effected.
But is close enough good enough? I tried various combinations of things in LTspice with temperature step, and it thinks it is the opamps that are the worst... nothing like the realworld tho.

Cheers.

SpmP

  • Newbie
  • *
  • Posts: 23
Re: Ghurd controller with 7805 or similar
« Reply #10 on: April 24, 2012, 08:21:42 PM »

Have a good look at how the hysteresis works... it is dead cunning, and may take a bit to figure out first off, but worth the look I feel. As the circuit changes state, it references different reference points (switching op amps on and off as the reference points) They cant cross interfere.

Thanks, I found this just after I posted the first link 8)
 It is dead cunning and I havnt got it quite sussed yet. Well, anywhere near it yet 8)
The question will be in regards to an analgous controller to Ghurds, whether it will switch state fast enough. Would it work with a normal opamp like LM324 I wonder. Maybe I should ask directly when I understand the circuit more.

ghurd

  • Moderator
  • Super Hero Member Plus
  • *****
  • Posts: 8059
Re: Ghurd controller with 7805 or similar
« Reply #11 on: April 24, 2012, 09:00:04 PM »
That's a lot to reply to in the time I have available at the moment.... but here goes...

"Feel sorry for the guy whos name is Xantrex."
Oh yea?  Try ghurd?

"If I understand a little of the operation of your circuit... in that it is of a high frequency, is this correct?"
Basically, yes.
I can make in run ~5MZ if I want to, but I do NOT because it is not a good idea.

"the switiching effect was akin to PWM"
Correct.
It is a self varing PWM, with a self regulated pulse width and frequency.
(a friend had that put into a GREAT sales pitch for something like ebay... if he wants to post it  LOL)

I set out to build a simple charge controller, not a clock.  The battery doesn't care what time it is, so there is not much related to time in the ghurd controller.

"zener... temperature drift... especially bad down the low end, 4.7/5/8V"
Actually, 5.1V is most stable from what I recall.
And a guy who does stuff with huge temp variations, in stuff bigger than campers rated in numbers I can not even imagine, warned me about using pots as variable resistors (instead of as voltage dividers).
The 5.1V Zener ZD2 in the ghurd controller is to stop flow in the body diode of the ZM.
The 5.1V ZD1 temperature effect is small because the pot (all 3 pins) is doing most of the calculating.

Zeners are everywhere.  Not sure I would bother removing a used one to try a circuit since they are so cheap.
If something does not work, how will you know if something salvaged did NOT come from an item smoked by lightning?
5.1V ZD must be common with all the USB powered devices on the market now?

"But is close enough good enough?"
I can not answer that for you.
I know it was not good enough for me.
(search the board for "bioelectrifier" or ifred?)

"spice" is for food.  LOL
Might give an idea where a SERIOUS problem exists, but I don't trust it... at all... for this kind of thing.
Then again, I never do anything "normal".
G-
www.ghurd.info<<<-----Information on my Controller