Storing Cookies (See : http://ec.europa.eu/ipg/basics/legal/cookies/index_en.htm ) help us to bring you our services at overunity.com . If you use this website and our services you declare yourself okay with using cookies .More Infos here:
https://overunity.com/5553/privacy-policy/
If you do not agree with storing cookies, please LEAVE this website now. From the 25th of May 2018, every existing user has to accept the GDPR agreement at first login. If a user is unwilling to accept the GDPR, he should email us and request to erase his account. Many thanks for your understanding

User Menu

Custom Search

Author Topic: Graham Gunderson's Energy conference presentation Most impressive and mysterious  (Read 193162 times)

Spokane1

  • Full Member
  • ***
  • Posts: 115
Oh, I forgot to annotate the signal path. The BNC on the right looks like a signal input for the 74ac14 hex inverter. It looks like the signal is going through 2 gates on one side and 3 gates on the other side, so the two brown wires coming from the 74ac14 are carrying cleaned up and squared-off in-phase and out-of-phase versions of the input signal. Where they go? I dunno, maybe to the H-bridge gate drivers. The other BNC at the top.... I dunno.

Dear TK,

Good work on annotating that photo. It takes some time to put those labels in there and re-post it. It looks like our conclusions are pretty much the same. We don't know what he has going on there. Fortunately were are pretty sure what the purpose of these components are and we can move ahead with our own design for this sub-system.

You are correct on the unknown voltages. I used +12 and +5 as reference points. The actual voltages are probably higher since most all of the IC's are CMOS.

You are also correct on the presence of off-board connections. This close up photo was taken after the convention when Graham was doing further experiments. I suppose he had a variable pulse generator and a frequency meter hooked to the board. Rieyuki has a nice photo of the logic board configured for demonstration mode. I shall draft up both versions. (In progress)

Thank you for your technical contribution.

Spokane1

Spokane1

  • Full Member
  • ***
  • Posts: 115
Dear TK,

Do you happen to have any idea as to what the function is of this sub-circuit is in the bottom half of the attached schematic? My first WAG is that it is some kind of fixed voltage reference network, if so that is a lot of components that could be replaced by a single voltage reference chip.

Thanks in advance.

Spokane1

TinselKoala

  • Hero Member
  • *****
  • Posts: 13958
Dear TK,

Do you happen to have any idea as to what the function is of this sub-circuit is in the bottom half of the attached schematic? My first WAG is that it is some kind of fixed voltage reference network, if so that is a lot of components that could be replaced by a single voltage reference chip.

Thanks in advance.

Spokane1
The whole circuit doesn't make sense to me. Did you sketch that from the photo of the breadboard? I admire your perseverance and determination!

But I've certainly never seen a dual comparator used like that. Checking the pinout of the ts372 and the schematic,  I notice that the inverting input of C1 comparator is connected directly to the noninverting input of the C2 comparator and to nothing else (Pins 2 and 5). And the resistor-diode string leading to the Pin 3 noninverting input of C1 is completely bypassed by a length of wire (the WT-OR wire on the right side). Not only that but the diode in the string is reverse-biased, which makes no sense, so no wonder it's bypassed. So let's say that there is a small voltage coming from the bottom part of the circuit. This voltage is then present at the Pin 3 noninverting input of C1. Since there is nothing to compare it to (no voltage at Pin 2) the C1 output at Pin 1 goes high and stays there. Meanwhile some small fraction of the voltage from the bottom part of the circuit goes through the diode-resistor string on the left and is presented to the Pin 6 inverting input of C2. Again, there is nothing to compare this voltage to (no voltage at Pin 5 which is only connected to Pin 2). So the output of C2 goes low and stays there.

If the voltage labelled Vcc supplied to the bottom part of the circuit is a constant DC voltage, then that part of the circuit can only supply a fraction of half the Vcc voltage to the comparator inputs, and this will also be constant, I think.  Why all the components there.... I have no idea. You are right, if it's supposed to be some kind of regulated voltage reference it would be far easier and simpler to use a TL431 circuit there.

This is just a first pass, I could be completely wrong. I may try to breadboard the circuit later on to see what actually happens. I don't have a TS372 on hand so I'll have to try it with a different dual comparator or op-amp though. It is possible to configure comparators as oscillators but I don't think that's happening here, and that connection between pins 5 and 2 has got me very puzzled. Comparators are supposed to _compare voltages_ coming in to their inverting and non-inverting inputs. If one or the other input has no voltage supplied to it -- what's the point? Generally one input is used to read a set reference voltage and the other input is the "test" or varying voltage that you want to compare to the reference. In this case the "reference" voltage for both comparators is floating, undefined, since pins 2 and 5 are simply connected together and to nothing else. Or maybe the "reference" voltages are coming from the bottom circuit, in which case they are being compared to nothing--the floating pins 2 and 5. Maybe it would oscillate randomly under those conditions. Makes no sense to me.

The connection between 2 and 5 is a common way of setting up a "window comparator" circuit but it is always connected to a reference voltage, not just to each other.

gotoluc

  • elite_member
  • Hero Member
  • ******
  • Posts: 3096
I've been playing with this a little and it looks like I can get a similar wave form with just using 36vdc chopped by a mosfet switch around 21.7kHz with a 33% duty cycle to a primary coil which has a 0.002uf capacitor connected in parallel tuned to resonance (21.7kHz).

Yellow is a x100 Voltage probe across primary coil, Blue is a Tektronic P6021 current probe on negative of primary coil, Red is the product (math) of V x I and Purple is the DC output of the secondary (after FWBR to a 27,000uf cap with 20 Ohm load).

Luc


Hi Ben,
nice to see you here

the coil was around 10mH with a ferrite core. Use your signal generator (at 33% duty cycle) to turn on and off the mosfet's current to the coil which has a .002uf  to .005uf cap connected in parallel across it. Tune the frequency till you get (single Wave Ring) Resonance.

Kind regards

Luc


Sorry Ben and any others that have been trying to replicate the sine wave (posts above) with the information I provided.
Looks like the effect only happens with a bi-directional switch which I was using at the time and did not realize it was that configuration allowing it to happen.
Today, I was trying to make it work with a single mosfet and got nothing close. Then connected the bi directional switch and it works. So it looks like you need 2 mosfet's triggered at the same time with one flipped so the current can go in either direction through your switch.

Sorry I missed that.

Luc

TinselKoala

  • Hero Member
  • *****
  • Posts: 13958
Is it possible that the .01 cap on Pin 6 actually connects over to Pin 5 instead of to Ground?

Still makes no sense though.

Spokane1

  • Full Member
  • ***
  • Posts: 115
Is it possible that the .01 cap on Pin 6 actually connects over to Pin 5 instead of to Ground?

Still makes no sense though.

Dear TK

That was certainly quick work, Graham said that only 25% of the components on that board were actually being used. There is a good chance that that IC was not being used during the demo or at the shop (since the connections haven't changed).

IF you are interested in digging deeper into this logic circuit send me an email to mmckay@simplexgrinnell.com and I shall send you the raw photos I took at Graham's shop. They are too big to post at full resolution.

I'm just about finished with the overall wiring schematic for the board. Next is to do the Logic layout so we can establish how the various timing signals were derived.

Thanks again for your time and expertise.

Spokane1

Spokane1

  • Full Member
  • ***
  • Posts: 115


Sorry Ben and any others that have been trying to replicate the sine wave (posts above) with the information I provided.
Looks like the effect only happens with a bi-directional switch which I was using at the time and did not realize it was that configuration allowing it to happen.
Today, I was trying to make it work with a single mosfet and got nothing close. Then connected the bi directional switch and it works. So it looks like you need 2 mosfet's triggered at the same time with one flipped so the current can go in either direction through your switch.

Sorry I missed that.

Dear luc,

No Problem with the oversight, is happens all the time, even at VW.

Would the requirement of a bi-directional switch in your circuit imply that the H-Bridge configuration will be a requirement in the Gunderson Circuit?

Spokane1

Luc

TinselKoala

  • Hero Member
  • *****
  • Posts: 13958
OK.... er.... hmmm.  I breadboarded the circuit, sort of, and got some pretty strange results. It's a weird sort of window-oscillator, kind of unstable, but I finally got some consistent results out of it.

First, I had to use a TL082 op-amp as the comparator. This is another high-input-impedance dual opamp with the same pinout as the one Gunderson used. I didn't have any 0.01 uF caps to spare so I used 0.1 uF caps instead. This may affect only the frequency of oscillation or it may have other effects different from the original. When I can scrounge some 0.01 uf caps I'll re-test.  And for the bottom part of the circuit I simply subbed a variable voltage source (a pot across the supply rails with the wiper to the bottom of the 1k resistor in the center of the schematic.) I supplied the circuit with a little over 5 VDC.

The scopeshots below show the results as I varied the voltage input from the pot. The Yellow trace is the output of Comparator 1 (Pin 1), the Blue trace is the output of C2 (Pin 7), and the Purple trace is the applied voltage from the pot to the bottom of the 1k resistor.

k4zep

  • Hero Member
  • *****
  • Posts: 650


Sorry Ben and any others that have been trying to replicate the sine wave (posts above) with the information I provided.
Looks like the effect only happens with a bi-directional switch which I was using at the time and did not realize it was that configuration allowing it to happen.
Today, I was trying to make it work with a single mosfet and got nothing close. Then connected the bi directional switch and it works. So it looks like you need 2 mosfet's triggered at the same time with one flipped so the current can go in either direction through your switch.

Sorry I missed that.

Luc

Hi Luc,

This is the best I can get with a single ended, actually just driving a circuit off a pulse generator.  The waveform is not symetrical due to that fact but it does work.  I think your
assessment is correct!!!

Ben

« Last Edit: July 28, 2016, 01:50:53 AM by k4zep »

TinselKoala

  • Hero Member
  • *****
  • Posts: 13958
Er.... nice work, but see what happens when you post a huge horizontal picture?

Please shrink it to 800 pixels wide and reupload it to replace the overly wide one, or simply use the scope's screen capture utility (stick a USB stick in the hole, press the green "print" button under the Help button.)
Not only will the screenshot look a lot better but it will be _much_ smaller in total file size, and it won't screw up the forum page from being so wide.

gotoluc

  • elite_member
  • Hero Member
  • ******
  • Posts: 3096
Yes Ben, we have to down size pictures here before we post them or else the topic page size gets to be a problem.
Just open the pic in Windows Paint and click on re-size and select pixels and make them 800 to 1000 max wide and save.


Attached is a 1000 pixel re-size


Are you going to try the dual (bi-directional) mosfet's?


Thanks for sharing


Luc

gotoluc

  • elite_member
  • Hero Member
  • ******
  • Posts: 3096
Dear Luc,No Problem with the oversight, is happens all the time, even at VW.
Would the requirement of a bi-directional switch in your circuit imply that the H-Bridge configuration will be a requirement in the Gunderson Circuit?


Spokane1


Seeing my oversite now, my best guess is this would better support the use of an H-Bridge in Graham's device.


Luc


BTW, you must of deleted " quote " at the bottom of my post and why your message is in the quote body.

Spokane1

  • Full Member
  • ***
  • Posts: 115
OK.... er.... hmmm.  I breadboarded the circuit, sort of, and got some pretty strange results. It's a weird sort of window-oscillator, kind of unstable, but I finally got some consistent results out of it.


Dear TK,

I might have missed a component (or connection) in my take off so it is amazing that you got that circuit to run at all on the first go around.

An oscillator makes good sense since it connects to the H-Bridge ring timer just above it.  I was wondering if I was going to find out where the master clock was hiding.

Good work. I know that these distractions are taking time from your main project.

Spokane1

TinselKoala

  • Hero Member
  • *****
  • Posts: 13958
Well, I didn't exactly get it to run "the first go round". And I'm not sure still whether I'm seeing the actual quasi-stable performance of the circuit or just some weird breadboard artifact.

But my "main project" at the moment is just trying to get some sleep. And I haven't been sleeping very well lately anyhow, so doing this stuff isn't really distracting. Much.

It's a weird way of implementing an oscillator though.

k4zep

  • Hero Member
  • *****
  • Posts: 650

Seeing my oversite now, my best guess is this would better support the use of an H-Bridge in Graham's device.


Luc


BTW, you must of deleted " quote " at the bottom of my post and why your message is in the quote body.
 

All,

Sorry about the oversize picture, forgot about downsizing on my MAC. 

Ben K4ZEP