Eldarion and Bruce's build of Bob's Energy Converter

[eldarion]

All,

It's probably time for an update here.  As you know, I am working on a custom SMD circuit board for the controller while I wait for the bricks to arrive.  I should have the bricks in hand by Friday, and have already received the UCC27322P driver chips.

A warning to all:  When purchasing items off of Ebay with PayPal, do not use your checking account!  It will take over a week to go through (the reason that I do not have the bricks already.)   I had to call up the surplus store directly and use a credit card instead.

Hopefully things will get more interesting here on the weekend!

Eldarion

[eldarion]

Well, the IGBTs finally came!

One of them is shot (drawing 1.5A when "off"), but the other three checked out OK.  That one will be going back to be replaced.

I went ahead and hooked everything up and ran some frequency sweeps.  I was using a 40W bulb for a load, and surprisingly it lit, albiet dimly, at 5000ns pulse width, drawing about 30W from my power supply.  As the frequencies were swept around, it grew brighter and dimmer, although it never approached normal 40W brightness in my short test.  Disconnecting my scope ground from the signal generator caused the bulb to become slightly brighter, although the signal generator should still have been grounded through the CRT cable.

Now to the reason the test was so short:  I had the bulb connected directly to the secondary winding, no HV potential or anything.  I wanted to see what would happen if I tied one end to Earth ground.  When I did so, the bulb nearly extinguished (why, I am not sure), and the current draw shot way up on the input.  Surpisingly, the bulb stayed barely lit, fluctuating a lot, and the core made terrible sounds.  On the next power-up, the MOSFET drivers blew, probably from all the HF HV coupled onto every wire near the core.  Unfortunately, I have not yet received my MAX627 replacements, so further testing will have to wait until I can build a board for the new UCC27322 chips or I receive the MAX627s.

As an aside, these IGBTs are amazing.  Even when the MOSFET drivers failed (which would have detonated my MOSFETs), they stayed stone cold.  Scoping across the primaries revealed far cleaner waveforms than before, which is probably why this test showed a bit more of the interesting stuff Bob has been mentioning.

I think I can safely assume that the reason this test was under-unity was because of the lack of the HV potential.  Bob has mentioned that the system is under-unity with a potential below 11.5VDC, and since my potential was at 0VDC, these null results are easily explained.

I am not sure why grounding only one end of the bulb did such nasty things...any ideas here would be appreciated.

Eldarion

[Bob Boyce]

Both sides of the secondary must be kept isolated from ground at all times. the only path to ground must be through the load ;-)

Bob

[eldarion]

Both sides of the secondary must be kept isolated from ground at all times. the only path to ground must be through the load ;-)

Bob

Hi Bob,

Thanks for confirming this.   I think I may have had far greater power flow into ground than I had through the light bulb based on the sparks I saw...maybe it was overunity after all!  I still have to get used to the peculiar aspects of the longitudinal energy coming out of that secondary.

I assume I can connect one end of the pancake winding to ground, as this is conventional DC TEM?

All,

Unfortunately I did not have a chance to check the longitudinal windings or anything before my driver board blew.  I have finished designing the Thor 1.1 output driver board and have sent it off to the fab; I should have it in a couple of weeks.  This board should be much more stable and robust than my crazy hand-soldered wiring jobs.   Bypass capacitors within 2mm of each MOSFET driver, oh yeah!

Signing off for now,

Eldarion

[Earl]

Eldarion,

ringing is usually caused by
leakage inductance
leads too long
current loops too long

a toroid usually has low leakage inductance.

most people have trouble understanding that a circuit operating at 67 kHz, still needs to be built as if it was operating at hundreds of MHz, even 1 GHz, if there are pulses with fast rise and fall times.  Circuits with fast edge transistions MUST USE VERY SHORT LEAD LENGTHS.  By short, I mean ZERO mm.  Since zero mm is not possible, do every thing in your power to approach zero as close as you can.  Spare no effort.

Notice that your gate voltage has ringing that even goes negative.

Remember my image of current loops.  Look at this image again and ask yourself if all 3 of your currents loops have a length close to zero?
Is the distance between the ground of the driver IC and the IGBT source = to zero?

Don't forget E = L * di/dt and since we want di/dt to approach infinity to have RE effects, L must approach zero.
Even nanoHenries of inductance can cause sufficient and undesired voltage bounces.
Why do you think PCB designs use maybe a dozen SMD ceramic caps around each microprocessor?  The multilayer PCB has one layer dedicated only to being a ground plane.
These caps and CPU ground are connected to through-holes directly to the ground plane.

Post some photos in your thread of your physical layout, and I can make some more comments.

Would like to see a scope photo full screen of just the area in the attached photo.

Regards, Earl