Well, I thought I might as well post the source for my current Spartan-3 based TPU controller.
Right now it is basically a computer-commanded precision pulse generator, although control functionality will not be difficult to add once we find out exactly what type of control is required.
Please keep in mind that it is a bit rough at this time, and there may be ways to crash it that I don't know of. ;)
I attached a screenshot of the computer interface to show the various options I have.
Eldarion
OK, I'll be posting schematics here over the weekend.
So far, I have attached the primary coil driver board and the dual-channel ADC. The firmware currently supports only one channel of the ADC, but I will add support for the second channel soon.
The primary coil driver board schematic is incomplete; I still have to add the HVDC generator and the LVDC magnetic bias PWM circuitry.
Also attached is a new version of my pulse generator firmware that allows for fixed width pulses, as well as the ability to generate Bob's pulse sequence.
Well, I do not yet have everything in a "releaseable" state, but I thought I'd post a few teaser pics of my new controller...
This board no longer requires a computer at all, except for initial programming (a one-time operation, barring firmware updates)!
The pulses can be of MUCH higher voltage, but my scope wouldn't have been able to display them. ;D
Very nice Eldarion!
Bob
Thanks! :D I hope it will get the job done...
One item that I am still in the process of adding is a highly-regulated 0-100V DC switching power supply.
This may take some time, as the FPGA control algorithms are rather complex to implement. When I am done, there will be another menu option to control that voltage.
The only external components that should be needed for the entire controller (so far) are two MAX627 driver chips, four IRF510 "RF" MOSFETs, an inductor for the switching supply, a capacitor and diode for that same supply, and a few filter caps. How's that for simple? ;)
Eldarion
Well, I thought I'd give an update since it has been a while...
I got the regulated HVDC supply working! It wobbles a bit (+/- 0.5V at 40V) but other than that it looks good. I may try to tune it better if the wobble becomes an issue.
Attached is a pic of the custom circuitry and a shot of the voltage selection screen and cable hookups. You will notice a little "VLOW" on the LCD; I am building in self-diagnostics. This particular message is telling me that I forgot to hook up the external power supply... ;D
Still to do is the TPU control loop. Also, I am finding the LCD screen rather cramped, so I will probably display more information on an external VGA monitor. No extra hardware is required (other than an old monitor), so I figure why not? ;)
Eldarion
Well, I thought I should post an update before the conference. It sure is tricky to take a picture of a CRT monitor!
Not everything is actually working (notably the "CONTROL>" prompt), but everything in the "System Status" window is fully functional.
I can use either the small built-in LCD screen and the rotary encoder or a VGA monitor and a PS2 keyboard to control this thing. I really like being able to see all the parameters at once! ;D
One of the next things that I will do is implement an automatic resonance sweep and try to plot the results on the VGA screen. That way, we can determine resonant peaks of the attached coils easily (and the software can auto-tune a new coil more easily!)
Best of all, since I am not adding any extra hardware, power draw is minimal for the whole setup (if you don't count the CRT power drain ::) But then we don't really need that for automated control...just for initial calibration and setup.)
If anyone wants a .bit file to download to their board, just post here and I will post the latest version for you.
Eldarion
Well, I have finally gotten around to drawing up the schematic--this is what I have so far on the main board; I have not yet calculated or built the voltage division networks for the add-on ADC modules.
Also, I have attached a screenshot of what I believe is the final GUI revision of the controller. Once again I apologise for the dismal quality of the picture; it is nearly impossible to take a picture of a CRT monitor and have it turn out correctly.
As you can see, I have built this controller to handle either one or two coils simultaneously. If an out of spec condition is encountered, a fault will be triggered and that coil will shut completely down temporarily. If a coil's ADC is not plugged in, the contoller will also shut down the pulse generator for that coil and signal a fault (NOT INSTALLED message on the second pane)
As always, a bitstream is available for those who want it. I do ask that both the schematics and the bit file remain confidential to the group; from what I can see this controller may be a significant step forward from previous versions and I haven't yet decided what I want to do with it. A copy of the software will always be available to the group's members for free, however!!! ;D
Tim
Hey Eldarion,
The control screen looks awesome! The screen provides instant analysis of everything going on. Also, the self diagnostics and fail safes are vital for safe operation. The name "Thor" is fitting for this controller indeed. I look forward to watching you fire the new coil up next week!
May you have only Thor's thunder, and never the lightning! ;)
Bruce
Quote from: btentzer on September 28, 2007, 11:29:39 PM
Hey Eldarion,
The control screen looks awesome! The screen provides instant analysis of everything going on. Also, the self diagnostics and fail safes are vital for safe operation. The name "Thor" is fitting for this controller indeed. I look forward to watching you fire the new coil up next week!
May you have only Thor's thunder, and never the lightning! ;)
Bruce
Thanks! I'd prefer not to have thunder in my basement either--that would mean something blew up! :D
Attatched is the finished schematic of Thor. The only reason that this would be modified is if something is wrong when I actually hook up the coil.
I'll see if I can get a better screenshot sometime in the future...
Eldarion
Hello All,
I have finished my first Bob Boyce coil! ;D
The first picture is of the six primaries, wound around the entire coil. I wound 25 turns per primary. 23 actually would have fit a little more comfortable. I am starting to tape over the primaries with #35 blue electrical tape.
The second picture is of the yellow motor tape wound over that.
The third picture is of the finished coil. All of the fiberglass sleeves are on, everything is properly labled and the final wind of #35 electrical tape. The core is solid, waxed and taped. No way any windings can move while in transit.
I then placed it into a special case, which will then be packed in sytrofoam peanuts and a box, and shipped Monday, to Eldarion, where it will be linked up with Thor! ;) Next week ought to be very interesting around here!
Also Monday night, I start winding my own coil.
Cheers,
Bruce
looks GOOD!!!!
Bruce
on to the next level
wer
I am in the process of designing a custom PC board for this system. My current thoughts are:
1.) It should be small enough to fit in the center of the toroid--probably long and skinny, 2 inches in width?.
2.) I will be including the ability to drive up to 12 primaries. Whether or not I even populate those parts is something to be decided later...
3.) Along those lines, the system would be able to drive two 3- or 6-primary coils for extra power output.
4.) I will be adding an Ethernet port for remote control and monitoring purposes if possible, in addition to the VGA and PS/2 ports.
I would then be able to sell a ready-to-run controller, just add your own IGBTs or MOSFETs. Would there be any interest in something like this? (I will be making up at least two PCBs for sure, one for Bruce and one for me).
Obviously I will wait to order any boards until I at least have the prototype operational. ;) I would like to keep the cost below $150 if possible.
Eldarion
hello Eldarion
that sounds very nice.
with 2 sets of controllable outputs you can use 2 cores.
as you know i am using Bob's pwm3f-hd
still working on the variable hv potential and a variable DC bias supply
i also had in the back of my mind to build a second core to
parallel with the first one. plan to mount in a metal box.
wer