Bedini Ferris Wheel Regauging Motor

[Les_K]

Hey Everyone!

Just wanted to post an update. We have separated the 3 circuits and gotten rid of the clip leads. We changes a few things as well and "think" we have it right. The paralleled transistors are still really cool, but we are having some timing issues and wanted to see if you guys had any ideas. The transistor for the hall is getting pretty warm. Not hot, but warm enough to warrant some questions. We have noticed that the timing of the hall is absolutely critical for everything. A couple of notes about that. We are using the south side of the magnets for the hall. I know we need to get some bigger timing magnets, maybe 1", and also have a bigger diameter. Any suggestions? It's taking roughly 15 seconds for the cap to charge up to about 77 V and dump, but it is dumping about 30 V. A far cry from from dumping 3 times per second. We are really eager to get the slaves done and finally get rid of those darn clamps.

Anyway, here is the video update!

If you guys have any suggestions, we are all ears! Thanks so much for all your help!

@John K,

Thanks for your suggestions. I am going to pick up some 10mm blue LEDS in the morning. We still have the circuit on breadboards, but I totally agree with everything you said! Once we have the circuit nailed down (We hope we don't have to wait until the conference), we will probably get a PCB done at PCBExpress.

Thanks,
Scott and Brent

EDIT: Wanted to post a picture of the setup.

I changed from the MPSA06 to a 8599 and had something like what you are seeing. check you pinout of the 8599. Because it is a PNP and the MPSA06 was NPN. The emitters are not in the same place. in the video it is not firing anything like mine and in your earlier vid it was firing much better as well. As far as the magnet spacing for the hall I have found it very important as well. I am not saying I have it the way it is supposed to be but I can tell you what I have found so far.
I had 3/4 inch magnets at about one inch apart. and it was pulling a lot of current and when I posted the scope shot John K said it was not quite right so I went to 1/2 inch magnets at the same distance from center as the first ones. They are much better and easier to control. When I calculated out my design I discovered that you can go closer to center with smaller mags and further out with larger and it makes no difference. Draw it out, you will see what I mean. Except in the physical size and specs on the specific hall you are using. I got some 3120 halls from mouser that have the same data as the 3020 halls JB had been using.
As to visual responses My meter will kick as high as three amps at first and then as it speeds up it will settle at a lower 700ma to one amp. I have a 75 volt 90 volt peak, 12,000uf cap and it charges within the second cycle of the wheel using tree lights as a load ( I am waiting on parts for the pulser). I tried a 300uf 250 volt cap with no load and after two pulses it was at 250 volts. I stopped the wheel quickly!
I have no idea how high it would have gone. I don't have any higher voltage caps to play with. So a word of caution on that. Anyway make sure the emitter is connected to the 220 ohm resistor or in my case I am using a 680ohm.

Hope that is helpful
Les

[John_K]

Hey guys,

I mentioned in the latest video, that Scott just posted, that I would post a schematic...

Here it is!


Thanks, Brent

Hi Scott & Brent,

Thanks for the HD video, very nice!

A couple of things - firstly, you are charging up the cap too high before dumping and wasting a lot of energy in the process. Ideally, the cap should be charged to 63% of it's capacity (47.25V). This is around 10V over the top of the battery voltage. So a 4.7V zener instead of the 36V. This should get you dumping 3 times per second.

Your hall transistor might be getting hot because you have a 12V zener on it. I was using a 15V zener and it was working fine and the MJE2955 I was using did not get hot at all.

I would also ditch the FWBR and replace it with a single STPS8H100D, which is a 8A 100V Shottky rectifier diode. Make sure the positive of the cap goes to the cathode of the diode. It only has a 0.7V forward voltage drop and switches real fast.

Apart from that, great work. You guys are really close!

Oh, also if you want slow the RPM down try to adjust the either the coil gap or the hall timing. You should be able to get it to slow down by adjusting the hall so it retards the timing in relation to the coil to rotor magnet position when the coil fires, so the coil does not give the rotor much of a push.


John K.

[txaggie00]

Les and John K, thank you for the replies!

We are going to run to the electronics store and pick up the suggested parts. I have a couple of quick questions. Are the coils supposed to be energized when the hall is over the magnet or the reverse? The way we have it right now, when the hall is over (using this term loosely) the magnet the coil switches off. Also, are we correct to use the south pole of the magnet for the hall switching? Do you recommend using the north? Les, we have tried different transistors for the hall (I think in the video we were trying the 93 b/c the 2955 was getting warm) and had it reversed (PNP vs NPN) once like you were saying, but we quickly found out that when that happens, the 94's get hot pretty quick, and since those never even get warm, we knew we had something wrong. Also, John K, we are going to try the 4.7V zener for the cap dump, but we both thought that JB had said to charge the cap to twice the battery voltage and that's why he went with the 75VDC, 90VDC surge capacitor. Did we misunderstand what he was saying? It's certainly nice to see ~35V get dumped into the charging batteries !

Thanks for your help!
Scott

EDIT: @John K. You may be interested in one of our little side projects involving a Renaissance Charge Rejuvenator. I can PM you the details if you're interested.

[RS_]

Scott, Brent,

It's Looking good......

Looking at your SCH, i have several recommendations, based on my work with Muilty Strand SSG's, etc.... and i think JB will agree with these.

The MPS8599 at 250ma Collector current, is too small to effectively drive the Bases on muilty MJL transistors. and a NPN is needed here so that you have the Coils ON when the Hall is over the magnet. if you have a BD243C transistor, it should work in place of the MPS8599, If it needs it, use a MPS size tranny to drive the base of the BD243C so that the Hall can handle the current.

Use a 22ohm resistor on each of the bases of the MJL transistors, as a current divider network, so that none of the trannys can hog all the current. Use your ohm meter to select a matched value set of 22ohm base resistors.

Hope this helps, and if ya'll need Help, Let me know, and I will come up to see you......

RS

[txaggie00]

Hope this helps, and if ya'll need Help, Let me know, and I will come up to see you......

RS

RS,

Good to hear from you! It's funny, b/c last night, Brent and I were talking about shooting you an email to ask if you could come give us a hand and see the setup! We certainly appreciate the offer! Thanks for your suggestions as well. Brent's out picking up some parts now, and we are working on setting up another timing wheel. I'll shoot you an email and maybe we can figure out a good time and whatever is most convenient for you.

Thanks,
Scott

[Les_K]

Scott, Brent,

It's Looking good......

Looking at your SCH, i have several recommendations, based on my work with Muilty Strand SSG's, etc.... and i think JB will agree with these.

The MPS8599 at 250ma Collector current, is too small to effectively drive the Bases on muilty MJL transistors. and a NPN is needed here so that you have the Coils ON when the Hall is over the magnet. if you have a BD243C transistor, it should work in place of the MPS8599, If it needs it, use a MPS size tranny to drive the base of the BD243C so that the Hall can handle the current.

Use a 22ohm resistor on each of the bases of the MJL transistors, as a current divider network, so that none of the trannys can hog all the current. Use your ohm meter to select a matched value set of 22ohm base resistors.

Hope this helps, and if ya'll need Help, Let me know, and I will come up to see you......

RS

RS,
I am using an MJ8599 that goes to a 680ohm resistor in place of the 220.
Still have two 470 ohm resistors at each base of the MJL's I get about 3 amps through the coils if I hold the wheel so the hall is steady on I should test the coils directly to the battereis and so how close this is. My coils are not the same so this may make the difference in the circuit but it looks like the MJL's are fully on and the 8599 never even get's warm just stays cold. I am not using any zener's on the front end like John K. As I am only running at 24 volts right now. And I only have three sets of matching MJL's right now.
Anyway As I was thinking about your change to the base resistors, and looking at the video John B. posted showing the high RPM window motor you can freeze the frame and see there are only a couple resistors inside the box. Unless there are some underneath the circuit board, but he is not one to build that way. I was always taught that you needed to add resistors to the emitter side to get feedback to the base to balance paralleled transistors but it looks like nobody is doing that.
In the end the goal is to get the MJL's fully switched on without wasting excess current through the base. So I am very, Very... interested in hearing more about what you folks do with the circuit so please post here as you go if possible.

Thanks
Les

[genessc]

[I would also ditch the FWBR and replace it with a single STPS8H100D, which is a 8A 100V Shottky rectifier diode. Make sure the positive of the cap goes to the cathode of the diode. It only has a 0.7V forward voltage drop and switches real fast. ]

You know, putting up to 4 in parallel will lower that drop to about .4V... same concept as paralleling anything else... drops the overall resistance. Or at least thats what it looks like on the diode measurement from the DVM. Should be pretty straight forward to personally verify.

Just fyi.
Gene

[RS_]

Les,

Using very small value resistors on the emitters is good for tuning the current to be equal through each of the parallel transistors, and if needed could be done on the Bedini/ Cole circuit. The Base resistors current divider network on parallel transistors, are so that no one tranny can hog all the base current for some reasion or another. Not quite the samething......

JB instructed me to do this Base current divider network, when I was building 8 and 16 strand SSG coil Driver circuits with the MJL's......

Is The window motor driver a muilty parallel transistor driver like this one, or single Bedini /Cole Driver...?

>>>"In the end the goal is to get the MJL's fully switched on without wasting excess current through the base. "

I agree, but go look at the Base current (Continuous) on the MJL Data sheet at 5 Amps DC, and then multiply by 4 parallel Tranny's

They need far more current to turn on fully, than the MPS8599 at 250ma can provide.......

RS

[John_K]

Are the coils supposed to be energized when the hall is over the magnet or the reverse? The way we have it right now, when the hall is over (using this term loosely) the magnet the coil switches off. Also, are we correct to use the south pole of the magnet for the hall switching? Do you recommend using the north?

EDIT: @John K. You may be interested in one of our little side projects involving a Renaissance Charge Rejuvenator. I can PM you the details if you're interested.

Hi Scott,

The coils are supposed to energize when the hall is over the magnet, not the other way around. Most halls use a South pole to trigger it, you may want to check the datasheet for the one you are using to be sure. Make sure you are using a unipolar hall and not a bipolar.

Feel free to PM me or email john_koorn at yahoo dot com dot au about the side project.


John K.

[BrentA929]

Hall sensor

John K,

Thanks for the clarity...

That is what RS is suggesting too and seems to make more since. You may have already made the adjustment on your schematic, but your last revision shows a PNP (MJE2955) instead of a NPN for the hall section. The post near the beginning of the thread and your schematic is where we have started to create the circuitry.

Also, is the bottom of the coil suppose to connect to the MJL21194 and the top of the coil to the MJL21193 transistors? That's how we have it now.

Also, thanks for clearing up the polarity on the halls. Ours is a ATS137-PL-A-A. I think that was one JB had mentioned.


Thanks as always, Brent

Hi Scott,

The coils are supposed to energize when the hall is over the magnet, not the other way around. Most halls use a South pole to trigger it, you may want to check the datasheet for the one you are using to be sure. Make sure you are using a unipolar hall and not a bipolar.

Feel free to PM me or email john_koorn at yahoo dot com dot au about the side project.


John K.

[John_K]

John K,

Thanks for the clarity...

That is what RS is suggesting too and seems to make more since. You may have already made the adjustment on your schematic, but your last revision shows a PNP (MJE2955) instead of a NPN for the hall section. The post near the beginning of the thread and your schematic is where we have started to create the circuitry.

Also, is the bottom of the coil suppose to connect to the MJL21194 and the top of the coil to the MJL21193 transistors? That's how we have it now.

Also, thanks for clearing up the polarity on the halls. Ours is a ATS137-PL-A-A. I think that was one JB had mentioned.


Thanks as always, Brent

Hi Brent,

No , the PNP (MJE2955) is correct. If you are using an NPN there that is possibly why your coils are on when they should be off and vice versa.

Yes, the top of the coil goes to the 93's and the bottom goes to the 94's. This will give you a North pole on the top of the coil when it is on, which pushes away the North facing magnet on the rotor.

The ATS137 is also what I have ordered. They should be here any day now. Some wire arrived yesterday, so I expect the rest of the parts won't be far away. As soon as they turn up I can come and play in the sandpit with you guys.

I've attached an updated schematic, in the previous the zener on the cap dump was backwards.


John K.

[RS_]

Brent, Scott,

Your Hall Sw part number, is a south pole activated , active Low output, that turns on the PNP , that turns on the MJL's , and is a correct operational circuit.... So, make sure you have the south pole out, on your timing wheel magnets.......

other wise, put the BD677A NPN darlington you mentioned elsewhere, in place of the MPS8599, and use the MPS8599 as a inverter base driver for the BD677A from the Hall Sw to get the Hall over the magnets and coil on condition.....

RS

[Les_K]

RS,
My Bad I am using a MJ2955 not a 8599 or MJE2955.

I am fascinated by the base current divider network I still don't see enough resistors in John B's video to account for that either....
your 8 and 16 strand coils were not paralleled though, Were they? didn't you have a single transistor for each strand?

The reason I am taking so much interest in this area is because I am just starting to get a grasp of something JB said about Ron Coles circuits that inductors in parallel didn't increase current through the circuit like you would think. and it drove the engineers nuts. So I am just trying to understand what is happening at that point when the impedance of the coil is matched. I am guessing but this must be a very low impedance coil with a very low inductive reactance and it must run at no more than about 28 RPM (16 *28 = 448/60 = 7.46 HZ) to hit resonance. and if you are not there then your CE current will be very low. And now that I have said all that I realize clearly why Scott has the only coil with the right impedance. And perhaps why my circuit is not manifesting the same heating problems....
However, I have noticed that as I have been running the machine for short periods of about 5 minutes each over the last few days I have seen the 94's get warm sometimes and then the 93's the next time. This went back and forth for a few days and today they all run cold to the touch. no heat anywhere in any of the circuit. None! but it runs the same. the input current draw has not changed. and the output is the same. the scope is the same....
Do these transistors beta self adjust? mmmmm

Teach me more

Les

Oh P.S. All my coils are in repulsion. I tested them one at a time on the battery before connecting them together and hooking up the circuit.

Les,

Using very small value resistors on the emitters is good for tuning the current to be equal through each of the parallel transistors, and if needed could be done on the Bedini/ Cole circuit. The Base resistors current divider network on parallel transistors, are so that no one tranny can hog all the base current for some reasion or another. Not quite the samething......

JB instructed me to do this Base current divider network, when I was building 8 and 16 strand SSG coil Driver circuits with the MJL's......

Is The window motor driver a muilty parallel transistor driver like this one, or single Bedini /Cole Driver...?

>>>"In the end the goal is to get the MJL's fully switched on without wasting excess current through the base. "

I agree, but go look at the Base current (Continuous) on the MJL Data sheet at 5 Amps DC, and then multiply by 4 parallel Tranny's

They need far more current to turn on fully, than the MPS8599 at 250ma can provide.......

RS

[RS_]

Les,

Yes, a single strand per transistor SSG, up to 48 strands on my big machine.....

But the Idea of using a Base current divider network is the same, if it's a tranny per strand, or several trannys in parallel hooked directly to several parallel strands, as to keeping one of the tranny bases from hogging all the base current

The difference that JB indicates as i understand it, is that the 1 strand per tranny setup, will pull far more current, than having several trannys in parallel, hooked directly to several parallel strands setup, for an equal number of strands / trannys.

RS

[Ecancanvas]

can you make up a schematic

Hey RS
can you post a schematic of how your adding the R22 to the multi MJL. are you just adding them to the base? Why dont you want to use a bridge? to slow?
Thanks for the help
ED