Spellman DF3 Power Supply Fusor Application

This forum is for specialized infomation important to the construction and safe operation of the high voltage electrical supplies and related circuitry needed for fusor operation.
Post Reply
User avatar
Anze A Ursic
Posts: 109
Joined: Thu Mar 10, 2022 9:28 pm
Real name: Anze A Ursic

Spellman DF3 Power Supply Fusor Application

Post by Anze A Ursic »

Hello everyone,

I have a DF3, the -60kV, 60mA power supply. I have built an entire remore-control setup via LabView and it all works great but our lab has yet to figure out how it can be used for fusion applications. What we are most confused about is the connection using the 3 pins (L, S and C).

We have tried several connections which all ended up arcing between C and L/S somewhere between 10-15kV. Could someone with experience kindly tell me how to connect this power supply to the fusor? Trial and error may yield results at some point, but we're afraid of breaking this power supply.

Any help greatly appreciated!

AAU
274692271_1461672110895269_393509053977776276_n.jpg
User avatar
Liam David
Posts: 312
Joined: Sat Jan 25, 2014 5:30 pm
Real name: Liam David
Location: Arizona

Re: Spellman DF3 Power Supply Fusor Application

Post by Liam David »

The potential difference between L/S and C should not be more than a few volts since it's a floating filament supply meant to deliver several amps. I've used a floating filament supply for several years by connecting only one of L/S or C and leaving the other terminal unconnected.

Are you sure it arcs between those points? What does your cable look like? Do you know your unit is fully functional and that it hasn't, say, arced through the isolation transformer or failed under its previous owners?
User avatar
Anze A Ursic
Posts: 109
Joined: Thu Mar 10, 2022 9:28 pm
Real name: Anze A Ursic

Re: Spellman DF3 Power Supply Fusor Application

Post by Anze A Ursic »

Liam David wrote: Mon Mar 14, 2022 1:29 pm The potential difference between L/S and C should not be more than a few volts since it's a floating filament supply meant to deliver several amps. I've used a floating filament supply for several years by connecting only one of L/S or C and leaving the other terminal unconnected.

Are you sure it arcs between those points? What does your cable look like? Do you know your unit is fully functional and that it hasn't, say, arced through the isolation transformer or failed under its previous owners?
Thanks for the input! It always arcs across the exposed wires at the end of the cable.

So far I've tried a few things. By leaving all 3 wires (L, S and C) separate, it doesn't make it past 10kV until L or S arc with C. Then I tried connecting L and S together (shorting them) and then adding a 10 Ohm resistor across them and C. That arced across the resistor. Then I tried connecting the same resistor between L and S and connected C to L. That also shorted across the resistor, between C and S.

When you say you connected only one of L/S or C, does that mean you just shorted together L or S with C and had the other terminal (S or L) unconnected?
User avatar
Liam David
Posts: 312
Joined: Sat Jan 25, 2014 5:30 pm
Real name: Liam David
Location: Arizona

Re: Spellman DF3 Power Supply Fusor Application

Post by Liam David »

I left the filament output floating, didn't even add the pin to the custom connector I made.

Something doesn't add up. Pin C is at HV and S/L are at HV+/-LV, where LV should be at most a few volts. Connecting L/S with 10 ohms to C and setting the filament program to max should warm up the resistor. Connecting L/C and with 10 ohms to S should just short the filament current with the resistor doing nothing. Arcing between the pins implies one or more is not at HV, and perhaps dead at ground. One check you might be able to do with everything completely discharged is to use a high-range ohmmeter to check the pins for connections to ground. There's a resistive divider for voltage feedback somewhere in there, but I would expect that to be 1-10G range. Anything less and you might have a problem.

Have you opened the unit? Where did you get it from and in what condition? Pictures of the setup perhaps?

Also, no need to use block quotes on this forum. We're all smart cookies here.
User avatar
Anze A Ursic
Posts: 109
Joined: Thu Mar 10, 2022 9:28 pm
Real name: Anze A Ursic

Re: Spellman DF3 Power Supply Fusor Application

Post by Anze A Ursic »

We got it on Ebay. I do not think there's an internal error. I've set the filament current to 0 before and ramped up the voltage to ~20kV without a cable connected and it was working well. So I'm almost certain it's my incompetence in connecting the cable properly. I've never used such a power supply, so it's a learning curve. I'll do what you recommended and report back.

Thanks for the help!

AAU
Rex Allers
Posts: 520
Joined: Sun Dec 30, 2012 3:39 am
Real name:
Location: San Jose CA

Re: Spellman DF3 Power Supply Fusor Application

Post by Rex Allers »

As Liam has described, for a fusor you only need one connection to C. You are not trying to light up the filament of an xray tube which is what the other two are there for.

What kind of cable are you using? What does your connection end look like and the other end too. Got a pic or two to post?

I suspect you are arcing to something grounded on your cable that is too close to the HV terminals in the supply.
Rex Allers
User avatar
Richard Hull
Moderator
Posts: 13898
Joined: Fri Jun 15, 2001 9:44 am
Real name: Richard Hull

Re: Spellman DF3 Power Supply Fusor Application

Post by Richard Hull »

I worry you do not have a proper cable. Measure between the pins with an ohmmeter the two that read low ohms are filament pins the other pin is the hot pin. That hot pin will arc to anything unless you have a proper cable for the supply.

Once you identify the hot pin, tie it to any one of the filament pins. I assume, again, that you have a bonafide proper cable in hand.

Once you have a non-arcing hot lead, that goes to the HV input terminal on the fusor and its grid and that supply case ground goes the fusor metal shell. Make sure you have that case ground well connected to the electrical wiring ground as well.

Richard Hull
Progress may have been a good thing once, but it just went on too long. - Yogi Berra
Fusion is the energy of the future....and it always will be
Retired now...Doing only what I want and not what I should...every day is a saturday.
User avatar
Anze A Ursic
Posts: 109
Joined: Thu Mar 10, 2022 9:28 pm
Real name: Anze A Ursic

Re: Spellman DF3 Power Supply Fusor Application

Post by Anze A Ursic »

Sorry for the delayed response, other things came up. I redid the test today and found the following. By making a custom cable with only a C connector (one pin only), the power supply starts ramping up but then gives me several errors. They are all posted below with their definitions copied from the manual.

1. mA REGULATOR ERROR: Indicates that the tube emission current regulation circuit cannot maintain current regulation. This may occur when tube filament problems exist.

2. FILAMENT CURRENT LIMIT: Indicates that the filament current has exceeded the programmed filament current limit trip level.

3. OPEN FILAMENT: Indicates the X-ray tube filament or connections to the filament are open.

The last one makes me the most confused. That error didn't show up when I had the other connections described a few replies ago (when stuff was shorting). Back then, I shorted the L and S wires at the end of the HV cable. However, I figured that it was a redundancy since the manual says L and S are internally shorted and that is true, I checked with a Fluke meter.

Someone mentioned I should do resistance reading from one connector to the other, so here are the results:

L-S measurement yields 0.8 Ohm, makes sense since they are supposed to be shorted.
L/S - C yields 477 Ohms... I am assuming that is no bueno.

Does anyone have the schematic to this specific power supply? I've found quite a few HV power supply schematics on this forums in the past few years, but never for this model. If anyone has it for the DF3, I'd very much appreciate if you could attach it.

Let me know what you think, and thank you all for your help!

AAU
User avatar
Dennis P Brown
Posts: 2599
Joined: Sun May 20, 2012 10:46 am
Real name: Dennis P Brown
Location: Glen Arm, MD

Re: Spellman DF3 Power Supply Fusor Application

Post by Dennis P Brown »

Well, the most obvious issue to me is that the unit's internal logic for the filament see's no voltage on said filament. Rather, that circuit/logic see's the filament system as an 'open' or burned out filament. All three errors appear to relate to that issue directly, I'd think.

Richard and others suggested that you tie one filament pin to the HV output pin. Then run that HV to the fusor cathode and be certain to ground the fusor case to the power supply ground. Have you tried this arrangement, yet?

The drawing you supply basically shows that arrangement, as well. The filament (if that x-ray tube existed) would offer essentially a short across itself so connecting those pins would essentially be the same thing, no? Of course, you will need a properly insulated connector/cable - that is, a cable for access to the output connector that fits your power supply. Maybe I'm missing something here?
User avatar
Anze A Ursic
Posts: 109
Joined: Thu Mar 10, 2022 9:28 pm
Real name: Anze A Ursic

Re: Spellman DF3 Power Supply Fusor Application

Post by Anze A Ursic »

Dennis, thanks for your input!

We've tried doing it the way the picture shows, in several configurations. L and S are internally shorted, and almost every configuration arcs pretty badly. When all 3 are separate, C arcs with L/S, then we tried adding a resistor to model the load but it again arced in several configurations at anything past 10kV. Shown below are the results. The unshielded wire is the C connection (this is a factory custom made cable, don't ask me why it's a bare wire). But you can see the charring on it.

Again, we have tried several different configurations but all either give us output errors which ultimately shut down the supply or stuff arcs. We have yet to successfully run this power supply.
20220327_152020 (1).jpg
User avatar
Dennis P Brown
Posts: 2599
Joined: Sun May 20, 2012 10:46 am
Real name: Dennis P Brown
Location: Glen Arm, MD

Re: Spellman DF3 Power Supply Fusor Application

Post by Dennis P Brown »

How are you attaching that wire to those pins? A picture of your cable attached would be necessary to further understand your issue. At those voltages the pins will arc between each other if one is a ground. You need to have properly insulated/designed connector that goes to the pins with a cable.

I am making some assumptions here so be careful as I have no direct knowledge of that unit or its internal wiring.

With that disclaimer, I will say that the high voltage is connected to pin L & S. So pin "C" is either acting as a ground or is a ground. At those voltages without proper insulation between those close pins arcing is guaranteed between them.

That high voltage cable you are showing really puzzles me - the yellow, green, and two bare metal wires - such a system can't work with a 60 kV supply if anyone of those wires are a ground. The thin insulation on those wires will never support that field without arcing over to another wire in that bundle. Also lost why there are two bare metal wires in that bundle with those other wires at all.

I might be able to suggest a work around but would need to see what you are trying to do with that cable connected and is that the original HV cable or something you are just trying to use? If you are grounding those bare metal wires then that is a major error (frankly, they are a danger regardless of their function since they are an arc over threat if they are ever near anything that could remotely act as a ground.)
User avatar
Anze A Ursic
Posts: 109
Joined: Thu Mar 10, 2022 9:28 pm
Real name: Anze A Ursic

Re: Spellman DF3 Power Supply Fusor Application

Post by Anze A Ursic »

None of the three pins are ground. I am posting the answer we got from a Spellman engineer about this:

"DF unit in negative output voltage so it has a floating filament, hence the use of the multi-pin Federal Standard X-Ray Connector. C = common, L = large filament and S = small filament. Filament current flows between common and Large/Small. The filament is a floating filament, one side is connected to the common so the filament if referenced to the high voltage output voltage. Filament is connected between the C and the L&S pins."

"The DF3 is a negative polarity/floating filament X-Ray generator. The DF was designed to operate X-Ray tubes under closed loop filament control. The customer provides a 0-10Vdc signal (from either the local front panel or remote interface) that represents 0-80mA and our internal closed loop emission control circuitry adjusts the output of the floating filament supply to provide the desired current emission current. There a timing sequence of bringing up the high voltage, ramping up the filament, over and under current faults, regulation error fault, filament open fault, filament short fault and other protective circuitry to assure proper operation. This was all done to protect the X-Ray tube. There are probably other issues that I’m not aware of as I’m not the DF product engineer. Trying to operate the DF unit in a non-X-Ray generator application will create some if not all of the above mentioned faults and perhaps other problems, preventing operation. This unit was never intended to be used in a non filamentary controlled X-Ray generator application. "

There is a ground connector on the back of the case, but none of the wires coming out of this cable are ground. In fact, I remember seeing somewhere that the voltage difference between the C and L/S pins should never exceed 10kV, which makes sense when you see how close together they are and how that wire in there is just bare.

Anze
User avatar
Anze A Ursic
Posts: 109
Joined: Thu Mar 10, 2022 9:28 pm
Real name: Anze A Ursic

Re: Spellman DF3 Power Supply Fusor Application

Post by Anze A Ursic »

Here is a picture representing the typical setup for this power supply. In some sense, it would seem also a no-brainer that this thing can work for a fusor, but almost every single possible configuration of the L, S and C pins have resulted in either arcing or errors that shut down the power supply. It's quite a quandary!

Anze
Capture.PNG
User avatar
Joe Gayo
Posts: 398
Joined: Sun Jan 06, 2019 9:34 pm
Real name: Joe Gayo
Location: USA

Re: Spellman DF3 Power Supply Fusor Application

Post by Joe Gayo »

Drive the HF transformer directly with a half bridge and remove all the other circuitry. The half bridge can be supplied with the rectified output of a Variac transformer (that will be the master control)

You can probably do this in a way that it will still read the voltage and current for you.
Rex Allers
Posts: 520
Joined: Sun Dec 30, 2012 3:39 am
Real name:
Location: San Jose CA

Re: Spellman DF3 Power Supply Fusor Application

Post by Rex Allers »

From your pic of one end of your HV cable I think I see a big problem that I was trying to warn you about earlier. But the details of that will be coming later in this post.

I have no experience with a supply designed specifically for driving a big xray tube and those 3-pin HV output connectors. I downloaded an Instruction manual from Spellman web pages for the DF3 supplies. I hoped to see a figure like the one you shared that shows how to connect the 3 pins to an xray tube. There wasn't one that I could find in this manual.

The figure you posted at the start of this thread seems to have come from a manual for the XLF supplies. One would think that all supplies for an xray tube with that 3-pin connector would have the exact same pin functions, but apparently not.

So what is this HV connector? Looking at the DF3 manual there is one picture of the supply backpanel with a tag on the connector that says, "Federal Standard X-Ray Connector to the X-Ray Tube." And another picture of just the connector with the caption, "FIG. 3.1 A 75kV FEDERAL STANDARD CONNECTOR PINOUT", but no other guidance about the actual usage of the 3 pins.

So I did a web search on "federal-standard xray cable" and started to find some information about what these cables and connectors look like. Here is a pic of a plug that would go into the supply...
Federal Standard Xray Connector.jpg


I added the 5" dimension to that pic, which I got from this document on the Dielectric Sciences web pages.
Federal-Standard.pdf
(70.57 KiB) Downloaded 30 times


That pdf also says the cable for the 75 kV version with 3 pins is type 2214. Here is a datasheet on that cable...
2214.pdf
(63.17 KiB) Downloaded 25 times


It shows the center conductor is composed of 4 wires, two slightly insulated (one yellow, one black), and two bare wires. I never found any guide on which wire should connect to which pin of the 75 kV Federal Standard plug but we might guess as we go along here.

A few years back I bought some (~15 ft) xray cable and it seems like it may be a version of the 2214. I stripped back just enough of one end to see all the component parts and took a picture, which I then labeled...
2214 maybe.jpg


In the center conductor section you can see the 4 wires, two slightly insulated (black and yellow) and two bare. My thoughts on how these are probably intended to be assigned is, the insulated wires would be for the filament drive voltage which can't short together or the filament won't light. Not much insulation is needed around them because the filament drive voltage is low, less than 20 V usually. This leaves the two bare wires which are for the HV output. These two wires short together along the run so applying HV to either one effectively applies it to the other. And, in fact, the two filament wires are running at the HV output voltage too, only they also have a few volts between them, too, which drives the filament.

So what are the actual output connections from the supply to the 3 pins of the Federal Standard connector. I searched and searched through the DF3 Instruction Manual and didn't find what I was looking for. There was no pretty picture like the one from the XLF supply manual. But finally I found something in an unlikely place. Near the end of the manual there is a section that lists and describes the various options that can be ordered on the DF supplies. One of these -- LARGE-COMMON LC -- describes the Standard way the 3 pins are assigned and then how it changes if you order this option. Here is the section:
  • 5.10 LARGE-COMMON LC
    The standard DF/FF output connections are
    Common = 60kV, L and S = filament output.

    The LC option ties L and C together and has the
    filament on S.

In above, I added a paragraph break to more clearly separate standard from the LC option. So if this is correct the DF3 supply doesn't have the same 3 pins assignment as the pretty picture from the XLF manual.

From these option description words, I created a picture with the two options this DF3 supply could have and how they might connect to an X-Ray tube.

DF3 Output.png

You should be able to determine if you have any options on the supply from the full model number on the supply label.

But for a fusor application, it doesn't matter. There is no filament. The only connection needed is the HV out on the 'C' socket of the output connector. I would use one or both of the bare wires as this connection. The other wires can be cut flush and left unused.

How are you now making connections to any of the 3 sockets in the supply connector? Your pic of the cable end doesn't show any kind of pin that might make the connection.

I would think you need a proper 3-pin plug for the end of your cable or a home-made version from some kind of plastic. In some fashion it would have a single pin at the end and a key on the plastic rod to align the pin to the C socket.

Now, the issue, why are you getting arcing? I think I see. Look at the end of your cable, or, at my picture of the end of my cable. I only cut back my cable enough to show all the component parts. The thick HV insulation is what keeps the HV on the center wires from breaking out. Just outside of that insulation is a shield of wire braid. This braid must be grounded. If the thick insulation of the cable should fail for some reason, the grounded braid will absorb the arc and keep things mostly safe. But at either end of the cable, the braid must be cut back many inches from the live HV which is inserted on the central wires. In air the 10's of kV can arc a long way between the end of the central wires and the braid. This is made worse by all the pointy ends of the braid wires toward the HV.

From my earlier pic of the Federal Standard plug, the plastic rod inserts into the connector 5 inches. The outer jacket of the cable and the braid must be cut back from the end of the cable at least this far. And the outer jacket a bit more to leave some of the braid exposed for a ground connection where it leaves the case of the supply.

I haven't seen one, myself, but I believe that in the proper Federal Standard plug and cable, the braid of the HV cable is crimped or otherwise joined to a metal ring at the back end of the plug. There is a big coupling nut that screws onto the threads of the supply's HV socket and firmly clamps the metal ring onto the supply case so the braid is firmly attached to the grounded supply case.

The other end of the cable (where it will feed the fusor) must also have the braid stripped back quite a few inches to avoid arcing there too.

So, my educated guess is that the arcing you are seeing is to the braid. Black plastic insulating tape is no match. Distance is needed. And I think you must get a good plug that fits into the supply's connector dependably. A proper plug and cable can be bought but I would expect at a lot of expense. I think others have made their own (maybe Liam?) but it will take a lot of planning and work to do it right. The braid must be grounded well for safety.

Be careful. This kind of voltage and power can easily kill.

Oh, one last thought. This DF3 supply is designed to drive an xray tube. The beam drive is speced at 60 kV and 60 mA or 3.6 kW. Wow. But it also is expecting to light up a filament with up to 14 V @ 5 A = 70 watts for the filament. I don't think you want to short this out. First, 70 W would be a lot to draw, second, it would probably trip some over current alarm and shut down.

Hence the suggestion to only connect to the 'C' socket.

But if the filament is open (I think you mentioned some kind of such alarm indication) I'm not sure if the supply will enable the HV output. Unknown to me and I didn't see that level of info in my quick scan of the instruction manual.

You do have, I think, a knob for filament current. Maybe best to have this all the way down.

Lastly, if you get everything connected and running, a fusor is an ugly load as plasma strikes. The info on the DF3 definitely says it has some level of arc detection. Richard has mentioned his experience with high tech HV switcher supplies because they are sensitive and prone to shut down. OTOH many people here have done good work with various switching supplies.

You need to get through the steady state issues first, to then see if there are other hoops to jump through.

Hope some of this helps.
Rex Allers
User avatar
Anze A Ursic
Posts: 109
Joined: Thu Mar 10, 2022 9:28 pm
Real name: Anze A Ursic

Re: Spellman DF3 Power Supply Fusor Application

Post by Anze A Ursic »

Rex, brilliant post, thank you very much! I actually ended up peeling back that wire mesh quite a bit, my research advisor saw it and said the same thing. It initially was actually shorting to that, good catch! But now it's shorting across the pins, we are certain of it because we have video. Also, we have the proper 5" federal standard 75kV plug, it's not a home-built solution or anything.

That being said, your comment about specialty-built DF3s may actually be correct. Our device doesn't have a model number! It just has a serial and parts of the information plate on the back are missing, leading us to believe this was a custom-built power supply. I just hope whatever configuration it is, it will be able to be used for our fusor.

I will keep everyone posted once further testing is done.

Anze
Rex Allers
Posts: 520
Joined: Sun Dec 30, 2012 3:39 am
Real name:
Location: San Jose CA

Re: Spellman DF3 Power Supply Fusor Application

Post by Rex Allers »

Anze,
Glad you found my post somewhat helpful.

A couple things I don't understand.

You say you have a proper 75 kV plug. Is that with the cable built in to it then? I only know what I've read about, not direct hands on, but I haven't seen any connector listed that you could solder your own cable into. It sounds good but I'm trying to understand what exactly that proper plug means.

What pins and where are you sure are shorting? You say you have video but I can't think where it might be that you can catch on video. Could you be saying just the bare supply socket without anything plugged in is arcing?

I'm puzzled.
Rex Allers
User avatar
Dennis P Brown
Posts: 2599
Joined: Sun May 20, 2012 10:46 am
Real name: Dennis P Brown
Location: Glen Arm, MD

Re: Spellman DF3 Power Supply Fusor Application

Post by Dennis P Brown »

Check the resistance between the pins and the case - hopefully, S&L aren't grounding to the case. Those pins are very close together and you say they are arcing - that couldn't happen if the S&L are supposed to be isolated and floating relative to the cathode pin. That tells me there is likely an internal source of grounding at elevated voltage that is making S and/or L see ground. Then you get the arc from 'C'. The fact that the system see's that the filament isn't present is an issue but not the primary one.
User avatar
Anze A Ursic
Posts: 109
Joined: Thu Mar 10, 2022 9:28 pm
Real name: Anze A Ursic

Re: Spellman DF3 Power Supply Fusor Application

Post by Anze A Ursic »

OK, back in the lab, so let me answer some questions.

1. Dennis: you mentioned tying C to either L and S and using that as the output. Issue is, when L and S are not shorted on the cable, we get that filament / current error because it thinks the filament is missing.

Also, neither pin is shorted to ground. Just checked. L, S and C all give a resistance measurement of 300-500MOhm when measured to ground. That's good! Based on the info Spellman gave us, it seems as though L and S heat up the filament with an AC current and C "brings it up" to the high voltage as a HV DC offset. This would also make sense why those four wires are so close to each other with next to no insulation, especially when we're talking about several tens of Kilovolts.

For everyone asking, below is our cable. We ordered this custom made, but we asked for only one side to be completed as an actual plug, obviously, since for our fusor input we just need a bare wire, which will connect to our massive ballast resistor and from then goes to the HV input on the chamber. We did, however, ask for the manufacturer to also add the uncompleted plug with our order, so we can make a custom cable if needed. The one I'm holding in my hand is the actual cable that arcs and you can see it plug into the DF3 power supply (the supply is off and had been disconnected from the outlet for hours, I wouldn't be touching it like this otherwise.) Next to it on the ground you can see just the plug with a single white wire coming out. That's the plug with just the C pin connected. Also I know the plug should be cranked down with the large coupling nut, but I took it off just to use with the other plug last week, but it's always there and screwed in when we do use this thing.
20220330_100541.jpg
2. Rex, I reread your lengthy post one more time. Thanks again for all the intricate analysis. The fact that the DF3 has no model number on the back makes me believe this is a custom built device with a different configuration, but I can't see how either of those setups you show in the picture would coincide with my measurements. L and S are shorted (R = 0.8Ohm across them) and measuring C with L/S if 477Ohms. You said to just use C, but again, that arcs to L/S when we do that. Further, the filament current IS set to 0, always, although it is interesting, whenver we turn the supply on, the filament amperes monitor always ends up reading 0.88A for whatever reason. Again, the know is turned all the way to 0 and is fixed at that position and reads 0 before the supply's HV output is activated... Another anomaly.

Also to answer your question "You say you have a proper 75 kV plug. Is that with the cable built in to it then?" Yes, from the picture above we have a federal standard 75kV plug. The cable is built into that. Arcing only happens across those wires at the end of the cable. I haven't seen shorting internally yet, but again, with no cable connected, the thing just shuts down because it sees no filament.

Thanks everyone!

Anze
2.
User avatar
Dennis P Brown
Posts: 2599
Joined: Sun May 20, 2012 10:46 am
Real name: Dennis P Brown
Location: Glen Arm, MD

Re: Spellman DF3 Power Supply Fusor Application

Post by Dennis P Brown »

The fact that L&S pins read 500 M-Ohms to ground does not mean these aren't the source of the grounding issue. When a wire is holding off 60 kV arcing internally is very possible but an ohm meter will never 'see' that ground source because it requires a high voltage to cause the jump to ground to occur. You would need to trace the circuit/wires of the filament system and confirm that none of that is permitting arcing with a safe high voltage source (say a NST - I use such a NST system to safely test HV diodes - very useful.)

I say this because you indicated that the high voltage is arcing from the 'C' pin to the other pins (or cable ends- same difference) - that absolutely means the filaments circuit for those pins(L&S) is 'seeing' ground when you reach a critical value of voltage.

As for the cable, I was only worried about arcing from the bare metal wires to a ground near the cable - not the cable but good you checked that. As Rex pointed out, that high field will find a way (the outer ground mesh in that case) so it is good you checked that aspect of the cable.
Rex Allers
Posts: 520
Joined: Sun Dec 30, 2012 3:39 am
Real name:
Location: San Jose CA

Re: Spellman DF3 Power Supply Fusor Application

Post by Rex Allers »

Great to have that pic of your cables and see that they are both professionally made and have the proper connector for this DF3 supply.

You said:
  • Based on the info Spellman gave us, it seems as though L and S heat up the filament with an AC current and C "brings it up" to the high voltage as a HV DC offset.

Yes, that is what my reading eventually found as the normal configuration and I drew up in my picture section labeled "Standard Default".

Just for my curiosity it would be nice if you could remove the 3-pin plug and cable and then ohm out (end-to-end) to see what wire each of the 3 pins (C, L, S) is connected to, and report back. This shouldn't change anything but I'd just like to know how a proper cable is connected.

In your picture, it looks like the white connector and thin cable sitting on the floor are a custom cable you had made with just the C pin installed and a single HV wire out. If I understand correctly, you can't use it because the supply, expecting an xray tube, senses an open filament and won't power on the HV. From your statement that I copied above, it seems you have some contact with Spellman. Maybe you can ask for help. Tell them you would like to use the DF3 just as a HV supply and ask if there is some setting or configuration where you can disable the filament alarm and just use the HV supply. Then you could use the simpler connector and cable.

Now back to your arcing. If I understand what you are saying, the arcing is happening at the end of the cable away from the supply, show in the top-center of the pic you just shared. It does not look to me that you took my advice to cut back the braid several inches. I would recommend at least 6" or 150 mm.

Here's one of the several sections from my earlier post where I mentioned this issue:
  • The other end of the cable (where it will feed the fusor) must also have the braid stripped back quite a few inches to avoid arcing there too.

    So, my educated guess is that the arcing you are seeing is to the braid. Black plastic insulating tape is no match. Distance is needed.

Today, I cut back the braid on the cable I have here and took pictures to show exactly what I mean. But, first, I was thinking there is another thing you should do before beginning my process, or maybe between steps 1 and 2. I suggest you cut back the thick center insulation by an inch or two. This will give a clean end of the insulation in case previous arcing has damaged it, and will give you a longer and clean section of the central 4 wires. You might want to cut off the original wire ends that have seen some arcing. Cutting a portion of this thick insulation is a difficult task. When I did it I first cut around where I wanted the new end but careful not to cut down to the center wires. Then I made a few cuts lengthwise to create sectors and removed small sections with a sharp knife and small cutting pliers. I worked to leave only a thin section of insulation around the center wires then cut and peeled that carefully, to not damage the center wires.

When removing a long section of the braid, as described below, be sure the 6 or more inches begins at the new cut insulation end just described.

Here are the braid cut-back steps...
step1.jpg
step2.jpg
step3.jpg
step4.jpg
step5.jpg

Optionally, you could then use a thin probe of some kind to unwind the wires of the remaining braid. then you could twist the wires into 2 or three pigtails and point these away from the HV wires at the end of the cable.
Rex Allers
User avatar
Anze A Ursic
Posts: 109
Joined: Thu Mar 10, 2022 9:28 pm
Real name: Anze A Ursic

Re: Spellman DF3 Power Supply Fusor Application

Post by Anze A Ursic »

Rex,

thanks again for the thoughtful response. I actually have ohmed out the plug to wires and I remember that the bare wire is C and the other two are L and S. I forget specifically which one is yellow and which is green but I wrote it down in the lab notebook so I can definitely let you know tomorrow.

I will do what you said with the 6" of braid. I will make sure to do it right and I'll remove the damaged part of the cable too. I'll take a picture.

In terms of help from Spellman, it's a no-go. We've been in contact with one of their engineers and he kind of refuses to tell us how to use this supply in any other method that it was not intended for. I get it, liability, but I'll try and atleast figure out if our model (based on Serial number) is something special in terms of configuration.

I'll leave all three wires flush and unconnected and try to power on the HV. I think a more systematic approach is warranted at this point. Otherwise we're just sticking random resistors across the different pins and smelling ozone and burnt plastic at this point.

Thanks for the help, I will do post some results of a systematic approach.

Anze
John Futter
Posts: 1826
Joined: Wed Apr 21, 2004 10:29 pm
Real name: John Futter
Contact:

Re: Spellman DF3 Power Supply Fusor Application

Post by John Futter »

Some cables like the one pictured use a conductive layer under the braid and that too has to be carefully cut away
User avatar
Richard Hull
Moderator
Posts: 13898
Joined: Fri Jun 15, 2001 9:44 am
Real name: Richard Hull

Re: Spellman DF3 Power Supply Fusor Application

Post by Richard Hull »

I was about to say that arcing at the end of the wire stubbed like shown indicates perfect operation, Rex beat me to it. you cannot test a 50-75 kilovolt cable hook up with out a huge separation of the hot leads. I guess I assumed the users knew that the wires had to be separated by several inches at the output end and far enough from the grounded braid so as not to arc. Oh well, I assumed too much. The photo helped.

Richard Hull
Progress may have been a good thing once, but it just went on too long. - Yogi Berra
Fusion is the energy of the future....and it always will be
Retired now...Doing only what I want and not what I should...every day is a saturday.
User avatar
Anze A Ursic
Posts: 109
Joined: Thu Mar 10, 2022 9:28 pm
Real name: Anze A Ursic

Re: Spellman DF3 Power Supply Fusor Application

Post by Anze A Ursic »

John, if you are talking about the black layer underneath the braided ground wire, it is nonconductive. I have tested it.
John Futter
Posts: 1826
Joined: Wed Apr 21, 2004 10:29 pm
Real name: John Futter
Contact:

Re: Spellman DF3 Power Supply Fusor Application

Post by John Futter »

Good
i've been caught before
User avatar
Anze A Ursic
Posts: 109
Joined: Thu Mar 10, 2022 9:28 pm
Real name: Anze A Ursic

Re: Spellman DF3 Power Supply Fusor Application

Post by Anze A Ursic »

Ironically, I got an X-ray of my knee today and their device used the same type of connector! I should've asked the tech to tell me how to connect it. :-)
User avatar
Richard Hull
Moderator
Posts: 13898
Joined: Fri Jun 15, 2001 9:44 am
Real name: Richard Hull

Re: Spellman DF3 Power Supply Fusor Application

Post by Richard Hull »

Sadly, many among us here could rip the top off that thing and search for the single little logic line that is at a logic one or a logic zero which alerts the over-watch circuitry that says "there be filament here!", and tie that line down forever to that "fils OK line.

Alas, we are not there and can't guide you in our remoteness.

Dare you try to fake it out with a well insulated, highly isolated air of oil cooled glow-bar resistor so it thinks it has a filament load??? (that resistor will be deadly electrified against ground to the full negative HV potential.)

Richard Hull
Progress may have been a good thing once, but it just went on too long. - Yogi Berra
Fusion is the energy of the future....and it always will be
Retired now...Doing only what I want and not what I should...every day is a saturday.
User avatar
Anze A Ursic
Posts: 109
Joined: Thu Mar 10, 2022 9:28 pm
Real name: Anze A Ursic

Re: Spellman DF3 Power Supply Fusor Application

Post by Anze A Ursic »

I've opened this thing apart and disassembled it to some extent before, but without circuit diagrams, figuring out which thing controls the filament error would be a pain in the neck. I suppose I could retrace what circuitry turns on the front lightbulb during this error, but I'm trying to avoid that for now.

Anze
User avatar
Dennis P Brown
Posts: 2599
Joined: Sun May 20, 2012 10:46 am
Real name: Dennis P Brown
Location: Glen Arm, MD

Re: Spellman DF3 Power Supply Fusor Application

Post by Dennis P Brown »

Wouldn't a cathode in a fusor be equivalent to a filament? That is, simply use a small wire cage in a vacuum chamber (metal) and place that under vacuum. At high vac it should allow full voltage stand-off and the said cathode would then be seen in the 'logic' as a filament?
User avatar
Anze A Ursic
Posts: 109
Joined: Thu Mar 10, 2022 9:28 pm
Real name: Anze A Ursic

Re: Spellman DF3 Power Supply Fusor Application

Post by Anze A Ursic »

Excellent news! Upon Rex's fine idea of stripping more of the wire mesh and black stuff from the cable, I did so. Pictures below.
cable.jpg
I actually only stripped the metal wire mesh but I figured I'd re-measure the resistance of the black mesh too and across that entire 6" length, it was only 10k Ohm! Last time I measured it was at 11pm on a thursday so maybe in my drowsiness I set the range to Giga Ohms and it read 0, but seeing as it is not a good insulator, I cut that off too and now I believe it could have been shorting to that! Because I removed the wire mesh before, although not 6", but never the black layer, so it was less than a quarter-inch removed from the bare C wire.

Then I tried two tests (so far). First was leaving all three wires disconnected from each other (L, S and C disconnected). It ramped up to 15kV normally with no arcing but I got mA Regulator error and Fil. Cur. Limit. So again, I figured with no load across L and S the thing doesn't see the filament and spits out these errors. So I shorted L and S, leaving C disconnected, which usually resulted in arcing, but this time it did not! In fact, all I got was the OPEN FILAMENT error and the supply shut down. For the sake of everyone, I am also attaching a screenshot of the manual's error subsection.

Overall, getting rid of arcing is fantastic, it just seems to me now that I have to connect L, S and C in such a way as to trick the supply thinking there is a filament there. Once that's done I will be ready to, dare I say, rock n' roll!
manual errors.PNG
User avatar
Richard Hull
Moderator
Posts: 13898
Joined: Fri Jun 15, 2001 9:44 am
Real name: Richard Hull

Re: Spellman DF3 Power Supply Fusor Application

Post by Richard Hull »

To answer Dennis' question about the cathode of the fusor acting as a filament equivalent.....

NO! A filament is a high current closed loop as in the x-ray tube. Two of those three wires are expecting a multi-ampere burning filament load and the supply checks for that load. If it is not there, the system figures the tube is absent or the filament is burned out, and thus, will not allow the HV to be brought up.

The cathode/grid in a fusor is not a filament in spite of it being part of a closed, conductive gas circuit for the high voltage in the milli-ampere range during normal operation.

Richard Hull
Progress may have been a good thing once, but it just went on too long. - Yogi Berra
Fusion is the energy of the future....and it always will be
Retired now...Doing only what I want and not what I should...every day is a saturday.
User avatar
Anze A Ursic
Posts: 109
Joined: Thu Mar 10, 2022 9:28 pm
Real name: Anze A Ursic

Re: Spellman DF3 Power Supply Fusor Application

Post by Anze A Ursic »

I think we figured it out. I connected L and C and left S disconnected and just cranked up the filament current to 0.9A. Since as mentioned before, the filament output screen keeps showing 0.88A no matter the settings (I always had the knob for filament current set to 0, I don't know where 0.88A comes from), so I figured just above that should work. Well it does! It went to -15kV with no errors or arcing and stayed there for 30 seconds! Hopefully this all works when we connect to the fusor.

For now, thank you all! I very much appreciate it!
Rex Allers
Posts: 520
Joined: Sun Dec 30, 2012 3:39 am
Real name:
Location: San Jose CA

Re: Spellman DF3 Power Supply Fusor Application

Post by Rex Allers »

Anze,

Good catch on the conductive black layer of the cable. I did not know and would not have expected it. I just measured the cable I used to take pics of the cut-back. I wrapped some wire around the black material near end (for good connection all the way around) and measured resistance through the ~6" of unbraided length to the remaining braid -- less than 2k ohm.

In retrospect, in that pdf for the cable I posted earlier, '2214.pdf', it looks like the black layer is 'D' in the drawing which has the text "SEMI CON EPR", so semiconductive, I can now gather. Great that you figured that out.

-----
On the C, S, L pins of the HV connector... It looks like I misinterpreted the manual's text that I found in an odd place. Today I was looking at Spellman web pages and found clearer wording in the data sheet 'DF-FF.pdf'.
  • Output Connector:
    75kV, 3 conductor Federal Standard X-ray connector. -60kV is connected to terminal “C”. Terminals “S” and “L” are jumped together. The filament output is connected between terminals “C” and “S”. Other configurations are optional. (On the FF3, all output connections S, L, & C are connected together).

So that HV wiring figure I drew in my first long post was wrong. It looks like the XLF one you put in your first message of this tread is correct for the DF3 too. Sorry for the bad info I provided from my misinterpretation of the wording in the DF3 Instruction manual. I would edit my earlier post to fix but seem to be past the time where I can do this.

-----
I'm not sure if you have noticed one short paragraph in the "OPERATING INSTRUCTIONS" section of the Instruction Manual about Presetting operation limits. This one:
  • J) Presetting: Depress and hold the X-ray
    off switch. The kV, mA, filament current limit
    and maximum tube power limit can now be
    preset. NOTE: No output will be generated
    even though the meters are displaying in the
    preset mode. It is suggested to set kV to a
    minimum value, (10 kV-20kV). If kV is set too
    low the kV MIN indicator will illuminate.
    Simply raise the kV level if this occurs. Set mA
    level to a minimum value (5mA). Filament
    current and tube power limit should be set as
    required by the X-ray tube manufacturer.
    Release the X-ray OFF switch.

So I am wondering, maybe the Filament current knob on the front panel only works, during this configuration process, to set this max limit? Maybe that is why you always see .88 A in operation. Maybe the max power knob also only works in this presetting mode.

It might be worth trying to enter this Preset mode and see what the present values are, then set Max Filament current to some very low value rather than waste a lot of filament power with the shorted connection you seem to need to get active HV.
Rex Allers
John Futter
Posts: 1826
Joined: Wed Apr 21, 2004 10:29 pm
Real name: John Futter
Contact:

Re: Spellman DF3 Power Supply Fusor Application

Post by John Futter »

Aha
again the black stuff comes to haunt
told ya
User avatar
Anze A Ursic
Posts: 109
Joined: Thu Mar 10, 2022 9:28 pm
Real name: Anze A Ursic

Re: Spellman DF3 Power Supply Fusor Application

Post by Anze A Ursic »

Yup, thanks John! I knew someone mentioned it. Appreciate it.
Really happy it works now!
User avatar
Anze A Ursic
Posts: 109
Joined: Thu Mar 10, 2022 9:28 pm
Real name: Anze A Ursic

Re: Spellman DF3 Power Supply Fusor Application

Post by Anze A Ursic »

I've figured out the supply, so I need no additional help (for now), but I will post all the data I have collected here. That way, it may be helpful for other people in the future. I have also had some good conversations with Spellman engineers so I am posting those too, in case you can infer some wisdom from them.

Plug on the back of the device has three pins (C = common, L = large filament and S = small filament). In my situation, the high voltage had the following color coding: L = Yellow, S = Green and C = two bare wires.
PIN OUTPUT.PNG
There is a ground post on the back of the supply, that is what the -60kV is with respect to. All three pins (L,C,S) measured an impedance of several Giga-ohms with respect to ground. L and S are internally shorted (stated in the manual and tested) and the impedance from L/S to C for my supply is about ~470 Ohms. When trying to configure this supply as a fusor supply, I ran several configurations of wiring of L, C and S. Below are the configurations and the errors I got with them. The definitions of errors are in the DF3 manual which is easily found online and I also posted the screenshot of errors in this thread a few posts back, so use that for your reference.

L, C, S disconnected with each other (current set to 0): mA RGLT ERROR, FIL. CUR. LIMIT

L, C, S disconnected with each other (current set to 1mA): mA RGLT ERROR, FIL. CUR. LIMIT, OPEN FILAMENT

Only L and S shorted (current set to 0): OPEN FILAMENT

Only L and S shorted (current set to 1mA): OPEN FILAMENT, FIL. CUR. LIMIT, mA REG. ERROR

What ended up working for us was L and C shorted and S disconnected, although since L and S are internally connected, that sort of just implies all three are shorted, which to some extent makes sense, when looking at the picture below provided by Spellman on how this supply operates.
df3 connection.PNG
Even with this configuration, (L and C shorted), we got the same mA REG. ERROR and FIL. CUR. LIMIT, what ended up working was cranking the "filament current" knob to 0.9A. Reason being, we noticed that no matter the settings, when testing the supply and the voltage was rising, the front panel filament current knob read 0.88A, so we just went slightly above that. I am not sure why that happens, nor do I care since that part is irrelevant as long as we have control of voltage and the mA. And, actually, writing this out, I am curious if that 0.9A filament current setting would also solve the issue in the previous configurations, I'll try next time in lab.

Here are the responses I got from Spellman that intricately describe how this unit operates:

"The DF3 is a negative polarity/floating filament X-Ray generator. See attached product data sheet and a couple of slides from an internal training class I teach. The DF was designed to operate X-Ray tubes under closed loop filament control. The customer provides a 0-10Vdc signal (from either the local front panel or remote interface) that represents 0-80mA and our internal closed loop emission control circuitry adjusts the output of the floating filament supply to provide the desired current emission current. There a timing sequence of bringing up the high voltage, ramping up the filament, over and under current faults, regulation error fault, filament open fault, filament short fault and other protective circuitry to assure proper operation. This was all done to protect the X-Ray tube. There are probably other issues that I’m not aware of as I’m not the DF product engineer. Trying to operate the DF unit in a non-X-Ray generator application will create some if not all of the above mentioned faults and perhaps other problems, preventing operation. This unit was never intended to be used in a non filamentary controlled X-Ray generator application."

"The DF/FF unit is used in both negative and positive output polarity.
In negative output polarity a floating filament is required so the additional pins on the Federal Standard 75kV 3 pin X-Ray connector come in handy.
In positive polarity a ground referenced filament is used, so we just tie the high voltage to all three pins, so they are all at the same potential."

"DF unit in negative output voltage so it has a floating filament, hence the use of the multi-pin Federal Standard X-Ray Connector. C = common, L = large filament and S = small filament. Filament current flows between common and Large/Small. The filament is a floating filament, one side is connected to the common so the filament if referenced to the high voltage output voltage. Filament is connected between the C and the L&S pins.

"FF unit is positive polarity so it has a ground referenced filament, connected on a rear panel terminal block. Since only one high voltage connection is needed all the pins are tied together."

Hopefully this helps other people with similar supplies. Overall, this thread became a wealth of information that helped me immensely and I hope it can do so for others. Thank you everyone!
Rex Allers
Posts: 520
Joined: Sun Dec 30, 2012 3:39 am
Real name:
Location: San Jose CA

Re: Spellman DF3 Power Supply Fusor Application

Post by Rex Allers »

Anze,

Thanks for sharing the details and glad you have found a configuration that works for you.

I'm wondering if you had a chance to see what 'Presetting' values are set in your supply? (I mentioned this, quoting from the manual paragraph, in my last message.)

If you don't have time to do any more looking now, I understand.
Rex Allers
User avatar
Anze A Ursic
Posts: 109
Joined: Thu Mar 10, 2022 9:28 pm
Real name: Anze A Ursic

Re: Spellman DF3 Power Supply Fusor Application

Post by Anze A Ursic »

The presetting values are just the values the supply is set to before it's turned on. You press the X-RAY OFF button to show them, set them, then press X-RAY ON button to turn on the supply and it will slowly ramp up to those preset values.
The filament current was always set to 0 in the preset. Once I went out of presetting mode and turned it on (X-RAY ON), it jumped to 0.88A. That's why it makes no sense. The Kilowatts knob under "preset" is set to the minimum, because if at 0, the supply automatically gives you a "OVER POWER" error on the front panel. Funny enough, just for a test, I cranked up the Kilowatt knob so it was a bit higher than usual, just to see if the filament current would change from 0.88A. It did not...
The kV and mA are preset via my LabView script.
Rex Allers
Posts: 520
Joined: Sun Dec 30, 2012 3:39 am
Real name:
Location: San Jose CA

Re: Spellman DF3 Power Supply Fusor Application

Post by Rex Allers »

Thanks for sharing the preset info and behavior. The filament current you see seems illogical but it is what it is.

I would think when you get to actually drawing current on the HV, you might need to set the kilowatt limit preset a bit above expected operating values.

I wish you good luck on getting to a situation where you can actually try it.
Rex Allers
User avatar
Anze A Ursic
Posts: 109
Joined: Thu Mar 10, 2022 9:28 pm
Real name: Anze A Ursic

Re: Spellman DF3 Power Supply Fusor Application

Post by Anze A Ursic »

Another update for those interested: this power supply, as mentioned before, can start with a minimal voltage of 15-16kV. Anything below it shuts it down on startup ang gives the kV min error. Anyway, the interesting part is, once you get to the 15-16kV and it stabilizes, you can then lower down to about 8.5kV before it shuts off with the kV min error. I will say, it's hard striking a plasma with 8.5kV minimum, it's easy for it to runoff with the current forcing me to shut it all down or I catch it and lower the pressure but then it stalls and extinguishes. Not to mention, once plasma ignites it creates like a mini-EMP or something which often shuts off a lot of our equipment, the camera in particular. Will do more testing though.
Post Reply

Return to “High Voltage - Fusor Input Power (& FAQs)”