Toshiba D-138B(s) X-Ray Tubes: Filament Voltage and Wehnelt Bias / Electromagnetic Radiation / Forums

Forums

4hv.org :: Forums :: Electromagnetic Radiation

« Previous topic | Next topic »

Toshiba D-138B(s) X-Ray Tubes: Filament Voltage and Wehnelt Bias

previous 1 2

Move Thread

LAN_403

Sigurthr

Fri Jan 08 2016, 10:00AM

Registered Member #4463 Joined: Wed Apr 18 2012, 08:08AM
Location: MI's Upper Peninsula
Posts: 597

Here's a whole bunch of new radiographs!

hboy007

Sun Jan 17 2016, 11:46AM

Registered Member #1667 Joined: Sat Aug 30 2008, 09:57PM
Location:
Posts: 373

Awesome pictures, Sigurthr!

I'll have to put a microfocus tube on my wishlist to inspect those BGA chips and SeaRay connectors.

Sigurthr

Fri Jan 29 2016, 12:59AM

Registered Member #4463 Joined: Wed Apr 18 2012, 08:08AM
Location: MI's Upper Peninsula
Posts: 597

I added a -140V Wehnelt Bias supply on to the x-ray machine today. I am quite surprised it did not aim the resulting x-ray beam any better. My tube's beam does not exit at the marked location, but rather far off in two axes. Adding the Wehnelt bias did double the x-ray flux though which is pretty nice! I can cut exposure times in half now.

I didn't do a full photographic setup to ensure a good radiograph, only a quick bare-bones setup to verify performance and check beam steering. You can see the bottom of the image has much better exposure, this is where the beam is actually aiming, despite the anode pointing above 45deg higher than this.

In addition to the -45deg (rotational axis), the beam is also offset by about 25deg towards the cathode end (it should be 90deg to the axis of the tube).

Anyone have some insight as to the strange beam aiming? I'm wondering if this tube was maybe a factory second hence the strange beam aiming.

klugesmith

Fri Jan 29 2016, 09:48PM

Registered Member #2099 Joined: Wed Apr 29 2009, 12:22AM
Location: Los Altos, California
Posts: 1714

You're way ahead of me in actual radiography.

Not sure if your beam-direction issue is just about normal geometric foreshortening.
The e-beam target area is not a round spot. It's extended in one direction to reduce heating power density. Goal is equally small height and width when viewed at 90 degrees from tube axis.
Here I played with a drawing program instead of finding an illustration on the Internet.

I think the target region is a lambertian emitter of X-rays, after accounting for self-shielding by anode. So the maximum X-ray intensity is in a direction offset toward the cathode, but the image focus is not as good over there.

How about making a pinhole [in Pb metal] camera image of your x-ray emitter size and shape? Stella (pbuh) discussed that here many years ago.

Please show us more about your electrical configuration. I was going to say that the bias for Wehnelt cylinder (often labeled Focus or Grid terminal) does not need a third power supply. A plain resistor can do the job, or a shunt regulator. How 'bout a 0A2 or 0D3 cold cathode VR tube?

In fixed-current systems, a resistor also reduces the HV current sensitivity to filament voltage

Sigurthr

Fri Jan 29 2016, 11:51PM

Registered Member #4463 Joined: Wed Apr 18 2012, 08:08AM
Location: MI's Upper Peninsula
Posts: 597

Thanks Klugesmith! I drew up a schematic and a diagram showing the beam shape. Hopefully it is understandable, I'm not trained in nor do I have 3D software, which would really have made this a clearer visually.

QCjJ3RH

How does one determine the resistance needed for a bias resistor? Obviously R = Vbias/I, do I assume I = anode current?

Those VR tubes look awesome, but I don't see any in quick googling that work down to 1mA, most are rated 5mA, and my HV supply is only 3mA max from my calculations. I haven't been able to directly measure the current due to RF interference from the inverter. Could a Zener diode be used instead?

Sigurthr

Sun Jan 31 2016, 09:45AM

Registered Member #4463 Joined: Wed Apr 18 2012, 08:08AM
Location: MI's Upper Peninsula
Posts: 597

Well, I tried tonight to use the resistive bias method and got mixed results. I didn't bother doing a radiographic setup, but instead just used my ion chamber and scope to observe results. With a 68k bias resistor the x-ray flux output was comparable to having the Wehnelt electrode tied to filament. Probing across the 68k resistor yielded a surprising 40V according to my scope, but it was noisy with a ton of AC coupled on to it. I tried using some capacitance to shunt the ac but had no luck. I know from when scoping the divided HV that any ac remaining ruins any integrity the measurement might have had, so that 40V number is pretty much meaningless. I tried using different resistance from 10R to 18k but got nonsense numbers each time. The closest I came to a believable number was 2.5mA, but I think it was just a fluke because there was superimposed ac on all measurements. Sigh, an analog current meter would solve all of this.

Likewise, a 10sec burn destroyed a 1W 68k resistor, and if it really was only 40V across that would have been 0.6mA, something like 25mW of power - no where near enough. If we assume 1W of dissipation (which we know was met or exceeded) that yields about 3mA, which seems more reasonable given the input power and measured HV voltage.

I've placed orders for a few 5W 68k ceramic power resistors and also a few 150V 5W zener diodes. I was going to get a VR tube but couldn't find any that worked at <5mA with a >100V regulation. I found some 85V VR tubes which worked down to 1mA, but they were pricey and I would need sockets and a lot of space to mount them. The zeners have a graph in the datasheet showing their regulation down to 100uA, and are rated for a maximum current of 30mA, so they should work well. I'll give both the resistor and zener tries when they arrive, including an actual radiographic test. I expect the zener to work best as the voltage would be independent of the current, and the wasted energy is 50% less than with the resistor.

previous 1 2

Moderator(s): Chris Russell, Noelle, Alex, Tesladownunder, Dave Marshall, Dave Billington, Bjørn, Steve Conner, Wolfram, Kizmo, Mads Barnkob