2
\$\begingroup\$

I recently bought a 0-35V 0-3A (105W) CC/CV linear power supply, made by Takasago in Japan, second hand. It's a nice unit, and inside of it is a massive power transformer along with three pass transistors on a very large heatsink (it is interesting why there are three transistors but I can only suspect that they are to reduce the load on an individual transistor), and it weighs a ton. I've fully tested it to 105W, although only for about 5 seconds as even my 2x25W resistors were getting too hot.

I originally had plans to put it on top of my HP 54501A, a digitising oscilloscope with a CRT raster monitor (unlike a normal CRT scope which draws a trace by moving the electron beam - this sets pixels on and off representing the trace.) However, I found that when turned on, whether above the scope or on the bench below, the CRT would have a weird ripple effect or would shift position randomly. It didn't matter if the supply was loaded down or not. The trace was not affected, just the display (even text and graphics). Why would the supply cause this? I suspect it has something to do with the transformer because that makes a humming sound in operation if you listen carefully.

\$\endgroup\$
6
  • \$\begingroup\$ Can you determine the frequency of the ripple? \$\endgroup\$
    – tyblu
    Jan 19 '11 at 0:54
  • \$\begingroup\$ @tyblu I've since moved it away from the scope... I'm guessing 50 Hz or 100 Hz (because I live in the UK) - I suspect the transformer is producing a magnetic field of some kind, but I'm not entirely sure on the operation of that. \$\endgroup\$
    – Thomas O
    Jan 19 '11 at 0:56
  • \$\begingroup\$ Sounds like a perfect excuse to get a shiny new digital LCD scope! \$\endgroup\$ Jan 19 '11 at 14:27
  • \$\begingroup\$ @reemrevnivek Why? This one works perfectly well. It is a digital oscilloscope. \$\endgroup\$
    – Thomas O
    Jan 19 '11 at 17:48
  • \$\begingroup\$ @Thomas - Because you need more tools? It was a joke. \$\endgroup\$ Jan 19 '11 at 18:43
7
\$\begingroup\$

The power transformer in the power supply is producing stray flux lines, which are causing the scope electron beam to deflect.

The magnetic flux in a transformer is never entirely constrained to the transformer core, and as transformers get bigger, it becomes progressively more involved to minimize the stray flux as much as possible. Also, the standard EI core used in power transformers is not the best type when it comes to low stray flux (that would probably be a pot-core).

Basically, the transformer is acting like an electromagnet, modulated at the frequency of the mains.

The effect is akin to waving a magnet around your TV, since the monitor in your DSO is effectively a little TV.

Even scopes which use electrostatic deflection to steer the electron beam are magnetic fields. since it's a beam of electrons in either case, and electrons are charged particles, rendering them influenced by magnetic fields.

I have the same problem with a bit 100 V 10A power supply I have a work. Whenever I turn it on, it drives my (analog) scope nuts.

\$\endgroup\$
3
\$\begingroup\$

Fake Name pretty much took care of the "Why", I'll try to help with the "What now".

Of course, since the field strength is inversely proportional to the distance squared, moving the scope and PSU away from each other could also do the trick. Similarly, changing the orientation of the two devices so the CRT screen is perpendicular to the magnetic field lines could reduce the distortion and flicker to something more readable. Finally, I'm guessing that this is a linear power supply with only low frequencies on the transformer, so if having the scope and the controls for the PSU in the same location or on top of each other is important, you could probably put the transformer in its own box, and run a cable back to the PSU.

Another easy (but expensive) fix is to get a scope with an LCD. Most new scopes are like this. Now your oscilloscope won't be affected by the magnetic field. However, remember that you still have a significant magnetic field at your workstation. You can forget about working with magnetometers.

The third option is to encase the transformer in a material with a high magnetic permeability, like mu-metal. This will form the equivalent of a Faraday cage (which should, analogously, have high conductivity) to block stray magnetic forces. Mu-metal is a nickle/iron alloy with traces of copper and molybdenum that is often used for this purpose. Other nickle/iron alloys will also work. Layers of encasement will increase the attenuation exponentially if a single layer is insufficient to do the job.

You can get shielding from aircraft supply companies in tape form, used to wrap the magnetos and prevent radio interference. LessEMF.com sells various other forms such as sheet metal, 1" tape, and a more flexible mesh; shields were more popular when CRT monitors and TVs were in wide use so there are probably lots of other sources. You are probably dealing with relatively week magnetic fields, though, so it might be worthwhile to try out cheap steel sheet metal before buying something more exotic and expensive.

\$\endgroup\$
2
  • 2
    \$\begingroup\$ The option which worked for me was to move the power supply to the other side of the bench away from the scope. No rippling now. \$\endgroup\$
    – Thomas O
    Jan 19 '11 at 17:49
  • \$\begingroup\$ Glad that the easy way out worked for you. \$\endgroup\$ Jan 19 '11 at 18:41

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.