Connect with us

HV Probe construction advice

Discussion in 'Electronic Repair' started by Damian Menscher, Dec 16, 2004.

  1. A pet project for the past several years is repairing an ancient
    Tektronix 674 oscilloscope. (The high voltage circuit was going into
    overload protection, I think as a result of leakage from the diodes
    (5647 tubes). Replacing them with solid-state diodes seems to have
    fixed that issue (I wish I could keep it in its original form, but I
    simply gave up after being unsuccessful for a few years.).

    Anyway, with the reduced load for lighting the tubes, I think this
    needs to be recalibrated (intensity suggests it's got too much HV
    now). The service manual indicates a HV test point, at which I
    should read 2.2kV. Problem is, I'm just a hobbyist with no HV probe.

    After reading specs on a 6kV Fluke HV probe, I realize they're really
    just voltage dividers (the Fluke uses a 75MOhm resistor and a 75kOhm
    resistor). So I figure I could save myself $75 by building my own.
    Off I go to radioshack.com, where I find 5-packs of 1MOhm, 1/2W
    resistors for $1. Recalling high school physics, I work out that one
    of these should be able to drop 707V at .707mA. So if I put 10 in
    series, I should be able to measure up to 7kV, and simply measure off
    the last one and multiply the voltmeter reading by 10.

    One problem: Rat Shack says these have a "max working voltage" of
    350V (see http://www.radioshack.com/search.asp?find=271-1134). Are
    they just arbitrarily throwing a fudge factor of 2 away, or are they
    mislabeling 1/4W resistors as 1/2W resistors, or what? (They list
    10Ohm 1/2W resistors as 350V max also, so I'm guessing this is safe
    to ignore.)

    Any other warnings I should heed before doing this? I realize I'll
    lose some accuracy (they're only accurate to 5%). I guess I also
    have to worry about affecting the circuit when I draw .7mA out of it?
    I think my measurement error should be small, since my multimeter has
    a 10MOhm imput impedence. Is that good enough, or should I use a
    smaller final resistor in the voltage divider to reduce measurement
    error?

    Finally, my crazy idea of protecting myself from electrocution is to
    encase this thing in the plastic barrel of a Bic pen. They seem
    about the right size, plastic doesn't conduct, and as long as I hold
    it at the low-voltage end, it seems reasonably safe. Am I asking
    for trouble here, or does that seem reasonable for measuring only up
    to 7kV? I suppose I could build a safer handle onto it if needed.

    I'd appreciate any advice that could improve my life expectancy. ;)

    Damian Menscher
    --
    -=#| Physics Grad Student & SysAdmin @ U Illinois Urbana-Champaign |#=-
    -=#| 488 LLP, 1110 W. Green St, Urbana, IL 61801 Ofc:(217)333-0038 |#=-
    -=#| 4602 Beckman, VMIL/MS, Imaging Technology Group:(217)244-3074 |#=-
    -=#| <> www.uiuc.edu/~menscher/ Fax:(217)333-9819 |#=-
    -=#| The above opinions are not necessarily those of my employers. |#=-
     
  2. NSM

    NSM Guest

    | A pet project for the past several years is repairing an ancient
    | Tektronix 674 oscilloscope. (The high voltage circuit was going into
    ....
    | One problem: Rat Shack says these have a "max working voltage" of
    | 350V (see http://www.radioshack.com/search.asp?find=271-1134). Are
    ....
    | Finally, my crazy idea of protecting myself from electrocution is to
    | encase this thing in the plastic barrel of a Bic pen. They seem
    | about the right size, plastic doesn't conduct, and as long as I hold
    | it at the low-voltage end, it seems reasonably safe. Am I asking
    | for trouble here, or does that seem reasonable for measuring only up
    | to 7kV? I suppose I could build a safer handle onto it if needed.

    1. I'd be reluctant to push the voltage too far. If one flashes over they
    all will. You're probably OK but be cautious.

    2. Don't even THINK of using a hand held probe. Hook the divider up to the
    circuit, turn the scope on and check the voltage. Don't touch anything
    except the scope on/off switch or however you control this setup.

    3. 5% is near enough.

    4. If you want to own a probe, buy the real thing on eBay. Remember: "One
    flash and you're ash"!

    N
     
  3. James Sweet

    James Sweet Guest


    They can flash over, that's your main worry. I've seen people make HV probes
    with simple resistors though, the best way seems to be to chain them
    together and then place them in a piece of plastic conduit then fill that
    with mineral oil. You can make a reasonably professional HV probe that way,
    and you're only measuring a few Kv so you should be fine. That said, I got a
    real Fluke 40 Kv HV probe on ebay for under $30 a couple years ago, I've
    seen the 6 Kv ones go for under $20 but with a 40 Kv you can measure CRT
    anodes.
     
  4. aurgathor

    aurgathor Guest

    In high voltage circuits like that there's not a whole lot
    current, and 10 Meg could be a significant load on it;
    however, that's probably no big deal.
    Not exactly. With voltages in that range theres's a possibility
    of a flash-over. You definitely want to make the divider so
    the resistors are phisically arranged in a series like:
    --####---####---####---
    and don't use your HV probe when the humidity is too high.

    Also, I think those high value resistors are composite, so put
    one or two metal films in there -- if something bad happens,
    the metal film will evaporate, so in essence, it will act as a
    fuse.

    If you ever get hit by a CRT, 2.2kV isn't that high of a voltage,
    but it can be very inconvenient, or deadly if the PS can supply
    enough juice.
     
  5. mike

    mike Guest

    Your resistors are too small. Too much load on a typical high voltage
    circuit. You're putting too much voltage on each resistor. When you
    buy resistors from RS, you have no idea what you're getting.
    When you use a pen as a housing, you have no idea what impurities are
    in there and whether it will arc over or through.

    If you have regular need for a HV probe, put a watch on EBAY for one
    and snag one that goes at a low price. Whole bunch of 'em have sold
    near the $10 mark recently.
    IF you only need this one measurment, borrow a probe and be safe.

    I wouldn't risk my life on a probe that cost me $8 to build when I could
    have bought a safe one at a swapmeet for $10.
    mike

    --
    Return address is VALID.
    Wanted, Slot 1 Motherboard
    500MHz Tek DSOscilloscope TDS540 Make Offer
    http://nm7u.tripod.com/homepage/te.html
    Wanted, 12.1" LCD for Gateway Solo 5300. Samsung LT121SU-121
    Bunch of stuff For Sale and Wanted at the link below.
    http://www.geocities.com/SiliconValley/Monitor/4710/
     
  6. He's talking about the miserable almost no current output of a scope
    high voltage multiplier! Everyone is going safety crazy. I'm all
    for safety - there are enough references to it in the FAQs but for
    this, a simple voltage divider really is sufficient and safe. At
    worst, it will fry his meter if something bad happens.

    This is NOT a microwave oven.

    This is NOT projection TV.

    It's only 2 or 3 kV at a few hundred microamps max.

    2 or 3 kV isn't going to flash over 6 inches!

    See: http://repairfaq.ece.drexel.edu/sam/hvprobe.htm

    However, I do agree that if you plan to do more HV stuff, be on the lookout
    for a proper HV probe.

    --- sam | Sci.Electronics.Repair FAQ Mirror: http://repairfaq.ece.drexel.edu/
    Repair | Main Table of Contents: http://repairfaq.ece.drexel.edu/REPAIR/
    +Lasers | Sam's Laser FAQ: http://repairfaq.ece.drexel.edu/sam/lasersam.htm
    | Mirror Sites: http://repairfaq.ece.drexel.edu/REPAIR/F_mirror.html

    Note: These links are hopefully temporary until we can sort out the excessive
    traffic on Repairfaq.org.

    Important: Anything sent to the email address in the message header above is
    ignored unless my full name is included in the subject line. Or, you can
    contact me via the Feedback Form in the FAQs.
     
  7. See: http://repairfaq.ece.drexel.edu/sam/hvprobe.htm

    Building a voltage divider out of 10, 10M ohm resistors (and suitable
    math!) is sufficient and safe.

    However, I do agree with the other posts that if you plan to do more HV
    stuff, be on the lookout for a proper HV probe.

    --- sam | Sci.Electronics.Repair FAQ Mirror: http://repairfaq.ece.drexel.edu/
    Repair | Main Table of Contents: http://repairfaq.ece.drexel.edu/REPAIR/
    +Lasers | Sam's Laser FAQ: http://repairfaq.ece.drexel.edu/sam/lasersam.htm
    | Mirror Sites: http://repairfaq.ece.drexel.edu/REPAIR/F_mirror.html

    Note: These links are hopefully temporary until we can sort out the excessive
    traffic on Repairfaq.org.

    Important: Anything sent to the email address in the message header above is
    ignored unless my full name is included in the subject line. Or, you can
    contact me via the Feedback Form in the FAQs.
     
  8. Ol' Duffer

    Ol' Duffer Guest

    350 is pushing the envelope. Most designers use 200 as a
    reasonable ceiling for common resistors, and it doesn't
    necessarily increase linearly with wattage. High voltage
    resistors are optimized for their task, i.e. long, sealed
    with low leakage substances, etc. In addition to creepage
    (flashover), there's a voltage coefficient that usually
    isn't mentioned because it doesn't amount to much at the
    voltages the resistors are designed for, but a typical
    figure might be a few percent per KV per resistor. So
    the more resistors you string together, the better.
    Add a percent or two for the voltage coefficient.
    I'm not familiar enough with the particular circuit, but
    the current sounds a bit high for high voltage. Also, I'm
    thinking their 2.2KV "test point" may already be the tap of
    a resistive divider, which will also limit current available
    and affect accuracy. In other words, what is the source
    impedance? Do they tell you what to measure this point with,
    give you a minimum input impedance or any other clues?
    10Meg across 1Meg equals 9% error, but it's a predictable
    error that you can factor out. *If* your meter is really
    10Meg (which some are). Or does it say >= 10Meg?
    I wouldn't hold it at all. Suspend it with electrical tape,
    dental floss, whatever. "Probably" not enough current available
    to kill you, unless you have a pacemaker or something, but you
    wouldn't enjoy the experience.
    Probably okay at the 2.2KV you are expecting, but I wouldn't
    push it much higher. Larger diameter and thicker tube wall
    and string more resistors together for 7KV.
     
  9. mike

    mike Guest

    And if that's what he'd proposed, I probably wouldn't have chimed in at
    all. Reread the original post. He said "10 'one meg' resistors in series".
    I maintain that the value is too low.

    I err on the side of safety simply because there are so many people here
    recommending generally unsafe behavior without sufficient explanation of
    why it might be OK in THIS case. Note that people (even famous repair
    gurus) often fail to read
    what's written, as suggested by your own comment above.

    Remember that these posts hang around in the archive for a LONG time.
    It's easy to imagine some kid reading a random post out of context,
    generalizing the snippet of information and killing himself off.
    Anybody who has to ask how to build a voltage divider shouldn't be using
    their first attempt on a 2KV measurement inside an instrument that has
    other sources of lethal voltages...especially when there are a bunch of
    commercial HV probes on ebay for cheap. Save your death wish for
    something that can save you some real money.

    While I'm on a rant, I'll suggest a general process to be used when
    posting advice to the web. You're familiar with the game "telephone".
    Tell someone what you're about to post. Tell them to pass it on.
    Come back a day later, pick someone 3 layers down the chain and see
    if what they heard bears any resemblance to what you said. If it's
    still safe, post it to the web. Soon you'll get to where you can do
    this in your head quickly by reviewing your proposed post with a blank
    mind. (I pride myself on my blank mind) The reader has no idea what
    you meant.
    All they know is what you wrote...then interpreted in their own unique way.

    Just cause I got away with it for 30 years,
    doesn't mean It's safe to recommend it to a newbie.

    mike


    --
    Return address is VALID.
    Wanted, Slot 1 Motherboard
    500MHz Tek DSOscilloscope TDS540 Make Offer
    http://nm7u.tripod.com/homepage/te.html
    Wanted, 12.1" LCD for Gateway Solo 5300. Samsung LT121SU-121
    Bunch of stuff For Sale and Wanted at the link below.
    http://www.geocities.com/SiliconValley/Monitor/4710/
     
  10. Jim Yanik

    Jim Yanik Guest

    I agree,and it WILL load the HV circuit.Even a 100 Mohm probe does.
    I used 4 22.5Mohm or 9 10Mohm 1/2w precision metal film resistors in series
    for a simple probe for a 10 Mohm input Z DMM. It works great.


    HV <------[90meg]------[DMM 10meg]-----/gnd

    If you do the true divider way,you have to compensate for the DMM input
    Z,if you want the probe to be reasonably accurate.
     
  11. Asimov

    Asimov Guest

    "Sam Goldwasser" bravely wrote to "All" (16 Dec 04 08:57:46)
    --- on the heady topic of "Re: HV Probe construction advice"

    SG> From: Sam Goldwasser <>

    SG> See: http://repairfaq.ece.drexel.edu/sam/hvprobe.htm

    SG> Building a voltage divider out of 10, 10M ohm resistors (and suitable
    SG> math!) is sufficient and safe.

    SG> However, I do agree with the other posts that if you plan to do more
    SG> HV stuff, be on the lookout for a proper HV probe.

    The high voltage divider in a broken flyback can be used in a pinch.
    There are high value resistors in it along with high value adjustable
    pots for the screen and focus. It may be hard to get the potted stuff
    out though...

    A*s*i*m*o*v

    .... You mean 15" sparks are SUPPOSED to come out of this thing?!?
     
  12. If it's a one time measurement, no real need to disassemble it.

    --- sam | Sci.Electronics.Repair FAQ Mirror: http://repairfaq.ece.drexel.edu/
    Repair | Main Table of Contents: http://repairfaq.ece.drexel.edu/REPAIR/
    +Lasers | Sam's Laser FAQ: http://repairfaq.ece.drexel.edu/sam/lasersam.htm
    | Mirror Sites: http://repairfaq.ece.drexel.edu/REPAIR/F_mirror.html

    Note: These links are hopefully temporary until we can sort out the excessive
    traffic on Repairfaq.org.

    Important: Anything sent to the email address in the message header above is
    ignored unless my full name is included in the subject line. Or, you can
    contact me via the Feedback Form in the FAQs.
     
  13. Jim Adney

    Jim Adney Guest

    674? What's that? Is there a typo there?
    That's a pretty low voltage for a Tek scope, so my guess is that this
    has already been divided down. This makes it more important that you
    make a pretty high impedance divider.
    Could you buy 10 MOhm resistors, and then just use your 10 MOhm meter
    as the last resistor in the chain? That would reduce the load on your
    test point by a factor of 10, which would be a good thing.
    That's just the voltage where they are likely to flash over. Since
    you're proposing to work with just 200 V on each one you should be
    fine. It's probably still a good idea to solder these together first
    and then wipe off the resistor bodys with alcohol to remove finger
    oils which can make surface leakage or flashover problems.
    Don't touch any of this while you're doing the measurement. .7mA is
    probably too much for this circuit; .07mA is likely to be fine.
    It's safe only until they flash over. Then it will destroy your meter
    and would probably not kill you but really make you wish you'd used a
    real HV probe.

    Damian Menscher
    --
    -=#| Physics Grad Student & SysAdmin @ U Illinois Urbana-Champaign

    There are probably 50 HV probes within a quarter mile of where you
    work. Ask around.

    -
     
  14. Sorry, it's the 647. And the tubes are 5642s, not 5647s. Apparently
    it doesn't work to remember numbers after it's been 2-3 years.
    This is the "HV test point" (-2.2kV). The anode of the CRT gets the
    higher voltage you're expecting: +11.8kV.
    Radio Shack doesn't have anything larger than 1 MOhm. Hence my idea
    to use so many of them. I suppose I could add more to the chain,
    though. Or I could just get a real HV probe....
    Thanks to you and everyone else for the advice. I'll certainly think
    more carefully about this (and probably just use a real HV probe).

    Damian Menscher
    --
    -=#| Physics Grad Student & SysAdmin @ U Illinois Urbana-Champaign |#=-
    -=#| 488 LLP, 1110 W. Green St, Urbana, IL 61801 Ofc:(217)333-0038 |#=-
    -=#| 4602 Beckman, VMIL/MS, Imaging Technology Group:(217)244-3074 |#=-
    -=#| <> www.uiuc.edu/~menscher/ Fax:(217)333-9819 |#=-
    -=#| The above opinions are not necessarily those of my employers. |#=-
     
  15. NSM

    NSM Guest

    | Thanks to you and everyone else for the advice. I'll certainly think
    | more carefully about this (and probably just use a real HV probe).

    If you ever have any doubts about the power of electricity, watch the
    episode of "The Beverly Hillbillies" where the bear eats granny's love
    letter and Jethro comes up with a plan to X-ray the bear.

    N
     
  16. James Sweet

    James Sweet Guest

    I would hardly think Beverly Hillbillies to be a good source of information
    on electricity safety, anything in a TV show like that is pure fabrication,
    it may coincide with real life but no more likely than what any random
    person would assume to be true.
     
  17. NSM

    NSM Guest

    |

    | > If you ever have any doubts about the power of electricity, watch the
    | > episode of "The Beverly Hillbillies" where the bear eats granny's love
    | > letter and Jethro comes up with a plan to X-ray the bear.

    | I would hardly think Beverly Hillbillies to be a good source of
    information
    | on electricity safety, anything in a TV show like that is pure
    fabrication,
    | it may coincide with real life but no more likely than what any random
    | person would assume to be true.

    After you see Jethro stick his tongue in the electric lamp socket (to show
    what the "dumb old bear" will do) you'll understand why the prospect of the
    electric chair fills him with dread.

    N
     
  18. mike

    mike Guest

    Great analogy to newsgroups.

    Yes, THIS is one of the better groups.
    Some very smart people hang out here.
    Some of them know what they're talking about.
    Some of those have RELEVANT practical experience.
    Some of those can communicate.

    Multiply all those factors to get and indication of the
    signal to noise ratio. Maybe we should call it the
    newsgroup Hillbilly factor.

    If you already know the right answer to
    your question, you can often find it in the thread.
    But if you already know the wrong answer, it's probably
    in there too. Sigh!
    mike
    --
    Return address is VALID.
    Wanted, Slot 1 Motherboard
    500MHz Tek DSOscilloscope TDS540 Make Offer
    http://nm7u.tripod.com/homepage/te.html
    Wanted, 12.1" LCD for Gateway Solo 5300. Samsung LT121SU-121
    Bunch of stuff For Sale and Wanted at the link below.
    http://www.geocities.com/SiliconValley/Monitor/4710/
     
  19. It's one of the best comedies in history though, and I think the
    previous poster was being somewhat tongue-in-cheek also. :)

    Tom
     
  20. Sheesh...it was bad enough when I was stupid and drunk enough to
    tongue-test a 10.6 VDC, 850 mA cellphone adapter...

    Tom
     
Ask a Question
Want to reply to this thread or ask your own question?
You'll need to choose a username for the site, which only take a couple of moments (here). After that, you can post your question and our members will help you out.
Electronics Point Logo
Continue to site
Quote of the day

-