Always Drain Your Filter Caps

Discussion in 'Amps/Cabs Tech Corner: Amplifier, Cab & Speakers' started by JacksonAmpworks, Mar 17, 2005.

  1. JacksonAmpworks

    JacksonAmpworks Member

    Messages:
    898
    Joined:
    Jan 29, 2003
    Location:
    Fort Worth, TX
    Just a quick reminder to ALWAYS drain your filter caps before your work on your amp. I have no SPECIAL reason to say this, just thought it would be a great reminder for everyone.

    On an completely unrelated note, I'm going to go change my underwear.
     
  2. onemind

    onemind Member

    Messages:
    3,588
    Joined:
    Oct 28, 2004
    Location:
    Oceanside, NY
    Make sure you drain everything before you change it...could be dangerous. :p
     
  3. trisonic

    trisonic Member

    Messages:
    13,163
    Joined:
    Mar 24, 2003
    Location:
    NJ out of England
    Brad,
    You may find camouflage underwear a help. Especially for us old boys........

    Best, Pete.
     
  4. LarryN

    LarryN Member

    Messages:
    959
    Joined:
    Jan 1, 2004
    Location:
    Santa Rosa, Ca.
    Is it OK to use your speaker for a cap drain by leaving the standby up?
     
  5. Wakarusa

    Wakarusa Member

    Messages:
    1,477
    Joined:
    May 7, 2004
    Location:
    Indianapolis, IN
    Look at the bright side. You're still alive to worry about it :AOK
     
  6. John Phillips

    John Phillips Member

    Messages:
    13,080
    Joined:
    Mar 17, 2002
    Location:
    Scotland
    No. The speaker is completely out of the high-voltage circuit since it is connected only via the OT.

    There are several ways to drain the caps, depending on what you know about the amp circuit.

    In many amps, especially higher powered ones, there are resistors across the first filter cap stage, because there are two caps 'stacked' in series. If you know for certain that the amp has these, all you need to do is to turn the power off while leaving the standby on*, and wait a couple of minutes. All the caps will discharge through the resistors.

    Some amps don't have these, but do have tube rectifiers (eg Fender Deluxe Reverb). Here, all you have to do is pull the rectifier with the amp on (it's actually better to turn off, let it cool, pull it, and then turn the amp back on) - leave the standby on* too. With the other tubes powered up, the caps will drain through them in a couple of minutes and can't recharge due to there being no rectifier.

    Some amps with solid-state rectifiers have no caps on the supply side of the standby switch (eg 50W Marshalls) - if so, just turning off* the standby while leaving the power on for a few minutes will drain the caps.


    If the amp isn't one of those types, or you aren't sure, it's best to be safe. You need to deliberately drain them. The way I do it is to connect an old analog voltmeter (fairly low resistance) across the main filter cap, and check the voltage. Turn the power off, and watch the voltage drop - it will discharge the caps through the resistance of the meter. You can click down through the ranges as it gets low, which increases the rate.

    If you don't have one of those, you either need to make a discharger - a 100K 5W resistor with wires and clips (well insulated) on both ends would do fine - or find the plate connection in one of the early preamp stages and connect that to ground - the caps will discharge through the B+ chain.

    DO NOT be tempted to short the caps to ground directly with a screwdriver or anything - you may damage the caps.


    *I mean "standby on" = play, "standby off" = mute, here. This can be slightly confusing, with "putting the amp on standby" meaning switching it off.


    Be careful. I've had a couple of nasty shocks from caps too...
     
  7. WailinGuy

    WailinGuy Member

    Messages:
    1,267
    Joined:
    Aug 30, 2003
    Location:
    Silver Spring
    I like this method; it's simple and effective. I use a wire with an alligator clip on each end. It's a good idea to connect the ground side first, before connecting the other end of the wire to the plate side (connecting to pin1 or pin 6 of tube socket) of the preamp plate resistor. And when disconnecting the wire, you should remove the side connected to the plate resistor first. Also, I recommend using a piece of solid core wire with bright color insulation which is long enough to stick well above the chassis opening while connected. This way, you're less likely to forget to remove the wire before powering the amp back on (which could cause the plate resistor and p/s dropping resistors to fry).
     
  8. Wakarusa

    Wakarusa Member

    Messages:
    1,477
    Joined:
    May 7, 2004
    Location:
    Indianapolis, IN
    Favorite tool here is a home brew 1K/3.75W resistor soldered into a probe. (barrel pulled back so you can see the resistor)


    [​IMG]
     
  9. JacksonAmpworks

    JacksonAmpworks Member

    Messages:
    898
    Joined:
    Jan 29, 2003
    Location:
    Fort Worth, TX
    Or if your lazy, just use a big screwdriver. POW!!! :) I'M KIDDING!!!!!
     
  10. Swarty

    Swarty Member

    Messages:
    1,143
    Joined:
    Dec 16, 2003
    Location:
    Michigan
    Wait a second, that is my method! ...the standby switch on a Fender and the HT Fuse holder on a Marshall. The POW from a Marshall Major and some older Mesa's is really spectacular... My screwdrivers have arc marks all over them.
     
  11. JacksonAmpworks

    JacksonAmpworks Member

    Messages:
    898
    Joined:
    Jan 29, 2003
    Location:
    Fort Worth, TX
    I have to admit that I have done it before. It's certainly VERY poor practice but I have done it. If you have a really good charge there, it's always funny how the screw driver sticks to the cap and you have to pry it off. :)
     
  12. Wakarusa

    Wakarusa Member

    Messages:
    1,477
    Joined:
    May 7, 2004
    Location:
    Indianapolis, IN
    You guys are scary :eek:

    'Course I seem to recall many moons ago a guy who looked just like me (only thinner and with more hair) discharging 480V/100HP motor controller caps with a big-ass screwdriver. Do it right and you get big enough chunks of screwdriver to make cleanup a breeze. :D
     
  13. Big Dan

    Big Dan Member

    Messages:
    197
    Joined:
    Dec 29, 2004
    I am just getting ready to start my first amp project and I've been very nervous about this. I have made a jumper wire with a resistor built in it. I used a 2 or 3 watt (can't remember which) metal resistor in the 10 ohm range. I saw that value on another forum or a book or something. Anyway, is that value way off? Will I get into trouble using that resistor to drain the caps?

    Thanks,
    Dan
     
  14. Big Dan

    Big Dan Member

    Messages:
    197
    Joined:
    Dec 29, 2004
    Hey Todd, how did you make that probe? Better yet, can you sell me one just like it?

    Dan
     
  15. Wakarusa

    Wakarusa Member

    Messages:
    1,477
    Joined:
    May 7, 2004
    Location:
    Indianapolis, IN
    The real trick is to find a probe body that's big enough and a resistor that's small enough so the whole mess fits together. After that, it's just a soldering job.

    To keep everything compact, the probe tip will usually have a solder cup. You trim the one resistor lead short, heat the cup (SMD tweezers work surprisingly well for this trick, but a vice and a regular iron will suffice), fill it with solder and plunge the resistor lead into the cup. Hold it all very still until things cool. Then you just solder a hunk of wire to the other resistor lead Western Union style. Alligator clip on the far end (after you slide the barrel on the wire ;))

    About the resistor choice. I actually prefer a 10K for big filter caps. I would not use a 10 ohm. The difference is in how fast the cap discharges -- too fast and you blow the resistor, the cap, or both. Without going too deeply into it, the formula you're after is basically: R * C * 5 = T where R is the resistor in ohms, C is the capacitor in Farads, and T is the discharge time in seconds.

    So a 70uF cap at 450V will take about 7/10 of a second to discharge thru a 10K, 70mS thru a 1K and 0.7mS through your 10. The upshot is that even with a 10K if you count "one Mississippi" the cap is discharged to safe levels.

    Playing a tad fast and loose here, figure too that during the discharge the average current through the probe is 5mA with the 10K, 50mA with the 1K and 5A with the 10 (with an instantaneous value as high as 16A at initial discharge and initial power dissipation of almost 3KW :)).

    So, to answer questions specifically:
    1) Yes, you could get in trouble with your 10 ohm probe.
    2) They're easy to make. Mine's kinda fancy, but you can just splice a resistor into the lead wire anywhere and cover the whole mess with shrink wrap or electrical tape. Use a 1K/2W or better resistor minimum for tube amp work.
    3) I can make one for you, but don't remember where I found the cheesey probe. If you'd really like one shoot me an e-mail.
     
  16. Swarty

    Swarty Member

    Messages:
    1,143
    Joined:
    Dec 16, 2003
    Location:
    Michigan
    Please explain why this is very poor practice. BTW, I've never had to pry the screwdriver off a cap.
     
  17. Wakarusa

    Wakarusa Member

    Messages:
    1,477
    Joined:
    May 7, 2004
    Location:
    Indianapolis, IN
    As for prying off screwdrivers, you can get high enough pulse currents on discharge to have micro welding between the cap and discharge tool. FWIW, I've never seen this anywhere but in big high voltage power work. Certainly never in an amp (doesn't mean can't happen, just means I haven't seen it). Could be he just has sticky stuff on his screwdrivers too.


    The poor practice has to do with the capacitor's ability to withstand current pulses. This will appear as a dV/dt rating on the spec sheet. It is not always listed. A high enough pulse current will destroy the internal connection from the capacitor plates to the external lead. Using a screwdriver (the electrical equivalent of wire) allows for much higher pulse currents than a probe with a resistor.

    That's the theory anyhow. YMMV.
     
  18. Swarty

    Swarty Member

    Messages:
    1,143
    Joined:
    Dec 16, 2003
    Location:
    Michigan
    Well, 100s of amps an no capacitor casualties yet... Not sure I understand how the rapid discharge could be detrimental to the internal structure. It is just a bunch of stored electricity waiting to be released, which is the function of the device. But then again, I'm the first to admit that I'm not the most technical of techs. (I'll take tone over math every time.)

    Most amps are somewhat disappointing in the pop, as they have bleeder resistors, but some of the old ones as I mentioned, particullary Marshall Majors and some older MKI and MKII Mesas have some seriously large caps (200uF+) with apparently no bleeders....
     
  19. Wakarusa

    Wakarusa Member

    Messages:
    1,477
    Joined:
    May 7, 2004
    Location:
    Indianapolis, IN
    The damage to the internal structure is due to the high initial discharge current through the joint where the internal plate structure connects to the external lead. Just like any fuse, too much too fast and things melt. As you've seen in practice, it may be pretty unlikely for this to actually happen in a tube amp's filter caps.

    Your "pop" comment reminds me of other issues, though. The pop, of course, is an arc between the cap lead and the screwdriver (or ground and the screwdriver, depending on what touches what first). The arc leaves behind pitting and carbon. Neither will destroy the amp, but neither is good for it.
     
  20. saros141

    saros141 Guest

    Hey Todd, thanks for the math behind the resistor. I made one up awhile back with a 15k 10W Sprague Koolohm, some 16 AWG wire and two alligators. I left it clipped between the + terminal and ground for 2 minutes, for each cap in the power supply. I guess that was overdoing it a bit, on a 5 watt amp. Good to know I can get to work a bit quicker next time.
     

Share This Page