SVT CL Tube Matching and Operating Temperatures

Discussion in 'Amps and Cabs [BG]' started by melodiousmartin, Feb 5, 2020.

  1. melodiousmartin


    Feb 5, 2020
    Hey Bass Peeps, I'm new to the site and this is my first post. I joined up because I've been reading many threads here and have found a lot of value in them.

    I have recently embarked on my ab initio journey into home amplifier repair. So far I diagnosed and fixed faulty solder joints on my JC120, a blown T1A fuse on a friend's Laney GH100L, and a fussy pre-amp tube on a Vox 50. My next project is my SVT-CL which I have owned since around 2009 and was manufactured in January 2001. I watched an excellent video on youTube regarding cleaning and inspecting the SVT and am ready to take her apart.

    My prior experience in electronics is mostly with smaller equipment such as general repair ranging from cars to laptops, Arduino maker projects, and small devices like my oscilloscope. I appreciate that life threatening voltages are present and understand how to bleed down capacitors safely and wear insulated gloves when needed.

    My Ampeg has always been a but grumpy when it comes to scratchy glitches especially when you first start to play despite warming it up. The crackles are worst on certain notes (mostly low e string). I took it for an assessment from a respected tube amp tech and he said it is a pretty good amp "for a St. Louis Music example" putting out 280 watts on his bench. The amp has historically been easy to adjust the bias on and I believe the Sovtek 6550's are original. I have never changed them. I have observed bad behavior on the bias lights seems to correlate with crackling trouble. Lately the biasing became a bit more difficult to achieve and I started to get motivated to address the problem with action.

    I measured the 6550 operating temperatures with an infrared thermometer and mapped their positions viewed from the rear of the cabinet.

    270 235 273 Front

    340 300 240 Rear

    One thing that is immediately apparent to me is the drastic difference in temperatures with the coldest tube at 235 and the hottest at 340 degrees f. I am tempted to conclude that either the set was never balanced very well, has had individual tubes changed at some time in the past, or there is a significant degradation in some of the tubes. Does anyone have a benchmark for the operating temperature for a 6550 with the volume and gain turned all the way up? The Laney quartet of EL34's all run a consistent 240 degrees f.

    I'm guessing the all the tubes in the amp are twenty years old by now. My general opinion of the sound in the years that I have owned the amp is that it has been a little muddy compared to all of my other amps. Perhaps I should have changed the tubes out a long time ago? I generally run the amp not much more than half way up on gain and volume. The gain when it's all the way up sound a bit hissy to me. I haven't done much research on getting a clean signal through the preamp section yet.

    The first order of business is a teardown, general inspection and cleaning, repair of any obvious faults, and re-testing.

    Thanks for reading, Martin
  2. agedhorse

    agedhorse Supporting Member Commercial User

    Feb 12, 2006
    Davis, CA (USA)
    Development Engineer-Mesa Boogie, Development Engineer-Genzler (pedals), Product Support-Genz Benz
    Envelope temperature will vary by physical location in the amp as well as quiescent current. They can vary quite a bit.

    The only way to accurately access tube balance is to measure either the plate current (dangerous) or cathode + screen current.
  3. Wasnex


    Dec 25, 2011
    I wouldn't be surprised if 20 year old output tubes are tired and starting to get noisy.

    The bias circuit indirectly measures cathode current. It basically samples the voltage drop across a resistor that is connected between the cathode and ground.

    When a tube makes noise, it changes the amount of cathode current, so if one of the tubes is on the green or red threshold, the noise can cause the LEDs to flicker.

    Here's a graphic I found that supposedly explains how the bias circuit works (I think it's accurate).
    As you can see the bias circuit doesn't require the tubes to be matched very well.

    I would clean out the amp and try a new set of tubes before I totally tore the amp down. It's fairly common to replace the two 12AU7 driver tubes when you replace the output tubes.

    If you decide to tear the amp down, make sure you observe and document the polarity of the connectors. There is a recent thread where it appears the connectors were installed backwards and this caused additional damage.

    I'll attach a SVT CL disassembly guide. It may or may not be helpful depending upon which revision of the amp you have.

    One thing to consider doing is inspecting the tubes sockets. Over time they can loosen up so the tubes are not making good contact. This can allow arcing that causes carbon deposits that can be a source of noise and other problems. Cleaning and retensioning the tubes sockets may be all you need, to get the amp up and running good as new. Discharge the capacitors before hand to ensure there is no voltage on the tube sockets.

    Good luck!

    Attached Files:

  4. agedhorse

    agedhorse Supporting Member Commercial User

    Feb 12, 2006
    Davis, CA (USA)
    Development Engineer-Mesa Boogie, Development Engineer-Genzler (pedals), Product Support-Genz Benz
    I would first verify if there really is a problem before changing anything.
  5. JimmyM

    JimmyM Supporting Member

    Apr 11, 2005
    Apopka, FL
    Endorsing: Yamaha, Ampeg, Line 6, EMG
    Well far be it from me to question you about all matters tech, but the signs do all seem to point to wonky tubes, no? Not biasing, bias lights flashing when it crackles, muddy tone otherwise...and honestly, after 19 years, I'd probably want to change out a batch of factory Sovteks solely on the basis that they're 19 years old and spent 8 years with someone else doing God knows what to them :D

    You know me...I'm very much a "don't fix it if it ain't broke" guy. But it's broke and it's probably time to consider new tubes all around regardless if there's anything else wrong with it. You could probably go through the pre and driver tubes and save a couple, but I would never trust those power tubes anymore at the very least.

    DISCLAIMER: Agedhorse is a world class engineer who designs amps for Mesa and I can barely operate a soldering iron.
    Microbass likes this.
  6. Wasnex


    Dec 25, 2011

    I think Andy is saying test the tubes before trashing them. Your probably right, but the problem could be loose or dirty tube sockets, or any number of other issues.
    JimmyM likes this.
  7. beans-on-toast


    Aug 7, 2008
    The push and pull trios combine to act as a single tube on each side. If the tubes are not perfectly matched, some will be working harder than others, drawing more cathode current. To minimize distortion, you want the currents of each trio to be about the same. The balance control of the SVT-VR allows the push and pull side to be matched by adjusting a pot, the SVT-CL does not have this feature.

    The push and pull trios are not grouped front and back but as in the image below. V1, V2, V3 are in one group, V4, V5, and V6 in the other group. Note the globby soldering in this example, I wouldn't stand for that. Although the bottle temperature can be corelated to current draw, as was mentioned, the temperature can vary due to other factors. For instance, position relative to the fan, airflow around the tubes is not evenly distributed. I'm not aware of any documentation relating current flow to bottle temperature for this amp. The temperature should be used as a general guide to see what is going on. You really are better measuring the actual cathode current rather than inferring. By the way, the GE data sheet lists the maximum bulb temperature at the hottest spot on the glass envelope at 250 degrees centigrade. I assume that this is at maximum plate dissipation.

    With a 2001 manufactured SVT-CL amp, you should check to see if it has the ground loop issue: Ampeg SVT-VR&CL ground loop hum solution

    As for the muddy sound, there are many possibilities. The speaker cabinet plays a roll. The particle board ones are characterized as sounding muddy. The amp sounds more open when the master is turned up full and the preamp volume used to set the level. Within the amp, check the easy stuff first: check the small signal tubes --a dull sound can be a tube, inspect the tube sockets and tube pins for oxidation, carefully clean with deoxit. Clean the sliding tone switches with deoxit. More time consuming: check the treble and bass tone stage capacitors and resistors to ensure if they are within spec.

    Note the metal hex standoffs under the power tube socket mounting tabs. They were not in early models and were retrofitted as recommended in a tech note. Without them the tube socket solder joints were cracking due to stress. Could lead to problems. Check the solder joints and reflow them if necessary. All later revisions had these standoffs in place.

    SVT tube board.jpg
    Last edited: Feb 7, 2020
  8. agedhorse

    agedhorse Supporting Member Commercial User

    Feb 12, 2006
    Davis, CA (USA)
    Development Engineer-Mesa Boogie, Development Engineer-Genzler (pedals), Product Support-Genz Benz
    Correct, replace them IF they need to be replaced, but if they are good there's no reason not to keep them. In fact, they have proven reliable.
    JimmyM and Wasnex like this.
  9. melodiousmartin


    Feb 5, 2020
    Thanks for the outpouring of awesome information! I'm really glad I joined the forum. I will report some progress soon. I did make some inquiries with Ampeg customer support to get their interpretation of my questions, which I will share in this thread if appropriate.

    Regarding the arrangement of the 6550 tubes, I did my best to group them in equal trios based on the temperatures that they run at. The temperatures I have recorded are consistent for for the individual tubes regardless of which socket is used so far.

    Thanks for the safety information, I am adept with the multimeter and will bleed the big capacitors down with a large resister and some rubber gloves for additional safety. I downloaded a schematic for this model SVT and the Ampeg user guide.
    beans-on-toast likes this.
  10. Wasnex


    Dec 25, 2011
    If you find the schematic labeled "PWR Amp Schematic" you can see the bias sense circuits for both trios of output tubes. The cathode of each 6550 is tied to ground through a 10 ohm bias sense resistor. (R35,R36,R37) is for one trio of tubes and (R38,R39,R40) is for the other trio of tubes. You can trace the connections back through J11 and J7 to the "Tube Bd Schematic" if you like.

    If you gain access to the resistors you can measure the voltage across them and derive the current each tube is pulling using Ohms Law E=IxR. You will know the voltage and resistance so solve for I. I=E/R. Or just remember that 10mV = 1mA of current.

    Target current for each tube should be around 24mA. 240mV = 24mA.

    Once you know the current of the individual tubes you can use logic to try sort them in a different order to get a better current match. They don't always behave the way you expect them to.

    Keep in mind you have two trios of tubes and each trio has it's own bias control. Probably he best way to sort the tubes would be to adjust adjust the bias controls so each trio of tubes has the same bias voltage on the control grid and is operating within a safe current range. In order to do this you will need access to the tubes sockets so you can measure the voltage on the control grids. Careful as the plates have almost 700V on them.

    It's a hassled to get to the test points so I have a test jig called a bias probe. Basically an additional tube socket is inserted between the tube and the amp's tube socket. All of the connections are pin to pin except the cathode. The connection between the cathode pins is broken. There are two different options here. Some bias probes insert and ammeter inline with the cathode pins and measure the cathode current directly. Other bias probes insert a precision 1 ohm bias sense resistor and you use a mV meter to measure the voltage drop across the resistor. 1mV corresponds to 1mA of current.

    Some words of caution if you decide to buy a bias probe. Some of the highest rated probes cannot tolerate the working voltage of an SVT. So do your research before buying. Also a bias probe expose you to shock hazard as the cathode voltage can potentially float up to the plate voltage if the probe malfunctions.
  11. agedhorse

    agedhorse Supporting Member Commercial User

    Feb 12, 2006
    Davis, CA (USA)
    Development Engineer-Mesa Boogie, Development Engineer-Genzler (pedals), Product Support-Genz Benz
    Before jumping into a repair, I suggest that you learn how the amp is supposed to work, and how each circuit stage is supposed to work first.
    Wasnex likes this.