Asher Guitars WD Music Products Amplified Parts Mod Kits DIY Nordstarnd Pickups Warmoth.com
Asher Guitars WD Music Products Amplified Parts Mod Kits DIY Nordstarnd Pickups Warmoth.com
Asher Guitars WD Music Products Amplified Parts Mod Kits DIY Nordstarnd Pickups Warmoth.com

Valco clone

Discussion in 'Shock Brother's DIY Amps' started by Prophetsnake, Sep 17, 2017.

  1. D'tar

    D'tar Tele-Holic

    946
    Jan 11, 2013
    WNY
    Cathode to ground should be ~= supp grid to ground, 4k7r

    Screen grid (p6), to ground at 276r is no good, this should be approx 122kr, confirm that one
     
    Prophetsnake likes this.

  2. tubeswell

    tubeswell Friend of Leo's

    Jul 1, 2008
    NZ
    I think I see the problem. The original schematic uses a 6J7, and you're using a 6SJ7. I should have clicked about this earlier. You should change the screen and plate supply to something that works better with a 6SJ7. Ditch the 330k, 100k and 22k resistors and install a 250k plate resistor, a 2M2 screen resistor and a 0.05uF screen bypass cap (as per the example in the Fender 5C1 schematic). Keep the 4k7||20uF at the cathode for now (and make sure you have the socket wired for the right pin-out for the 6SJ7)
     

  3. Prophetsnake

    Prophetsnake Tele-Meister

    134
    Aug 2, 2016
    ireland
    Sure. But the socket is wired for a 6SJ7, and comparing the two data sheets, I can't see a world of difference between them aside from the pin arrangement. Where's the power going? The pin that feeds the plate seems to be getting healthy voltage with the tube out, and even when the tube is in place the power supply is good all the way the resistor, so where is it going afterwards? Can one of the grids suck that much out of the plate?
    This is the data sheet I used.
    http://www.r-type.org/pdfs/6sj7.pdf
     

  4. Prophetsnake

    Prophetsnake Tele-Meister

    134
    Aug 2, 2016
    ireland
    Right. I did the modification to basically turn the pre-amp circuit into the 5C1.
    LOUD!

    So, that issue is sorted. Thanks Tubeswell!

    Now, the voltage at the supply end has lept up considerably - 273. ? I take it that the previous setup was simply leaking a lot of that back to earth? The voltage at the plate is 178 now, which I suspect might be a bit high for the 6SJ7, butI suppose I can tweak that at the supply end with a slightly higher value resistor between the supply nodes.

    New problem, it's pretty hissy and there's a bit of hum as well. I'm aware that the 6SJ7 isn't reputed to be the quietest of tubes, so I'll explore that and post more detailed voltages, etc later on when I'm not so tired.

    Now, can anyone tell me why the plate voltage soared like this? I'm at a loss here. I presume all that voltage drained away someplace. Where?
     
    Last edited: Nov 10, 2017
    tubeswell likes this.

  5. tubeswell

    tubeswell Friend of Leo's

    Jul 1, 2008
    NZ
    What are your overall VDC High Tension supply node, plate, screen and cathode voltages throughout now?
     

  6. Prophetsnake

    Prophetsnake Tele-Meister

    134
    Aug 2, 2016
    ireland
    The nodes are:
    355
    325
    274

    The plates
    V1 - 6SJ7 171
    V2 6SN7 plate#1 155
    V2 Plate #2 160
    V3 6v6 346
    V4 6v6 347

    Cathodes
    V1 2.85 Volts
    V2 #1 7.3
    V2 #2 7.3
    V3 18.3
    V4 18.3


    Screens

    V1 - 0
    V2 #1 0
    V2 #2 29 mv
    V3 355mv
    V4 75 mv


    Seems a bit odd to have the second stage at lower voltages than the first, but I have to say, the thing sounds great. Hissing aside, it's the equal of my DeLuxe which, until today was the best sounding amp I'd ever played through.

    By the way, the hiss is present whether the instrument is plugged in or not, and gets louder with the volume turned up. Haven't bumped into anything that makes any difference with a chopstick yet.

    Upate:

    Tried substituting the old 'suspect' 6SJ7 with the broken centre pin tube I originally thought was defective, and the hiss went away completely. There is very little hum or noise of any sort at anything up to about 60% of the volume travel, and then it squeals like mad. The other tube only hissed continuously all the way up and down the volume, no squeals, so I'm now thinking I have two duff tubes.
     
    Last edited: Nov 10, 2017

  7. tubeswell

    tubeswell Friend of Leo's

    Jul 1, 2008
    NZ
    The screen voltages all look wrong. Can you re-check those? (Pin 4 on a 6V6, and Pin 6 on a 6SJ7)

    Re: the 6SN7 plate voltages, each triode is running a 100k load resistor, which is probably why there isn't much plate current* or voltage. If you experimented with decreasing the plate resistors to (say) 47k each, you would find those voltages changing

    *160V-7.3 = 152.7. (7.3V/3300R)/2 = 1.1mA per triode. 1.1mA x 152.7V=0.17W. Each 6SN7 plate is rated for 5W maximum (when the triodes aren't run in parallel), so they can easily take a bit more heat.

    Hiss can be from a number of sources, including bad tubes (as you found out), or load resistances that are too high. Lowering the 6SN7 triode plate resistors to 47k each could also help with this. But also, having metal film resistors throughout the amp on all resistances that are in the signal path** can help. Furthermore, using resistors that are otherwise rated for higher power dissipation can also help. (e.g. 1 or 2W resistors instead of 1/4 or 1/2W resistors)

    ** plate resistors, grid leak resistors, input resistors, grid stopper resistors

    The hum could simply be a microphonic tube (e.g. most likely the 6SJ7, since this will have relatively high gain). You can 'shock-mount' the 6SJ7 socket - with a rubber o-ring between the socket and chassis, and with nylon grommets for the bolt holes - to dampen excess mechanical vibration, and/or you can slip a couple of o-rings over the 6SJ7 envelope.
     

  8. Prophetsnake

    Prophetsnake Tele-Meister

    134
    Aug 2, 2016
    ireland
    Sorry, I gave the voltage for the wrong grid.

    The 6SJ7 is 52V and the 6V6 are both 327V

    There's no hum at all. One tube produces a hiss that increases with volume, which I assume means it must be coming from the pre-amp. The other tube doesn't hiss or hum at anything less than about 60% -70% of volume travel, above which it produces a squeal. That squeal appears a bit earlier in one tube than the other.

    I tried cleaning the pins and socket, but that made no difference. I'll try shock mounting as well and the rings. I'm not sure the silicone ones I have will fit the slimmer tube, though. I'll see if I can rig up something as an experiment though. See i it makes a difference.
     

  9. Prophetsnake

    Prophetsnake Tele-Meister

    134
    Aug 2, 2016
    ireland
    Oh, and all the resistors are metal foil throughout, excepting some power resistors and that one 2M I replaced, which is carbon. I went beefy on everything. There isn't a single 1/4 watt in there, and wherever I was in doubt, I went large.
     
    tubeswell likes this.

  10. tubeswell

    tubeswell Friend of Leo's

    Jul 1, 2008
    NZ
    That looks better
     

  11. tubeswell

    tubeswell Friend of Leo's

    Jul 1, 2008
    NZ
    The 6SJ7 is biased quite cold at 2.78V. (In the 5C1 it uses grid leak bias, which can only get to about 1V). You could probably reduce the bias voltage a bit. I'd start by experimenting with decreasing the cathode resistor to something like 2k2 or 1k8 to see where that lands (and then remeasure all the voltages again). Reducing bias like this make the tube run a bit hotter, which will also lower the plate voltage a bit.
     

  12. Prophetsnake

    Prophetsnake Tele-Meister

    134
    Aug 2, 2016
    ireland
    In the 5C1, the cathode runs straight to earth. At the moment I still have the 4k7 (actually it's a 5k, since that's what Dave sent to me) so would it be an idea to try bypassing it with a clip and see what happens?
     

  13. Prophetsnake

    Prophetsnake Tele-Meister

    134
    Aug 2, 2016
    ireland
    Kay, I tried bypassing the 6SJ7 cathode, and it basically just sucked all of the sweetness out of the tone.
    Then, I replaced the 100k resistors on the 6SN7 with 47 K and that did, indeed, cut out a bit of the squealing. I spent some time poking everything inside of the chassis and on the control panel with a chopstick and found nothing wrong, and then I bumped one of the leads to the volume control and got a bit of a squeal. The way I have the chassis twisted so that I can get inside while it's powered up had the volume leads too close to one of the 6V6 bases. I'd say that's my problem and that should go away when I shift everything to where it's going to live inside the cabinet. However, the plan was always to make those leads as well as the input shielded, so I'll do that before I start arranging everything. Would it be better to earth those at the chassis, or on the control panel - to the pot and input jack? The latter would be a lot easier and probably safer as the chassis has become a bit crowded.

    Oh, and the swop to those 47k resistors has made the thing sound incredible. I won't describe it now, but when I get it put all together I'll record a snippet and post it and you can tell me!
     
    tubeswell likes this.

  14. tubeswell

    tubeswell Friend of Leo's

    Jul 1, 2008
    NZ
    No that's not what I suggested. In the 5C1, the 6SJ7 is 'grid-leak biased'. That type of biasing requires the cathode to be grounded, and the grid employs a fairly highish (5M) grid leak resistance to create the bias voltage. However, the maximum voltage you can achieve using the grid-leak bias method is about 1V.

    Cathode biasing involves using a cathode resistor in series with the cathode and the ground to create the bias voltage - it enables much better control over the bias voltage, and it provides more consistent biasing as the tube ages. The valco amp has the input tube cathode-biased with a 4k7 resistor. However, with the 6SJ7, this has resulted in the 4.7V bias voltage you currently have, which leaves the 6SJ7 biased very cold. What I suggested was swapping that 4k7 (~5k) for 1k8 to 2k2 to bring the bias voltage a bit closer to centre-bias, which will enable the tube to conduct harder. This should improve the tone of the amp and enable to you get more unclipped volume with less hiss.

    No, I didn't mean bypassing the cathode resistor with a jumper. That will make it run more poorly. See the above response.

    Yes that's because a 47k load brings the 6SN7 triode closer to its ideal load line. 33k or 22k will make it run even sweeter, but you'll probably need 3W rated plate resistors to do that, and you'd also probably want to change the cathode resistor to get the bias more centre-biased. A 6SN7 is the 'octal equivalent' of a 12AU7. It will sound pretty sterile with a 100k load. If you're happy with the 47k load resistors then just leave it that way
     
    Last edited: Nov 11, 2017

  15. Prophetsnake

    Prophetsnake Tele-Meister

    134
    Aug 2, 2016
    ireland
    Yes, thanks. that all makes sense. I wasn't saying that you suggested that I bypass the cathode resistor. It was just something that occurred to me to try. I decided it couldn't do any real harm so I did, and just as you say, it sounded quite sterile. I think I understand what you mean about grid leak as opposed to cathode biasing, but I have to say, I'm still at a complete loss as to where all that voltage at the 6SJ7's plate was going. Simply leaking straight to Earth via the 22k resistor, and the rest going to the screen?


    I installed shielded leads to the volume and input and re-routed them well away from tubes. The squeal and most of the other weird noises are gone. I have some humm remaining, too loud to live with. But I should be able to kill that simply by tidying up.


    Thanks so much for the help!
     
    tubeswell likes this.

  16. teleman1

    teleman1 TDPRI Member

    90
    May 16, 2003
    Arizona
    How are they affected? You hear it and/or is it putting extra ware on the tube not having a shield?
     

  17. Old Tele man

    Old Tele man Tele-Afflicted

    May 10, 2017
    Tucson, AZ
    With preamp tubes, it's a form of magnetic-field feedback, with the moving magnetic field modulating the electron-stream that flows from the cathode to the plate. When the magnetic field is 'in-phase' with the electron stream, the electrons are deflected 90-degrees towards a singular spot on the plate, creating a hot-spot that can eventually melt a hole into/thru the plate. When the magnetic filed is 'out-of-phase' with the electron stream, the electrons are deflected 90-degrees in the opposite direction causing the same problem, but on the opposite side of the plate.

    With power tubes, it depends upon whether they're beam tetrode or true pentode.
     
    Prophetsnake likes this.

IMPORTANT: Treat everyone here with respect, no matter how difficult!
No sex, drug, political, religion or hate discussion permitted here.