Princeton Reverb - Choosing a Grounding Scheme

Discussion in 'Shock Brother's DIY Amps' started by Lemon, Jun 19, 2021.

  1. Lemon

    Lemon TDPRI Member

    Posts:
    42
    Joined:
    Feb 10, 2018
    Location:
    Pale Blue Dot
    Hi everyone,

    I'm starting my first tube amp build - a Mojotone kit Princeton Reverb, and would like to decide on a grounding scheme before populating the chassis.

    I've already drilled a hole for the bias pot, next to the bias board, and will also drill a hole for a dedicated safety ground / earth bond - probably for an M4 or M5 size bolt. Preamp ground bus will terminate at the input jack.

    What I'm not really sure about is which way to go regarding the other ground points - the CT, 6v6s 8 pins, bias board, bias resistor, can cap, etc. as there are multiple options.

    Are some of them objectively better than the others or is it simply a matter of preference? What are some considerations to be aware of?

    This is the Mojotone version: [​IMG]
    Pins 8 are grounded to the sockets' fastening bolts, 6.3V CT connects to the bias board, the rest to the can cap's bolts.


    Next is StewMac's approach: [​IMG]
    Safety ground is connected to one PT bolt, everything else to a second one.




    RobRob's version:


    [​IMG]



    and some more options involving the bias pot, based on RobRob's layout:


    V1[​IMG]

    V2[​IMG]

    V3[​IMG]



    I realize that I might be over complicating things here and that all options will probably work, but as a complete newbie, there are things that I know, things that I know that I don't know, and most importantly (or dangerously) - things that I don't even know that I don't know, so it would be foolish of me not to consult.

    Any comment is appreciated.
     
  2. 2L man

    2L man Tele-Holic

    Age:
    61
    Posts:
    582
    Joined:
    Nov 23, 2020
    Location:
    Finland
    I recommend you don't solder anything to potentiometer cases!!! They operate as a Shield and pot case get electrical connection with shassis when it is screwed to chassis which electrical role is Shield as well so it is not good practise to make any current flow thru chassis of potentiometer cases.

    If your board is not designed for solid ground bus I recommend a Star grounding. There B+ and B1 has one wire to Power Supply filter capacitor which supply OPT and power tube Screens and its because Anode and Screen current flow thru Cathode(s) and 1 ohm bias measure resistors where ground wire starts.

    For B2, B3, B4, B5...etc. can be consideress as "sub power supplys" and on Star grounding each has own ground wires to PS electrolyt and this wire leaves from each stage filter capacitor negative where all stage current is first collected.

    Ground is brought to Earth potential connecting input stage ground to chassis using non insulated steel jack or using a screw mounted lun next to input. This is best place to null potential error because amp input is most sensitive stage.

    Amp is made electrically safe connecting Mains cable Earth wire next to its enter to Chassis but this earth wire nulls potential to Earth as well
     
    Paul G. likes this.
  3. King Fan

    King Fan Poster Extraordinaire Ad Free Member

    Posts:
    6,654
    Joined:
    Jan 1, 2013
    Location:
    Salt Lake City
    You’re right: This amp is complex enough; don’t complicate the simple stuff. Where's my standard text block? “I like Mojo kits a lot, but when Mojo or StewMac differ from Rob's layout, go with Rob.” Rob’s ground scheme is just better at every point including the household safety ground. And it’s also much easier.

    2L man is right about the pots; a big reason I go with Rob. He and I sometimes differ on the need for complex grounds; I know I'd have found star grounding confusing back when I built my PR. But luckily, on a Mojo kit I think you can easily run a bus bar or wire where Rob shows his ground bus. There are enough ground connections that a bus bar (or stripped Romex core, or stripped 18ga solid wire) can basically float right in front of the board. My Rob-style bus actually floats solidly in space, but closer to the board would work well too.

    0314F74A-7EF1-490B-B21F-9FA99C8B11EC.JPG

    As far as the bias pot, your first Rob version would work.

    If you want a *harmless* option to puzzle over, you can try this anchor for the red/yel HV center tap. :) Rob tells me he's not sure it'd quiet the large ripple return current from the first filter cap, but he agrees it might.

    IMG_0738.jpeg
     
    theprofessor likes this.
  4. Phrygian77

    Phrygian77 Friend of Leo's Gold Supporter

    Posts:
    3,471
    Joined:
    Apr 30, 2016
    Location:
    Crawfordville, FL
    I've run the bias ground, and power tube cathodes, all the way to the power supply chassis ground point, but I don't think it really makes any difference at all. The most important thing is that the main PT secondary center tap and the can cap connect to the chassis at the same point. And, unless you ditch the can, you can't make the one grounding change that really will make a difference.
     
    King Fan likes this.
  5. 2L man

    2L man Tele-Holic

    Age:
    61
    Posts:
    582
    Joined:
    Nov 23, 2020
    Location:
    Finland
    This is excellent Schematic by Merlin Blencowe where Star Grounding is drawn very clear and it is how trained electric technicians read all schematics even when they are not as good as this is so it is very beneficial to study it if grounding principle is not clear yet.

    Sub power supply is where a stage operative voltage is dropped using a resistor and then filtered using a capacitor and they are drawn below of schematic because of clarity. On capacitor tops there are B+, B1, B2... and most tube amp builders do not have any problem building positive side just like it is drawn on schematic but negative ground side there often is and it seems because often ground is understood to mean Chassis but electrically they should be separate components.

    Capacitor negatives have numbers which match each stage ground and it means that on build all stage current is first collected to capacitor negative. On schematic negatives there are a white star signs and it means to each comes a wire which connects to the black star which is on first two capacitors negatives and it is where all return current is collected before power transformer sucks the current.



    SEL_MerlinHG-FX-1030x633.png
     
  6. schmee

    schmee Doctor of Teleocity Ad Free Member

    Posts:
    15,817
    Joined:
    Jun 2, 2003
    Location:
    northwest
    I recommend following the Fender scheme exactly except on the PT end it doesn't matter too much as long as it's near the PT.. Having not their scheme on a couple amps I think there's a reason they settled on what they did. (noise)
     
    Phrygian77 likes this.
  7. 2L man

    2L man Tele-Holic

    Age:
    61
    Posts:
    582
    Joined:
    Nov 23, 2020
    Location:
    Finland
    Single solid Ground Bus is just a simplified Star grounding method where one bus is built which begins on power transformer Center Tab or when a solid state bridge rectifier used GB starts from its negative. All circuit grounds are wired to this GB and it is finished to chassis at or next to input jack. Chassis is not used on amp circuit when GB is connected to chassis only at input jack and there is no possibility for current to escape to chassis.
     
  8. Phrygian77

    Phrygian77 Friend of Leo's Gold Supporter

    Posts:
    3,471
    Joined:
    Apr 30, 2016
    Location:
    Crawfordville, FL
    Since we have naysayers here (I'm looking at you, @King Fan) concerning a ground buss on the pots, I'm going to speak up tell you that it's perfectly okay. The only legitimate argument against it is that it creates little ground loops. That would matter more in a circuit with more gain (although I did it without issue in a Trainwreck Express preamp) than any Fender circuit.
     
  9. 2L man

    2L man Tele-Holic

    Age:
    61
    Posts:
    582
    Joined:
    Nov 23, 2020
    Location:
    Finland
    Mojotone and StewMac layouts are poor because they use chassis and also use potentiometer cases to flow return current. When amplifier amplifies frequencies there comes impedance phenomenon and return current gets confused and flow where it is easiest. Some frequencies current might flow to chassis thru potentiometer screws and some thru input jack.

    Chassis is metal and has lots of area and surrounds amp circuit so it is bad idea to make it an antenna which spreads electromagnetic noise to circuit.

    That Robrob layout in first post make pre amp current flow to chassis and it goes thru input jack which is worst place.

    Search RobRob Advanced 5E3 lauout and use it because it has good ground layout!
     
  10. Lowerleftcoast

    Lowerleftcoast Friend of Leo's

    Posts:
    3,989
    Joined:
    Dec 30, 2019
    Location:
    california
    If you want to learn about ground schemes, read through this:
    http://www.valvewizard.co.uk/Grounding.pdf

    Try to understand Fig 15.6
    This shows how each filter cap is connected to each load (tube). The filter cap and load make a filter node. Each node should be separate. The power should flow from noisiest to quietest.
    Fig 15.6 also shows the *ground* is not quiescent. The dashed line "ground impedances" just denotes the ground path, ie *chassis*, *bus*, *wire* whichever is used for the return path. It is not suggesting adding any impedance/resistor at these locations. It is showing the ground path has *some* impedance. The noise (ripple current) can be transferred to the more sensitive preamp nodes when additional paths are available (ground loops, ground planes).

    Try to understand Fig 15.10
    Figure (a) shows the grounds attached to the bus as opposed to figure (b) where the grounds are grouped into nodes.
     
    King Fan likes this.
  11. Phrygian77

    Phrygian77 Friend of Leo's Gold Supporter

    Posts:
    3,471
    Joined:
    Apr 30, 2016
    Location:
    Crawfordville, FL
    @2L man there's nothing really bad about the Mojotone layout. It's roughly the same as the original Fender scheme. The only thing wrong with it is that they still have a switched neutral for the mains, but that has nothing to do with the ground scheme.

    There's not a lot of benefit to doing anything else, unless you're going to change other things too. First and foremost would be to separate (edit: isolate, not separate) the main filter and CT.
     
  12. King Fan

    King Fan Poster Extraordinaire Ad Free Member

    Posts:
    6,654
    Joined:
    Jan 1, 2013
    Location:
    Salt Lake City
    Heh, @Lemon , you see why asking for advice on TDPRI is fraught with peril. :) Too many smart people with informed opinions -- that just happen to differ. As the guy asking, that's confusing, but it's also good news; it tells you there are different paths through the forest.

    But -- for every right path there are many wrong paths. You'll get strong opinions from people who can't cite a reputable source or well-known principle. News flash: "This is the right path because it's my path" is not rare on the internet. And that's before you enter the Marshes of Incorrect But Widespread Belief.

    To play my end of the pitch a moment, @Phrygian77 is right that the ground loops from soldering to the pots are small and not very noisy. (Merlin, in that good chapter from @Lowerleftcoast , actually admits this). But I still don't do it because it's 1. totally unnecessary 2. kinda hard to solder to pots and 3. really hard to unsolder pots if you need to change a shunt cap or replace a pot. Merlin also shows you the idealized single ground bus, which as noted by @2L man, is a simplified star. But if you look at 100 builds here, you may find only one or two that reach full Merlin nirvana; Rob's split grounds are just as useful in practice. And to circle around one more time, @Phrygian77 is right that you may not need explicit grounds from the cathodes (I didn't do it, FWIW).

    But mainly my defense falls back on that World Cup goalie, Robrob; his designs are practical, well-proven, widely built, and based on great research and experience. For a new builder, sometimes the smart plan is to follow a single, simple, safe map, like Rob's layout.
     
    Phrygian77 likes this.
  13. Lemon

    Lemon TDPRI Member

    Posts:
    42
    Joined:
    Feb 10, 2018
    Location:
    Pale Blue Dot
    Thank you for all the great information and for taking the time to explain. It's greatly appreciated.

    This is an amalgamation of your suggestions: [​IMG]

    HV CT connects directly to cap can ground terminal, which connects to chassis ground along with 6.3V CT and bias resistor. Power tubes connect to socket bolts. I will also attempt the floating ground bus bar.

    Should I drill a dedicated hole for the chassis ground connection or do you think that Mojotone's method of using the 4-40 cap can bolt is adequate?
    I think that they avoid drilling extra holes in their chassis to keep it as historically correct as possible (hence the inclusion of the ground switch), but I don't mind drilling if it's called for.
     
    King Fan likes this.
  14. Lowerleftcoast

    Lowerleftcoast Friend of Leo's

    Posts:
    3,989
    Joined:
    Dec 30, 2019
    Location:
    california
    The ground scheme for the Princeton Reverb makes for an interesting study. It breaks all the rules for a good ground scheme. The filter caps are all contained in the cap can. The cap can has a shared ground terminal. Each power node has a long distance route back to that shared ground terminal. The chassis was used as a ground path. Fender didn't even have a ground to chassis connection near the input jack. Even with breaking the rules many of the amps were quiet.:) Not all of them.:(

    You have asked the Shock Bros for advice about ground schemes. You have found a topic where we disagree. Currently we are split in our opinions.
    2L man
    suggests organizing the ground nodes and having one reference to ground near the input jacks.
    King Fan suggests Rob's layout.
    Phrygian77, and schmee suggest the Fender ground scheme. Although Phrygian77 hints at modifying the ground scheme.

    I agree with 2L man. That said, you are a first time builder and the kit would have to be modified. I think your build can be quiet without the extra effort.

    I urge you not to use the back of the pots for any grounds. So that leaves us with Rob's layout.
     
    2L man and Phrygian77 like this.
  15. Lowerleftcoast

    Lowerleftcoast Friend of Leo's

    Posts:
    3,989
    Joined:
    Dec 30, 2019
    Location:
    california
    I would drill another hole.

    If you are up for a small ground scheme/filter modification, another filter cap can be added for the preamp section. Two of the sections of the cap can can be paralleled to up the value to 40uF for the reservoir. The additional cap 10uF to 22uF can be added on the board for the preamp section. This should stiffen the bass response and reduce the wiring distance of the last filter node keeping it in close proximity to the preamp side.

    AA1164_Princeton_Reverb_Layout_DIYLC_LI.jpg

    The green near the cap can would be the dropping resistor. It could have a tag strip to hold the end of the resistor.
    The black on the board would be the additional cap.
    The red wire shows the two caps in the can parallel. Any of the cap sections could be used here.
     
    2L man and King Fan like this.
  16. King Fan

    King Fan Poster Extraordinaire Ad Free Member

    Posts:
    6,654
    Joined:
    Jan 1, 2013
    Location:
    Salt Lake City
    Dude, you're going to the head of the class. First build, really? Not only are you a quick study, but you're quick *and* good with DIYLC.

    Your guess about them avoiding extra holes for authenticity may be right, or maybe they don't want to assume people can drill holes in galvaneal. :) As a side note, they do try to modernize -- they used to leave in the death cap, but not wire it. In fact, note they're reading robrob and Merlin on the HV CT. Also to their credit, the StewMac version still contains some older Mojo ideas many of us avoid. Too bad Mojo still switches the neutral.

    Since you are good at DIYLC, I'm gonna guess you can drill holes for #8 bolts -- and at that point, I'm with LLC. Drill a (UL standard) separate hole for the safety ground (Leo's left sidewall location is great) and another for the power amp ground neatly located between the bias board and cap can.

    Coupla bias details. If you're ordering some extra components, it's nice to upsize that first bias 'knockdown' resistor to 2-3W, and if you're using a 10KL pot (which is a tried and true solution, shared by Doug Hoffman as long ago as the Dot Matrix Age), it often works best on the PR with a 27K 'tail' resistor. Doug's version:

    upload_2021-6-19_14-38-58.jpeg

    My DIYLC of that area:

    upload_2021-6-19_14-49-32.jpeg
     
  17. Phrygian77

    Phrygian77 Friend of Leo's Gold Supporter

    Posts:
    3,471
    Joined:
    Apr 30, 2016
    Location:
    Crawfordville, FL
    What I would avoid is using the can clamp mounting point for the mains safety ground, just because it should be close to the entry point, and that wire sould be left long enough that it would be the very last wire to break if the power cord were ripped from the chassis.

    Using the cap clamp mounting point for a solder terminal (or terminal strip) for the CT and can ground is fine. What you will usually hear adamantly is not to use the transformer studs. This is because vibration in the transformer itself can loosen the nuts over time. That's not so much a problem with the can clamp, but having a dedicated mounting point for a ground terminal is ideal.

    I would absolutely not go out of my way to run the bias voltage divider resistor directly to anyplace other than the back of the bias pot. @King Fan

    :p

    [​IMG]

    I'm just stirring the pot. @King Fan is correct about the bias resistor. A 22k may not get you enough negative voltage to bias some tubes cool enough. It's happened to me. I don't see any reason for the 2-3 watt recommendation though. The max amount current it will see is essentially fixed because it's connected to ground.

    A quick over guesstimate...

    50V ^ 2 = 2500
    2500 / 27000 = .093 watts
     
    Lowerleftcoast likes this.
  18. 2L man

    2L man Tele-Holic

    Age:
    61
    Posts:
    582
    Joined:
    Nov 23, 2020
    Location:
    Finland
    If you extend this "floating" ground bus so that you can connect it to the B+1 and B+2 capacitor negative terminal where you connect the PT CT and connect 1 ohm cathode resistors to it as well (and all other grounds) the ONLY chassis connection is at input jack there does NOT flow any current thru Chassis.

    When chassis is not used as a part of amp circuit it is easy to modify "ground lift" where two parallel but opposite diodes, 100 ohm resistor and 0,1uF capacitor are connected between ground and chassis. This can be useful when pedals increase noise.
     
  19. Phrygian77

    Phrygian77 Friend of Leo's Gold Supporter

    Posts:
    3,471
    Joined:
    Apr 30, 2016
    Location:
    Crawfordville, FL
    Then you may as well float the whole thing with isolated jacks (except the reverb driver). You need a cap to chassis for the input shields, including the reverb recovery, so that RF has lower impedance path to the chassis. Of course, you need return paths for the reverb recovery and foot switch jacks, and the speaker jacks (since we have NFB). Also, anti parallel diodes with a current capacity higher than the mains fuse for fault protection.
     
  20. Lowerleftcoast

    Lowerleftcoast Friend of Leo's

    Posts:
    3,989
    Joined:
    Dec 30, 2019
    Location:
    california
    I agree. I wouldn't go out of my way for the resistor on the bias pot but since I have recommended a hole to be drilled to avoid using the clamp bolt as a ground point. Positioning the hole between the bias pot and cap can would provide a gathering point for all of the grounds on the PT side of the chassis. More betterer ime.

    Of course if a terminal strip were placed in that hole, the grounds could meet on the terminal strip. The ground bus could meet there as well to fulfill the suggestion by @2L man . Which would give a single reference to ground at the input side of the chassis. It might be worth considering though I doubt it would be noticeably quieter one way or the other.
     
IMPORTANT: Treat everyone here with respect, no matter how difficult!
No sex, drug, political, religion or hate discussion permitted here.