Options for input wiring, shielded cable, grid stopper location?

Discussion in 'Shock Brother's DIY Amps' started by King Fan, Sep 25, 2021.

  1. King Fan

    King Fan Poster Extraordinaire Ad Free Member

    Posts:
    7,201
    Joined:
    Jan 1, 2013
    Location:
    Salt Lake City
    Let's try to keep this simple, KF. It's well known that the ideal place for a V1 grid stopper is right on the socket pin. But I don't see many amps built that way (tho I'm aware of several smart builds here that do).

    Instead, a lot of us use shielded input cable as far as possible, hanging the (Fender) 68K grid stoppers on the input jacks to make the shield as complete as possible. Bit of a hassle, if I'm honest, including bringing out the shield to ground.

    upload_2021-9-25_11-47-18.jpeg

    So a few related questions.

    1. Anyone got examples of tying the grid stopper right to the V1 grid pin? Do you lay it on the chassis, or float it in space? I couldn't find pics of these methods; I think I've seen both?
    2. Has anyone used an 'isolated terminal post' like this, to anchor the resistor at both ends?

    upload_2021-9-25_11-52-39.jpeg

    3. And, if you mount the grid stopper straight to the socket pin, could you omit the shielded cable, since the stopper should be rejecting the noise the shield was absorbing?
     
    theprofessor likes this.
  2. tubedude

    tubedude Tele-Holic Silver Supporter

    Posts:
    535
    Joined:
    Jul 26, 2013
    Location:
    east georgia
    Mass production and circuitboard amps minimize off board components to simplify the production process.
    Stoppers don't reject noise, their resistance prevents the tubes capacitance from causing oscillation.
    I wouldn't use shielded cable there unless it was a very high gain amp.
     
    jsnwhite619 likes this.
  3. tubedude

    tubedude Tele-Holic Silver Supporter

    Posts:
    535
    Joined:
    Jul 26, 2013
    Location:
    east georgia
    If your using one guitar, why build an amp with 2 input jacks? We might as well have mic and accordion inputs also.
     
    wadeeinkauf likes this.
  4. King Fan

    King Fan Poster Extraordinaire Ad Free Member

    Posts:
    7,201
    Joined:
    Jan 1, 2013
    Location:
    Salt Lake City
    I was working from Merlin:
    "A resistor should always be added in series with the grid of the input valve (and others stages too). This is known as a 'grid-stopper', and it can serve several purposes:
    Reduce radio frequency interference;
    Reduce treble response (high cut);
    Prevent local parasitic oscillation;
    Limit grid current (which reduces blocking distortion and...."
    I agree it may not be necessary; OTOH even in low-gain amps, that's the most-amplified spot in the circuit, and RF does exist in there, so I'm not sure getting rid of it can hurt.

    Interesting, never heard that before. First off, I like my 5E3s to look and act vintage. But also, whether or not the 'Mic' channel there has any great utility by itself, and altho you can argue the 'hi' and 'lo' are also unneeded, those four little holes certainly give up some great sounds when jumpered.
     
    Lowerleftcoast likes this.
  5. King Fan

    King Fan Poster Extraordinaire Ad Free Member

    Posts:
    7,201
    Joined:
    Jan 1, 2013
    Location:
    Salt Lake City
    Now I'm trying to recall who I've seen mount the grid stopper right on the V1 pin(s). I'm gonna phone a friend; was it you, @Nickfl ?
     
  6. Mexitele Blues

    Mexitele Blues Tele-Afflicted

    Age:
    43
    Posts:
    1,845
    Joined:
    Mar 7, 2018
    Location:
    Westminster, CO
    I did this on my parallel 6V6 build, at your recommendation.

    20210925_134811.jpg
     
  7. tubedude

    tubedude Tele-Holic Silver Supporter

    Posts:
    535
    Joined:
    Jul 26, 2013
    Location:
    east georgia

    Stoppers don't reject noise, their resistance prevents the tubes capacitance from causing oscillation.
    I was working from Merlin:
    "A resistor should always be added in series with the grid of the input valve (and others stages too). This is known as a 'grid-stopper', and it can serve several purposes:
    Reduce radio frequency interference;
    Reduce treble response (high cut);
    Prevent local parasitic oscillation;
    Limit grid current (which reduces blocking distortion and...."

    Yes, reduces RFI, prevents oscillation, can cut hi end. But doesn't reduce noise.

    "I agree it may not be necessary; OTOH even in low-gain amps, that's the most-amplified spot in the circuit, and RF does exist in there, so I'm not sure getting rid of it can hurt."

    If you are having RF interference issues then 2" of shielded wire may help.
    The best way to reduce noise in the first stage is to limit series resistance. Using a smaller value resistor with a tiny cap shunted to ground and using metal film resistors on V1 will drop noise significantly.

    In the case of copying an old design to be authentic, all is fair game, carbon resistors, electrolytics, waxed cloth wire, etc. And leaving the grid stopper in the wrong place.
     
  8. King Fan

    King Fan Poster Extraordinaire Ad Free Member

    Posts:
    7,201
    Joined:
    Jan 1, 2013
    Location:
    Salt Lake City
    I like your elegant implementation. Heh, typical me, suggesting it to others before I even do it myself. Seriously, it is common knowledge; in fact it had been suggested to me on my very first build, and the pros recommend it. Aiken: "...put a 10K-68K resistor on the grid pin of the first tube, right at the socket." Paul Ruby: "The key thing is that the input stage is a very sensitive node for noise. This is especially true once past the 68K ohm grid stopper resistor in series with your input lines. As such, the 68K grid stopper really should be mounted right on the tube socket with ZERO lead length between the socket lug and the resistor."

    Ruby goes on BTW to argue for shielding *plus* pin-mounted grid stoppers on input lines: "Note, I use shielded input lines on ALL my amps and mount the 68K resistors on the tube socket as the first thing when getting rid of hum. I don't even bother to debug first. It's cheap, easy insurance against hum, noise or feedback (squeal). Just do it!" And: "Your shielded wire then runs from the other side of the resistor to the input jack. The shield of the cable is only grounded at the jack end. Cut the shield off the tube end and use a hunk of heat shrink to make sure it doesn't short to anything near by."

    So why haven't I done it? My amp-smart buddy on that first amp told me it was fine to hang it as an extension off the pin with no anchor at the other end. I recall his reasoning, too: "It isn't like there're actual demons who get in there and swing on your wiring." But before I thought about terminal posts (or tag strips) I had also seen more than one apparent expert saying it wasn't cool to hang it in space, or lay it on the chassis, and just solder to it. I'll be honest; I figured somebody would see my build pic and point out how lame it was to have the resistor hanging off the pin. At the same time, I still kinda figured my buddy was probably right. So now, years later, I'm here asking these questions.
     
    Mexitele Blues and dan40 like this.
  9. Tom Kamphuys

    Tom Kamphuys Tele-Holic

    Age:
    42
    Posts:
    871
    Joined:
    Sep 18, 2018
    Location:
    The Netherlands
    Guilty as charged!

    IMG_20181203_225356086.jpg

    This is my JTM45 clone. This was my first build and I went a bit overboard with wirewound gridstoppers mounted with double sided tape. The grid stopper does not stop noise, it actually is the biggest cause of noise (according to Blencowe). Wirewound have lowest noise and I used Blencowes preferred input circuit with just a 10k gridstopper (for noise reasons).

    IMG_20200726_190507866 (1).jpg

    And here is my JCM800, now with 10k 2W metal film gridstoppers mounted between the tube socket and a terminal.

    IMG_20210405_155205803 (1).jpg
    And here is my stereo hifi amp where I used 3W metal film resistors. Larger power, lower noise. They are a bit hard to find: they are the red/brown large resistor mounted between the socket and a terminal.
     
    Mexitele Blues, dan40 and King Fan like this.
  10. King Fan

    King Fan Poster Extraordinaire Ad Free Member

    Posts:
    7,201
    Joined:
    Jan 1, 2013
    Location:
    Salt Lake City
    Oh, nice!!! Thanks, Tom. Elegant and well-thought out, as we'd expect from you. I see you also keep the shielding as Paul Ruby suggested and @Mexitele Blues did, so I'm going to consider my question 3 as answered (I mainly asked since this being the internet, I did see a posted opinion somewhere sometime that the downstream grid stopper made shielding unnecessary).

    I'm glad you mention the *resistance* of the grid stopper, which Tubedood also noted. In fact I shoulda included that in my list of "well-known input ideas most of us ignore" (hey, but you don't!!) since Merlin covers it in that same primer: "...the input grid stopper adds the most amount of Johnson noise (hiss) of any resistor in the amp. (A 68k stopper generates at least four times more hiss than a typical 12AX7!) Can we use a smaller resistor to reduce noise, but still keep the bandwidth the same? Yes, it's easy. We simply add a little extra capacitance to make up for the lower resistance. I recommend a 10k resistor, which will make the amp much less hissy, together with an extra 100pF to 470pF capacitor from grid to ground."

    So we all agree the grid stopper adds hiss. But am I getting confused on what we mean by 'noise'? Merlin's #1 item on his list of grid stoppers is rejecting RF. Is RF not a form of noise? Ruby uses the grid stopper to fight hum. Is hum not a form of noise? Anyway, they (and you) know more than me, but if it can reject RF and hum, even before it fights oscillation etc, I'll be OK.
     
  11. Tom Kamphuys

    Tom Kamphuys Tele-Holic

    Age:
    42
    Posts:
    871
    Joined:
    Sep 18, 2018
    Location:
    The Netherlands
    There are multiple types of 'noise'.
    RF is an inaudible undesired signal from outside the amp.
    Hum is an audible signal from inside the amp. I don't see how a grid stopper could stop that. It is around the frequency of the lowest note on a (detuned) guitar.
    Resistor noise really is noise, random.
     
  12. King Fan

    King Fan Poster Extraordinaire Ad Free Member

    Posts:
    7,201
    Joined:
    Jan 1, 2013
    Location:
    Salt Lake City
    Good points. I had thought some RF was in the audible range, but that was just an impression. Radio stations get dumped at the input jack, then? And inside the chassis, I guess if it's out of audible range, we just want to reject it to keep the signal electrically clean? (Tho I would've thought of that as "free of ultrasonic noise?") But again, I'm learning here. As for hum, Paul Ruby may be wrong, or I may be imputing more to his grid stopper than he was. At any rate, these are nice arguments for keeping the shield on the input wiring (for whatever noise including hum that shielding can prevent).

    Also, on closer inspection, Tom, I see you are a pro at mounting resistors on sockets! Score a bunch of points for tag strips.

    upload_2021-9-25_15-30-9.png
     
    Last edited: Sep 25, 2021
  13. dan40

    dan40 Friend of Leo's

    Posts:
    2,375
    Joined:
    Aug 19, 2015
    Location:
    Richmond Va
    I have also gotten in the habit of mounting the stoppers directly at the socket. I will use tag strips if space allows. If space is tight, I will connect directly to the input wire and then tape the wire down securely to the chassis to prevent movement.
     
    King Fan likes this.
  14. NSB_Chris

    NSB_Chris Tele-Holic

    Age:
    56
    Posts:
    634
    Joined:
    Sep 2, 2017
    Location:
    New Smyrna Beach Florida
    For all my later amps, i tie the resistor right at the socket and I don't bother to anchor the resistor with a separate lug. I either run the wire naked or run a shielded cable depending on the amp and the length of the run. For an unshielded wire, there is plenty of stability of the resistor hanging off of the socket lug. For a shielded cable, a wire tie somewhere reasonably close is sufficient. Just my two cents.
     
    dan40 and King Fan like this.
  15. Paul G.

    Paul G. Friend of Leo's

    Posts:
    3,936
    Joined:
    Mar 17, 2003
    Location:
    Rhode Island
    If you're using shielded cable, the grid stoppers can be mounted on the jacks because you're not gonna pick up any noise or RF in the cable run. This way, you need only one wire run.

    If you're using plain cable, run two wires below the board, flat against the chassis for shielding and put the stoppers right on the pin. You put them on terminal strips, but just soldering them to the end of the wire and shrink wrapping 'em is just fine.

    Sometimes I will use shielded cable AND put the grid stopper right on the pin, but I know that's obsessive and totally unecessary. I can't help myself.
     
    dan40, Lowerleftcoast and King Fan like this.
  16. sds1

    sds1 Tele-Afflicted

    Age:
    45
    Posts:
    1,874
    Joined:
    May 4, 2017
    Location:
    Orlando, FL, USA
    I have floated Rg between the wire/pin before but to be honest this never sat well with me. Feels janky.

    Now I will mount a terminal strip now to accommodate Rg, right between the socket and board. Sometimes using the socket mounting hardware even. Just like @Mexitele Blues ' pic.
     
    King Fan likes this.
  17. King Fan

    King Fan Poster Extraordinaire Ad Free Member

    Posts:
    7,201
    Joined:
    Jan 1, 2013
    Location:
    Salt Lake City
    These are all really helpful answers. Thanks, everyone. I'm starting to see that my survey of build pics wasn't an accurate sample of builds -- so many of you guys are experienced builders. It's we less frequent fliers who're more likely to take pictures of all the clouds out the window.

    I like the "anchored if possible, but if not, that's OK" vibe I see here. Lots of ways to skin this cat.

    A secondary question; I've tended to think like you, @Paul G. , that if I run shielded wire I don't need to anchor to the pin. But maybe shielding is not the whole story? Does right-on-the-pin also do more to fight oscillation, or have other advantages? Heh. All these questions are just telling you, Paul, you're not alone in being a bit obsessive.
     
    Paul G. likes this.
  18. chas.wahl

    chas.wahl Tele-Meister

    Posts:
    248
    Joined:
    Dec 8, 2012
    Location:
    NYC
    Here's an example by someone whose workmanship I admire:
    Screen Shot 2021-09-26 at 09.55.07 .jpg

    Direct flyover, no shielding; seems to work well.

    Here is my (at this point, theoretical) similar implementation, with shielded cables and 2-terminal tag strips mounted to edges of board to both support resistors, terminate them to cable, and also support the cables over by the jacks.

    Screen Shot 2021-09-26 at 10.00.04 .jpg

    Through a reference on this forum, I think, I've found some mil-spec aviation-grade shielded cable 22 gauge being only 0.078" diameter (single conductor) and also multi-conductor varieties, at Aircraft Spruce. Gauges available from 24 up to 6 -- in single-conductor anyway. Cable has ETFE sheath, stranded conductors (like most shielded cable I've come across), braided shield, and the multi-conductor ones are twisted. Part numbers are 11-14422 (mil-spec M27500-22TG1T14) for single and 11-05650 (M27500-22TG2T14) for double.
     
    Last edited: Sep 26, 2021
    King Fan likes this.
  19. King Fan

    King Fan Poster Extraordinaire Ad Free Member

    Posts:
    7,201
    Joined:
    Jan 1, 2013
    Location:
    Salt Lake City
    Thank you. Nice examples. And your circuit drawings are great, as I’ve said before.

    Both those pics nicely emphasize how the problem is, um, amplified by the four leads and four resistors in a 5E3. My lazy solution was to stay with the 68Ks on the jacks so there're only two shielded runs. But I need to think about this after some more coffee...
     
  20. chas.wahl

    chas.wahl Tele-Meister

    Posts:
    248
    Joined:
    Dec 8, 2012
    Location:
    NYC
    Thank you for the compliment, @King Fan -- 40 years in a design occupation ought to be worth something. I admit the complication of four cables across the (narrow enough, in the case of gopher coffins) chassis, and am considering whether two 2-conductor cables might be a more elegant solution practically, but wonder whether there would be any downside to a twisted pair such as some unwanted interaction between the two inputs of each channel. Twisted pair is, I grant, used for isolation/shielding in compact communications wiring, but that's typically for send/return for a single signal, no? Rather than two different signals.
     
IMPORTANT: Treat everyone here with respect, no matter how difficult!
No sex, drug, political, religion or hate discussion permitted here.