Grounding, again...

Discussion in 'Amp Tech Center' started by moosie, Mar 13, 2019.

  1. moosie

    moosie Doctor of Teleocity Silver Supporter

    Age:
    62
    Posts:
    15,398
    Joined:
    Jul 18, 2010
    Location:
    Western Connecticut
    Hey guys. Sorry to beat the dead horse...

    Assume a modern build, with a three prong AC cord, and the green (3rd) wire firmly bolted to the chassis. The chassis is now at the same potential as the human reaching over to touch it. Zero volts. Grounded.

    Best practice multi-star grounding is observed, ala Merlin (or a bus, or a hybrid approach). Fiber shoulder washers isolate the chassis from the jacks. Common return paths (aka 'grounds') are considered based on the currents they serve, and differing current levels are kept apart.

    I'm good with all of this, until the last bit: connect the main star point (or end of bus) to one location (and one location only) on the chassis. Either near the power supply, or near the input jack.

    I get the rationale for one-and-one-only... but why do it at all? Oh wait, the chassis won't be able to function as a RF shield. OK, so we use a death cap at the input jack, which shunts off any high freq AC to the chassis before it gets into the preamp.

    Can we stop there? Beyond this, is there any functional reason to connect the chassis to our carefully planned return path layout? I can't think of one, so I'm guessing it's about safety.

    What are the failure scenarios in which this connection to the chassis will help? In other words, are the fuse and the AC safety ground wire enough?

    I've read a lot of sources, and a lot of details, but this one point tends to be more of a given, and thus not well explained. Some imply it's all about the shielding, which the cap corrects. I suspect it's something of a belt-and-suspenders thing, which I'm all FOR. Just trying to understand the details.

    [ EDIT: the one reason I read about which seemed to make it necessary is to keep the chassis' of multiple devices, which are connected together, at the same ground potential. Not sure I understand why this matters, but yeah, if the chassis of the amp and the external preamp, or the outboard reverb unit, etc, are each entirely floating, it doesn't really matter if they're precisely at zero volts, but it might matter that they be the same. And they probably wouldn't be. ]

    Thanks.
     
    Last edited: Mar 13, 2019
  2. Phrygian77

    Phrygian77 Tele-Afflicted

    Posts:
    1,723
    Joined:
    Apr 30, 2016
    Location:
    Crawfordville, FL
    What did you do in the Revibe?
     
  3. moosie

    moosie Doctor of Teleocity Silver Supporter

    Age:
    62
    Posts:
    15,398
    Joined:
    Jul 18, 2010
    Location:
    Western Connecticut
    Pretty much as described above, but with a small hum block network at the chassis connection point. Especially to reduce ground loop issues because the Revibe for sure connects to an amp which has it's own ground connection.

    But I just followed 'instructions', and still don't truly understand it. Based on my thinking above, I suspect the Revibe could have remained entirely floating. Again, no functional issues, except perhaps the one described in my edit, that I found on some audiophile site. (And we know how they can be... LOL). But maybe very definite safety issues. That's what I'm trying to find out.

    For instance, with a floating chassis, with the only connection to the circuit being a) death cap for RFI, and b) AC safety ground... what can possibly occur to make that chassis hot? Since it's grounded, won't it always be at zero volts? Perhaps causing the fuse to blow, which is the desired thing...
     
  4. clintj

    clintj Friend of Leo's

    Posts:
    4,919
    Joined:
    Apr 4, 2015
    Location:
    Idaho
    Bonding to the metal chassis ties the whole circuit to a single, common zero voltage reference point. In theory, you could use a plastic chassis and tie all your ground bussing to the mains ground to achieve that same zero volt reference, but as you suspect it wouldn't make a very good shield against outside noise.

    Sent from my Pixel 2 using Tapatalk
     
  5. peteb

    peteb Friend of Leo's

    Posts:
    4,088
    Joined:
    Apr 25, 2003
    Location:
    Cascadia
    if the chassis is safety grounded it is not floating.

    If the chassis is safety grounded it stays at zero electrostatic potential and is safe.

    If the chassis is safety grounded the chassis should provide RFI sheilding.





    Is the question that the safety or RFI sheilding could improve?






    I don’t think the safety could improve, but could the RFI sheilding improve with the addition of an anti RFI cap?
     
  6. moosie

    moosie Doctor of Teleocity Silver Supporter

    Age:
    62
    Posts:
    15,398
    Joined:
    Jul 18, 2010
    Location:
    Western Connecticut
    Hi @clintj, thanks for chiming in. When you say 'mains ground', do you mean the AC safety ground coming in from the wall? Or the PT secondary?

    If you mean the PT secondary, yeah, I need to complete the path, or it simply won't work.

    But if you mean the wall ground, why?

    Let's assume I'm using a .01uf cap from (isolated) input jack to (steel) chassis, to get the shielding effect.

    Why do we need more than that? What's the value of having the reference level be the same for the (isolated) circuitry, as it is for the chassis and wall wiring?

    Is it the point I alluded to about having all potentially connected devices' circuits share the same reference? Or is it more than that?

    Is floating the circuit (except the RFI cap) dangerous?
     
  7. peteb

    peteb Friend of Leo's

    Posts:
    4,088
    Joined:
    Apr 25, 2003
    Location:
    Cascadia
    Moosie,

    I don’t understand what you are trying to accomplish that isn’t already standard practice.



    Could you please restate your goal?
     
  8. clintj

    clintj Friend of Leo's

    Posts:
    4,919
    Joined:
    Apr 4, 2015
    Location:
    Idaho
    The power cord ground.

    The cap to the jacks would provide an AC ground reference, but not a DC ground reference. There's a difference there, and is the reason a death cap equipped amp will give you a tingle when the switch is in the wrong position.

    With the jacks isolated, there's still a connection from the circuit to you. Your guitar interfaces with the amp's grounding system through the input stage grid leak path, the 1M resistor on most input jack networks. That point needs to have a DC ground reference, or you become the ground reference through the strings.

    Guitar amps are also special in that the user is normally touching some part of the device in use, usually the guitar but also the chassis itself when you reach for a knob and make incidental contact. This is why the safety ground's importance is two fold. It provides the DC reference for the whole circuit via the grounding and chassis, and provides a way to drain off dangerous voltage if there's a problem.

    Does that clarify any, or are we still talking past each other?

    Sent from my Pixel 2 using Tapatalk
     
    moosie and D'tar like this.
  9. moosie

    moosie Doctor of Teleocity Silver Supporter

    Age:
    62
    Posts:
    15,398
    Joined:
    Jul 18, 2010
    Location:
    Western Connecticut
    It does help. Thanks!

    I had forgotten all about the player, connected to the DC circuit via the strings. Doh.

    I have another question related to this, but first... can you confirm what happens in a failure scenario with the amp wired as I'm supposing?

    Let's say a power tube shorts. Normally, the fuse would blow. Will it still blow without the chassis connection to the circuit? Will voltage rise to melt-levels in the B+ circuit? Will that roughly-near-zero floating common float ever higher, putting high voltage on the strings of a plugged in guitar?
     
    Last edited: Mar 14, 2019
  10. moosie

    moosie Doctor of Teleocity Silver Supporter

    Age:
    62
    Posts:
    15,398
    Joined:
    Jul 18, 2010
    Location:
    Western Connecticut
    Further discussion may have clarified this for you, but in case not, I may have thrown out a red herring in my OP. I think I said 'floating chassis'. I meant 'floating circuit, with respect to chassis and wall'. Does that clear up the question?
     
  11. peteb

    peteb Friend of Leo's

    Posts:
    4,088
    Joined:
    Apr 25, 2003
    Location:
    Cascadia

    yes, thanks
     
  12. clintj

    clintj Friend of Leo's

    Posts:
    4,919
    Joined:
    Apr 4, 2015
    Location:
    Idaho
    I need to draw the supposed circuit out and walk through it for that one. Should be an interesting exercise for after the kids go to bed.

    Sent from my Pixel 2 using Tapatalk
     
  13. peteb

    peteb Friend of Leo's

    Posts:
    4,088
    Joined:
    Apr 25, 2003
    Location:
    Cascadia
    I get you. The question is could the chassis be safetey grounded, and the circuit be floating, completely isolated from the chassis?


    Maybe the one solid attachment is so that the circuit is not floating, for safety reasons, and also for RFI grounding.
     
    Last edited: Mar 14, 2019
  14. peteb

    peteb Friend of Leo's

    Posts:
    4,088
    Joined:
    Apr 25, 2003
    Location:
    Cascadia

    If the high voltage in the power tube shorts to ground, then fuse will blow. If it shorts to the chassis, and the chassis is grounded, then yes, the fuse will blow. Is that what you mean? That the tube shorts to the grounded chassis?




    Or do you mean it shorts to the circuit (ground) and gets on the guitar? If the high voltage DC shorted to something floating, then it might not blow the fuse.








    Connecting the circuit ground to the safety grounded chassis is safer and should be quieter.


    Is there any rational for not connecting circuit ground to chassis ground?
     
    Last edited: Mar 14, 2019
  15. moosie

    moosie Doctor of Teleocity Silver Supporter

    Age:
    62
    Posts:
    15,398
    Joined:
    Jul 18, 2010
    Location:
    Western Connecticut
    Careful using the word 'ground'.

    If the circuit floats, ground is just a non-referenced common return path back to the PT, nothing more. It's my understanding the fuse blows because a short allows current to flow, unloaded. Racing back to the power company fast as it can. Generating too much heat, burning up. The fuse prevents it from getting out of hand. But what if there IS no connection back to the power company, because the circuit on the secondary side of the PT floats, and the short occurs on this side... I can't see the fuse blowing.

    In which case... what happens?


    Assuming I use the RF cap as described above, floating won't be noisier.

    In fact the rationale is noise reduction. Not cutting this sole chassis connection per se, but the whole idea of controlling and isolating return path currents.

    Ground is usually just treated as this constant black hole. But the currents flowing are quite varied. "Ground" is just the other half of any circuit loop. High current circuits like power supplies, and the OT, push a lot of current through the local ground, since it's just the same flow as on the 'hot' side, trying to complete the loop. If that local ground is something shared, like a chassis used to ground things randomly, or a wire that shares ground connections with other areas, then it's likely the high current flow from power supply etc will stomp all over the preamp, or whatever. Noise.

    I love vintage Fenders, but I don't want to ground to the chassis like they did. I'm shooting for a quieter amp, partly just because it makes sense to me to try. And that means controlling those ground currents.

    For examples of how to do this, Merlin's grounding treatise is a good starting point, IMO. EDIT: I attached it.

    Will I do this? Float the circuit? No. I never planned to, because I knew there must be a good reason not to, but I wasn't seeing it. Which meant I didn't understand something. Which can be dangerous. So, I'm drilling down on it. I do plan to almost-float it, though. Well organized and isolated currents, and possibly a hum block network at the chassis connection point. But that raises another question, I think...
     

    Attached Files:

  16. peteb

    peteb Friend of Leo's

    Posts:
    4,088
    Joined:
    Apr 25, 2003
    Location:
    Cascadia
    C.
    If it’s not shorted to ground, to real ground, earth ground or the power company, it’s not really shorted. (Thru the earth it will get back to the power company). I agree with you, if high power ‘connects’ to any floating thing, I also can’t see the fuse blowing, and that’s a dangerous and unwanted situation. What happens? It’s just bad, it’s hot, it’s dangerous.


    I see where you are going. It’s similar to the old fender 2 prong design with the chassis floating to DC but not AC. Your basic idea is the chassis is safety grounded, but the amp circuit is AC grounded only and floating to DC. The hazards would be the same hazards as the 2 prong design, except your design would be safer because the chassis is big and all around everything and there is a lot more chance of the high power DC contacting it than the amp circuitry, even though that would still be possible due to the proximity.




    Anyway, you could definitely try the completely DC floating amp circuit safely, to hear what it sounds like. The key is, if you are not grounded, you can’t get shocked.
     
  17. peteb

    peteb Friend of Leo's

    Posts:
    4,088
    Joined:
    Apr 25, 2003
    Location:
    Cascadia

    I see that you are going deep into the noise reduction area, which is of interest to me but not my most familiar area.



    I see that you are interested in the noise reducing benefits of the completely floating circuit. I suggest, if you want to answer this question for certain, try completely DC floating the circuit, just to test it out.
     
  18. peteb

    peteb Friend of Leo's

    Posts:
    4,088
    Joined:
    Apr 25, 2003
    Location:
    Cascadia
    Will it be safe?

    Shock hazard is an area I am somewhat familiar with and take an interest in and read any threads on the subject.



    This is something I noticed.



    Every time people get shocked from amps, it seems, it’s the wall power. The AC. I’ve never heard of anyone getting shocked from the DC inside the amp.






    I was unsure what happens when a tube shorts out. I looked into it.

    Apparently the most common tube short is the internal tube short between plate and cathode, over running the grid. In that case, the fuse will blow whether the chassis or circuit is floating or non floating.
     
    Last edited: Mar 15, 2019
  19. moosie

    moosie Doctor of Teleocity Silver Supporter

    Age:
    62
    Posts:
    15,398
    Joined:
    Jul 18, 2010
    Location:
    Western Connecticut
    Merlin's the first and last word. I've read that chapter at least twenty times. Initially not much made sense. Eventually, over time, a lot of it is coming together.

    Yep, that's the short I was thinking of. OK, I'll bite. What will cause the fuse to blow, in a floating circuit amp?

    Well, @clintj reminded me of a case... If I float the circuit, it won't be zero referenced. It might be close, but it'll float. The strings are part of that DC path. In a normal case it's just annoying, but there would be a tingle as the earth uses me as the conduit to equalize the voltage level between zero and ... whatever. Which is why I'm interested to learn if it can float quite high in the case of a short. That would be no fun, to have several hundred volts on the strings. But just think of the TONE... :lol:
     
  20. clintj

    clintj Friend of Leo's

    Posts:
    4,919
    Joined:
    Apr 4, 2015
    Location:
    Idaho
    Shorted tube will blow the fuse on overcurrent. The tube, when it arcs, will draw excessive current which is drawn from the power supply. That spike in current flow is what will blow the fuse.

    I'm still having trouble visualizing the fault scenario. Instinctually, I want to say it'll come to rest at an elevated point based on what I've seen in damaged house wiring, like floating grounds, loose neutrals, etc, but proving and explaining it are being difficult.

    Where's a practicing EE when you need one?

    Sent from my Pixel 2 using Tapatalk
     
IMPORTANT: Treat everyone here with respect, no matter how difficult!
No sex, drug, political, religion or hate discussion permitted here.


  1. This site uses cookies to help personalise content, tailor your experience and to keep you logged in if you register.
    By continuing to use this site, you are consenting to our use of cookies.