Rob Robinette mod I don't understand

[pompeiisneaks]

So someone's asking for some help on Reddit, and I tried pointing them here but I don't think they're going to, but they did point me to a mod they did that seems really confusing to me:

https://robrobinette.com/5e3_Modifications.htm#Switched_Negative_Feedback

(Rob if you're around help! :D)

I can't understand how three way switching that turns off the cathode resistor is ever good, you need that to keep the tube biased right?  Shouldn't the switch only switch the cap in and out and the NFB in and out?  as is it seems to me that you have 3 options, 1, cathode resistor, 2. cathode cap, and 3, nfb.  I would think you'd need the cathode resistor in all cases no?

If not, help me understand what's going over my head.

~Phil

[SILVERGUN]

Take a look at the picture on that same page.
The cath. R is always in the circuit because it's lead is wired to the same terminal as the wire coming from the tube socket pin #3

[pompeiisneaks]

OK that makes sense, that is not what it seems to show on the schematic from what I cant tell on that same page.  Thanks!  That's exactly how I imagined it being wired in my head so the cathode resistor was always in circuit. 

~Phil

[jjasilli]

I can't understand how three way switching that turns off the cathode resistor is ever good, you need that to keep the tube biased right?
No.  If you take the K resistor out of the circuit, the tube is turned Off.  This is one way to implement a standby:  the heaters will be On; the plate will be charged with B+; but but no current will flow until the cathode is reconnected. The tube does not need to be biased when it's Off. 

[pompeiisneaks]

But that wasn't the case in the schematic, it was either resistor, or capacitor or nfb, so in the case of the NFB I think that would be the 'off' position, but the schematic was wrong, the layout was right.  If you just have a cap, then it can conduct the signal to ground, via the cap, as well as via the anode but you hve no resistance and that would cause bias problems, or so I think anyway.  It was drawn as a three way toggle between the above three options instead of a three way toggle between resistor and cap, and resistor and nfb. and just resistor

~Phil

[pompeiisneaks]

Here's the schematic.

~Phil

[pompeiisneaks]

and my poorly drawn attempt at fixing it. The thicker red line should be a permanent jumper.  (that is how it's done in the real layout, just not right here)

~Phil

[Tony Bones]

I think this is what Rob intended.



I'll send him an email.

[2deaf]

If you just have a cap, then it can conduct the signal to ground, via the cap, as well as via the anode but you hve no resistance and that would cause bias problems, or so I think anyway.

With only a capacitor on the cathode, the capacitor will charge up until the tube reaches cutoff - - a few volts, give or take, depending on the anode voltage.  The fun starts when you give the tube some signal.  The positive side of the signal allows the tube to conduct again and the cathode capacitor will charge up some more.  If you give the tube a very large signal, as we are prone to doing, you can actually charge that capacitor up to something like 100V.  Now if that capacitor is rated at 25V, it generally explodes in a cloud of what is undoubtedly carcinogenic dust.

[jjasilli]

But that wasn't the case in the schematic, it was either resistor, or capacitor or nfb, so in the case of the NFB I think that would be the 'off' position, but the schematic was wrong, the layout was right.  If you just have a cap, then it can conduct the signal to ground, via the cap, as well as via the anode but you hve no resistance and that would cause bias problems, or so I think anyway.  It was drawn as a three way toggle between the above three options instead of a three way toggle between resistor and cap, and resistor and nfb. and just resistor
~Phil

Yes, but I was not addressing the schematic, which was addressed by others.  I was specifically responding only to the singular point I quoted, which point was pompeiisneaks' faulty premise about cathode resistors.


Moreover, if there's only a cathode bypass cap but no cathode resistor, I submit that the tube cannot operate.  Because a cap blocks DC, B+ current cannot flow through the tube; whereby the tube will be Off.

[robrob]

You guys are right, the schematic is wrong. I'll fix it when I get home. The layout is correct with the cathode resistor always in circuit.

[PRR]

.

[2deaf]

Moreover, if there's only a cathode bypass cap but no cathode resistor, I submit that the tube cannot operate. 

Sure it can, just not in a linear fashion.

Because a cap blocks DC, B+ current cannot flow through the tube; whereby the tube will be Off.

A capacitor will block static DC, although it appears to pass DC when the circuit is first completed.  It appears to pass DC when presented with a dynamic DC circuit, also.  The triode presents a dynamic DC circuit to the cathode capacitor.  The complicating factor is that the triode will cease to conduct when the cathode is sufficiently more positive than the grid.  But if you then give the grid a positive signal that exceeds that threshold, current will flow through the triode until the capacitor charges up enough to cut off that current.

You can turn the tube off completely by giving it a very large signal so that the cathode charges up very positive then suddenly reduce the signal so that the positive portion is smaller than what is necessary to overcome the cathode-to-grid voltage.  If you leave the input signal at a constant magnitude, the positive portion always causes a change in the triode current and the consequential inverted signal at the anode.         

[jjasilli]

I'm not getting this, nor can I find any authorities on the topic of a cathode cap instead of a resistor.  Do we agree that a diode cannot operate if its cathode is "connected" to ground through a cap only?.  If so, then it seems that a triode cannot operate for the same reasons.  Can you refer us to authorities that say otherwise?


BTW: I don't doubt that the tube might make a noise at turn-on.  I'm talking about actual operation.

[robrob]

Tony Bones schematic above is correct. It's a three position switch with the middle position cathode resistor only.

[SILVERGUN]

Tony Bones schematic above is correct. It's a three position switch with the middle position cathode resistor only.

I would consider PRR's version to be 'more' correct because it shows that the switch has 3 contact points (3 position)

1 dot difference 

[sluckey]

The switch is a simple SPDT with center off. There are only three contacts on the switch. Tony Bones drawing is correct. PRR's drawing is electrically correct but still a bit confusing. It shows the switch wiper (common) jumpered to a non-existent center contact. The reality is the wiper ***IS*** the center contact.

I think all the confusion has been because the schematic shows four contacts on the switch.

[tubeswell]

PRR's schematic gets a further option of 1k5 unbypassed with no NFB.

[sluckey]

PRR's schematic gets a further option of 1k5 unbypassed with no NFB.

So does Tony Bone's drawing and Rob's layout.

[pompeiisneaks]

But that wasn't the case in the schematic, it was either resistor, or capacitor or nfb, so in the case of the NFB I think that would be the 'off' position, but the schematic was wrong, the layout was right.  If you just have a cap, then it can conduct the signal to ground, via the cap, as well as via the anode but you hve no resistance and that would cause bias problems, or so I think anyway.  It was drawn as a three way toggle between the above three options instead of a three way toggle between resistor and cap, and resistor and nfb. and just resistor
~Phil

Yes, but I was not addressing the schematic, which was addressed by others.  I was specifically responding only to the singular point I quoted, which point was pompeiisneaks' faulty premise about cathode resistors.




Moreover, if there's only a cathode bypass cap but no cathode resistor, I submit that the tube cannot operate.  Because a cap blocks DC, B+ current cannot flow through the tube; whereby the tube will be Off.

Yeah I get it now, that makes sense, I was thinking the ac/guitar signal, but you're right, it needs to conduct DC to actually function, so it would blow up like mentioned earlier, thanks for setting me straight :)   BTW I'm pompeiisneaks in both of the replies you're referring to, unless you like referring to me in third person when talking to me lol

~Phil

[Tony Bones]

PRR's schematic gets a further option of 1k5 unbypassed with no NFB.

I just tried to edit Rob's schematic to make it what I think he intended it to look like in the first place. There's a note that the switch is 3 position, on-off-on. Even thought the diagram doesn't clearly show a center off position, it's sort of suggested by the fact that the lever is in a centered position, not connected to either the cap or NFB.

In any case, all of this is discussed in the text around the diagram on Rob's website, so anyone that reads it will not be confused.

BTW: I think this is one of the few really worthwhile mods to a 5E3 clone.

[jjasilli]

@pompeiisneaks:  BTW I'm pompeiisneaks in both of the replies you're referring to, unless you like referring to me in third person when talking to me lol
That reply of mine was to 2deaf.  I think I'm going to use the @ symbol more often; sorry for the confusion!

[pompeiisneaks]

I'm dizzy anyway :)

~Phil

[ululufut]

You guys are right, the schematic is wrong. I'll fix it when I get home. The layout is correct with the cathode resistor always in circuit.

Rob - I recently did the 3-way negative feedback switch with help from an engineer partner - I think we may have followed the schematic and not the wiring diagram. I sold the amp to a friend of mine out of state, and he has had a series of issues, starting with bad distortion/cutting out, and now with smooth but drastic volume fluctuations that occur at a single volume setting.

Is there any chance the issues could be related to the schematic discrepancy? Unfortunately my engineer partner is in Boston, I'm in Chicago, and my friend the buyer is in L.A. I'm just wondering if I should recommend that he take it to a local tech and make the change to the switching circuit if necessary. I have limited troubleshooting skills, so actually the prospect that this issue might in some way be related would actually be welcome news to me, because I'm running out of ideas!

By the way, I'm relatively new to this forum, and it's great to see you're here. Your mods are terrific - we did several of them! Thanks for providing all that great information.

[robrob]

ululufut, I know I'm very late answering this but no, the schematic discrepancy shouldn't cause your symptoms. It would cause the amp to not function (be silent) in one of the switch positions.

Your symptoms could be caused by a dying tube, a cold solder joint or many other issues. I'd start with trying a known good tube in every position and then chopsticking the amp.

[Diverted]

This is a great mod that I also found at Rob Robinette's invaluable site. I have put it in most of the similar amps I've built over the past few years and it really does make it much, much more versatile.
Here's a schematic for one I just built; I pulled the MV, feedback and reverb circuit mod from Rob's site as well. Though I think I drew the switch schematic wrong too