• Short answer --- - Paul Joppa 11:13:00 09/23/02 (0)
  In Reply to: Doc, PJ . . . posted by eedork on September 23, 2002 at 06:57:25:

  We have indeed discussed and researched this question. Here's the short answer, why we have not tried to make it a product:

  1) It's not that big a deal. With cathode bias resistors, there is no possibility of extreme currents, DHT filaments heat very quickly, and the loss of lifetime is just not that great. You are more likely to kill a tube by thermal and mechanical stresses, leading to an open-circuited filament. Relatively few tubes actually die by loss of emission, too many other things go wrong earlier.

  2) It's very difficult to do it right, that is both safely and reliably. If you have VALVE back issues, look at the "Buddhafied Afterglow" notes for some more information. Really, the output transformer should also be protected during startup and shutdown transients, as long as you are going to have a sequencer of some sort.

  3) The Bottlehead amps have always been as simple as possible, to keep the costs down and so that they are easy to experiment with. You pay a big price for complexity -- sonically, economically, and in reliability. Every so often, we fantasize about all the things that would be nice to have in a particular product, and it always adds up to a vastly higher price. Heck, nobody even wants a Foreplay with DC filaments, if it costs more!
  
  

  ---
• Re: Doc, PJ . . . - eedork 09:02:59 09/23/02 (12)
  In Reply to: Doc, PJ . . . posted by eedork on September 23, 2002 at 06:57:25:

  I tried to reply to someone that posted in reference to this thread but both my reply and original message I was replying to got deleted. Weird.

  To whomever asked about standby switches, please post your question again and I'll respond.

  Thanks,
  -Matt
  
  

  ---

  • Curious... - csown 10:30:01 09/23/02 (0)
    In Reply to: Re: Doc, PJ . . . posted by eedork on September 23, 2002 at 09:02:59:

    This just occurred to me - I'm also curious what the effect of repeatedly applying and then de-applying B+ to a circuit would be. Not many have this power so it is not documented. I remember reading some accounts of why it's bad NOT to have B+ up when heater is up (discounting the turn-on cathode stripping case) for reasons that I don't recall right now. I think it was something to do with material evaporation onto the electron source, decreasing efficiency of the thoriated surface. I would be interested in seeing if there is some other of stress/fatigue effects when one repeatedly cuts the B+ instantaneously, whether it be several times a day or rapidly in succession.

    -Chris
    
    

    ---
  • 'Twas me... - csown 10:10:24 09/23/02 (10)
    In Reply to: Re: Doc, PJ . . . posted by eedork on September 23, 2002 at 09:02:59:

    Hi Matt,

    I did, but I noticed your ( thread below and that sort of answered my question (hence I deleted it until I gathered more information). I thought a switch would survive without serious arcing below 200V or so, hence I was not particularly worried. But the idea of failure with the full B+ exposed to the switch handle (and not the switch housing, which should be grounded) is scary. I hadn't thought of that, and there's not much you can do to prevent a nasty shock then. I'm intrigued by the discussion on the fender standby switches, but the bottom line advantages/disadvantages weren't really well-elucidated so I'm still evaluating that idea. A thermistor is a rather elegant solution, but low currents are hard to implement this with, and there's a fair bit of magic involved in getting a proper thermistor delay implmeentation so it's out for DIY.

    Anyway, I'm interested in an elegant <5 components solution that could be implemented after the PSU filtering. I suppose that keeping the supply charged is a ticking time bomb with a high-voltage, higher-powered amplifier. However, inrush current across the switch is limited if the circuit consumes little current. Perhaps a switch with wiping contacts would be better to clean off that arc residue every time you actuate it (AKA - high-voltage Frankenstein knife switches complete with a shower of sparks!).

    -Chris
    
    

    ---

    • it is an interesting problem - eedork 11:43:30 09/23/02 (9)
      In Reply to: 'Twas me... posted by csown on September 23, 2002 at 10:10:24:

      I've been thinking about it for a long time and there is no easy solution. I may just buy the boards from Alexander Rubli and be done with it.

      -Matt

      
      
      

      ---

      • What if you reconsider... - csown 15:03:15 09/23/02 (8)
        In Reply to: it is an interesting problem posted by eedork on September 23, 2002 at 11:43:30:

        Hi Matt,

        I just read this post and am wondering if maybe everybody is being toooo safe, perhaps it's a safety warning that has gotten blown somewhat out of proportion. If you think about it, a great number of amps in the past had standby switches - as we all know, this is something that would be very useful. Consider the failure modes:

        1. The primary cause of switch failure is arcing. Part of the contact becomes covered with oxide with each successive switch and eventually the switch fails. As effective contact area decreases, more and more current passes through a smaller area, and heating and more arcing accelerates the oxide formation until finally the switch dies.
        Consequence: dead switch.
        Remedy: you have to go in and replace it (pain in the ass).

        2. See page 4 of a generic toggle switch datasheet . If the polymer wedge-shaped "nub" connected to the handle that provides the rocking action becomes damaged, there is the possibilty that high voltage will reach the metal handle. Switch designers design against this, making it nearly impossible for a short except in the case of a freak accident (i.e., heavy mechanical use, breaking off the handle and shoving it inside the switch body. The question is - how easily can this happen in a well-designed switch. Has anybody had this happen to them? For double-pole or more, I can't see how this can happen.
        Consequence: Switch obviously doesn't work. It might be always-on, or might not work at all. VERY low chance of shorting with a contact except for freak accident.
        Remedy: replace the switch, or, (in case of death due to electric shock) none.

        3. Rocker damage, i.e., the see-saw becomes bent and the switch becomes permanently on or is permanently broken. This is rare, but some poorly-built switches can do this.
        Consequence: Dead or always-on switch.
        Remedy: Replace or just leave it on all the time.

        
        So I think the biggest failure in our application is due to 1 - contact bounce/arcing, which is deadly to the switch but not-so-deadly to the operator. These switches are designed for safety, and you may die from touching an AC-shorted switch as well as a high DC voltage. The deadliness of the DC depends on the voltage and capacitance involved. I don't see how the switch can become shorted to chassis except in rare instances of case #2. More often one would have B+ on the switch handle only, which would be nearly never at all. If you used only plastic rocker switches, which do not have metal handles, or a good dielectric cover, you ought to be safe.

        So as long as one can remove that arcing with, i.e., a good snubber like that described in the post I mentioned, PLUS, use a nylon switch that has got a pretty good voltage rating, I figure you should be OK.

        As a mechanical/materials engineer, this is my take on the failure modes of a typical see-saw action switch. Other more seasoned EE folks please give me input on this. True, it is a safety issue, but we all know that sometimes what the risk management folks say can be a bit exaggerated. Everybody uses switches. While the vast majority of higher voltage applications are used to switch AC due to common-sensical safety, proper implementation of a high voltage DC switch would yield considerable convenience. I think it would be useful to contact a switch manufacturer to get more information on failure modes for a more conclusive picture of this risk.

        -Chris
        
        

        ---

        • Re: What if you reconsider... - Jim Carlon 07:51:38 09/24/02 (4)
          In Reply to: What if you reconsider... posted by csown on September 23, 2002 at 15:03:15:

          Yikes! The cap and resistor mentioned in my post should be connected in SERIES across the switch contacts!

          For what its worth, in 6 years of servicing Fender and other guitar amps, I have never seen a 'standby' switch failure, other than cases of actual physical damage. This was usually damage to the toggle/handle. As far as failure due to arcing, I simply have not seen it happen, and indeed, this is on equipment that has been in service for 30-40+ years.

          While Fender did not use a snubber across the contacts, the switch was in a rather good spot in the B+ line, after the first cap and before the choke, in a CLCRCRC filter. I imagine that the L and Rs relieved a good deal of inrush stress to the switch. As I have mentioned previously, B+ was typically in the range of 465-485 Vdc at this point, and the switch was a common Carling SPST toggle rated at 250 Vac, 7A. A similar switch was used for AC mains.

          A great deal has been written about cathode stripping and cathode poisoning over the years, but I must say that I have really never seen wany empirical data, ie: stripping reduces useful life by x hours, poisoning causes failure in y hours, etc. As far as guitar amps are concerned, I don't think that this was even considered---remember, this was called a 'Standby' switch, and its intended purpose was to provide a quiet-but-warmed-up state for times when the musician was not playing. As such, the amp may have been left in this state for a 20 minute break, or for an hour before the show.

          Personally, in a new design, I very much like the idea of using a damper diode tube, ala Bluesmaster, for a simple, fool-proof slow turn-on. Nothing to remember for the casual user, no mechanical or SS devices to fail, and Svetlana still makes one.

          If it is an existing piece, I still think a good quality switch is the simplest solution.

          Jim
          
          

          ---

          • Re: What if you reconsider... - csown 09:14:24 09/24/02 (3)
            In Reply to: Re: What if you reconsider... posted by Jim Carlon on September 24, 2002 at 07:51:38:

            Hi Jim,

            
            Thanks for your reply I am getting sort of crazy about this because we just had a vacuum pump power switch failure (due to arcing) in my lab and that got me thinking about how these things are operated and under what conditions they may fail. This is a high amperage switch and was used to stop/start the hefty motor which no doubt draws amps and amps at turn-on. EVERY SINGLE contact attached to it was oxidized/charred and the inside of the switch was toast. Not sure how it happened - whether it was oxidation over time, a short, or what. But the failure was completely safe - the switch innards were enclosed, the chassis was grounded, and the switch was nylon with a rubber cover.

            All this has come up because I wanted to find a way to place a standby switch on a Foreplay I'm working on. Since it has an external supply that I'd like to have hidden away, it's been difficult to figure out a way to implement this without adding something like a relay back inside the PSU with the switch action running all the way back through the umbilical. I've been playing with some ideas and the cleanest would require yet another power supply just for the relay, so it's been frustrating, and there's the issue of waiting for the B+ to energize. The switch was an afterthought after I had designed and built most of it, so I guess one could say that this is an existing piece. Furthermore, a fender clone is in the works for my girlfriend's brother, and he is rough with equipment so I want it to be safe!

            > Yikes! The cap and resistor mentioned in my post should be connected in SERIES across the switch contacts!

            OK I am getting confused here. When the switch is open, should it be OPEN OPEN, or should there be a bleeder resistor and cap across it? Do you have a diagram? I've seen conflicting recommendations. Please clarify as to what the goal of the RC network is.

            > A great deal has been written about cathode stripping and cathode poisoning over the years, but I must say that I have really never seen wany empirical data, ie: stripping reduces useful life by x hours, poisoning causes failure in y hours, etc. As far as guitar amps are concerned, I don't think that this was even considered---remember, this was called a 'Standby' switch, and its intended purpose was to provide a quiet-but-warmed-up state for times when the musician was not playing. As such, the amp may have been left in this state for a 20 minute break, or for an hour before the show.

            Yes, this standby business may have been taken a little far, and is probably now somewhat an institution in the audiophile nervosa canon. It is an interesting problem in and of itself, however. It does take a good amount of time for warm-up (in the sense of power supply stabilization) and the incentive to turn tube amps off and then on again day in and out is far greater than that of SS devices. Hence I think it is still a worthwhile cause, even if there is marginal benefit, as there MUST be (or have been) cases in the real world where you can't wait for your supply to charge up to the full voltage, and you need full high voltage instantaneously. How do high voltage test supplies do it? I'm too young to have used one :).

            I am glad you mentioned the 7A rating for the Fenders. I was looking for this and couldn't find it. That is a good deal larger than most mini-switches available nowadays.

            The difference between the "after-filtering", "between-filtering", and "before-filtering" methodology is great in terms of the current the switch must pass, but I think I've seen some where the switch is after filtering (not saying I understood all the circuits completely, but I'm fairly certain I saw more than one). Off the top of my head, see Super Twin Reverb for example. The standby switch (010703) switch is located after the two 220u caps and switches the full 500V and all the current they provide. There is no "soft switch" RC circuit at all. Instantaneously, the 20u RCRC supply demands current, as does the rest of the circuit.

            For a circuit that requires very low current such as the Foreplay, I would expect it to actually be safer for switch and diode longevity (in terms of current inrush) to put the switch after the filtering rather than after. In terms of safety, you run higher risk of electric shock.

            Anyway, if it was above 300V, I wouldn't even be considering this. It's kind of a gray area I guess. The Foreplay uses "low" voltage for a tube and is less likely to arc nastily, so I kinda like the idea of soft-actuation via a snubber of sorts. Nevertheless, any voltage with a reservoir of current available is potentially lethal. I think I'm going to run this by some switch manufacturers to see what they think.

            Thanks,

            -Chris
            
            

            ---

            • Re: What if you reconsider... - Jim Carlon 12:28:30 09/24/02 (2)
              In Reply to: Re: What if you reconsider... posted by csown on September 24, 2002 at 09:14:24:

              Interesting tale about the pump. Switches that operate heavy-duty motors get a double dose, because the collapsing magnetic field that occurs when the motor is shut off can throw quite a spike of its own.

              I believe it was Quest who recommended a minimum of 250 Vac at 10 A rating for a suitable B+ switch, and I think that this is a very wise recommendaion.

              For the contact smubber, the cap and resistor are wired in series, so one switch contact will have a free cap lead connected to it, while the other contact will have a free resistor lead attached. Remember that this is for B+, so when the switch is open, the cap will block DC trying to bypass it.

              I prefer to use metallized film caps in this service for their self-healing properties, and metal oxide resistors for ruggedness. If this is for Foreplay B+, one could easily get by with a 1 Watt resistor, as the current is teeny.

              I've really never seen one of these fail in such a way as to energize an exposed part of the switch, though I suppose it is theoretically possible. A switch with a plastic handle or a plastic cover over a metal toggle would be another margin of safety.

              All in all, though, I have to say that nothing succeeds like success. If a 40 year old piece of gear is still operating correctly and safely with its original metal toggle, I have no reason to doubt that a Foreply B+ switch will do the same.

              Jim
              
              

              ---

              • A very useful/interesting article... - csown 09:08:07 09/25/02 (0)
                In Reply to: Re: What if you reconsider... posted by Jim Carlon on September 24, 2002 at 12:28:30:

                See pages 40-48 and also page 53.

                I like this pessimistic passage:

                "Unless the switch is specially designed to withstand closure on a short circuit, the switch life can be predicted to be zero."
                

                • http://content.honeywell.com/sensing/prodinfo/basicswitches/technical/010172.pdf

                
                

                ---
              • Is this what you mean? - csown 07:39:31 09/25/02 (0)
                In Reply to: Re: What if you reconsider... posted by Jim Carlon on September 24, 2002 at 12:28:30:

                Requires an SPDT:

                
                ` +------0 <- B+ in
                ` |
                ` o
                ` | <- switch
                `o o
                `| |
                `R C
                `| |
                `+-+
                ` |
                ` +------0-> load
                

                A benefit to making R slightly inductive?
                
                
                

                ---
        • Re: What if you reconsider... - Paul Joppa 18:35:40 09/23/02 (1)
          In Reply to: What if you reconsider... posted by csown on September 23, 2002 at 15:03:15:

          Ott ("Noise Reduction Techniques in Electronic Systems") has a good discussion of contact protection. There are two discharge mechanisms, glow and arc. Here are some quotes:

          "To avoid a glow discharge, the voltage across the contacts should be kept below 300v." (This is with an air dielectric. Buddha had a story of a misfiring tank cannon due to this. He solved it with a vacuum relay. You really do not want your tank cannon going off unexpectedly ... -pj)

          "A set of contacts can normally handle a much higher ac than dc voltage ... a contact rated at 30vDC can, therefore, typically handle 115vAC ... contacts are normally rated by the maximum values of voltage and current they can handle feeding a resistive load ... if the load is not resistive, the contacts must be appropriately derated or protected" (Think of the inrush current of power supply capacitors, for example. -pj)

          I agree with Matt, I do not like to use devices outside their ratings.
          
          

          ---

          • :) - csown 09:18:05 09/24/02 (0)
            In Reply to: Re: What if you reconsider... posted by Paul Joppa on September 23, 2002 at 18:35:40:

            That tank scenario is hilarious (in a twisted way). The tank firing circuit depends on one measly little switch - that is the line between life and death for many. Will it discharge or not? That is a very real and sobering example of the russian roulette we play when we introduce a switch.

            -Chris
            
            

            ---
        • good analysis and points . . . - eedork 16:49:19 09/23/02 (0)
          In Reply to: What if you reconsider... posted by csown on September 23, 2002 at 15:03:15:

          you may be right, but I still feel odd using a component outside of its rated specs. I'm torn.

          -Matt
          
          

          ---

This post is made possible by the generous support of people like you and our sponsors:

Schiit Audio

---

[ Top of Page ] [ Contact Us ] [ Support/Wish List ] [ Copyright © 2025, Audio Asylum, all rights reserved ]

---

[ General ] [ Speakers ] [ Tubes ] [ Vinyl ] [ Digital ] [ Hi-Rez ] [ Video Asylum ] [ Cables ] [ Tweaks/DIY ] [ Music ] [ Films ]

---