Announcement

Collapse
No announcement yet.

Amp switcher

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • Amp switcher

    I'm getting to the point in my amp building and life where I'd like to have an amp switcher, partly for fun while playing and partly to compare amps in a hurry. I recall reading threads about it before where people claim it's a bad idea and yet when I lived in Nashville I saw them in studios- a rack full of amps, all fired up and running and a rack unit to switch them on the fly to a single cab in an iso room as needed.

    I figured I'd use RG's hacked 74c373 circuit as a "one of x" selector and power low and high power relays to select the inputs and outputs respectively. Large wirewound resistors would be wired across the returns from the amp power sections to help prevent major inductive kickback issues during switching; hopefully they'd be large & high enough in value to not get particularly hot while playing. I figured this way I could have a fender into a 4x12 for clean stuff and with a single button press pull up a marshall into a hotplate or a DC30 or something. It could be integrated into a pedal switcher without too much effort. A microcontroller could handle both without breaking a sweat.

    It sounds fun and easy but I might be missing something.

    Jamie

  • #2
    Try searching the Aiken Amplifier site via the way back machine, there used to be some designs for amp and speaker switchers.

    Comment


    • #3
      I've thought of doing something similar a few times. The 74C373 will work if you can find the chips, but they're getting hard to find. 74HC and 74HCT chips in the 373 or 393 flavors work, but you have to tinker the RC networks on them to get the latch enable to stay up long enough to latch in the data. No biggie, just a bit fussy.

      Your approach is good - get a clean logic signal that says "this amp on or not" unambiguously to each amp, then use the relays to satisfy the amps' needs.

      The complications on the amps all center around keeping a load on them, and on ground-hum issues. The grounds need to be separated unless you've pre-tested and solved any issues so the amps play well with their grounds connected. For random amps, you need to interrupt the grounds, never connecting two amp-grounds together.

      If you can get the amps to work well with a modest-value resistor, maybe 200 ohms or so, that's great; if not, use your relays to switch a load onto each deselected amp.

      It's worth considering that the "relay" connecting a load to each deselected amp could be made up of two MOSFETs connected source to source, appropriate gate drives. You're not listening to the deselected amps, so if the MOSFET relays introduce some distortion into the silence, so be it.

      And yeah, if you can stand doing the programming, the uC lets you do advanced things like a "mute all" while the switching takes place, and millisecond-stepped switching if you need sequencing.
      Amazing!! Who would ever have guessed that someone who villified the evil rich people would begin happily accepting their millions in speaking fees!

      Oh, wait! That sounds familiar, somehow.

      Comment


      • #4
        Thanks for chiming in, RG. I work for a company that ended up with vast quantities of military contract return 74 series chips so I long ago kludged your circuit to work with much faster IC's- I used little RC networks with a smaller cap on the reset pin than on the set pins so that it always falls first. Works like a charm. As long as the 74 is in the box with the footswitches I don't even bother with the hex inverters on the input and I've seen no failures yet.

        With relays it should be possible to wire a switcher such that your pedalboard/guitar are only connected to one amp at a time- switch the grounds and the signal. This keeps the user safe (all amps are still AC grounded) and cures and ground loop issues by only connecting a single amp at a time.

        I, too, had expected around 200 ohms would be a good value and readily available. Perhaps use a lower value and I could just leave the amp outputs floating (not grounded) into the resistor when the amp isn't selected. With the input of the amp grounded I suspect a tube amp would play nicely. I don't have any solid state amps so I'm not worried about how they would behave (and how their output filter wouldn't like such switching).

        Jamie
        Last edited by imaradiostar; 10-15-2013, 04:02 PM. Reason: /in be

        Comment


        • #5
          Good. I get a lot of mail from people who ignore the RC timing stuff.

          Solid state amps are usually not concerned with any larger loading than they're normally connected to. Most of them only have "output filters" to protect against abusively capactive loads. You can open them with impunity for almost all. Don't know about Class D, which may have more manly filters.
          Amazing!! Who would ever have guessed that someone who villified the evil rich people would begin happily accepting their millions in speaking fees!

          Oh, wait! That sounds familiar, somehow.

          Comment


          • #6
            Ted Weber used to make the FourHead switcher...and I think they still make one that does a couple amps at a time. I happen to have the last FourHead they made and I sent it back to TA a year ago or thereabouts so they could blueprint it in case they wanted to make more. I have no idea how they do the switching in these but it works well though you can't instantly switch two amps at once. You have to switch one off then the other on with two different switches.

            Greg

            Comment


            • #7
              The inability to switch on the fly seems like a real shame. Otherwise it sounds like a neat box. I wonder if Ted himself designed it?

              jamie

              Comment


              • #8
                Originally posted by imaradiostar View Post
                The inability to switch on the fly seems like a real shame. Otherwise it sounds like a neat box. I wonder if Ted himself designed it?

                jamie
                I believe Ted did design it. You could ask TA about it. I think they have one now that can switch up to two amps at a time, but again not in real time with one switch. TA borrowed the FourHead back from me to blueprint it because I got the last one apparently and Ted didn't keep any notes around on it, so if they wanted to make more they would have to re-engineer it or get the info from mine.

                The FourHead works for me because I built an amp that is 3 separate amps in one box...started life as a Conn organ power amp chassis...with 3 separate power amps and transformers and PI's in it. So I modded that, and added a separate preamp chassis with 3 preamps in it. If I want to use it live, I either need to bring 3 cabinets, or loads for two amps and manually switch when I want to use a different one, or use the FourHead through one cabinet. It would be nice to have on the fly switching but it is ok for what I need.

                Greg

                Comment


                • #9
                  Any news on Weber's re-engineered FourHead? Is it just a bunch of relays controlled by a simple logic circuit?

                  Comment


                  • #10
                    What about this:

                    Head Master amp switcher - $279.00 : MetroAmp, Amplifier Kits and Parts Online Store

                    Looks like pretty much what I was suggesting, but only for two amps.

                    Jamie

                    Comment


                    • #11
                      Are these valve amps being switched (with OT's)?

                      I've used source-source AC signal power FET switches - they are great - as the sources are common, the gates can be driven on/off by the same voltage, which I found easiest to do with a small dc/dc isolated switcher (eg. the old Newport LME or NME range).

                      Comment


                      • #12
                        Originally posted by trobbins View Post
                        I've used source-source AC signal power FET switches - they are great - as the sources are common, the gates can be driven on/off by the same voltage, which I found easiest to do with a small dc/dc isolated switcher (eg. the old Newport LME or NME range).
                        I've always shied away from the series-MOSFET switches largely because I've not had time to mess with them. What - if anything - does it do to perceptible sound quality?
                        Amazing!! Who would ever have guessed that someone who villified the evil rich people would begin happily accepting their millions in speaking fees!

                        Oh, wait! That sounds familiar, somehow.

                        Comment


                        • #13
                          Ahh well, I only used FET switches for DC and AC power switching (a controlled load bank) - the FETs were higher current discrete devices. Both FETs are fully enhanced on, eg. its easy to use 12V (or 5V for logic FETs) dc/dc output for Vgs drive, so the switch is effectively just 2x Rds-on, and hence negligible for most FET choices.

                          In my view, the switching would be same as for optomos devices, which I think are used for signal switching sometimes (?). In earlier days these devices were starting to move out of telecom relay use, and now I see Clare with fantastic power optomos to suit anything with speaker ratings (eg. check out CPC range such as CPC1918, although lower blocking voltage devices would need some tvs or similar over-voltage transient protection).

                          Comment


                          • #14
                            I've done the same for AC and DC power. All the evidence I've seen is that you can make a darn good speaker switch out of - as you say - a couple of discrete high current MOSFETs hooked source to source and with the gates enhanced by something floating.

                            I've used LED/Photovoltaic isolators to run isolated MOSFETs. They're designed to do this, and to fully enhance power MOSFETs floating by 1000's of volts. They work GREAT for AC or DC power, but I've never had the time and motivation at the same time to try this with speaker signals.

                            I guess another way to say it is I know it will switch the audio to a speaker on and off much like a relay contact. What I don't know is if one can hear any audio degradation. I believe Steve Connor has tried it and thinks it is OK. I was looking for confirmation.

                            To avoid the price and complication of an isolated power supply, look at:
                            http://www.panasonic-electric-works...._en_an_031.pdf
                            http://www.vishay.com/docs/83469/vom1271t.pdf
                            Avago ASSR-V622 2-Channel Photovoltaic SSR MOSFET Driver. Avago Technologies is a leading supplier of Solid State Relays (SSRs), automotive digital isolators, logic interface optocouplers, and IPM interface optocouplers.
                            http://www.irf.com/product-info/data...a/pvi5013r.pdf
                            Photovoltaic Output Photocouplers - Toshiba Semiconductor & Storage Products Company

                            I used them to simultaneously turn off the positive and negative sides of a high current bipolar power supply to an audio power amp. It worked... GREAT.

                            But I think they'd work for audio switching too, maybe.
                            Amazing!! Who would ever have guessed that someone who villified the evil rich people would begin happily accepting their millions in speaking fees!

                            Oh, wait! That sounds familiar, somehow.

                            Comment


                            • #15
                              Those are certainly neat opto-"PV" drivers (they appear to have come out after I went on to other things), and as you say have huge isolation levels and so much more cost effective than little dc/dc - ahh technology advancement! Apart from having to drive 2x Cgs, I can't see the turn-on transition time through linear operation of Rds to be an issue for this application.

                              On the thread topic, my only concern would be hot switching valve amps to speakers. Either each valve amp needs very good internal over-voltage protection to cope with fast turn-off of speaker load, or the switching mechanism needs to ensure some make-before-break loading on the valve amp prior to disconnection from speaker.

                              Comment

                              Working...
                              X