A EE friend of mine and I are planning to build a switch box that will allow me to toggle between my Yaqin MC-30L tube amp and my Integra 50.1 receiver to power my speakers.
The current plan is a very basic setup that would require the tube amplifier to be shut down before switching so that it isn't left without a load.
The utopian goal would be to be able to switch between the two amplifiers with a single source feeding both (i.e. the Yaqin has an Apple Express for a source and the Integra has an Apple TV so both could be running from iTunes on my Mac) without risk of damage to the tube amplifier.
I'm assuming that this would be complicated at best and probably isn't worth it, but thought I should as before we purchase parts.
It's actually very simple, just have a load resistor, of sufficient wattage, subbed for the speakers when you flip the switch. That way the amps can always see a load when they are not attched to the speakers and there is no chance for accidents happening.
As far as the source goes, you should have a look at the output impedance of the source and the input impedance of the paralleled amp inputs. If it is 10-20X you should be good to go. If not, all is not lost, you can use an output buffer at the output of the source and this should present a high impedance to the source and a sufficiently low output impedance to easily drive the paralleled inputs of the two amps.
Keep in mind that, with some amps, when they are turned off, the input impedance is significantly lowered, so I would do my experiment with one amp off to see if you need the output buffer or not. But they are cheap and fairly transparent so you should have no issues.
My Yaqin MC-30L is rated for about 50 watts per channel is ultralinear mode. So using a 100 watt load resistor should do the trick, right?
I'll clarify that when using my Mac and iTunes as the source, the source of the Yaqin is really a DAC fed by an Apple Express and the source for the Integra is an Apple TV. iTunes just has the option to send the music to more than one zone. The switch would only swap the speakers leaving both amplifiers with a source signal at all times.
I think we have the details worked out on the switch circuit. We'll be using an 8 Ohm 100W Non-Inductive Dummy Load Resistor from Part Express to load the tube amplifier when the speakers are switched to the receiver. I just need to measure the space in the cabinet so that we can get the right sized project box. I think all internal wiring will be 10 AWG to match my speaker cables.
"Should I be concerned with the tube amp not being loaded for a instant when the switch is toggled?" That depends, even though IF the tube amp has some sort of protection circuit built in, I won't risk it. Same thing on the ss amp, even though it has short protection, I wouldn't risk it either.
Imhififan - Your responses and link to the other thread made me dig into things a bit more. Thank you.
The revised plan is to use a heavy duty 4PDT toggle for each channel. The solid state receiver will be on/off, the tube amplifier will be off/on, and the 8 ohm 100W non-inductive dummy load resistor will be on/off. To bridge the instant between the speakers and the dummy load resistor on the tube amplifier we will also add 150 ohm 5W or 10W resistors on each channel based on the recommendation from the Niles manual that was linked. I'm not worried about my receiver as I often have it passing sound to the TV via HDMI with the speakers disconnected and hooked to the tube amplifier.
I'm interested to hear how the final switch box "sounds" as it will obviously add components to the signal path. Being completely objective may be a challenge in contrast to the negative placebo of knowing that the signal path has been compromised.
If I were doing it, I would use a make and then break switch that would put the two loads in parallel and use a load resistor of much higher impedance then you are using so that in parallel with the speaker load it would still be within the acceptable load range for the amp.
I agree that the load resistor could be made significantly higher than 8 ohms, which would allow a somewhat lower wattage resistor to be used. I recall reading somewhere that as a conservative rule of thumb guideline 32 ohms or thereabouts would be low enough to protect most or all tube amps from the possibility of damage, even if they are operated for considerable periods of time with signals going through them. (The possibility of damage from running with too light a load or no load is greatly reduced if no signal is being processed by the amp, although various scenarios can be envisioned in which a brief transient "signal" can be put into an amp's output stage even when no music is being played).
However I see several potential issues with using a make-before-break switch.
First, when the switch is being thrown there will be a brief instant in which each amp will be loaded by the parallel combination of the speaker impedance, the resistive load, and most significantly by the output impedance of the other amp (that being very close to zero ohms in the case of the solid state amp, and perhaps 1 or 2 ohms in the case of the tube amp). Obviously not a good idea, especially if signals are going through the amps. At the very least, it wouldn't surprise me if self-protection mechanisms in some amps would end up being triggered as a result.
Second, if the solid state amp has a small amount of DC offset in its output, say something like 20 millivolts, which I believe would not be uncommon in a properly functioning solid state amp, that would be applied to the secondary of the tube amp's output transformer for a brief instant when the switch is being thrown. That would cause a brief transient to appear on the primary side of the transformer, whose amplitude might very briefly be perhaps several hundred millivolts as a result of being stepped up by the turns ratio of the transformer.
If the tube amp is not powered up at the time, the energy of that transient would not have any place to go, since the output tubes would not be functioning. The result potentially being a large voltage spike due to inductive kickback, which is exactly the potential damage mechanism the dummy load is intended to prevent.
If the tube amp is powered up at the time that transient occurs, it would be introduced into the amp's feedback loop (if it has one), and offhand I'm not entirely sure that would be healthy under all circumstances either.
Almarq - I was hopeful that you would provide input as I value your opinion. The revised plan is to use the dummy load resistor (a parts express item) on both inputs and also put 150 ohm resistors on both inputs to make it a fail safe design for a tube amplifier. I'd rather not have to keep track of the tube safe input. Each channel will have its own switch because it would take at least a 6P or 8P switch to do what we want.
Marty, note this statement in one of Mceljo's earlier posts:
To bridge the instant between the speakers and the dummy load resistor on the tube amplifier we will also add 150 ohm 5W or 10W resistors on each channel based on the recommendation from the Niles manual that was linked.
And in his most recent post:
The revised plan is to use the dummy load resistor (a parts express item) on both inputs and also put 150 ohm resistors on both inputs to make it a fail safe design for a tube amplifier.
So as I understand the plan neither amp will ever see a load that is lighter than 150 ohms, even for an instant. That seems reasonable to me, although I suppose it would be good practice to not throw the switch when music is playing (since the possibility of damage resulting from having a load that is too light is very small if no signal is present).
My current solution is to have two sets of cables at my speakers that I manually switch depending on the amplifier that I want to use. To this point the tube amplifier has always been shut off (though not always at room temperature) before disconnecting the speakers. The receiver has been used to pass the audio to the TV via HDMI without the speakers connected many times.
My ultimate goal from a fun perspective is to be able to do live A/B switching as any other delay makes it virtually impossible to accurately "remember" what it sound like for comparison. It will likely be a gimmick type of event at first that eventually becomes rare. I
The long-term improvement from a practical standpoint will be that I can have the tube amplifier on for listening, switch to the receiver for watching something with my wife, and then go directly back to the tube amplifier without having to turn it off. As a general rule I don't turn the amp back on until it's cool and because the sound is much better with everything warmed up short power cycled listening just isn't the same.
To be able to do live A/B switching, this is an idea: A Shorting Contact Style 3 Deck 3 Position Rotary Switch, Deck 1: pole connect to speaker +, position A to AVR + output, position B open, position C to Tube Amp + output Deck 2: pole connect to speaker -, position A to AVR - output, position B open, position C to Tube Amp - output Deck 3: pole connect to 8ohm 100w Dummy Load Resistor to Tube Amp - output, position A & B jumpered and connect to Tube Amp + output, position C open Hope this help
To be sure it's clear, I would reword the description of the Deck 3 connections as follows:
Deck 3: pole connect to 8ohm 100w Dummy Load Resistor; other side of that resistor to Tube Amp - output; position A & B jumpered and connected to Tube Amp + output; position C open.
Also, if this approach is pursued it might be desirable to use something like a 16 ohm 50W resistor, or even a 32 ohm 25W resistor, rather than an 8 ohm 100W resistor. The reason being that when the switch is thrown between positions B and C there will be a brief instant when the tube amp will see the loading of both the speaker and the resistor. That may not be a problem anyway, but I can envision the possibility of a significant thump occurring with an 8 ohm resistor, especially if the amp uses feedback.
Also, Mceljo, as opposed to the approach you described earlier, with this approach you would of course have to provide markings or some other means of identifying which specific terminals each specific amp should be connected to.
Marty (Viridian), thanks for the nice words and the good wishes, which I reciprocate.
This sounds like a great alternate solution that someone else may choose to use. I have some of the parts in hand and almost everything else is in the mail. The only missing piece is some 10awg cable from Blue Jeans Cable as the stuff that we got from Parts Express didn't turn out to be what we wanted.