A-B Testing

[DQ828]

Hi Guys

As a bit of a beginner I'm not sure what I can get away with re connecting things together. I was hoping to do some A-B testing o two amps, using a separate Squeezebox Touch to supply each amp & connecting both amps to one set of speakers at the same time, with the idea of playing music through each amp one at a time. This would save me building a switching device.

If I cannot do that, does anybody have a simple design for a switch?

 

[luka]

I wouldn't drive other amps output with first, I would use switch or relay there

 

[DQ828]

I wouldn't drive other amps output with first, I would use switch or relay there

I was hoping to drive the speakers

Wouldn't electrical relays cause some horrible noises when switching?

I assume I would need two double pole relays and wire one amp out put to the two Normally On side of the two relays & the other amp to the Normally Off side of the relays, you can see I'm a genius at all this. )

 

[luka]

well if you would turn down the volume, then I don't think there would be any noise... if you need to keep constant volumes on both amps, then you will just have to go and switch when you need to... maybe I would also use 10R resistor, for each amp ch to drive when it doesn't have any speaker on

Maybe I am going to far with this, but..

 

[sampleaccurate]

Hi Guys

As a bit of a beginner I'm not sure what I can get away with re connecting things together. I was hoping to do some A-B testing o two amps, using a separate Squeezebox Touch to supply each amp & connecting both amps to one set of speakers at the same time, with the idea of playing music through each amp one at a time. This would save me building a switching device.

If I cannot do that, does anybody have a simple design for a switch?

You can't parallel (connect together) the outputs of power amps - you'll probably damage one or both amps and it will sound horrible at best.

Why not just purchase a cheap speaker selector switch? That will allow only one amp at a time to be connected to the speakers. If any of the amps are tube amps you'll need to make sure you either place a resistor across the output terminals of the amp (20 ohm would probably not interfere with the sound but would prevent over voltage) or just be careful to keep the volume at zero until the speakers are connected.

 

[sprinter]

I also would like to A/B two Tripath amps on one set of speakers, but as I understand it, Tripath amps cannot share the right and left negative terminals, as most other amps typically do. I have been looking but have yet to find a speaker selection switch that does not connect the negatives together to save space and cost.

Anybody have a solution to that problem? I'm about to build a switch of my own with a 4-pole DT switch of some kind.

From an eBay site selling the Lepai's:

The machine is BTL output, that is to negative side of speaker out is not public grounded (the two speakers cable must independent connect, no grounding together), we must strictly in accordance with the marking of plus or minus cable connection.

The TA2020 is a 20W/ch continuous average two-channel Class-T Digital Audio Power Amplifier IC using Tripath’s proprietary Digital Power Processing™ technology. Class-T amplifiers offer both the audio fidelity of Class-AB and the power efficiency of Class-D amplifiers.​

​

 

[DQ828]

I also would like to A/B two Tripath amps on one set of speakers, but as I understand it, Tripath amps cannot share the right and left negative terminals, as most other amps typically do. I have been looking but have yet to find a speaker selection switch that does not connect the negatives together to save space and cost.

Anybody have a solution to that problem? I'm about to build a switch of my own with a 4-pole DT switch of some kind.

From an eBay site selling the Lepai's:

I found this online A-B Box For Amplifier Comparisons it talks about sharing as well, but the design could easily be changed to switch both + & -

 

[sampleaccurate]

I also would like to A/B two Tripath amps on one set of speakers, but as I understand it, Tripath amps cannot share the right and left negative terminals, as most other amps typically do. I have been looking but have yet to find a speaker selection switch that does not connect the negatives together to save space and cost.

Anybody have a solution to that problem? I'm about to build a switch of my own with a 4-pole DT switch of some kind.

From an eBay site selling the Lepai's:

I seem to recall coming up against the same problem when looking for a speaker selector switch so I just made my own. It's about as simple of a DIY project as you can get.

 

[sprinter]

I seem to recall coming up against the same problem when looking for a speaker selector switch so I just made my own. It's about as simple of a DIY project as you can get.

So what did you use for switches? I guess the cheapest thing would be two common AC rated DPDT toggle switches and just throw them simultaneously. But a multi pole rotary would be nicer. Would you have suggestions? It's been a long time since I've parts shopped. Or how about linear multiple-pole push-button switches? A quick glance at available rotary switches, they seem pretty expensive at required ratings.

Ideally, I would like to eventually be able to switch between two amps and also two sets of speakers. Not sure what kind of switches would be suitable for that.

 

[DQ828]

Well here it is. I haven't tried it yet, but it is testing fine. I need to make some split RCA leads or else I can us my two Squeeze-boxes, synchronised I guess, then wont need the split leads. Any thoughs on what would be the best approach?

I'll be testing an Integra Class A-B amp against a DIY D class, I not sure If I will need a dummy load for the amps when they are not linked to the speakers, but have bought some 8.2ohm 10w wire wound resistor for that purpose if needed. Any thoughts guys on if the dummy load will be needed?

I have a illuminated switch, that switches two 12V 30amp relays, there is a LED so you know when the unit has power, the who thing is powered by my computer backup drive power supply.

I plan on torturing the kids & brothers (or anyone silly enough to get to close to the house) by making then listen to the amps & giving there opinion on which sounds the best to them, they wont know which amp is on, but I will know because the illuminated switch will be way of keeping track. When I try it myself I will try to make my assessment before confirming which is which.

My thanks to the designer, Phil Allison, A-B Box For Amplifier Comparisons

 

[sprinter]

Well here it is. I haven't tried it yet, but it is testing fine. I need to make some split RCA leads or else I can us my two Squeeze-boxes, synchronised I guess, then wont need the split leads. Any thoughs on what would be the best approach?

I'll be testing an Integra Class A-B amp against a DIY D class, I not sure If I will need a dummy load for the amps when they are not linked to the speakers, but have bought some 8.2ohm 10w wire wound resistor for that purpose if needed. Any thoughts guys on if the dummy load will be needed?

I have a illuminated switch, that switches two 12V 30amp relays, there is a LED so you know when the unit has power, the who thing is powered by my computer backup drive power supply.

I plan on torturing the kids & brothers (or anyone silly enough to get to close to the house) by making then listen to the amps & giving there opinion on which sounds the best to them, they wont know which amp is on, but I will know because the illuminated switch will be way of keeping track. When I try it myself I will try to make my assessment before confirming which is which.

My thanks to the designer, Phil Allison, A-B Box For Amplifier Comparisons

Looks excellent. Could you provide source and part numbers for the parts? I may build a double version of this so I can also A/B speakers with the same box.

Be aware that for useful subjective testing, it's imperative that gains for each amp be set up first for identical spl levels. A louder amp, even just a little bit, will always sound better at first blush. An SPL meter would be ideal if it's sensitive enough. They don't have to cost too much, or maybe you could borrow one.

Load resistors depend on the amps in question. I'd do it in any case. Some amps just don't like to be unloaded although decently designed mass-consumer grade amps are usually designed to be idiot proof.

The safest way of course is to turn the volumes down before switching, but then you can't match the levels that way unless you can think of some way to get around that, and you have to take extra time which makes good A/B'ing harder.

I'm not sure what you mean my splitting leads on the RCA's. Are you using RCA jacks for speaker terminals?

Keep us informed of progress.

 

[DQ828]

I'm not sure what you mean my splitting leads on the RCA's. Are you using RCA jacks for speaker terminals?

Keep us informed of progress.

I need to send the signal from my source (Squeezebox) to both amps, so I need RCA cables that have 2 plugs on one end & 4 on the other. I'll gather a parts list but as I bought most of the bits from Jaycar in Australia I'm not sure how useful it will be to you.

 

[sprinter]

I need to send the signal from my source (Squeezebox) to both amps, so I need RCA cables that have 2 plugs on one end & 4 on the other. I'll gather a parts list but as I bought most of the bits from Jaycar in Australia I'm not sure how useful it will be to you.

You can just get "wye" splitter adapters or cables. Very common RadioShack type items. Sharing output impedance of the squeezebox shouldn't be a problem as long as the input impedance of your amps are fairly high, as most are. You will lose some gain on each one, but I've never noticed the difference.

I'm mostly curious about what relays you're using.

 

[DQ828]

You can just get "wye" splitter adapters or cables. Very common RadioShack type items. Sharing output impedance of the squeezebox shouldn't be a problem as long as the input impedance of your amps are fairly high, as most are. You will lose some gain on each one, but I've never noticed the difference.

I'm mostly curious about what relays you're using.

This link should get you to the relay SPDT 30A Horn Relay 12V - Jaycar Electronics. I dont know a lot about these things, but it looked like the most suitable from what the store had to choose from. Unfortunately the stores are closed today and because I am impatient I'm going to modify my good RCA cables to do the job

 

[sprinter]

This link should get you to the relay SPDT 30A Horn Relay 12V - Jaycar Electronics. I dont know a lot about these things, but it looked like the most suitable from what the store had to choose from. Unfortunately the stores are closed today and because I am impatient I'm going to modify my good RCA cables to do the job

Hey, you're a diy'er. Whatever works...

These relays are 12 vdc (contacts as well as coils) for cars. I think that 120 VAC rated relays would be better. You may get away with it if the levels are kept modest, but I really would go for beefier relays. Also, if this is what you actually got, they are single pole. You need to end up with two DPDT devices if you're going to keep all the (-) poles separate. You mentioned a class D amp, but I don't know if it is a Tripath type or if it requires the (-) speaker leads to not be common. That was my big problem with finding a store bought switch.

Also, don't forget that diode on the coils. They're for protection from switching transients from the coils' induction (di/dt)

 

[DQ828]

Hey, you're a diy'er. Whatever works...

These relays are 12 vdc (contacts as well as coils) for cars. I think that 120 VAC rated relays would be better. You may get away with it if the levels are kept modest, but I really would go for beefier relays. Also, if this is what you actually got, they are single pole. You need to end up with two DPDT devices if you're going to keep all the (-) poles separate. You mentioned a class D amp, but I don't know if it is a Tripath type or if it requires the (-) speaker leads to not be common. That was my big problem with finding a store bought switch.

Also, don't forget that diode on the coils. They're for protection from switching transients from the coils' induction (di/dt)

I'm thinking 12V 30amp should carry the load ok. I'm not switching the (-) they are joined together as per the design. I hope the D class don't need separate negatives its not a T Class.The diode is in place just a did hard to see. I will keep the gain down for testing.

 

[sprinter]

I'm thinking 12V 30amp should carry the load ok. I'm not switching the (-) they are joined together as per the design. I hope the D class don't need separate negatives its not a T Class.The diode is in place just a did hard to see. I will keep the gain down for testing.

Ok. Good luck with it, It looks good.

BTW, I'm no expert on them, but there is a lot of confusion about Tripath, "class t", class d, etc. You probably know this already, but "class D" is a true class using switching technology and that's what Tripath chips are. There is no "class T", that's just what some people call Tripath designs. I really can't comment on your D amp without knowing more.

I don't wish to second guess your design, but I feel obligated to caution you about the relay rating. The contacts on the relay are designed to allow only 12 volts between them. Peak AC voltage from complex music signals can easily exceed that, depending on the usual factors of load impedance, applied power, etc. However there are other factors here. If you have no load offered to an amp, look at ohms law (power=E^2 / Z) and use infinity for Z and see what happens. Even assuming 8 or 16 ohm Z, depending on how much power you are developing from the amp, peak voltages can be quite large, and it's peak voltage we need to be concerned with here, not RMS. You probably want to end up with something you can use in a variety of situations in the future. I would hate to see any damage to your system because of relay arcing and shorting.
FWIW, of course.
Anyway, let me know how AB works out for you because I intend to do the same thing.

 

[DQ828]

Well I'll be Stuffed

Two amazing things have happened.

1. It worked first go & the switching is seamless except for the clicking of the relay which doesn't give anything away anyway.
2. I cannot tell the difference between the amps & yes I checked it is all working properly, I check because I couldn't believe what I was hearing, no difference that is. It really leads me to believe a lot of what we think we hear in in the mind.

Let me set the scene, I have built a six channel Coldamp to run my active speakers. Up until now I have been running them through my Integra Receiver set on multi channel DVD direct output. At this stage I don't have the setup to run a six channel test as I would either need a second MiniDSP setup or a six way switch box for the MiniDSP output, I figured a two channel would do & run through my Lukasz Frikus P17 passive speakers which I think sound great for the music I play at the levels I play it. In fact my prototype active speakers are struggling be beat them.

In the six channel mode my impressions where the Coldamp was not reaching the quality of the Integra & thought I was wasting my time building the switch as I already knew the outcome. Well in the two way setup through the P17's the music sound great from both, what can I say.

I still have some listening to do & some Crash Test Dummies to torture before I can give the final verdict on the two channel test.

I mostly play classical & was listening to sections of Handel's Messiah, the only problem was I get caught up in the music & forget to switch

The only way to be sure now is build the six way switch, I guess.

 

[DQ828]

Ok. Good luck with it, It looks good.

BTW, I'm no expert on them, but there is a lot of confusion about Tripath, "class t", class d, etc. You probably know this already, but "class D" is a true class using switching technology and that's what Tripath chips are. There is no "class T", that's just what some people call Tripath designs. I really can't comment on your D amp without knowing more.

I don't wish to second guess your design, but I feel obligated to caution you about the relay rating. The contacts on the relay are designed to allow only 12 volts between them. Peak AC voltage from complex music signals can easily exceed that, depending on the usual factors of load impedance, applied power, etc. However there are other factors here. If you have no load offered to an amp, look at ohms law (power=E^2 / Z) and use infinity for Z and see what happens. Even assuming 8 or 16 ohm Z, depending on how much power you are developing from the amp, peak voltages can be quite large, and it's peak voltage we need to be concerned with here, not RMS. You probably want to end up with something you can use in a variety of situations in the future. I would hate to see any damage to your system because of relay arcing and shorting.
FWIW, of course.
Anyway, let me know how AB works out for you because I intend to do the same thing.

If you find some better & more appropriate relays please let me know, I'd be happy to upgrade

The speakers I am using for the testing are fairly sensitive in the mid to high 90's I believe so it doesn't take as many AC volts to get them going. I ran a 400Hz signal through the amps with True RTA & measured the output with the multmeter (as recommended by the designer) and set the levels within 1%, when switching there was no perceptible volume difference.

Thanks for your input always welcome.

 

[sprinter]

Two amazing things have happened.

1. It worked first go & the switching is seamless except for the clicking of the relay which doesn't give anything away anyway.
2. I cannot tell the difference between the amps & yes I checked it is all working properly, I check because I couldn't believe what I was hearing, no difference that is. It really leads me to believe a lot of what we think we hear in in the mind.

Let me set the scene, I have built a six channel Coldamp to run my active speakers. Up until now I have been running them through my Integra Receiver set on multi channel DVD direct output. At this stage I don't have the setup to run a six channel test as I would either need a second MiniDSP setup or a six way switch box for the MiniDSP output, I figured a two channel would do & run through my Lukasz Frikus P17 passive speakers which I think sound great for the music I play at the levels I play it. In fact my prototype active speakers are struggling be beat them.

In the six channel mode my impressions where the Coldamp was not reaching the quality of the Integra & thought I was wasting my time building the switch as I already knew the outcome. Well in the two way setup through the P17's the music sound great from both, what can I say.

I still have some listening to do & some Crash Test Dummies to torture before I can give the final verdict on the two channel test.

I mostly play classical & was listening to sections of Handel's Messiah, the only problem was I get caught up in the music & forget to switch

The only way to be sure now is build the six way switch, I guess.

Well, if you can't figure it out,being right there, I wouldn't hazard a guess. Except, I just must ask the obvious question: Are you absolutely positive that the two amps are actually getting switched to the speakers? How did you confirm that? Because it sure sounds to me like you were actually listening to the same amp both times. You really should have heard some difference. Can you turn up the volume on one of them and then do it again?

Another thought. Could the two amps have gotten wired together (speaker outputs) in some way? That would do it.