I have a simple technique that will allow you to use a balanced source with any tube amplifier. Here's how it works:
The easiest way to understand this is to set it up on a single-ended triode amp. You do this by first installing the XLR connector on the amp.
Pin 1 is ground and ties to the ground of the RCA connection. Pin 2 is the same as the center pin of the RCA.
Pin 3 is where it gets more interesting. In some amps, there will be a cathode bypass cap, in others, the value of this cap will have to be calculated and then obtained. In the case of the bypass cap, the negative side, which is tied to ground, gets moved to pin 3 of the XLR. For single-ended operation, a jumper is installed in the XLR between pin 3 and pin 1.
If the cap is being added, the minus side ties to pin 3 and the plus side to the cathode of the input tube. The jumper from pin 1 to pin 3 is not needed if you are running single-ended.
That's it- it really is that simple. The input is differential and does not change the character of the amp. It does make it more flexible though, as you can install the amp near the speaker and run a longer interconnect if needed, which is easy if you have balanced line operation
The easiest way to understand this is to set it up on a single-ended triode amp. You do this by first installing the XLR connector on the amp.
Pin 1 is ground and ties to the ground of the RCA connection. Pin 2 is the same as the center pin of the RCA.
Pin 3 is where it gets more interesting. In some amps, there will be a cathode bypass cap, in others, the value of this cap will have to be calculated and then obtained. In the case of the bypass cap, the negative side, which is tied to ground, gets moved to pin 3 of the XLR. For single-ended operation, a jumper is installed in the XLR between pin 3 and pin 1.
If the cap is being added, the minus side ties to pin 3 and the plus side to the cathode of the input tube. The jumper from pin 1 to pin 3 is not needed if you are running single-ended.
That's it- it really is that simple. The input is differential and does not change the character of the amp. It does make it more flexible though, as you can install the amp near the speaker and run a longer interconnect if needed, which is easy if you have balanced line operation
You have probably introduced DC to your input cable, as the cathode is used for bias. You may have disturbed a feedback loop (local or global), which often goes to the cathode of the input stage, and so may seriously change the character of the amp. The result is not balanced, as the impedances of the two sides will be very different.
At best you have described a pseudo-balanced technique which may work to a limited extent with some amplifiers, and create all sorts of problems with others.
There is a general rule for mature technology - if you think you have spotted something which others have not then:
1. they probably have spotted it
2. they probably know the snags so don't use it
3. so they don't bother to mention it
4. so newbies think nobody else has spotted it
At best you have described a pseudo-balanced technique which may work to a limited extent with some amplifiers, and create all sorts of problems with others.
There is a general rule for mature technology - if you think you have spotted something which others have not then:
1. they probably have spotted it
2. they probably know the snags so don't use it
3. so they don't bother to mention it
4. so newbies think nobody else has spotted it
Atmasphere is hardly a newbie. But your main point is correct- this is differential (more or less) but not at all balanced. I would not try this on an amp with feedback unless I really knew what I was doing; there's a definite danger of shorting out the feedback loop, which will not make your day a pleasant one.
But your main point is correct- this is differential (more or less) but not at all balanced. I would not try this on an amp with feedback unless I really knew what I was doing; there's a definite danger of shorting out the feedback loop, which will not make your day a pleasant one.I think in short you are saying to float the first gain stage's ground on XLR's pin 3.
As others have said, tha problem is if some other signal also goes into this stage. Typically the negative feedback does go to this stage and it is referenced to ground.
Clasically there are two way to go
1) Use a transformer in the XLR input to convert balanced to unbalanced. this works perfectly. But results depend on the quality of the transformers and good ones cost almost $100
2) Design a true balanced amp. In #1 above the amp remains single ended. A tru balanced amp has a gain stage connected to each side of the balanced signal. This doubles the number of tubes required. But does it? No phase splitter is required. In most Hi Fi amps being able to loose the phase splitter more than makes ou for the added preamp tube. This design was not un-common in high end professional audio gear not not used in consumer HiFi likely because of the added cost.
This seems like an area were DIYers should be experimenting. Rather than building the same 50 year old designs, lets see what a true balanced amp sounds like
Here is a design of such an amp from the 40's. Using current terminology this is a tri-amped powered speaker with balanced input. You can ignore most of it. Look at the top "row". Where it reads "input from console" (g1 and g2) and follow it to the speaker on the right. It is very simple with no phase splitter. Negative feedback goes from a special secondary to the power tubes. This system when new sold for about 2X more than a new car and was about the size of a home type washing machine It was used in churches to reproduce organ music. A church would typically have two or more of these units.
http://www.captain-foldback.com/Hammond_sub/schematics/PR40_QR40_1.gif
In modern usage, we can copy a lot from this. Loose the spring reverb. Now you have a bi-amp where each amp has customized components for bass or treble (mostly a special bass or treble output transformer) and very short speaker cables but long XLR interconects going back to the source.
Many sources are inherently balanced, like phono cartridges and microphones. So you can keep it balanced all the way out to the speakers
As others have said, tha problem is if some other signal also goes into this stage. Typically the negative feedback does go to this stage and it is referenced to ground.
Clasically there are two way to go
1) Use a transformer in the XLR input to convert balanced to unbalanced. this works perfectly. But results depend on the quality of the transformers and good ones cost almost $100
2) Design a true balanced amp. In #1 above the amp remains single ended. A tru balanced amp has a gain stage connected to each side of the balanced signal. This doubles the number of tubes required. But does it? No phase splitter is required. In most Hi Fi amps being able to loose the phase splitter more than makes ou for the added preamp tube. This design was not un-common in high end professional audio gear not not used in consumer HiFi likely because of the added cost.
This seems like an area were DIYers should be experimenting. Rather than building the same 50 year old designs, lets see what a true balanced amp sounds like
Here is a design of such an amp from the 40's. Using current terminology this is a tri-amped powered speaker with balanced input. You can ignore most of it. Look at the top "row". Where it reads "input from console" (g1 and g2) and follow it to the speaker on the right. It is very simple with no phase splitter. Negative feedback goes from a special secondary to the power tubes. This system when new sold for about 2X more than a new car and was about the size of a home type washing machine It was used in churches to reproduce organ music. A church would typically have two or more of these units.
http://www.captain-foldback.com/Hammond_sub/schematics/PR40_QR40_1.gif
In modern usage, we can copy a lot from this. Loose the spring reverb. Now you have a bi-amp where each amp has customized components for bass or treble (mostly a special bass or treble output transformer) and very short speaker cables but long XLR interconects going back to the source.
Many sources are inherently balanced, like phono cartridges and microphones. So you can keep it balanced all the way out to the speakers
A few comments:
Pin 1 on an XLR is not a signal pin and should not be used as such. It's just a shield return.
A proper balanced output can handle pin 3 being grounded. The easy way to do this is connect pin 2 to the RCA center pin and pin 3 to the RCA shield. Pin 1 can be left grounded at the driving end only to avoid ground loops if necessary.
Proper balanced inputs and outputs have balanced impedance to ground. This is where the noise immunity comes from. Balanced vs. single ended should not have a "sound" of it's own.
Cheers,
Michael
Pin 1 on an XLR is not a signal pin and should not be used as such. It's just a shield return.
A proper balanced output can handle pin 3 being grounded. The easy way to do this is connect pin 2 to the RCA center pin and pin 3 to the RCA shield. Pin 1 can be left grounded at the driving end only to avoid ground loops if necessary.
Proper balanced inputs and outputs have balanced impedance to ground. This is where the noise immunity comes from. Balanced vs. single ended should not have a "sound" of it's own.
Cheers,
Michael
I like the transofrmer approach. It satisfies the differential requirement while providing isolation. I personally find transformers with high imput inductance work better. Low input inductnce transformers behave like 6db per octave high pass filters resulting in a bottom end thats cut off. Not a typo, i am talking about inductance rather than impedeance.
but we can get creative. use a pentode at the input stage and wire the XLR runs using ultralinear concepts. Avoids the issues below.
but we can get creative. use a pentode at the input stage and wire the XLR runs using ultralinear concepts. Avoids the issues below.
two resistors tied in series, pin 1 to the center tie point, one open resister end to pin 2, the other open resistor end to pin 3.
normal unblalanced use signal from pin 2
Pin 2 = Signal +
Pin 3 = Signal –
Pin 1 = Ground
want to bridge an amp, normal pin 2 to one unbalanced channel, pin 3 to the other unbalanced channel, speaker connect across + terminals of each channel. nothing on the amp - terminals.
normal unblalanced use signal from pin 2
Pin 2 = Signal +
Pin 3 = Signal –
Pin 1 = Ground
want to bridge an amp, normal pin 2 to one unbalanced channel, pin 3 to the other unbalanced channel, speaker connect across + terminals of each channel. nothing on the amp - terminals.
Apparently my first post was not clear!
I am not suggesting that the ground of the input section is connected to pin 3 of the XLR
A lot of amps have a cathode bypass cap on the first stage of gain. If so, you lift the cap at its ground point and tie it to pin 3. If you plan to use the amp with a single-ended source, pin 3 will have to be tied to pin 1 (ground) for best results (else you will have reduced gain).
If the amp lacks a cathode bypass, you have to calculate what that value would be and then install it between pin 3 and the cathode of the amp. In this case, no shorting jumper is needed for single-ended operation.
In the case of a feedback loop in the cathode, there will be no bypass cap. Therefore the latter approach is used.
I have used this system on a variety of amplifiers in the last year. It works well on all of them I have tried. It sounds better than using an input transformer: you get better (deeper) bass with smoother mids and highs. Of course you don't get the same CMRR, but you do get a fairly decent amount, somewhat dependent on the gain of the input circuit.
The reason I tried this is I have a balanced preamp. Like a line transformer with no CT, the output of the preamp is only between pin 2 and 3 of the XLR and ignores ground. So with an amp with a single-ended input, the preamp will seem to buzz, but all that is happening is that the unused output is floating, so there is no circuit.
This technique is a solution for that, as it also ignores ground, so you can use a balanced preamp with an amplifier that otherwise would only have a single-ended input. You don't get *all* the advantages of a proper fully-differential balanced setup (such as power supply noise rejection in the amp, and the CMRR is not as good) but it does work quite well, certainly better than using the single-ended input.
This technique is not new. It was first used by George Philbrick in his early opamps of the late 1940s.
I am not suggesting that the ground of the input section is connected to pin 3 of the XLR
A lot of amps have a cathode bypass cap on the first stage of gain. If so, you lift the cap at its ground point and tie it to pin 3. If you plan to use the amp with a single-ended source, pin 3 will have to be tied to pin 1 (ground) for best results (else you will have reduced gain).
If the amp lacks a cathode bypass, you have to calculate what that value would be and then install it between pin 3 and the cathode of the amp. In this case, no shorting jumper is needed for single-ended operation.
In the case of a feedback loop in the cathode, there will be no bypass cap. Therefore the latter approach is used.
I have used this system on a variety of amplifiers in the last year. It works well on all of them I have tried. It sounds better than using an input transformer: you get better (deeper) bass with smoother mids and highs. Of course you don't get the same CMRR, but you do get a fairly decent amount, somewhat dependent on the gain of the input circuit.
The reason I tried this is I have a balanced preamp. Like a line transformer with no CT, the output of the preamp is only between pin 2 and 3 of the XLR and ignores ground. So with an amp with a single-ended input, the preamp will seem to buzz, but all that is happening is that the unused output is floating, so there is no circuit.
This technique is a solution for that, as it also ignores ground, so you can use a balanced preamp with an amplifier that otherwise would only have a single-ended input. You don't get *all* the advantages of a proper fully-differential balanced setup (such as power supply noise rejection in the amp, and the CMRR is not as good) but it does work quite well, certainly better than using the single-ended input.
This technique is not new. It was first used by George Philbrick in his early opamps of the late 1940s.
Last edited:
CrisA mentioned units used for church organs (.Hammond B tone cabinets). Got one amplifier on ebay. Also found old Shure mic mixer $35 ,it has 5 of the balanced line in/out transformers too.(jensens cost $100 @) , I just use it's balanced output .Mixer has 30-20k response and sounds great. $135 for both. Seems like these units are around but most people don't know what to do with them.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.