Sansui Alpha x-balanced series 'hot' and 'cold' service manual terminology

smurfer77

Super Member
I think the terminology of 'hot' and 'cold' used in the bias measurement of alpha SMs is misleading. The 'Cold' output, is in fact electrically hot too. Seems trivial, but this means you need to take some care when measuring power etc, not to accidentally ground the negative speaker out terminal

The amp outputs should be called '+' and '-' (I like this since '-' in a mathematical sense implies pi (180deg) phase change), or something similar. So moving forwards I'm going to refer to the terminals as "+ out ('hot')" and "- out ('cold')". I added 'out' because I don't want to confuse '+' with a DC supply rail. What do you think?

This has caused confusion to others too. This caused me some confusion the first time I measured the power output on my first Alpha without checking the topology because after I saw 'cold' I figured the negative speaker terminal was tied to ground as usual. A rookie mistake for those working on Alphas (and car amps :) but this happened to me when I got my first Alpha.

Once you take a proper look at the schematic it is clear that topology is balanced output stage with + and - terminals both driven (out of phase), via 2 amp sections per each channel (one section for 'hot' and one for 'cold' (180 degrees phase shifted with respect to 'hot'). i.e. a type of bridged setup. Again, a simple thing and not that uncommon, but I'm sure I'm not the only one who dived into an Alpha without checking the schematic and topology properly
 
Yep, it makes these amplifiers a real handful for beginners if they have a fault. Also understanding the internally bridged design with neither speaker terminal at ground potential, and the number of, what seem like strange adjustments in some cases. I have an example what I think was the first incarnation by Sansui of that design as in the AU-D11 II - an absolute GEM of an amplifier once it is properly sorted out and upgraded. Mine is a 'cold dead hands' piece and no mistake, and as such it receives periodic attention from me adding tweaks that I have either learnt here, or devised for myself. ;)

It's in my system now, it's right up there jockeying for position with my AU-919 and AU-X1. :D
 
Last edited:
Crown with their bridged amplifiers like the Macro-Tech VZ series use the term 'High Side' and 'Low Side', and also 'Emitter Positive' and 'Emitter Negative'.
I think the Hot and Cold is not a good reference. The Way Crown do it makes it obvious that both the output sections are above earth.
Although there is a snag as the Crown topology is a "grounded bridge"...wont go into that here though as we'll be off topic in no time...
 
Crown with their bridged amplifiers like the Macro-Tech VZ series use the term 'High Side' and 'Low Side', and also 'Emitter Positive' and 'Emitter Negative'.
I think the Hot and Cold is not a good reference. The Way Crown do it makes it obvious that both the output sections are above earth.
Although there is a snag as the Crown topology is a "grounded bridge"...wont go into that here though as we'll be off topic in no time...

I don't really like 'High Side' and 'Low Side' (i mean, nothing is really 'low', except perhaps position on schematic ;)) but 'Emitter Positive' and 'Emitter Negative' sounds super clear to me and is what I think I will call them from now on.
 
Yep, it makes these amplifiers a real handful for beginners if they have a fault. Also understanding the internally bridged design with neither speaker terminal at ground potential, and the number of, what seem like strange adjustments in some cases. I have an example what I think was the first incarnation by Sansui of that design as in the AU-D11 II - an absolute GEM of an amplifier once it is properly sorted out and upgraded. Mine is a 'cold dead hands' piece and no mistake, and as such it receives periodic attention from me adding tweaks that I have either learnt here, or devised for myself. ;)

It's in my system now, it's right up there jockeying for position with my AU-919 and AU-X1. :D

Yes, I think AU-D11 II is the first Sansui with this bridged/balanced output design.

And in other news... after years of avoiding anything newer than an AU-X11, I've started collecting alphas... 607MR, 907Limited and 607NRAII so far... I'm only 2 weeks into this modern Sansui journey and its already been very very interesting. Will start up another thread or maybe add in the ye old 'AU-777A vs all contenders' thread...
 
I don't really like 'High Side' and 'Low Side' (i mean, nothing is really 'low', except perhaps position on schematic ;)) but 'Emitter Positive' and 'Emitter Negative' sounds super clear to me and is what I think I will call them from now on.
High side is positive, Low side is negative.
Must be in reference to which way the amp is pushing when looking at it on a scope...
I've got used to it, as they reference High Side Bias and Low Side Bias in the service data...
I prefer the Emitter Negative and Emitter Positive terminology...
 
And in other news... after years of avoiding anything newer than an AU-X11, I've started collecting alphas... 607MR, 907Limited and 607NRAII so far... I'm only 2 weeks into this modern Sansui journey and its already been very very interesting. Will start up another thread or maybe add in the ye old 'AU-777A vs all contenders' thread...

In my opinion, the alphas sound better than any other Sansui series, I'm in love with my X-901. I jump from my 717/777 when I want to listen to "hard" and strong music to one of my alpha when I look for detail and music up close:music:
 
High side is positive, Low side is negative.
Must be in reference to which way the amp is pushing when looking at it on a scope..

But when viewed in scope the 'low side' (pulling) becomes high side (pushing) for every alternative half cycle. :p. Maybe low side or negative is labeled so because it is the side out-of-phase with input (i.e. is negative when the input signal is positive)... Will confirm next i have one on the bench
 
Last edited:
But when viewed in scope the 'low side' (pulling) becomes high side (pushing) for every alternative half cycle. :p. Maybe low side or negative is labeled so because it is the side out-of-phase with input (i.e. is negative when the input signal is positive)... Will confirm next i have one on the bench
Not on a balanced amplifier.
 
But when viewed in scope the 'low side' (pulling) becomes high side (pushing) for every alternative half cycle. :p. Maybe low side or negative is labeled so because it is the side out-of-phase with input (i.e. is negative when the input signal is positive)... Will confirm next i have one on the bench

Have now confirmed this as per alpha607MR manual; it does actually refer in one place to "Hot (non-inverting)" and "Cold (inverting)".
 
This thread helped me a lot in understanding my AU-G77X. I haven't worked on many amps yet, and this is by far the most complex (with by far the least helpful schematics...) The terms "hot" and "cold" confused me a lot, and prevented me from understanding the amp (until I found this thread.)

After I got the amp working (er, mostly - see thread...) a funny thing happened. I was adjusting bias and offset per the SM and the very last adjustment wasn't acting quite right. For some reason I checked the collector voltage on one of the output transistors and got about 1V ???!!! I checked its partner and got about -65 V. What is amazing is that the amp still worked perfectly on both channels.

After a lot of testing of components and divide and conquer, I found the problem. A teensy speck of solder was shorting the output of the "cold" amp to ground. In an ordinary amp, that speck of solder (or something valuable) would have turned to smoke, but the differential, ground-agnostic design of this amp made it a non-issue. At the very input to the driver amp, the (ground referenced) signal is converted to a difference. No more ground. That difference (d) is multiplied by +G/2 ("hot" output) and -G/2 ("cold" output) and this voltage difference is applied across the speaker: (G/2)*d - (-G/2)*d = G*d, where G is the overall gain of the amplifier. That's why "+" and "-" would be more sensible names (but still confusing.)

Now the thing is, the power supply to the output section is ground-agnostic, too. Almost. A single filter cap straddles the B+/B- rails, and only a weak resistor (6.8k ?) connects each rail to chassis ground, tending to keep B+/B- symmetric. (The transformer secondary is not center-tapped.) So even with the solder shorting one emitter pair to ground, the amp continued to multiply the input voltage difference by G, the power supply maintained a voltage difference of (B+ - B-) across the output section, the feedback loop got closed, and the output sounded great. As far as I can tell, the only part of the circuit that "cared" was that pair of resistors in the power supply : one would get hotter, the other cooler. Probably something in the circuit would increase the distortion from .003%, but my golden ears didn't pick up on that.

Of course, I didn't crank the volume way up, and who knows what kind of weird electrical properties a loosely touching speck of solder has, so I wouldn't dare try to repeat this little experiment. But it made me spend a lot of time thinking about how this very cool amplifier works.
 
Last edited:
This looks like a specialized BTL amplifier. These are super common in automotive applications, where a designer desires to get more power out of a given supply voltage - I can see why this one would give a novice fits.. if you connect your scope ground and signal generator ground as you would normally, it'll short out the "cold amp".

For troubleshooting this type of amp, one work-around I would suggest, is to connect your load bank as usual, but then connect the primary of an audio output transformer, like from a tube amp, between the hot and cold outputs. Then, the secondary can be connected to the input of an oscilloscope, and you can proceed to troubleshoot the amplifier as usual. You will be able to observe what's going on in the amplifier, without shorting anything out.

Whenever I work on an unfamiliar amp, one of the first things I do before connecting test equipment - is make sure that the black speaker terminal is actually at chassis ground. Then, I ground my scope to the chassis, and not to the speaker terminal. I do this as a matter of course, just so that the one time I'm not paying attention, I don't end up causing issues.
 
Back
Top Bottom