YAMAHA A-S701 - tear-down, thoughts about the internals & few measurements

[Willem]

And it comes with dsp room eq.

 

[Cahudson42]

output power stage is created around the well-known complementary bi-polar Japanese SANKEN transistors 2SA1694 / 2SC4467 with 2 paralleled pairs per channel, each transistor being able to sustain a continuous 4A with peaks up to 8A,

Why would the designer do this compared to the more usual 15A single pair? Better heat disapation with the surface area of 4 tabs vs. 2 giving more tab surface area per Watt? Or?

I would think 4 devices/channel costs more than 2 of double current - correct? Plus balance issues?

Just curious - compared to my 'new' RX-596 I'm just starting to play with..

 

[trl]

AFAIK, Yamaha is not using anymore 15A transistors in their amps these years. 2SA1694 / 2SC4467 is what they use now, or the powerful MOSFET's from their high-end amp series (A-S1x00/S2x00/S3x00).

In RX-596 power dissipated on 2SA1492 or on 2SC3856 would be max. 130W (with infinite heatsink), while using tow pairs of 2SA1694/2SC4467 will dissipate 160W (with infinite heatsink), hence the increased power on the A-S701 (2x180W/4Ohms @approx. 0.015% THD+N) over the RX-596 (2x130W/4Ohms @approx. 0.015% THD+N).

Placing 15A output buffer means a completely different PCB, with much thicker traces, then the serialised resistors needs to get doubled in power etc. In the end yes, two heatsinks will better dissipate the heat inside the case.

 

[hyperknot]

@trl I see so many of these AS series amps on eBay / Hifishark in a non-working state. Is this generally true for them to break down or it's just that there is a lot of them being bought / now sold?

 

[trl]

I don't know, sorry, mine still works and is powered every day between 2h and 8h (mostly kid is watching cartoons and listening to the big speakers, so low volume). However, most amps today are having a microprocessor inside that if it ceases to work then entire amp becomes useless, unless someone who know the schematic can try to find a workaround and modify it as a power amp and add a pre-amp on top.

So, basically, if today's amps will break and manufacturer will ask for motherboard replacement on customer's money, it would be cheaper to get the amp to an electronic shop and ask an EE from there to modify it as a power amp and, of course, to remove all the protections it has. It's not easy, but seems doable.

 

[Willem]

They are a good traditional design from a reputable manufacturer. They have been around for quite a while as well so there must be many around. I have never seen any references to specific problems and I would imagine any problems were ironed out a long time ago.

 

[ElNino]

@trl I see so many of these AS series amps on eBay / Hifishark in a non-working state. Is this generally true for them to break down or it's just that there is a lot of them being bought / now sold?

Yamaha service manuals are readily available (HifiEngine, etc.) and these are often easy to fix if they do break, and they're great values if you can make them work, so I'd imagine these are easier for sellers to move in non-working condition than more niche gear.

 

[trl]

Yamaha service manuals are readily available (HifiEngine, etc.) and these are often easy to fix if they do break, and they're great values if you can make them work, so I'd imagine these are easier for sellers to move in non-working condition than more niche gear.

There are at least 3 part-numbers you might not find them: the two microprocessors and the potentiometer. I'm sure there are others too, but I do hope it will never break.

 

[trl]

I was measuring THX Reference Level today at listening position (2m in front of speakers) for a 2V RMS input level (when 0dBFS applied), but using -20dB audio files, REW + Superlux ECM999 and some closed over-the-ears headphones so I don't get deaf. Of course, CD-Direct was pressed.
My CANTON stereo speakers are having about 90dB/W@1m.

Seems that THX Reference Level for this amplifier is somewhere around 1 o'clock, although I don't think I ever pushed this amp above 12.5 o'clock with normal music (probably recorded around -3dB...0dB).

 

[exitonly]

how much power this gives headphone 120ohm/300/600 ? is there any calculation formula for this

 

[trl]

It depends on headphones sensitivity; with 100dB/1mW headphones you will only need 100mW to get to the insane level of 120dB SPL (usually if a headphones gets to 115dB SPL for peaks it might be considered as having a very good dynamic).

 

[exitonly]

i just sart to use a+b mode bi-wiring my b&w602.5s3 8ohm speaker. the instruction manual tells you that a + b mode must use low mode. Someone says you can use it without worry in high mode. and somebody else says use low to not damage the amplifier. what is a better way, keep low or high

 

[trl]

It all depends on speakers impedance, so if you have 8...16Ohms speakers, you can probably use both A+B speakers output, but you really need to find speakers datasheet, because for some manufacturers 8 Ohms speakers are actually having 3.5 Ohms below 100 Hz.

However, for high output power (2V RMS at inputs and volume knob below 1 o'clock) you might want to double check dissipated temperature on top of the case.

 

[audio_tony]

Volume is adjusted via an ALPS motorized RK16811MGA09 "1-gang" potentiometer. No, I don't think I'm wrong, it looks indeed like a "1-gang" pot, yes...a mono potentiometer.

@trl It is actually a standard stereo pot. And looking at the schematic, a standard motor driven one.

The pin layout is quite weird though!

 

[trl]

RK16811MG is 1-gang, while RK16812MG is 2-gang.

The volume seems to be controlled by the internal processor, and not in the old fashion way by the analogue pot, but of course, I might be wrong, so posting the picture again below:

 

[audio_tony]

RK16811MG is 1-gang, while RK16812MG is 2-gang.

The volume seems to be controlled by the internal processor, and not in the old fashion way by the analogue pot, but of course, I might be wrong, so posting the picture again below:

Maybe I misunderstood your original post.

Yes, the pot (motor) is indeed controlled by the CPU, however it will still be controlling an analogue system via the usual (analogue) method of a variable resistance - hence it's not a mono pot, as per my post above it's stereo (but looks mono because there is only a single row of pins).

If it was a digital control, there would be a chip like a PGA2310 / PGA2311 etc. and no motor would be required.

Does that help?

 

[trl]

I see that the IC502 takes care of Volume UP and Volume DOWN via pins 91 and 92. I think that the opamps that take care of the audio are inside this chip, but of course I might be wrong, so please analyze the schematic in thorough and let me know your opinion after.

 

[somebodyelse]

Maybe I misunderstood your original post.

Yes, the pot (motor) is indeed controlled by the CPU, however it will still be controlling an analogue system via the usual (analogue) method of a variable resistance - hence it's not a mono pot, as per my post above it's stereo (but looks mono because there is only a single row of pins).

If it was a digital control, there would be a chip like a PGA2310 / PGA2311 etc. and no motor would be required.

Does that help?

Your explanation is contradicted by the schematic which shows a single gang pot that's connected between 5V and ground, not the audio signal. This is consistent with an analog control input on a DSP or microprocessor, as is often used when the designer wants to present a conventional control interface to the user despite using logic internally.

 

[audio_tony]

I see that the IC502 takes care of Volume UP and Volume DOWN via pins 91 and 92. I think that the opamps that take care of the audio are inside this chip, but of course I might be wrong, so please analyze the schematic in thorough and let me know your opinion after.

My apologies you are correct.

I didn't notice VR851 on the schematic earlier. That is such a strange setup, as the main CPU (IC502) is producing the up/down control via pins 91/92 as you say. This then rotates the pot via the motor, which then returns a variable voltage to CPU via pin 90.

So there must be opamps within IC502 to provide the amplification as well. In fact it looks as though IC502 might be a DAC or ADC as well, although I'm struggling to find the analogue inputs / outputs (can barely read the pin designations).

I wonder why they didn't just use a rotary encoder then?

 

[zepplock]

I recently fixed (with help from this forum members!) A-S-701's cousin: R-S-700. The insides look identical, except it has no dac, no motorized volume but has a tuner. If you are looking for used ones, R-S line is worth looking into too.