I'd be careful about relocating the Vbe multiplier transistor. You'll need to monitor bias response vs. temperature. While Sansui's method is FAR less accurate and frowned upon by not only contemporary texts on amplifier designs but my Circuit Analysis professor as well, it is likely there is some compensation factor in the design that may cause bias to drift out of spec if the transistor is moved. Try it, but be ready to shut it down or adjust the bias to a new set point if need be. The design itself is bad because driver temps and output device temps are not necessarily correlated, especially in installations where the external heatsink may not be well ventilated or when difficult to drive speakers are used with highly dynamic signals. There is indeed a rough trend between driver and output temp but this type of thermal tracking is imprecise and worse still, very slow. This leads to imprecise bias control and higher output stage temperatures. Sansui compensates for this by using absolutely huge heatsinks but remember that it takes time for heat to flow from the transistor die to the heatsink. This is where you get into trouble, especially if the heatsink is already very warm. Most of the Sansuis that show up on my bench with blown outputs are either one owner or lightly used units and most have no signs of ever being serviced. I don't think it's due to "destroy it yourself" types, rather it is my opinion based on talks with my old professor and a couple books on the subject that these amps are flawed and less durable than they could be as a result. Anecdotally, I will also say that I definitely see more blown Sansui outputs than Pioneer outputs even though I get a good sampling of each brand. Pioneer of course directly monitors the output stage heatsink. Pioneers usually come in for the usual dirty switches and pots and thermal stress failures on the under chassis mounted regulator boards used on many models....another bad design by the way. (heat rises!) Still, design flaw aside, these are nice receivers and obviously they can last a while if properly maintained and used. Just keep that heatsink cool and stay away from speakers with wild dips in the impedance in the low frequencies. I definitely think they're worth fixing up and obviously their owners agree.

Thanks for the detailed response and explanation, I am doing a resto on my 3rd QRX9001 at the moment and this amp along with the 9090 series of amps are all the same too as they mount the VBE transistors on the driver heatsinks, I would really like to know why they did this on so many of their amplifiers.. Of course you are spot on with your assessment of the common failure mode. I have come across this too. A lot of Sansui designs were DC to RF by design at this time, everything was very high speed, direct coupled as you know, output transistors on the G9000 series are 25mhz, they were getting pretty ambitious, I think the 8700/9700 are 25mhz too so perhaps a little prone to going unstable....it seems the slightest flaky bias trimmer and its all over. I have a G7700 which had a very badly blown output section when I got it, it was a major rebuild, I ended up using a VBE multiplier as the STV's were totally gone, I designed it so if the trimmer does go open then the bias goes Minimum instead of Max. I also think other designs like the Pioneers are a lot more forgiving when it comes to that sort of thing. Their weakness seems to be regulated supplies mounted in silly places, as per the SX1980. It is a shame Sansui never made the G series able to handle higher current/low impedance, this would have unleashed a whole new potential because they are excellent sounding amplifiers. I may experiment with a remote mounted VBE on the main heatsink of my 9700, I am a tech so I do know when its going to all turn bad, if I find anything meaningful, I will report back.

Oh, and really, I did enjoy your videos, excellent work, it can't be easy working and filming at the same time!!

I have one that I brought over from Japan, its awesome and works as new with the case still looks new with no damage, thinking about selling it, how much should I put this bad boy on the market? I also have the Sansui Speakers 2 9700 and 2 1100 (big ones).

@@lamaatlanta 3,000. Do you have 2 AU1100 integrated amps?

Comparable modern units? None. Nobody makes 200W/ch, analog tuned receivers anymore. Restorations like this are about $600 USD.

I may be preaching to the choir here with you on this, but I wanted to speak to a broader audience with this comment as well so bear with me and my rant. The "monster receivers" were more of a product of the marketing departments than the engineering departments of the large Japanese firms of the late 70s and early 80s. If you need much over 100-120W/ch, the best designs will typically be found in the separates. The best value is also found in separates. The prices on vintage receivers, especially large ones, have been driven to record highs due to the demand for the retro look and exclusivity of these rare receivers. I sell them when I can because people love them despite their flaws but I still recommend buying separates instead if I sense the customer is not totally enamored by the looks and is instead looking for performance. (Happy customers are repeat customers!) The funny thing is, that for the price of one of these highly coveted monsters, you can go with a good tuner, preamp and power amp and get MORE POWER, less real world THD+N, better reliability and yet still end up with more money in the bank afterwards! If you end up not needing over 100W, and almost nobody does, the world of relatively inexpensive integrated amps and less extravagant receivers opens up. Most of these smaller receivers feature cosmetics similar to their monster brethren in a smaller enclosure. They also usually have more conservative design margins and are considerably less troublesome. When they do give trouble, the service costs are far lower. If you need 200 watts, I'm rather fond of the Yamaha C-2 and M-2 combo. That will cost you around or slightly less than a used G-9700 and it will play harder, last longer and sound better in every way. (I'm a Technics man, but the cost of the SE-A3, no doubt driven sky high by its beautiful meters, and its proprietary transistors makes the Yamaha the one to have.) Or you could go with just about any quality preamp or tuner combo you want and a Crown PS-400. That wouldn't be as pretty as the C2+M2 combo mentioned above but it would cost less and it too would utterly outclass the G-9700 in overall reliability, dynamic ability and stability into a low impedance load. If you played your cards right, such a system could be had for under $700. Good luck finding a good refurbished G-9700 for less than about $1200. Oh and if you want a "cheat code" for the vintage audio game, here's "god mode": Most, if not all, of the tonal characteristics of a vintage solid state receiver are in the tone control stack! Get a preamp from the manufacturer you like the sound of and pair it with a modern amp from the likes of Outlaw Audio or Emotiva. You'll get the tone you're after and you'll get a nice modern, reliable, low noise power amp to go with it. You can also get a small receiver, like a Pioneer SX-626, and use its pre-out jacks to send signal to a big power amp. I have a few customers that do just that with fantastic results.

+Andrew Kwiecinski I entered the '100W +' league with a kit amplifier named 'Playmaster series 200', which had mosfet output transistors (2 pairs (TO-3 case), Hitachi 2SJ49, 2SK134). At the time it retailed (component form) for A$549.00, from Dick Smith Electronics, (Electronics Australia), 1987. After a few weeks, where I used 15A mains cable (from an off-cut of an air conditioner cord) for earthing, it was working, with thermistors and driver transistors attached to the heat sinks of the output stage. I do not know of its whereabouts now, but noted from its phono setting, at full volume (no record playing) a very quiet 'hiss' with close proximity to the speaker... It was rated at 100W rms per channel into 8 ohms, and 160W rms per channel into 4 ohms...

Isn't it odd? They all did it, even Marantz! Not to be out done, Technics unleashed a new form of horror...the twist lock speaker connectors!

I should also mention that the smaller G8700 uses binding posts! It makes no sense!

Contact me at my firstname.lastname at gmail dot com.

I usually reside in Coppell, TX but I'm *temporarily* in Abilene working on a family home during COVID. Periodic visits to Coppell are still happening but with infection control precautions in place. (curbside pickup/drop off only, etc)

The 9000 is a notch up over the G9700 with nearly 5kg more mass! The 9000 has over twice the capacitance in the power supply for more dynamic headroom but the 9700 is just plain *bigger* despite having less headroom over the RMS ratings so the comparison is not an easy one to make as far as to which one will be louder. Into an 8 ohm resistance or 8 ohm nominal speaker with a relatively benign impedance curve (an "easy to drive" speaker), the 9700 likely will give more power due to its high voltage power supply. Into real-world varying impedance or just plain low impedance speakers, the more conservative design of the 9000 can allow you to pull ahead with its superior current reserves. (The 9700 will fall on its face into 4 ohms, don't try it; it can easily overheat!) The 9000 also has higher quality connectors, better knob feel, better soldering, and the 9000 will hold value better. To be honest, if you find a big Pioneer like an SX1050 or SX-1250, get that instead if it's in the budget. The build quality is higher still and you actually get proper thermal tracking of the output heatsink. The Technics SA-5670, Kenwood KR-9600 and Yamaha CR3020 are also something to look at if you like your receivers rare and good.

My budget is 900 bucks, but its a bit harder since I'm trying to buy locally instead of paying out the ass for shipping