Leestereo's Restoration and Upgrade of a Sansui AU-777A

[Leestereo]

Part 3: POWER AND PROTECTOR BLOCK (F-1153)

The F-1153 board contains the regulated power supplies as well as the protection circuit.

The stock rectifiers, D002 and D003, were replaced with 2A soft recovery types. The voltage dropping resistor at R007 was a 330ohm/3W cement type and was replaced with a 330ohm/ 5W low TCR wirewound. The 220µF/50V capacitor at C007 was replaced with a 470µF/63V low ESR type. Similarly, the 220µF/75V (C006), 470µF/35V (C005) and 220µF/50V (C003) capacitors were replaced with low ESR types rated at 330µF/80V, 1000µF/50V, and 680µF/50V, respectively. The 0.5W resistors: 1.5kohm (R001), 3.9kohm (R002) and 1.0kohm (R003) were upgraded to 2W metal film types. Similarly, the 1.0kohm/0.25W resistor at R004 was upgraded to a 0.5W metal film type. The 220ohm/5W cement resistors at R006 and R007 were upgraded to high stability vitreous enameled wirewound types rated at 7W (note that these were installed with their bodies well away from the board).

 

[Leestereo]

Part 4: DRIVER AMP. BLOCK (F-1170)

The Sansui AU-777A contains two F-1170 boards, one for each channel. The 470µF capacitor at C817 was replaced with a Nichicon Fine Gold type. The ceramic decoupling capacitor (C829) was upgraded to a 0.47µF polyester film type. The open-frame trimmer at VR803 was upgraded to a cermet multi-turn type (note that the corrected schematic above shows the proper locations of VR803 and R861). The emitter resistors, R851 and R853, were replaced with 0.68ohm/5W low TCR wirewound types (the resistance was increased from 0.5ohm as per the Sansui Service Bulletin).

 

[kray]

Hi Ben - would you say there are difference in tone between upgraded 555, 666, 777? or should they be similar just power output higher?

 

[Leestereo]

IMO, once restored and upgraded, the AU-555A and AU-777A do sound very much alike (especially with speakers that are easy loads and are relatively sensitive); this is largely due to their similar amplifier topology, in particular the use of a capacitor-coupled output stage. Differences between the 555A and the 777A may be more acute with difficult to drive speakers; the more elaborate and robust power supply of the 777A would allow it to drive more demanding speakers. The AU-666 uses a fully complementary direct-coupled output stage and this does give this model a different sound from the 555A and 777A; it has been described as more neutral/hi-fi sounding. The published power ratings of the 555A, 666 and 777 are 25 WPC, 35 WPC and 30WPC, respectively and can be considered comparable in actual use, i.e, a 5 WPC difference is not likely to be noticeable per se.

 

[Leestereo]

Part 5: TONE CONTROL AMP. BLOCK (F-1207)

The input high-pass filter capacitors (C701/C702) for the tone control stage were originally 1.0µF polarized electrolytic types and these were upgraded to 1.0µF WIMA stacked film types. The 10µF polarized electrolytic capacitors at C703/C704 and C719/C720 were upgraded to 22µF Nichicon ES bipolar types. Similarly, the 3.3µF polarized capacitors at C713/C714 were replaced with 3.3µF Nichicon ES capacitors. The stock polyester film capacitors at C705-C712 were upgraded to equivalent capacity polypropylene film types. The 47µF emitter by-pass capacitors (C715/C716) were replaced with same value Nichicon FG types. The failure-prone 2SC458 transistors (TR701-TR704) were replaced with KSC945C(G) transistors (hFE matched pairs for L/R channels). The output coupling capacitors (C749/C750) for this stage are installed on the solder side of the board; the stock 3.3µF polarized electrolytic capacitors were upgraded to WIMA polyester film types.

 

[Leestereo]

Part 5: POWER SUPPLY AND CHASSIS-MOUNTED CAPACITORS

The stock 4000µF/80V capacitor at C002 was upgraded to a 10,000µF/80V Nichicon KG type. Similarly, the capacity of the capacitors at C004 (1000µF/35V), C008 (470µF/50V), C819 (2,200µF/35V) and C820 (2,200µF/35V) were upgraded to 2200µF/50V, 1200µF/63V, 4700µF/50V and 4700µF/50V, respectively. The output-coupling capacitors at C825 and C826 (1000µF/63V) were upgraded to 6,800/63V Nichicon KG types. The stock diode bridge (D001), which had developed a hairline crack near one of the mounting holes, was replaced with a 10A/600PIV one.

 

[Leestereo]

Part 6: CHASSIS-MOUNTED CAPACITORS (MINIATURE) AND EXTERNALLY-MOUNTED TRANSISTORS

All input signals pass through a pair of coupling capacitors installed on the input selector switch itself. The stock 1µF electrolytic capacitors (C619/C620) were upgraded to 1µF stacked film types. The 0.033µF PIO capacitor (C001) that shunts the power switch was replaced a proper X2-Safety Class capacitor. The transistors (TR807-TR8814) installed on the rear chassis panel were unmounted, cleaned and reinstalled with new mica insulators and thermal paste.

 

[stopkidding]

Part 3: POWER AND PROTECTOR BLOCK (F-1153)

The F-1153 board contains the regulated power supplies as well as the protection circuit.

The stock rectifiers, D002 and D003, were replaced with 2A soft recovery types. The voltage dropping resistor at R007 was a 330ohm/3W cement type and was replaced with a 330ohm/ 5W low TCR wirewound. The 220µF/50V capacitor at C007 was replaced with a 470µF/63V low ESR type. Similarly, the 220µF/75V (C006), 470µF/35V (C005) and 220µF/50V (C003) capacitors were replaced with low ESR types rated at 330µF/80V, 1000µF/50V, and 680µF/50V, respectively. The 0.5W resistors: 1.5kohm (R001), 3.9kohm (R002) and 1.0kohm (R003) were upgraded to 2W metal film types. Similarly, the 1.0kohm/0.25W resistor at R004 was upgraded to a 0.5W metal film type. The 220ohm/5W cement resistors at R006 and R007 were upgraded to high stability vitreous enameled wirewound types rated at 7W (note that these were installed with their bodies well away from the board).

What were the calculations to uprate the capacitances on this board? Because it's a power supply board are the caps on this used for smoothing the voltage? Can similar changes be made on the Ripple filter boards of the AU-666? Those have 100uF and 220uF 50V cape.

 

[Leestereo]

What were the calculations to uprate the capacitances on this board? Because it's a power supply board are the caps on this used for smoothing the voltage? Can similar changes be made on the Ripple filter boards of the AU-666? Those have 100uF and 220uF 50V cape.

There were no calculations per se for increasing the capacitance of the F-1153 board. If one considers that the tolerance of electrolytic capacitors manufactured up to the ~mid-70s was typically -10%/+50% or even -10%/+80%, it appears that there is considerable latitude in choosing appropriate replacement capacitors. Generally, there is no problem increasing the capacitance rating of a local decoupling/filtering capacitor (especially for small values less than 1000µF); IMO, a 2X-3X increase is acceptable. In these cases, the extra load on the power supply transformer and diodes is inconsequential. Further, since modern capacitors often measure 90% their stated value, the use of replacement capacitors that are rated at the nominal capacitance of the original component may result in a DC supply that is at the lower limit of the design parameters (e.g., wrt AC ripple, reserve capacity).

More often than not, the original capacitance rating specified during manufacturing was often limited by capacitor technology of the day, the cost of component and the physical size (finite space available on the PCB). For example, the driver board local decoupling capacitors specified in the Sansui AU-7500 are 100µF/50V (12.5mm diameter), whereas those in the more powerful AU-9500 are only 47µF/80V (12.5mm diameter); the determining factor in this example is the available physical board space with then current capacitor technology rather than cost since the AU- 9500 was a TOTL model.

On the F-1153 board, the C003, C005 and C007 capacitors filter the DC voltage and the replacements have 2-3X capacity of the originals. For the capacitor at the base of a pass transistor (C006 on F-1153), it is recommended that it be no larger than 470µF. Regarding the AU-666's Ripple Filter Boards (F-1231, F-1268 and F-1277), increasing the 100µF and 220µF capacitors that are on the base of the pass transistors to 220µF and 330µF, respectively, should be considered.

 

[Leestereo]

Part 7: WOOD CABINET

This particular AU-777A arrived with a wood cabinet; it was in good structural condition, but its peeling, dull finish left much to be desired. The case is not an original Sansui one (an OEM case is shown in this post by @smurfer77). The tell-tale differences were the absence of wood runners on the interior bottom and the unit can only be secured within the case via two screws on the bottom rear (there were no openings for fasteners on the case bottom near the front).

To allow the unit to be secured in a manner similar to that of an OEM cabinet, two hardwood strips were installed and holes were drilled to allow screws to secure the unit at the front as well as the rear of the unit. The old finish was removed with a chemical stripper and the bare wood was lightly sanded. A tinted oil finish was then applied to the case and once cured, it received several coats of wax. The wood cabinet isn't flawless but it is nevertheless quite handsome; its imperfections are testament that it is indeed ~45 years old.

 

[greymatter]

Nice job on the cabinet Ben .

 

[smurfer77]

Part 7: WOOD CABINET

This particular AU-777A arrived with a wood cabinet; it was in good structural condition, but its peeling, dull finish left much to be desired. The case is not an original Sansui one (an OEM case is shown in this post by @smurfer77). The tell-tale differences were the absence of wood runners on the interior bottom and the unit can only be secured within the case via two screws on the bottom rear (there were no openings for fasteners on the case bottom near the front).

To allow the unit to be secured in a manner similar to that of an OEM cabinet, two hardwood strips were installed and holes were drilled to allow screws to secure the unit at the front as well as the rear of the unit. The old finish was removed with a chemical stripper and the bare wood was lightly sanded. A tinted oil finish was then applied to the case and once cured, it received several coats of wax. The wood cabinet isn't flawless but it is nevertheless quite handsome; its imperfections are testament that it is indeed ~45 years old.

Looks good to me and the ventilation slots is in the right area, over the driver boards.

 

[rasenthiran]

wow...Nicely done

 

[SansuiJavier]

Ben, I admire your work and your patience in explaining it here.
I notice in my 777 (not A) that when I turn it on it sends power to the speakers and the woofer moves with a low sound. Does it also do so after restoring the 777A? Is it dangerous for the speakers? Can it be avoided in any way? Thank you !!

 

[pete_mac]

Ben, I admire your work and your patience in explaining it here.
I notice in my 777 (not A) that when I turn it on it sends power to the speakers and the woofer moves with a low sound. Does it also do so after restoring the 777A? Is it dangerous for the speakers? Can it be avoided in any way? Thank you !!

I'd suggest turning the speaker selection switch to 'off' before switching older amps such as the 777A (which don't have a soft-start delay protection circuit) on or off.

 

[Leestereo]

...I notice in my 777 (not A) that when I turn it on it sends power to the speakers and the woofer moves with a low sound. Does it also do so after restoring the 777A? Is it dangerous for the speakers? Can it be avoided in any way?...

With amplifiers that utilize a capacitor-coupled output stage, at turn-on, the output capacitors charge up and this is the cause of the woofer movement and sound you noted; it is normal behavior and is not typically dangerous for speakers. In some Sansui models, including the AU-777A, this "charging-up" effect is not apparent since there are 150ohm/5W resistors (R829/R830) shunting the output capacitor to ground. The earlier AU-777 (non-A) model doesn't have these resistors, but if you installed them, this would mitigate the turn-on thump.

 

[pete_mac]

With amplifiers that utilize a capacitor-coupled output stage, at turn-on, the output capacitors charge up and this is the cause of the woofer movement and sound you noted; it is normal behavior and is not typically dangerous for speakers. In some Sansui models, including the AU-777A, this "charging-up" effect is not apparent since there are 150ohm/5W resistors (R829/R830) shunting the output capacitor to ground. The earlier AU-777 (non-A) model doesn't have these resistors, but if you installed them, this would mitigate the turn-on thump.

Great info!

 

[SansuiJavier]

With amplifiers that utilize a capacitor-coupled output stage, at turn-on, the output capacitors charge up and this is the cause of the woofer movement and sound you noted; it is normal behavior and is not typically dangerous for speakers. In some Sansui models, including the AU-777A, this "charging-up" effect is not apparent since there are 150ohm/5W resistors (R829/R830) shunting the output capacitor to ground. The earlier AU-777 (non-A) model doesn't have these resistors, but if you installed them, this would mitigate the turn-on thump.

Excellent information. I will do it, and I will comment on the change. Thank you !!

 

[smurfer77]

With amplifiers that utilize a capacitor-coupled output stage, at turn-on, the output capacitors charge up and this is the cause of the woofer movement and sound you noted; it is normal behavior and is not typically dangerous for speakers. In some Sansui models, including the AU-777A, this "charging-up" effect is not apparent since there are 150ohm/5W resistors (R829/R830) shunting the output capacitor to ground. The earlier AU-777 (non-A) model doesn't have these resistors, but if you installed them, this would mitigate the turn-on thump.

Please correct me if I'm wrong but that seems only partly right to me. Read-on, but essentially a connected speaker is much more efficient path to ground than the 150 Ohm resistor. I.e. the resistor may (*) enable defeating the thump when using the speaker-OFF start-up method but won't help much (~10%) for Speaker A or B start-up position. Forgive me if that's what you meant (it's just a charging path for speaker OFF position, not a solution for speaker A or B start-up), but thought it's best to be clear on the matter before people start opening up their amps.

Whatever DC thumps are presented at the top of the 150 Ohm resistors must also be presented to the speakers, and most of the current will go through the speaker due to it's much lower impedance, say of the order of 10 Ohm. I.e. the addition of the 150 Ohm resistor will only shunt away about 1/10 of the thump. I.e. It would be more accurate to say the speaker is the shunt to ground, rather than the 150 Ohm resistor.

The 150 Ohm resistor does help remove switching thumps, and does for example to allow you to set speakers to OFF during start-up, have the coupling cap charge up and then switch to speaker A or B and avoid the thump. Without such a resistor the cap charge and thump might just be delayed until you complete the circuit by switching to speaker A or B.

For those wanting to add such a resistor, for example in the 777, but don't want to open up the unit, you can connect the resistors externally over speaker B and use speaker B for start-up then switch to your actual speakers on A after some time. Alternatively, set to OFF, and connect a pair of headphones that you don't care about (that will also complete the output circuit as headphone is just knocked down from outputs by resistor).

*However, be aware that such thumps are sometimes due to other start-up transients upstream, and non-transient DC offset can even cause this too. I.e. if switching speakers to OFF during start-up solves your thump problem (even when later switching to speaker) in a model without the resistor, it may not be output cap charge-up related and addition of the resistor may not solve the issue.

 

[kevzep]

Just turn the speaker selector off at power up an power down.
The world is not perfect and neither are these amps, I would hate to think what horror show would happen if there was output to the speakers on some of the bigger amps when they power up.