Sansui 661 refurbishment October, 2024
This Sansui 661 is for sale in the Netherlands, fixed price 300 Euros, cash and Local Pickup Alkmaar area as I do NOT ship,
I only answer if you provide your phone number also, see the moving mail address on my main page.
Gemakshalve is deze pagina verder in het Engels geschreven.
This page is written using English language only.
This Sansui 661 was bought reported it worked perfectly until one day smoke came out.
There were no visible problems, but it smelled right away after power up, the heat source was found. After finding the reason on the power board I decided to give it a recap, and fix defective lighting, assuming the tuner and amp circuits worked.
The innards of the receiver looked clean, for its age.
All information contained on this page assumes interested readers to provides themselves with the schematics and manual as to be found on the internet.
The innards of the not-yet-refurbished receiver from the top.
Note: the cover of the main amplifier circuit card is removed.
The sansui 661 F-1500A Power Circuit Board
This board contains the transformer connections and power distribution, rectifiers, and stabilization circuits.
Additionally, it carries the main amplifiers circuit card F-1499A on a circuit card, and the associated power transistors including their mounting/heatsink, as well the relay operated loudspeaker protection circuit
On the picture below, there were two capacitors looking extremely hot according my infrared thermometer device. On the enlarged part on the bottom, a defective cracked diode seemed to be the cause.
The enlarged part on the left reveals a hot spot on a ceramic cap and the impression was this could also be because of a hot TO220 type 2SD330 regulator transistor.
The ZD2 zener diode was missing, it was found on the bottom of the card, since the holes drilled did NOT match the schematic.
One of the soldered-in fuses was defective, also.
The ZD2 diode was found on the bottom of the card. Af funny detail is when the board was still powered on and the zener measured, one more piece of glass cracked off but it still worked.
I decided to depopulate the card, after having repaired and tested it first.
I replaced the three diodes, as well as the one supposed to snub the relay switch-off effects, by BYV96E ones.
The old rectifier diode double pairs were replaced by 1N5408 diodes, the 3300 Microfarad power smoothing capacitors by Kemet ALP22 4700 Microfarad ones, and the 1000 uF one for the regulated supply by a 2000 uF Panasonic FR type.
Also other electrolytics were upped for capacitance if assumed approppriate. The soldered in fuses are socketed ones now, a hole on the right position was drilled to have ZD2 on the good side of the board.
Although the old relay worked, as part of the update I decided to mount a new one. The power transistors did get new thermal paste. I decided on a more robust driver transistor for the relay, a KSC2632, driven by a KSC945, now.
Both power regulating transistors did get a heatsink, the 2SC875 later on replaced by a more rugged 2N3440, to accomodate a lighting modification demanding a then unestablished yet current increase.
There was a hole to accomodate a pin at the 23 Volts regulator output, it was not used since it was just wired at the bottom. A pin was put in to feed the lighting modification.
This all makes the Power Circuit Board look much different, although it functions the very same!
Both power regulating transistors did get a heatsink, the 2SC875 later on replaced by a more rugged 2N3440, to accomodate a lighting modification demanding a then unestablihed yet current increase.
The extra "heat sink" on the 2SD330 actually simply consists of a heat conducting ceramic mounting plate. Additionally the power resistor over there was raised up for better heat flow. In the end, it looks everything stays pretty cool, now.
There was a hole to accomodate a pin at the 23 Volts regulator output, it was not used since it was just wired at the bottom. A pin was put in to feed the lighting modification.
The power amplifiers card, F-1499A "Driver Circuit Board"
I decided to recap it but not a total rebuild.
An electrolytics recap was done. As usual, the small amps input capacitors are film type, now.
The input transistor pairs were replaced by closely matched pairs of KSA992.
Due to the offset trimming design, its trimmer potentiometer is much too sensitive, due to the diodes as a reference one can swing the offset to 0.6 Volts!
I drilled holes and put the trimming potentiometer taper to ground using small 560 Ohms resistors. Also two small 33 uF capacitors were added to stabilize any fluctuation because of power supply changes or noise of the trimming pots. (see the arrow on the picture) Now, one can easily trim the amplifiers offset voltages to withing some millivolts.
Evaluating the power amplifiers, things which still COULD be done is adding a "speedup" capacitor over the "VBE multiplier" transistor, and do the "baxandall diode modification".
Depending on whether this receiver stays, I may do this later and measure distortions differences to get a grasp whether how much beneficial it really is.
F-1501 Preamp Circuit Board.
This board was recapped, using assumed suitable components.
A nice service-friendly feature of the receiver is the fact one easily can remove the tuner section, so working on the preamp is easy, without needing to remove it.
Observe, there are no sign of spray on potentiometers, this board is still quite clean.
The F-1491A Tuner Circuit Board.
Two tarnished trimming potentiometers were replaced by square looking yellow ones.
All the electrolytic capacitors were replaced, most of them are 4.7 microfarad BC ones, now.
This was the only work done on the tuner section.
The receiver lighting.
The old lighting was partly defective. The meter backlighting consisted of one backlight tubular lamp and two tiny 5 Volts lamps in holes in the plastic back of the meter. Below, the picture of the disassembled meter shows some debris caused by one of the little lamps witch apparently burned. I had to take the meter apart and clean it.
The scale backlighting comprised of three 7 Volts tubular lamps, also they were replaced by a LED string.
Next to it, is the picture of the type of LED string, including size, spec and partnumber.
Unfortunately, I forgot to make pictures of the work on the lighting. On the tuner section above, one can see the additional parts mounted on the back of the lighting module.
The stereo lamp remained as is.
I used a big round 2N1486 transistor, a 16 Volts BZT03C16 Zener diode, a 2.5K trimming potentiometer to feed the LED strings for the meter and scale backlighting. (see them mounted on the tuner section pictures).
On the picture on the right one can see the previously unused hole on the circuit card, where the 23 Volts is taken, using an additional wire going with the wiring to the lighting module.
Some pictures of the receiver.
Ga naar Gerards page / go to Gerards page ---->>>