Sansui 4000 Problem 72.2 volts on the main rail?

[murfman]

Long time lurker but retired and decided to purchase a solder/desolder station from parts express, a recap kit off that site (hard for a newbie to order parts) replaced all the caps and lo and behold everything worked perfect and the 4000 sounded great!! I got the bug now!!!!!

Purchased a Sansui 800 off Craigslist for 75 bucks and with excellent write up and parts list from stereofun I ordered all the parts from Mouser with an additional 100 extra KSC1845, and KSC1815 transistors. Guess what that little Sansui really does sound better than it should. Thanks stereofun!!!!

Well after the recap of the 800 I realized my 4000 also has the dreaded 2SC458 transistors 2 on the protector block and 1 each on the driver amp. Replaced the protector block with KSC1815 transistors and the driver amp with KSC1845 transistors. Less than a week later unit not working properly. Volume distorted, one channel cutting out all kinds of problems. Let it cool down and turn on sounds good for a bit? Maybe coincidence with transistor change maybe not I sure don't know.

So back to this site figure I had better do adjustments on the current and output from the 4000 service manual. Found I could not quite get the current up to the 10mA on the right channel so I switch the KSC1845 transistors for KSC1815 transistors and can now just get to the 10mA.

Move now to the output adjustment (Bias) and do not have a scope (wouldn't do me much good anyway) I use the measure the voltage a c007/c008 and get 72.2 volts? Suppose to be 66 volts? I still use the 1/2 -.5 volts and vr801/802 to get 35.6 volts. Unit sounds fine but haven't ran it long as I know 72.2 volts probably not a good thing? Long story but here's my questions.

1. Could my transistor swap effect the rail voltage?
2. If not suggestions for where to start to get voltage to 66 per spec
2. 1815 or 1845's for the protector block and driver boards?
3. Good repair shop in the Metro Detroit Area (Ya know I'm eventually going to need it)

THANKS FOR ANY AND ALL HELP GREAT SITE!!!!!!!

 

[Hyperion]

I am guessing you are in the US? - Does the unit have a voltage selector? if so, what is the mains input voltage of the unit set to?

If the unit is 'single voltage' (i.e. 117V) What does the mains voltage where you are run at?

Answers to these questions could explain where the extra voltage is coming from.

The high voltage secondary winding of your transformer is outputting ~52V (for ~72V rectified), when it should be outputting ~47V for ~66V rectified. The secondary winding output is proportional to the primary winding input.

 

[murfman]

Correct live in the US, unit does have a voltage selector and set to 117/127. Just checked line voltage and running at 122.

 

[stereofun]

Right on Hyperion. I just adjusted a "US" 4000. The selector was set for 117V - this is the best match to my home in Arizona where the mains are around 120-122V.

Some 4000's are set for 110V - the option one step down - so check those things first. As for measurements, I still have them out on the bench and I had exactly 69.7 on the supply caps.

On Bias I had 17 / 25 which I adjusted to
20/20.

P.s the 10mA you adjust for across the empty fuseholders via VR803 are bias ! Sure you turned VR803 ? 10mA is on the low side.

The adjustment where you measure the single rail and divide by 2 is the centre voltage where you split the single rail evenly between the two channels.

Not sure what's causing the popping when heating up....it could be a marginal solder joint that when amp heats up, it just begins to loose connection. 1845 vs 1815 are unlikely to cause this unless you really busted the amp rating. I think one is 50 mA and the other 150, so just make sure that they are not less than what they replaced. Again not likely to be culprit, but good practice to go over work and check these things including the actual solder work.

Glad you liked the 800 - impressive little thing for sure !

 

[murfman]

Thanks for the responses. Decided that I'd better check my readings with the comment that 10mA is on the low side so went to check and blew the meters 500 mA fuse!!!!! Maybe I was in the wrong range and was reading 100mA??? Or 1000mA??? Trip to Ace hardware, Lowes, and Home Depot a bust so trusty Amazon will have them delivered to my door tomorrow! Ordered 10 cause you never know. Will get new readings tomorrow and go from there but pretty confident that I will still have 72.2 volts instead of my 66??? Can put the old 458 transistors back in to see what happens but they sure looked bad when I took them out!!!

 

[stereofun]

No worries about the 72v for now - I had 69.7 - we are both higher than spec, but within 3% of one another. Mine runs superb.

While I think you should replace the drivers 1845 (50 mA rating) for 1815 (150 mA rating) to better match the outgoing 458's (100 mA) I don't think the problem lies there. I do wonder if the bias adjustment went wrong and you are running a very high bias causing the amp to begin distortion and popping as it quickly begins to run hot (noticed hot sinks ?)

Get the bias checked out first thing - if any of the VR803 pots are almost fully to one side (I cant remember which side is max vs min bias) just put them in the center position for now. Remove both fuses. Move your meter probe to the mA slot and select DC mA. PLEASE use alligator clips so you can secure the probes across the fuse holder before turning on amp. Turn on amp with no signal, check its not on tuner) and watch bias like a hawk. Make sure you have already identified the correct pot to begin turning (match fuse to driver board, both have L/R markings) - let us know what you get ?

 

[murfman]

Fuses for meter arrived and able to only get 11mA on the left channel with the pot maxed out. Set both left and right for 11mA . Spec calls 10-15 so good enough I hope. Rail voltage 72 so half would be 36 so should I adjust for 36 or 33 which is half the 66 expected. Till I hear differently I'm going to split the difference and target 34.5 and let her rip and see what happens!!!! THANKS for all the help!!!!!

 

[Hyperion]

Till I hear differently I'm going to split the difference and target 34.5 and let her rip and see what happens!!!! THANKS for all the help!!!!!

The setting should be half of whatever rail voltage you have, so the setting in your case should be 36V.

 

[raj12345]

The setting should be half of whatever rail voltage you have, so the setting in your case should be 36V.

Where shall I measure this 33 or 36 volts for the half of rail voltage ??

 

[EastPoint]

Set your multi meter to DC volts. For each channel there are two output transistors. Put your black probe on the chassis ground and your red probe on the metal case of either of the output transistors for each side. One will measure about 70 V and the other one will measure about 35 V. That’s where you measure the DC balance.