Discussion in 'Exclusively Sansui' started by SanthoshA, Sep 21, 2017.
Is this a resistor directly sharing a trace with a leg of a trimpot? I never heard this term before.
Yes it is basically the resistor connected in series to the trim pot.
Padding resistors are mainly used for attenuation if i am not wrong.
The value of this resistor in the AU-X11 and the AU-X1 are the same, the resistors are rated at 5.6kΩ, and denoted R17 in the AU-X11 and .(R73,R74) in the AU-X1.
Basically decreasing the resistance using the trim pot, does not help us attain the required calibration value which is -11V.
With the trim pot set to null or zero resistance the maximum value you get across test points in my case at least is -10.5V and it never fails to increase beyond that..
So we decrease the value of the resistor in series with the trim pot, as the circuit sees a net resistance from trim pot along with the resistor in series.
Decreasing it to 3.9kΩ just as John did, is a good way to go about this, it will compensate for the voltage, and then you can increase the pot to have precise control and set it to the required -11V.
Guys, after the upgrade i gave a listen to the amp for couple of hours.
There is a day night difference between the restored channel, and one that is untouched, i think this is an attribute of the Metal Film Resistors used for the upgrade.
And also the restored channel runs On Semi Transistors opposed to Original transistors in the unrestored board.
I can clearly say even with On Semi's, the restored channel is superior when compared to the unrestored channel.
Apart this i seem to have unravelled another mystery.
I have A968B and C2238B gain rating (O) for both of them in the left channel/unrestored board.
Whereas i have A968B and C2238B gain rating (Y) for both of them in the right channel/restored board.
@Hyperion: John, sorry to drag you in, but can you please by any chance let me know the gains of the transistors used in the AU-X11 you are currently working on, if by any means the section is clearly accessible for you. I just want to know if the transistors used in left channel driver board and right channel driver board are using the same gains. But yes i am not sure if this change actually makes a big difference,you could advise me of the same.
I think it is these transistors that cause an imbalance in the output stage.
I noticed the following with a 1KHz sine wave sent to the amplifier.
1. With preamp volume full, and gain pots set to max, the voltage across left and right channels are almost the same.
2. When i decrease the preamp volume to half, the difference in AC output voltage across left and right channels is around 0.5V to 1V.
What i noticed is that the difference in AC voltage across left and right channels varies with change in preamp volume know, in certain positions the difference ranges between 0.5V to 1V and in certain positions the difference is quite marginal.
Can it be due to differential gains on the pre-driver transistors(A968B and C2238B) that cause this problem, i had the same issue even before restoring the amp, and this is all i can make out for now.
Basically the Y rating of the transistors have superior gain than the O rating.
Abstract from data sheet above.
Any pointers to what causes the output difference, is it mainly limited to pre driver transistors or is this something else i need to look into.
The challenge is voltage measurements in driver board, this merely not practical, with it mounted on the amplifier, and the short length of wires.
Do you guys think it is a good idea to remove both driver boards and compare them back to back with the other, and also verify all transistors used in them ?
As for now i have noticed the differential gain used in pre-driver transistors.
Meanwhile i have removed the protector board to replace the faulty Hitachi Reed Relay and also to recap the board, luckily there are no trim pots in the board, but the board has been oxidised and the PCB traces have becomes rusty.
I will post pictures and appropriate documentations when free, but the above issue mentioned is bothering me a lot .
Thanks for helping me out !
A couple of things...
How are you altering the input signal to the pre-amp? it seems like you are using the front panel gain controls - try using the signal generator output level control. Note that the differences you see could be caused by dirty power amp level control pots?
Also didn't you say that in one channel you have the MJ211xx sacrificial OP transistors, and in the other channel the original OP transistors? - I don't think this imbalance you see is of much consequence - at least not until you have completed the reconditioning - lots of possible reasons.
You can't compare 'restored' with 'unrestored' and expect there to be no difference!
Yes I truly agree it is not a wise comparison between the restored channel and the unrestored channel.
The reason I wanted to check the voltages across left and right channel is because I faced the same problem before I started with the restoration.
I picked the Right Channel driver board for restoration as it was slightly weaker than the Left and I have the same results before/after restoring the Right Channel driver board .
I leave the power level knobs to full, and vary the bigger preamp knob alone, signal fed to aux in .
I have also cleaned all the gain pots and also the preamp volume knob(bigger dial).
I will try using external power amp inputs and then vary the signal in the signal generator and see if i can rule out the preamp knob(inactive when using external power amp inputs).
But like you say there are too many changes in one driver board compared to the other !
We have On Semi transistors on one channel and Sansui transistor in the other channel.
For now I will note down the voltage results for reference and probably check again after completing the restore.
Hoping the Left Channel Drive board restoration will solve this issue, fingers crossed .
Good News !.
The headphone out is working fine now , i have replaced the Green Hitachi Relay with NEC.
Thanks to @Kale for helping me with the spare parts.
Power Supply And Protector Board (F-3462):
In order to remove the board completely, all plug on wires were removed and i had to de-solder two wires attached to the resistors shown below.
I had to also remove the power LED connector harness for ease of removal.
Then re-route the bus wires so the board can be worked on.
Power Supply And Protector Board (F-3462):
Note a diode and a resistor are connector in series as shown in above image.
As you can see the rear of the PCB is in pretty bad shape, i managed to clean up the mess a little bit and also finished the recap.
Unforunately, i do not have a better pic of the completed board apart from the one below.
I have replaced the 4 bigger Elna capacitors with Nichicon KZ, the third leg of the old Elna capacitors is just a dummy leg to hold it in place. You can see the newer capacitors are much smaller, and i have used the same values and voltage ratings here.
1. Headphone out
2. Small static that was there when touching the knobs/front panel is now no longer present.
I believe the second problem, got solved be re-soldering all the joints in the board(or cleaning the ill fated glue residues all over the PCB), but i cannot say for sure, but the static is no longer present now.
Baby that new reed relay. They are not the most rugged component. I burned mine out by experimenting with FETs on my old AU-X1's EQ boards.
A replacement for one is dicussed here.
Nice work Ron !
I will keep this point noted and be extremely careful in not damaging it by making mistakes .
The amp is sounding so much better than before that I hardly don’t get the feel to tear it apart for reworking the other pending driver board .
It is quite saddening that I have still left it untouched .
I am also going to compare the NMA1718 and NMC1718 with the sankens used in the X1.
I will post the results sooner when I have some time.
Without the help from you guys I would have never made it this far !
The results have been very much satisfying and I am very delighted to complete the full restore .
This is a very common effect.
You have posted interesting pictures of the underside of the PSU/protection board.
I have yet to get a proper look at the underside of the one I am doing.
I hope it is helpful, another point which i want to add is that there is a trace cut on underneath the PCB.
If you look closely, the area boxed in Red has the trace cut.
It is quite tough to spot in the full image added in posts #187 above.
I have another update, the left channel driver board is removed now.
Unfortunately it is pretty bad shape, one of the ground lugs was previously pulled off, or something i don't know what they tried, but the PCB traces, have come off in the particular area.
The ground lug was soldered to the pin beneath it, as you can see from below picture, i have solder wicked the entire area clean.
The white ground wire was hardly clinging onto the second through hole socket, that it came off so easily when i desoldered the joints.
I was a pain to see the PCB board in this condition, after removal, i never had the heart to post this picture, but then here we go.
The PCB, has also chipped off along the edges in multiple areas, i guess this is due to slipping from the helping hand or third hand when the board has been worked on.
Any pointers in fixing the issue with the missing traces/half copper pad on the second joint.
I am reluctant to go ahead in using the same method, the ground wire lug can be soldered back, but i want to see if there are options to fix the PCB board in this condition.
Solder Re-Work on the underneath of the PCB, can be seen in multiple areas, the pre-drivers have been reworked, removed and may be re-greased/replaced, the black flags are not present this is all i could check very very quickly.
Overall, this board is in pretty bad shape/very bad shape.
I struggled to remove the output transistors mounted on the heat sinks, the screws were jammed, and i finally managed to get them all out .
You can use de-soldering braid to achieve a workable repair for broken traces like your one. Solder a length of braid onto the remaining track, then stretch the braid over the two lead-wire holes, fashion holes in the braid at the right locations by spreading it (not to the full size) and then attach it to the board with super glue. Insert the pins which will stretch the braid and then solder pins to braid. It's never going to look like an original circuit board but it is functional.
Hope this helps.
I'd go the Kale route, using copper shielding foil and J-B Weld Twin Tube™ original,
J-B ExtremeHeat™, or J-B HighHeat Epoxy Putty™ adhesive.
Here is a Kale PCB repair from http://audiokarma.org/forums/index....-110-volt-9090-receiver-to-the-220vac.714437/
Clean the area to receive the adhesive.
Apply the adhesive.
Apply the copper patch.
Apply low tack masking tape over the patch.
Over the tape, put an eraser or whatever can evenly apply pressure to the area.
Then clamp the eraser (or whatever is used to even out the pressure) in place.
Wait 24 hours.
Remove the clamp, eraser, and tape. Carefully clean / scrape away excess adhesive.
Looks neat and clean, as usual you seem to be the database of information when it comes to a restore.
Really appreciate your effort in putting forth the detailed steps .
I will try this procedure , thanks for tip Ron !
The second driver has been re-conditioned:
Following have been replaced/introduced:
- Carbon Films -> Replaced with MultiComp Metal Film Resistors
- Non Flammable Resistors -> Replaced with Vishay Dale Non-Inflammable Resistors
- Trim Pots -> Replaced with Bourns 3296P Series
- Wire Wound Emitters -> Replaced with Vishay LVR Non Inductive Wirewound
- Base Resistors -> Replaced with Vishay Non-Inflammable Resistors
- Vishay Polypropylene 0.22uF Bypass capacitors were introduced to bypass square transformer input voltage
- All electrolytics replaced with Nichicon Fine Gold with same capacitance and voltage ratings
- Original Output transistors removed, and On-Semi MJ21193G and MJ21194G have been put back in their place.
- Replaced all pre-drivers(both boards) with ones i had in stock, to match the uniformity between both driver boards.
Summary of how it was earlier:
Right channel driver board: A968B and C2238B(Y) rating, one of the A968B's had very low Hfe gain and poor measurement response.
Left Channel driver board: A968B and C2238B(O) rating, all of them were in pretty good condition.
Replacement used: A968B(O) and C2238B(Y) for both channels.
All transistors re-greased, and new thermal paste from Aavid Thermalloy applied.
I have replaced the pre-driver transistors (both boards) with stock which i had.
It currently uses a Y Gain for C22338B and O Gain for A968B, i wanted to keep both boards in sync, as the other combination which i had earlier was pretty much odd.
Finally the picture of the restored driver board:
I have some tidying up of the wires to be done, and the PCB fix for the rear.
I found couple of out of spec carbon films, and have replaced all of them,unfortunately i did not make a proper note of them.
Diodes In the Driver Board:
1. All the circle shaped diodes with a white dot marking are 3 junction diodes.
2. All the brown boats are also 3 junction diodes with an exception for D7 which is a 4 junction diode in boat package.
Well...ta freakn da!
So the X11 is ready to rumble?
Same Q here. (Kinda)
Separate names with a comma.