Opinions on state of AU-20000 sought

[richardpp]

4. A comparison between a BA-3000 and the amp section of the AU-20000 is intriguing. To my eyes, the amps are the same (same design, same boards) yet for some reason bias on the BA-3000 is set at 20mA (20 mV measurement), while the AU-20000 is set at 35mA (35 mV measurement). The wording on the service manual pages for setting bias in each unit is identical (down to misspellings!) except for the bias current. The only difference I can see is a difference in the value of the dropping resistors on the BA-3000 (330 Ω) vs. 220 Ω on the AU-20000. In @Leestereo 's restoration thread of a BA-3000, the dropping resistors are mounted on the PS board, but the original design only used a single 330Ω/5W resistor vs. a combination of 3 resistors on front and back of the board in the AU-20000. I assume the lower bias current setting in the BA-3000 leads to less heat generated, but it would be interesting to have some quantitative data. I know I'm missing something regarding what's happening with F-2508 in the two units. The schematic shows ±85V input to the board and ±65/±35V output from the board in both SM, so I'm not understanding why the difference in dropping resistors and bias settings. OT's are the same, power out put is the same.

The difference in the dropping resistors can be explained by the differing demands placed on the regulator. In the AU-20000, the entire control/ preamp section i.e. phono, eq, tape, filters etc., runs on 35V. The BA-3000 in comparison has far lighter duty, i.e. filter and meter circuits.

The BA-3000 has 330Ω/5W resistors in the low voltage regulator. The schematic indicates a theoretical voltage drop across the resistors equal to 11V, representing a current draw of ~33mA. This compared to the AU-20000 which has a measured current draw of ~120mA. To adapt the regulator to the increased current draw Sansui decreased the value of the dropping resistors. The schematic show the value at 270Ω, (150Ω + 120Ω). In actual fact, with the addition of 1.2KΩ parallel resistors on the conductor side of the board, the value became 220Ω.

The bias setting of 20mA comes from the BA-3000 service manual. Although the development of the BA-3000 moved forward in conjunction with the AU-20000, to my knowledge the manual was never updated. Developments included the 'dash one' driver board (F-2507-1), a revision which routed unregulated power to the pre-drivers in place of the regulated supply. Another significant development was a new emitter resistor board (F-2497) for the AU-20000 with a physically larger layout than the older (F-2509) belonging to the BA-3000. In short, I think the 35mA bias spec of the AU-20000 could apply to any BA-3000 with the above production changes but otherwise, without them, I would keep it to 20mA.

Finally, both models got an updated driver board, re-designated F-2507, that saw the bias compensation transistor moved from its old location on the output heatsink to a new location on the driver heatsink. I found this change to have a substantial improvement in thermal stability. I think the tendency for the bias to creep in the older units meant that they ended up running quite hot. I'm still trying to find out when this change was made. I suspect it was fairly early on looking at the number of examples on the web that have the new boards. The development of the AU-20000 from the BA/CA 3000 combo may have been a little troubled in the beginning, but I think by early (to mid) production I would give the AU-20000 its due.

Hope this helps.

 

[Sansuiman]

@richardpp, good info there. I gather that was gleaned mostly from your personal experience and observations, as I've never run across any factory tech bulletins concerning these changes, as was sometimes seen. Sansui appears to have been inconsistent in their issuance of service bulletins and service manual updates.

It's been quite a while since I was inside my BA-3000, so cannot recall if it has the original or the updated design and board compliment. I do recall setting bias at 20mA, per the manual.

 

[richardpp]

@Sansuiman, I've never come across any technical documentation apart from the relevant service manuals, so yes mostly from experience and observation. The manuals are like a snapshot in time and as flawed as they sometimes can be we would be lost without them!

 

[KeithD]

The difference in the dropping resistors can be explained by the differing demands placed on the regulator. In the AU-20000, the entire control/ preamp section i.e. phono, eq, tape, filters etc., runs on 35V. The BA-3000 in comparison has far lighter duty, i.e. filter and meter circuits.

The older BA-3000 has 330Ω/5W resistors in the low voltage regulator. The schematic indicates a theoretical voltage drop across the resistors equal to 11V, representing a current draw of ~33mA. This compared to the AU-20000 which has a measured current draw of ~120mA. To adapt the regulator to the increased current draw Sansui decreased the value of the dropping resistors. The schematic show the value at 270Ω, (150Ω + 120Ω). In actual fact, with the addition of 1.2KΩ parallel resistors on the conductor side of the board, the value became 220Ω. I believe this became standard for the BA-3000 as the power section of the two models converged.

The bias setting of 20mA comes from the BA-3000 service manual. Although the development of the BA-3000 moved forward in conjunction with the AU-20000, to my knowledge the manual was never updated. Developments included the 'dash one' driver board (F-2507-1), a revision which routed unregulated power to the pre-drivers in place of the regulated supply. Another significant development was a new emitter resistor board (F-2497) for the AU-20000 with a physically larger layout than the older (F-2509) belonging to the BA-3000. In short, I think the 35mA bias spec of the AU-20000 could apply to any BA-3000 with the above production changes but otherwise, without them, I would keep it to 20mA.

Finally, both models got an updated driver board, re-designated F-2507, that saw the bias compensation transistor moved from its old location on the output heatsink to a new location on the driver heatsink. I found this change to have a substantial improvement in thermal stability. I think the tendency for the bias to creep in the older units meant that they ended up running quite hot. I'm still trying to find out when this change was made. I suspect it was fairly early on looking at the number of examples on the web that have the new boards. The development of the AU-20000 from the BA/CA 3000 combo may have been a little troubled in the beginning, but I think by early (to mid) production I would give the AU-20000 its due.

Hope this helps.

Richard, your explanation makes a lot of sense, thanks for those clarifications, and I really appreciate you sharing the research you've done!

Early on in this discussion I felt that Sansui must not have taken heat generated by the dropping resistors into consideration in the early design of the AU-20000, but now I'm thinking maybe they did - using three resistors instead of two: one on the back of the board and two in series mounted perpendicular to the board. Looking at a few BA-3000 restoration threads, it appears the 330 Ω resistors were mounted horizontally on the PS board. The restoration pictures I saw showed a fair amount of heat scoring like seen on early AU-20000 PS boards. So, even though there is less current going through the dropping resistors on the BA-3000, there seems to have been a similar issue with heat dissipation. Did Sansui relocate the dropping resistors in later BA-3000 production?

 

[richardpp]

Keith, I don't believe that Sansui relocated resistors to F-2512 in the BA-3000. I stated in my earlier post that they did but that's not correct. It appears to be just in the AU-20000 that the resistor changes were made. Apologies for the error and thanks for making me aware of it. I agree that heat dissipation was already a problem in BA-3000 well before the introduction of the AU-20000 as you've pointed out.

I've edited post #81.

 

[KeithD]

Replacing filter capacitors will be next. I bought Nichicon LNT2A223MSE 22,000µF/100V, which are the right size to fit in the brackets and space. Anything I need to watch for in the replacement? Assume there will be a significant current draw at first power up and wondering if I should use or not use a DBT for first power up? Looks like a straight forward replacement.

 

[w1jim]

I don't think there will be any additional current draw just because they're new.
No different then powering it up after turned off for a minute or two.

 

[Hyperion]

Personally, I power up with DBT for the first time when I have installed new main PSU capacitors. I believe there is a forming and conditioning process that happens (just once), when such large capacitors are first put into service. This action also guards against the possibility of a faulty new component or a wiring error, maybe this is just my OCD self having a moment, but I feel much happier doing it this way.

 

[KeithD]

Thanks for the insights @w1jim and @Hyperion . New filter caps installed.

 

[KeithD]

Been listening to this for the past week. Noticeable improvement in the bass, more punchy and fuller. I think a good decision to replace the filter caps. Old ones tested only 10% high. I wasn't really expecting to hear much difference, but there definitely is.

Third parts order placed to restore the protection board and pre-amp boards. Moving ahead!

Also, I found this picture at some point on the net. I saved it because I have the same cassette deck, though mine has a wood case: Teac A-640.



Just happened to look at it today, and noticed the amp under the cassette deck! I guess "the boss" had great taste in audio gear!

 

[richardpp]

Hi Keith, good to hear about the improved bass response from the new filter caps and nice photo of "the boss" with the AU-20000, running hot no doubt!

Third parts order placed to restore the protection board and pre-amp boards. Moving ahead!

Have you repaired the pass through connections on the emitter resistor boards? This is important because they are known to crack.

 

[KeithD]

Hi Keith, good to hear about the improved bass response from the new filter caps and nice photo of "the boss" with the AU-20000, running hot no doubt!



Have you repaired the pass through connections on the emitter resistor boards? This is important because they are known to crack.

Yeah, hope he didn't "cook" the amp! The A-640 is a beast itself, weighing 30 pounds without the wood case.

I plan to inspect the pass through connections on the F-2497 boards. All output transistors and bias transistors were replaced before the previous owner, so possibly the pass throughs were re-done at that time. It's on my list of things to do. Progress is likely to slow substantially as fall semester rapidly approaches.

 

[w1jim]

If the bias transistors are mounted on the power amp heat sink then you could be looking at trouble.
The later models mounted them on the drivers.
That's what cooked my amp.

 

[KeithD]

Got back to the 20000 today. Pulled the protection board F-2511. One thing I noticed after removing the electrolytic caps, this board does not have C11, while later boards do. Also, R25 is not present. It appears R25 was used on BA-3000 units, but not the AU-20000. Also looks like TR4 was replaced somewhere along the way. In some cases C905 is mounted on the foil side, but on this unit it was in the R22 position, as many recommend for the replacement. Finally, C10 (10 µF/50V) had an additional cap mounted on the foil side with value 4.7 µF/50V. I assume this is not original, as I have not seen that mentioned in other restoration threads.



There is something odd on the back of the board. A 33 µF/50V cap is seen in the above picture. Here is a close up of the foil side.



It seems the ceramic cap and diode are correct for the foil side, but the e-cap was added. Also, inside the plastic housing under the ceramic cap is a diode and resistor in series. You can also see two points are bridged with a wire inside a plastic protector. I have no idea what this is or why/when it was done. Looking at other restoration threads the ceramic cap/diode is present in that position, but I've not come across the other components that are mounted on the foil side. Anyone have thoughts/insights?

One more surprise was in store for me, a long trace on one side of the board is completely exposed and lifted from the PCB. I was lucky I did not inadvertently snag that thin copper and break it. Need to get the epoxy out!

 

[KeithD]

Here's a full picture of the foil side with the cap across C10 removed.

 

[richardpp]

As you pointed out, R25 is used in the BA-3000. It's for the bi-coloured LED protection indicator. The AU-20000 has a single on/off LED indicator so R25 is redundant.

The 4.7µF electrolytic cap on the foil side of the board has been added at the factory. I have it on my boards too. It parallels C10 (10µF) on the component side and increases the speaker relay delay time (from ~4 secs to ~7 secs). I replaced both with a single 15µF capacitor on the component side.

The ceramic cap and small signal diode on the foil side suppress any switching transients caused by the pre-main relay, (connected to TM-03). The resistor/ diode combination in the plastic sleeve and the 33µF capacitor on the foil side have been incorporated as D21 (1S2473D), R27 (220Ω 1/4W) and C11 (33µF 50V) on my boards.

The lifted foil is not a nice find. Because there are no fuses in the power supply rails, overload current shows up as damage to traces. I don't think this would ever happen under any normal circumstances, more likely a result of an output stage failure.

 

[KeithD]

As you pointed out, R25 is used in the BA-3000. It's for the bi-coloured LED protection indicator. The AU-20000 has a single on/off LED indicator so R25 is redundant.

The 4.7µF electrolytic cap on the foil side of the board has been added at the factory. I have it on my boards too. It parallels C10 (10µF) on the component side and increases the speaker relay delay time (from ~4 secs to ~7 secs). I replaced both with a single 15µF capacitor on the component side.

The ceramic cap and small signal diode on the foil side suppress any switching transients caused by the pre-main relay, (connected to TM-03). The resistor/ diode combination in the plastic sleeve and the 33µF capacitor on the foil side have been incorporated as D21 (1S2473D), R27 (220Ω 1/4W) and C11 (33µF 50V) on my boards.

The lifted foil is not a nice find. Because there are no fuses in the power supply rails, overload current shows up as damage to traces. I don't think this would ever happen under any normal circumstances, more likely a result of an output stage failure.

View attachment 3829037View attachment 3829038

Thanks so much for all that info Richard!

Thanks for confirming the 4.7 µF cap across C10. I don't have anything near 15 µF, so I'll need to replace with a 10 + 4.7.

Thanks for the clear image of your F-2511 showing the capacitor/diode/resistor places. I also found cut traces on either side of the connector pin.



Your board has C11 marked and places for the other components while mine does not. I'm assuming what happened here, is in early AU-20000 production the factory used boards designed for the BA-3000 before the F-2511 design was modified for the AU-20000? Do you have a schematic that reflects the later F-2511 board design?

Agree with you about the damage. This unit had all of the OTs replaced at some point, so I'm assuming there was a failure requiring replacement of many or all of the OTs.

 

[richardpp]

Hi Keith, apart from what you have already brought to light the only other differences I'm aware of are a couple of extra resistors labelled R13 and R21. I've marked those on the first schematic.

The modifications in your last photo, including trace cuts, appear consistent with the inclusion of R27, D21 and C11 on the newer board. I've mapped the changes onto the second schematic so you're able to confirm this. The aforementioned small signal diode and .01µF ceramic capacitor are not shown. They are across A_B.

 

[Hyperion]

That's an extraordinary level of detailed knowledge displayed there.

Someone doing a restore on another one of these added 'in-line' main rail fuses, (x4 IIRC) as there was no protection otherwise from that 'track damage' problem/possibility.

 

[KeithD]

Hi Keith, apart from what you have already brought to light the only other differences I'm aware of are a couple of extra resistors labelled R13 and R21. I've marked those on the first schematic.

The modifications in your last photo, including trace cuts, appear consistent with the inclusion of R27, D21 and C11 on the newer board. I've mapped the changes onto the second schematic so you're able to confirm this. The aforementioned small signal diode and .01µF ceramic capacitor are not shown. They are across A_B.



View attachment 3829149


View attachment 3829150

Thanks so much Richard! The modifications you show on your second schematic match what I'm seeing. I did a crude markup of that section of the schematic last night showing the mods I'm seeing, and they match what you have. The yellow line represents the cuts to the traces.