Nakamichi TA-3A Bias Question / Opinions

[mrwasabihead]

Greetings, I'm currently trying to save a Nak TA-3A receiver that was going into protection at start up.
I'm familiar with the issues that heat can cause with this unit in regard to solder joints so those were addressed first.
At first power up the PS relay engaged but the speaker relay clicked multiple times and into protection it went. With the unit on a DBT it came out of protection. The the voltages from the supply board were a little low so I recapped the power supply board which brought all voltages up to acceptable levels. After checking the voltages on all TR's on both amp channels I found the left channel voltages fine and the right channel were almost all not fine. The problem was right channel R309(R), which was open. I replaced R309 on both channels. Now both channels TR voltages are in the green except the connected collectors of Q316 (L&R), which the SM shows should be [+40mV] but measure +5mV. This also puts the base of Q317 [+40mV] at +5mV and the emitter [+29mV] is +80mV, farther away from spec than I'd like but it comes out of protection off of the DBT. I also checked the Q206-Q211 "18 Volt" circuit and all voltages look good.
There is only a Bias current adjustment for the amp section of this receiver and it's supposed to be set at 25mV +/- 5mV. I found a three pin computer connector to attach to CN-32 connector as recommended so both channel measurement points can be taken easily and made the 2 RCA shorting plugs for the process. After substantial warm up, with the bottom and top cover installed, both channels stay at a pretty steady -4 mV and do not respond to bias resistor adjustments.
I've measured both channels' variable resistors as well as all emitter resistors out of circuit, as well as R's 391 / R310 / R321 / R322 / R330 / R344. I also replaced Zener diodes zd301,302,303 (both channels) & D304 as well as electrolytic caps C301/C306 both channels. The voltage drop over TH301 is correct and TH301 measures approx 92K ohm (in circuit) room temp - and drops to zero ohms once the unit warms up.
There is a LOT here on AK about problems with the Nak TA2/3 which I have attempted to read all of but nothing seems to shed light for me on what my particular problem might be. The unit passes audio just fine and sounds great but the bias current not being able to be adjusted just doesn't sit right with me. I'm happy to post any voltages and answer any questions about what I've done so far and would appreciate any input and or suggestions.
Thanks!
M->
https://www.hifiengine.com/manual_library/nakamichi/ta-3.shtml

 

[northpaw]

I'll say up front that I don't know what causes this problem so I can't provide knowledgeable advice on the fix, but since the voltages you measured in the L and R amp circuits are now all fine excepting the collectors of Q316 R and L, have you pulled them to test them out of circuit?

You state that you replaced the 22KΩ R309 in both channels, after finding one open. The original spec was 0.5W. Perhaps you know this, but as has been noted elsewhere, this is a common failure point in the TA series. Subbing in a 1W or 2W resistor would be prudent.

If you haven't done so already, I'd also check the voltages on the pins of U301 and in all the components between Q317/Q318 (you've already stated the voltage on the base of Q317 is off) to RY301 on the schematic. They may be off if they are dependent on the wrong voltage sourced from Q316 (at least, I think Q316 would be the source) to the base of Q317, but it seems it would be good to know the state now, possibly for comparison if you end up replacing Q316, if it (one or both) proves faulty.

 

[mrwasabihead]

I'll say up front that I don't know what causes this problem so I can't provide knowledgeable advice on the fix, but since the voltages you measured in the L and R amp circuits are now all fine excepting the collectors of Q316 R and L, have you pulled them to test them out of circuit?

Hey "northpaw" and thanks for the reply. I have pulled both Q316 - 2SA90(BL) and tested them. They both tested good but their hFE's were pretty far apart. So I replaced them with KSA992F and the result was the same.

You state that you replaced the 22KΩ R309 in both channels, after finding one open. The original spec was 0.5W. Perhaps you know this, but as has been noted elsewhere, this is a common failure point in the TA series. Subbing in a 1W or 2W resistor would be prudent.

I currently have 1/2 W in there but just received 1W today, so they'll go in tomorrow.

If you haven't done so already, I'd also check the voltages on the pins of U301 and in all the components between Q317/Q318 (you've already stated the voltage on the base of Q317 is off) to RY301 on the schematic. They may be off if they are dependent on the wrong voltage sourced from Q316 (at least, I think Q316 would be the source) to the base of Q317, but it seems it would be good to know the state now, possibly for comparison if you end up replacing Q316, if it (one or both) proves faulty.

Yes, Q317's base is directly connected to both channels Q316 collectors. Q318 voltages are within the acceptable range. I also pulled and checked Q317 and Q318 and they both pass. I'll be checking all of voltages at and around U301 tomorrow and will report with the findings. Thanks again.
M

 

[mrwasabihead]

So, I've measured the voltages around U301 and they are as follows: (spec)
Pin 1: +9mV (+25mV)
Pin 2: -105mV (-120mV)
Pin 3: +1.0mV (Ground)
Pin 4: +2.42V (+2.61V)
Pin 5: 1.1mV (Ground)
Pin 6: 760mV (740mV)
Pin 7: +2.22V (+2.17V)
Pin 8: +3.33V (+3.30V)
Aside from pin 1 everything looks decent. Pin 1, by the way, is connected to the emitter of Q317 via R335 / R336. I'm about to pull and check those to see if one has gone of out spec. The relay/protection circuit is working but I like to get everything as close to the recommended voltages as is reasonable.And of course my original issue concerned the un-setable bias and I'd really like to get some input from the amp Gurus among us.
More to follow..

 

[northpaw]

Here are a few things to consider.

Have you measured the DC offset on both channels? It appears (as best as I can determine), that the Protector circuit involving U301 compares what is on Pin 1 (overload det.) and Pin 2 (zero level det.) to determine if there is excessive DC on the speaker outs. You are observing a difference of 114mv there, whereas the nominal difference on the schematic is 145mv. That is a fair difference, and it must have a cause somewhere.

Q316 (L and R) is one of the critical connections between the amp and the protector circuit. Via its connection to Q317 and a network of R335, R336, and C313/C316 (bypassing to ground), it determines what is delivered to Pin 1. On this circuit, since you have already replaced Q316 and get the same results, let’s assume that is ok; Q317 itself could be the next most likely suspect., and bears checking, or replacing if you have something suitable. For the Pin 2 connection, it ties into the amp circuit via R333 (L and R). I guess I’d look carefully at it and around Q314 and Q315 and other nearby resistors and zeners; I know you have already looked at at least some of these components.

I’ll also note that transistors can sometimes check ok on a simple tester, but misbehave under load in a circuit. It is just something to keep in mind as you go through the testing. Sometimes it might make sense to just go ahead and replace one if other (passive) components check out.

You’ve stated that the bias is only 4mv for each channel, and does not vary with adjustment to the bias pots. With that level of bias, the heat sinks should not be warming up much. So after the unit has been on for, say, 20 min, are you sensing any warmth on the heat sink (check both the L and R sides, so you also get an idea if there is any imbalance)? With normal bias (i.e., ca. 28mv), they should run what I call “pleasingly warm”. If that is the case, I’d suspect your bias measurements are faulty (see below as an alternative way to measure bias). If the heat sink is cool or barely warmish, that would confirm that the bias is actually as low as you are measuring. If the bias is that low, the sound would tend to be on the thin side (at least compared to what this amp can produce).

Now, as to that unvarying bias you measured at 4mv via CN-32, to make sure there is not a problem with this connector, you could take the measurements directly off the emitter resistors themselves. In the TA-3, two 0.47Ω/5W resistors in series are used on each emitter pin of the output transistor pairs of Q312 and Q313 (a L and R for each). That gives 4 pairs of those resistors (R324-R345 and R325-R348, one pair for the L and one pair for the R). Using mini-grabbers on their leads, measure DCV across each of these 4 pairs and see how these compare to your previous measurements (the TP or CN-32 pins only measure one of the pairs of each channel, but you can measure both by doing the above). As you are doing these measurements, you can also try adjusting the bias pot to see if it has any effect (do this for sure if you are getting measurements that differ from the 4mv you got via CN-32).