Yes, I followed service manual directions for bias measurement - speaker negative and built in test point. .
There are 4 output transistors per channel - two NPN and two PNP. Two had "blown" twice before. Both times, an NPN and a PNP. I an fairly confident all eight output transistors are healthy as I replace all of them twice and currentl the DBT goes dark immediatly.. But I have not checked their emitter resistors.
Hopefully I can locate them easily,
Q411 (2SC678) had turned to slag. It was shorting and the reason why DBT stayed bright. I replaced it and Q311 yesterday with KSA1015. DBT goes dark quickly now.
I rechecked voltages at Q403 base and emitter. I am now seeing 27V at Q403 base not 30.45V and am now seeing 30.43V at Q403 emitter not 27.5V. Those numbers may have gotten reversed the first time I measured them?
Voltage drop accross R410 measures 0,0mv.
R407 / 406 / 408 are bang on the resistances they are supposed to be R407 & R406 are each 3.3K and R408 is 1.2K.
I think C48 is missing a digit.
I must retire for the evening. Everthing I did today was fatiguing and then this. Now my eyes hurt.
I'll complete more of the above tasks tomorrow.
So from base to emitter on Q403 is about 3 V? Then Q403 is blown. That shouldn't be more than 0.6-0.8 V. It's a PN junction after all.
I wouldn't be surprised if Q404 is dead too.
I suggest that you leave the working channel alone until you get the non-working fixed. That way you have a known good reference. You can mod later.
Tom
I wouldn't be surprised if Q404 is dead too.
I suggest that you leave the working channel alone until you get the non-working fixed. That way you have a known good reference. You can mod later.
Tom
Don’t let it discourage you too much. I had a DC offset pot intermittently go open, causing thunderstorms and giving my woofer yoga lessons. It could have easily been a bias potentiometer.
After that, I replaced all the pots with modern o-ring sealed types. I’m going to add a resistor as mentioned by BSST on the bias pots from their wiper to the “low” direction just as a precaution.
You need to take a step back here...
The correct point is from speaker positive, not negative. If you measured from the negative then you are simply measuring from ground and if the emitter resistor were OK then you are just measuring DC offset... which at 25 volts shows a problem.
Here is what I would do.
1/ Add a shorting link across collector and emitter of the vbe multiplier (Q304 or Q404 depending on channel). That forces zero bias in an otherwise OK amp and takes any issues with bias out of the equation. The amp will work normally like this.
2/ Check all four emitter resistors for being open or very high in. If they read OK in circuit then that is good enough at this point.
3/ Measure the DC voltage from ground to the speaker positive output. If you have 25 volts or in fact anything more than -/+100mv then you have a basic fault giving an offset that needs attention.
If any output transistor is short or any emitter resistor open then for test purposes build the amp up with just one pair of output transistors (one NPN and on PNP) and their respective emitter resistor. As long as one of each is fitted it should all work as they are just parallel pairs.
See where that gets you first
I took a close visual look at Q403 last night before bedtime and I noticed I made a mistake.
As mentioned before, the original transistor at Q303 (2SA678) was faulty and I replaced it with a KSA1015. Both have the same pinout. What I did not see was that 2SA798's has printing on both sides. That failure to notice caused me to install the KSA1015 backwards. B-C-E instead of E-C-B.
I waited until after dinner tonight to correct this blunder. I removed the KSA1015 and installed a fresh one. The DBT was in place and I intended to make those voltage measurements again but I could not as within 15 or 20 seconds, the DBT went full bright.
A very disapointing development as I was hopefull that correcting my transistor install faux pas solved the problem.
Pic is both sides of a 2SA678.
The correct point is from speaker positive, not negative. If you measured from the negative then you are simply measuring from ground and if the emitter resistor were OK then you are just measuring DC offset... which at 25 volts shows a problem.
Here is what I would do.
1/ Add a shorting link across collector and emitter of the vbe multiplier (Q304 or Q404 depending on channel). That forces zero bias in an otherwise OK amp and takes any issues with bias out of the equation. The amp will work normally like this.
2/ Check all four emitter resistors for being open or very high in. If they read OK in circuit then that is good enough at this point.
3/ Measure the DC voltage from ground to the speaker positive output. If you have 25 volts or in fact anything more than -/+100mv then you have a basic fault giving an offset that needs attention.
If any output transistor is short or any emitter resistor open then for test purposes build the amp up with just one pair of output transistors (one NPN and on PNP) and their respective emitter resistor. As long as one of each is fitted it should all work as they are just parallel pairs.
See where that gets you first
Thank you for pointing out the correct speaker terminals for bias measurement.
The amplifier is not near my computer so I was following a printed copy of the SM instructions. The clarity is not good and that (+) sign by "speaker terminal" was hard for me to see and I overlooked it. My vision has gotten noticably worse over tha past year or so.
So I can refer to them in the room where the amplifier is, I printed (and enlarged the print) of the above steps that you posted. I will start removing again all output transistors tomorrow. Fortunately they are socketed so no potential solder pad damage from another desoldering.
Tonight I did measure the emitter resistors. Except for R319, all of them measured over 1 ohm. 1.1 ohm to 1.4 ohm
R319 measured 0.4 ohms. I took the photo below to show what type of resistors they are. The SM parts list describes them as 0.47 ohm 2W metal oxide but I did not expect metal oxide resistors to look like these.
Should those emitter resistors be replace with new 0.47 ohm 2W or is 1.1 to 1.4 ohm good enough?
If they are genuinely high in value then they should be replaced but first be absolutely sure they really are out of range. Low value resistors are not always easy to measure because any oxidation etc on the leads can stop the meter probes making a true 0.00 ohm connection. I would check them again making sure the meter probe contact is good.
With the probes shorted you should read the lead resistance (say 0.1 ohm) and that has to added to the result.
If you can measure from the solder side then press really hard into the solder to get a good contact and if they are high then they all should be replaced.
Today I was occupied with chores at hand so all I was able to attend to amp wise was re-measuring the emitter resistors on the solder pad side of the board. The readings obtained were all 0.6 to 0.7 ohms. And touching meter probes together produces a 0.1 ohm reading.
I have yet to proceed with the instructions from your post #25 except the above resistor measuring. Can I assume that my 0.6 to 0.7 ohm readings means the resistors are still suitable?
Tomorrow I will try to at least begin step 1, step 3 and the supplemental.
I have yet to proceed with the instructions from your post #25 except the above resistor measuring. Can I assume that my 0.6 to 0.7 ohm readings means the resistors are still suitable?
Tomorrow I will try to at least begin step 1, step 3 and the supplemental.
I am bothered by Q403 so before beginning on those suggested steps, I looked at it again. Formerly a 2SA678 now a KSA1015 both of those transistors have an E-C-B pinout. I would like to confirm that I have the transistor oriented correctly.
The first photo is the unoccupied Q403 spot and E is clearly marked at one hole. Comparing Q403's silkscreen image to other images where a 2SA678 is present, they differ somewhat and I am uncertain as to which hole is C and which hole is B.
The 2nd photo is of a 2SA678 which has a beveled corner. That beveled corner is represented in the silkscreen diagram. As positioned, in the photo, pinout of the transistor, left to right, is E-C-B.
With no question which hole is E, I did have B going to the top left hole and C going to the bottom hole. But a transistor could easily be positioned to reverse B & C.
The first photo is the unoccupied Q403 spot and E is clearly marked at one hole. Comparing Q403's silkscreen image to other images where a 2SA678 is present, they differ somewhat and I am uncertain as to which hole is C and which hole is B.
The 2nd photo is of a 2SA678 which has a beveled corner. That beveled corner is represented in the silkscreen diagram. As positioned, in the photo, pinout of the transistor, left to right, is E-C-B.
With no question which hole is E, I did have B going to the top left hole and C going to the bottom hole. But a transistor could easily be positioned to reverse B & C.
What are the holes on the board - the tracks and connecting parts (resistors and capacitors etc) will tell you (regardless of the silkscreen prints).
What is the pinout of your transistor - your DMM will tell you (unless you have invested some small money into a LCR-T4 style component tester).
After that - 2+2 is usually 4.
What are the holes on the board - the tracks and connecting parts (resistors and capacitors etc) will tell you (regardless of the silkscreen prints).
What is the pinout of your transistor - your DMM will tell you (unless you have invested some small money into a LCR-T4 style component tester).
After that - 2+2 is usually 4.
The coating on the trace sice of the board is heavily applied and it's thickness varies. It was hard for me to see some traces at all in some areas and overall hard to see period.
I needed to compare the trace side diagram in the service manual with the photo I posted above. That allowed me to orient the E-C-B holes, relative to the other components surrounding the Q404 transistor.
When satisfied I had the orientation correct, I installed another fresh KSA1015 in place of the original 2SC678. I wished to see if that permitted the DBT to dim right away after the amp was switched on but it did not. Used in conjunction with a variac, the DBT glowed steadily beginning at about 50V and grew brighter if I turned up the variac.
Next, I linked Q403 Collector & Emitter. It did not affect how the DBT behaved. Getting progressively brighter from 50V on up from the variac. It did not seem prudent to start looking for offset voltage if the amp was not getting full AC power.
The shorting link is still in place.
To proceed with the latter part of Step 3/, I removed output transistoes one at a time, checking DBT after extracting each one. The DBT behavior did not change. It began getting bright with the variac at 50 volts and it did so after each output transistor was removed. After all 8 output transistors were gone, the DBT continuued to get bright beginning with the variac at 50 volts.
EDIT: Forgot to mention that I was watching right channel offset voltage between speaker (+) and ground. When the DBT was noticabley bright (variac about 50 volts) my meter was displaying 0.0mv. Variac voltage did not matter. I saw 0.0mv from 0 volts AC to 50 volts AC.
Does this not dictate that a short exists somewhere on the amp board and has nothing to do with output transistors or emitter resistors?
I have not pulled and checked every transistor on the board but have done so with some. Originals and newer replacements installed a few years ago. I am tempted to think of replacing all remaining orignal transistors with modern substitutes.
Last edited:
No... you are applying full supply voltage to the driver inputs doing that. Depending on channel it is Q304 or 404 as here:
I'll look in later
As you noticed, I confused Q403 with Q404. It seems I can add late onset dyslexia to my list of infirmaties.
I removed the C to E jumper at Q403 and bridged C to E of Q404. Do the output transistors need to be in place? I removed them all yesterday.
As it is, with no output transistors, DBT starts to brighten when I increase the variac. I have yet to re-check the output transistors. I will do that today after I mow the grass and swap my winter wheels for summer wheels.
I removed the C to E jumper at Q403 and bridged C to E of Q404. Do the output transistors need to be in place? I removed them all yesterday.
As it is, with no output transistors, DBT starts to brighten when I increase the variac. I have yet to re-check the output transistors. I will do that today after I mow the grass and swap my winter wheels for summer wheels.
Why would you check the output transistors if the DBT lights up without them?
Bulb should not brighten with no outputs fitted because there is no low impedance path for current to flow. You have something else amiss somewhere, perhaps the other channel at this point.
If nothing burns or smokes then try faultfinding by voltage measurement. Check you have no voltage across the 0.47 ohm emitter resistors in the other channel that I assume has outputs in place. If you have then try turning bias down on that channel while you work on the amp.
Just to learn if I can reinstall them. A quick check with a component tester just now identified them all as BJT transistors.
Twice, earlier on, an NPN and a PNP in the right channel did get damaged. The same tester displayed either "diode" or "unknown or damaged part" when assessing those.
As mentioned, all 8 output transistors from both left and right channels would seem to be OK.
My thought was to remove all eight (four left channel and four right channel) output transistors so as to take them out of the equations and then see if the DBT stayed bright or dimmed with no output transistors present. To me, that seemed a logical way to determine that the fault absolutely preceeded the outputs.
Nothing has burned or smoked yet but I have not been applying full power to the amp for the past two or three days.
Should I reinstall the left (the good) channel output transistors then measure voltage accross those emitter resistors?
Easier to just lift one end of each of the 0.47 ohm resistors. That will isolate them.
Try and determine why the bulb is lighting. If the other channel is good then then the most likely reason is just that the good channel is drawing 'normal' bias current and that is lighting the bulb. Turning the bias down on that channel should stop that.
Depending on the normal background current draw of the amp some glowing of the bulb may be normal.