• The move to the new server is done. There are some software and database maintenance updates in process. This has us passing the hat around to help out. We appreciate any donations. Seriously, even a dollar helps. The payment page may be found here - https://www.audiokarma.org/support.html

SX-1050 rebuild of working unit

Pins 1, 2, 3, & 4 are good.
Pins 15, 16 & 17 are OK, (10%). 16/17 a bit high will make 15 a bit high. Check voltage cathode D1, sb +14vdc.
Pin 13/14 is too high, first thought D3 is shorted.
Pin 8 is high (>10%) Since pin 7 (-13.3vdc, D2 zener) is a bit high (>10%), you might consider selecting a slightly lower voltage D2 also. have you changed Q5, Q6 & Q7?
Pins 9, 10, 11 related to pin 8. You still have 4.4kΩ for R16?
Pins 5, 6 too high, check voltage on cathode D4, sb +6.1vdc, check Q9, 10, 11 again, sb Q9b +13.5vdc, Q10b +12.9vdc, Q11b +6.7vdc
Pins 18, 19 too high, related to pins 5, 6. That is the lamp circuit. If you have burned out lamps that ripple killer (Q12 , 13) may drift high, anyway FWIW mine does until I select the un-burned-out lamp switches. They are 8v lamps so it won't really hurt them, but they are brighter.
 
Pins 1, 2, 3, & 4 are good.
Pins 15, 16 & 17 are OK, (10%). 16/17 a bit high will make 15 a bit high. Check voltage cathode D1, sb +14vdc.
Pin 13/14 is too high, first thought D3 is shorted.
Pin 8 is high (>10%) Since pin 7 (-13.3vdc, D2 zener) is a bit high (>10%), you might consider selecting a slightly lower voltage D2 also. have you changed Q5, Q6 & Q7?
Pins 9, 10, 11 related to pin 8. You still have 4.4kΩ for R16?
Pins 5, 6 too high, check voltage on cathode D4, sb +6.1vdc, check Q9, 10, 11 again, sb Q9b +13.5vdc, Q10b +12.9vdc, Q11b +6.7vdc
Pins 18, 19 too high, related to pins 5, 6. That is the lamp circuit. If you have burned out lamps that ripple killer (Q12 , 13) may drift high, anyway FWIW mine does until I select the un-burned-out lamp switches. They are 8v lamps so it won't really hurt them, but they are brighter.

Yes, i still have 4.4k for r16. I haven't purchased any replacement resistors yet.
Q5,6,7 have been replaced

OK - I have a bunch of questions

“you might consider selecting a slightly lower voltage D2”. I’m not familiar enough with how to determine a lower voltage D2. Is voltage part of part number?
I don’t know what “ripple killer” means
“If you have burned out lamps”. ... “may drift high, anyway FWIW mine does until I select the un-burned-out lamp switches”. I’m very unclear what this means.
How would I “select” “un-burned-out lamp switches”? All of my lamps seem to be working, so they are ‘un-burned?’

does it make sense for me to replace all diodes to stated part numbers and then remeasure voltages and decide where to go from there?
 
I guess the other question I have is
Should I try different transistors for Q9, 10, 11 if they appear to what is driving the higher currents?

I did find a 4.8k resistor which I replaced R16 but it didn’t change the voltages
I have not replaced any of the zener diodes yet
 
Also, when I turned it on live, it takes about 6 seconds for the relay to engage. About the same amount of time as it takes while on 100w DBT
 
I guess I was on a roll :rolleyes:.
About selecting D2 , components tolerances vary from one to the next. With zeners the 'zener voltage' is the primary spec, like 13vdc +/- 5% tolerance, so a voltage from 12.35v to 13.65v would be in spec. You can try them in circuit, but that can be hard on the pcb with all the soldering and un-soldering. Or you can make a test jig with a resistor and a couple of 9v batteries and your meter (for any zener or pn junction):

ZenerTest3.jpg

Zeners are cheap, I bought 100 each of the common values used in Pioneer amps, so I can use a test jig to select from them.
The regulated voltage is proportional (but not linear, because of the gain in the regulator) to the differences in the reference zener voltages, so a slightly high one raises the regulated voltage, a slightly low one lowers it. You can 'tune' a voltage regulator that way, to some extent. The voltage in the jig will not exactly match the voltage in circuit unless you get exactly the same current through it, but again they track. If you measure D2 in the jig and then find one from your stock that is lower, it will be lower in circuit also. I may be may overthinking this but, it does work, I have done it before.
You can also select resistors in the feedback divider but that is much more work. This may be way more than you want to get in to.


If your lamps are all working they are 'unburned' On my SX-1050 the Speaker A lamp quit working many years ago (burned out) and the AUX lamp is also dead. When those switches are selected the load on the lamp circuit is reduced and the voltage rises, so the other lamps get brighter. I can see the voltage rise on pin 18 and 19 of the -103 pcb. The circuit is intended to have a minimum load no matter which switches are elected, IF all the bulbs are good.
If I select an input other than AUX and Speakers other than only A the minimum load is satisfied, the voltage is correct and the lamps all have equal and correct brightness.


Ripple Killer, often called a 'capacitor multiplier' is much like a voltage regulator.

The attached schematic has the 'voltage regulators' in blue boxes. Their reference voltage is the zener diode and their outputs are intended to have a constant dc voltage with low ripple, regulated down from a higher (less stable) dc voltage with much larger ripple. The circuits in green boxes are often called 'capacitor multipliers'. Their input is an already regulated voltage, their job is to kill ripple caused by a varying load. Their primary function is killing ripple without using a large expensive capacitor, and providing a stable lower dc voltage from a higher regulated voltage, rather than providing a regulated voltage from an independent reference like a zener diode. The two circuits are very similar, but not quite the same. The output dc voltage of the ripple killer is dependent on its input voltage, where the voltage regulator's dc output voltage is independent of its input voltage (unless its input drops too low).

AWR-103regulator.png


End of lecture, hope that helped and was not too long winded . . . :rolleyes:

Six seconds seems OK to me for the protection relay, that may change after the power supplies, dc offset and bias are set correctly, don't worry about it until then.

If you used Q9, 10 and 11 replacements recommended by MTF or one of the other long time techs current BOMs they should be fine, unless one is off a bit. The transistor whose emitter is connected to the zener is the one that (its gain) controls the regulator voltage primarily, it need to be a high gain fast transistor, the other (one or) two are there for current gain and power handling.
Hope that all helps rather than confuses. But don't hesitate to ask :thumbsup:
 
Last edited:
Cathode D4 = 6.2v

Q 9: B: 20 : E: 19.3 : C: 21.7
Q10: B: 20.5 : E: 20 : C: 21.7
Q11: B: 9.6 : E: 6.3 : C: 20.5
 
Q11: B: 9.6 : E: 6.3 : C: 20.5
The Q11 base emitter junction in your post is 3.3v, it should be about 0.6v.
I think Q11 is either bad or installed wrong.
Is it a KSC1845FTA?
 
I guess I was on a roll :rolleyes:.
About selecting D2 , components tolerances vary from one to the next. With zeners the 'zener voltage' is the primary spec, like 13vdc +/- 5% tolerance, so a voltage from 12.35v to 13.65v would be in spec. You can try them in circuit, but that can be hard on the pcb with all the soldering and un-soldering. Or you can make a test jig with a resistor and a couple of 9v batteries and your meter (for any zener or pn junction):


Zeners are cheap, I bought 100 each of the common values used in Pioneer amps, so I can use a test jig to select from them.
The regulated voltage is proportional (but not linear, because of the gain in the regulator) to the differences in the reference zener voltages, so a slightly high one raises the regulated voltage, a slightly low one lowers it. You can 'tune' a voltage regulator that way, to some extent. The voltage in the jig will not exactly match the voltage in circuit unless you get exactly the same current through it, but again they track. If you measure D2 in the jig and then find one from your stock that is lower, it will be lower in circuit also. I may be may overthinking this but, it does work, I have done it before.
You can also select resistors in the feedback divider but that is much more work. This may be way more than you want to get in to.


If your lamps are all working they are 'unburned' On my SX-1050 the Speaker A lamp quit working many years ago (burned out) and the AUX lamp is also dead. When those switches are selected the load on the lamp circuit is reduced and the voltage rises, so the other lamps get brighter. I can see the voltage rise on pin 18 and 19 of the -103 pcb. The circuit is intended to have a minimum load no matter which switches are elected, IF all the bulbs are good.
If I select an input other than AUX and Speakers other than only A the minimum load is satisfied, the voltage is correct and the lamps all have equal and correct brightness.


Ripple Killer, often called a 'capacitor multiplier' is much like a voltage regulator.

The attached schematic has the 'voltage regulators' in blue boxes. Their reference voltage is the zener diode and their outputs are intended to have a constant dc voltage with low ripple, regulated down from a higher (less stable) dc voltage with much larger ripple. The circuits in green boxes are often called 'capacitor multipliers'. Their input is an already regulated voltage, their job is to kill ripple caused by a varying load. Their primary function is killing ripple without using a large expensive capacitor, and providing a stable lower dc voltage from a higher regulated voltage, rather than providing a regulated voltage from an independent reference like a zener diode. The two circuits are very similar, but not quite the same. The output dc voltage of the ripple killer is dependent on its input voltage, where the voltage regulator's dc output voltage is independent of its input voltage (unless its input drops too low).




End of lecture, hope that helped and was not too long winded . . . :rolleyes:

Six seconds seems OK to me for the protection relay, that may change after the power supplies, dc offset and bias are set correctly, don't worry about it until then.

If you used Q9, 10 and 11 replacements recommended by MTF or one of the other long time techs current BOMs they should be fine, unless one is off a bit. The transistor whose emitter is connected to the zener is the one that (its gain) controls the regulator voltage primarily, it need to be a high gain fast transistor, the other (one or) two are there for current gain and power handling.
Hope that all helps rather than confuses. But don't hesitate to ask :thumbsup:
Thank you..... the lecture is greatly appreciated and I will need to read it several times to digest it. I need these lectures as I go because my knowledge is basic. The diagram is very helpful. I started taking a basic electronics class Just yesterday, so I am in the classroom mode.
Q11: B: 9.6 : E: 6.3 : C: 20.5
The Q11 base emitter junction in your post is 3.3v, it should be about 0.6v.
I think Q11 is either bad or installed wrong.
Is it a KSC1845FTA?
it is a KSC1845, new. I’ll replace it as I have a bunch. I’ll double check the orientation
 
I have an oscilloscope But I haven’t yet learned how to hook it up. If you can walk me through that I would be grateful
This is already written, I read it, it is well done:
Oscilloscope basics | Audiokarma Home Audio Stereo Discussion Forums
Make sure you have good probes, and be careful what the probes' ground clip touches, it is hard grounded through the scope to the ground from the wall. If you clip it on (or touch it to) a voltage that point WILL short to ground.
 
Q11: B: 9.6 : E: 6.3 : C: 20.5
The Q11 base emitter junction in your post is 3.3v, it should be about 0.6v.
I think Q11 is either bad or installed wrong.
Is it a KSC1845FTA?
You nailed it. Q11 was wrong orientation. Not sure how I did that as I checked it several times but it was wrong

I replaced Q11 and all 4 zener diodes. Much better but I am still a bit high (<10%) on pins 8 -> 16. My tuner is working again. I will likely replace some resistors as they are all dark as if burnt. Here are my pins with the target voltage
P 1 : +23.2 : +24v
P 2 : +7.2 : +76v
P 3: -77.1 : -76v
P 5/6: 13.3 : +13v
P 7: -12.7 : +13v
P 8: -64.5 : -60v
P 9: -32.4 : -30v
P10/11: -26.4: -25v
P13/14: +26.7: +25v
P15: +32.8 : 30v
P16/17: +63.4: 60v
P18/19: +5.2 : +5.4v
 
I think I did exactly the same thing with the same transistor when I did my -103, I thought I checked it three times, but the voltages did not come up. Went back with the meter and it was in wrong and shorted. Replaced correctly, and voila, good voltages.

P 2 : +7.2 : +76v
I am assuming this is a typo? Otherwise it all looks OK!
 
This is already written, I read it, it is well done:
Oscilloscope basics | Audiokarma Home Audio Stereo Discussion Forums
Make sure you have good probes, and be careful what the probes' ground clip touches, it is hard grounded through the scope to the ground from the wall. If you clip it on (or touch it to) a voltage that point WILL short to ground.
OK. My question is apparently so basic that it wasn’t answered or I miss it. My question is..’What are we measuring? Voltage? Amperage? Shape of current wave at a point in the circuit? I thought a tech wound instigate a signal in some way into the receiver are pick that signal up on the scope to examine it and perhaps compare to original signal. I would have thought touching the circuit board with the probe would allow too much current into the scope and damage it.
I think I did exactly the same thing with the same transistor when I did my -103, I thought I checked it three times, but the voltages did not come up. Went back with the meter and it was in wrong and shorted. Replaced correctly, and voila, good voltages.


I am assuming this is a typo? Otherwise it all looks OK!
yes...75.2
 
Last edited:
Aha, you are right, it is so basic it is overlooked. An oscilloscope produces a visual representation of changing voltage during a certain amount of time. The vertical excursion is voltage, the horizontal sweep is time. Both have range settings that can be changed to show what you are looking at more clearly. The trick comes to learning to interpret the meaning of what you are seeing, based on what you "know" (or expect) should be there and what you are measuring.
The input itself (not the ground clip) is very high impedance, usually 1 Megohm, then a x10 probe, the most commonly used, is 10 Megohms, so you can measure several hundred volts without risk. The high impedance also keeps from loading down the signal. The probe is also calibrated to the scope, a simple adjustment, on the probe, from a signal produced by the scope. For higher voltages you need different (higher impedance) probes. There are other specialty probes available.
The maximum frequency that can be measured is a function, primarily, of the oscilloscope itself, and the probes.
Scopes are incredibly useful for measuring things that cannot be measured by a digital multimeter, but are not a magic crystal ball (Clarke's Law you quoted). In an analog audio amp, usually if the dc voltages are correct, the ac takes care of itself. Also, I have seen a lot of ingenious ways to use a meter on AK, that I would never have bothered with as a Bench Tech with a scope handy, that overcome its absence.
The reason I asked if you had a scope is because I have had trouble diagnosing voltage regulators that had excessively large input ripple passing through the regulator, making it look like the regulator was bad, when it was the big filter cap before the regulator being open, and not supporting the required dc input to the regulator. This happened on my SX-1050, and I have seen it catch other techs on AK also. The main negative cap, C2, was open, and the dc voltages were way off. I figured it out different way, but a quick check with a scope would have shown it right off, as a half wave rectified, unfiltered signal. :blah: I got carried away again.
 
Last edited:
Aha, you are right, it is so basic it is overlooked. An oscilloscope produces a visual representation of changing voltage during a certain amount of time. The vertical excursion is voltage, the horizontal sweep is time. Both have range settings that can be changed to show what you are looking at more clearly. The trick comes to learning to interpret the meaning of what you are seeing, based on what you "know" (or expect) should be there and what you are measuring.
The input itself (not the ground clip) is very high impedance, usually 1 Megohm, then a x10 probe, the most commonly used, is 10 Megohms, so you can measure several hundred volts without risk. The high impedance also keeps from loading down the signal. The probe is also calibrated to the scope, a simple adjustment, on the probe, from a signal produced by the scope. For higher voltages you need different (higher impedance) probes. There are other specialty probes available.
The maximum frequency that can be measured is a function, primarily, of the oscilloscope itself, and the probes.
Scopes are incredibly useful for measuring things that cannot be measured by a digital multimeter, but are not a magic crystal ball (Clarke's Law you quoted). In an analog audio amp, usually if the dc voltages are correct, the ac takes care of itself. Also, I have seen a lot of ingenious ways to use a meter on AK, that I would never have bothered with as a Bench Tech with a scope handy, that overcome its absence.
The reason I asked if you had a scope is because I have had trouble diagnosing voltage regulators that had excessively large input ripple passing through the regulator, making it look like the regulator was bad, when it was the big filter cap before the regulator was open, and not supporting the required dc input to the regulator. This happened on my SX-1050, and I have seen it catch other techs on AK also. The main negative cap, C2, was open, and the dc voltages were way off. I figured it out different way, but a quick check with a scope would have shown it right off, as a half wave rectified, unfiltered signal. :blah: I got carried away again.
That is excellent and answers my basic question and I am interested in hooking it up.
I’ve played with the scope with a signal generator (not sure if I connected it properly) so I am familiar with changing the time and the amplitude, having pattern move and freezing it. My scope is a Kenwood CS-4125. It does not have a ‘Astig’ adjustment that I can find. I’m assuming there is another way to adjust the roundness.
The other post’s main comment said an alligator clip on the ground lead will work fine. A follow up post said to ground unit, receiver in my case, chassis to scope chassis, in which case the ground cable will dangle as we probe. That seems very problematic to me
There are several connectors on the scope. For a simple “probing” we would not be using any except the main probe input, correct? And I would be looking for a smooth wave pattern, correct?
 
I’ve been giving this some additional thought. If I would take a reading at pin 1 of the -103, which is the raw positive DC input voltage to the board and I would look to ensure the wave form is smooth, then I would then read pin 5 and see that it is also smooth, then we would conclude all components in that circuit are functioning as they should. If I then read pin 17 and see a distorted wave, then I would conclude one or more components in that circuit are faulty and I would then probe different components to isolate the problem item

is this how I would use a scope?
 
On most scope probes the 'ground clip' can be removed so it does not dangle and risk bumping and shorting out your circuit. With very small signals the ground clip is attached and used to eliminate noise pickup.
On pin 1 of the -103 you will have a large DC voltage (compared to the ripple), so the scope input would have to be set to AC coupled or the signal you are looking at will be off the screen. The DC component can be measured with the DMM, then the AC ripple can be seen with the scope. The ripple on a full-wave rectified supply will be a 120Hz rounded saw-tooth like signal. Its amplitude will vary, depending on load. Those power supply regulators it feeds are the lamp circuit and the tuner circuit, neither of which will have a load varying enough so that it will likely be seen at the raw dc supply at pin 1. If you set the scope trigger to 'Line' the ripple will be steady on the display. Raw (unregulated) filtered by a cap, (C11 on -104 in this case) DC will always have some ripple. Pin 5 (regulated dc) will have much smaller ripple and noise, with very little (almost none) of the 120Hz from the raw dc input. The difference between the input ripple and the output ripple is the regulator's "input regulation ratio", or how much to regulator rejects the input ripple and voltage change.
Look at section 6.1 Linear Regulators and 7.1 in this link: https://en.wikipedia.org/wiki/Voltage_regulator
Also see https://en.wikipedia.org/wiki/Rectifier#Half-wave_rectification about Rectifiers There may be way too much info in those links.
 
Back
Top Bottom