SX-1250 Restoration

[the_burmese]

You put a 100µF cap in for C4? That and R16 are the primary timing components for the startup delay.
Lets not get offtrack, finish the power amp stuff, then come back to Protection if necessary.

Yeap I did replace C4 with 100uF 25v.

Sounds good. I am gonna order the rest of the transistors for AWH-048.
In the meantime, i am gonna try to test the VR2 as you have described - playing for a while to get it warm and checking it.
Detoxit all of the pots before I hit the sack tonight.

 

[the_burmese]

Run the unit for a while so that the output transistors are luke warm to the touch. (Ambient temperature in the garage is not helpful to get them warmer!)
Turned down the volume all the way, disengaged speakers from the selection switch, VR2 on both channels measured for about 95mV.
Dialed them down a little to 92mV to give more headroom during summer time.

The left channel is still weaker than right. It is more obvious as the unit warms up. Any typical reasons for this?
While I am waiting on the remaining transistors for power amp board, I plan to start working on the front end of the unit - flat, tone and equalizer amps to see if the flat amp would address the weak left channel.

 

[merlynski]

I am pointing at the damaged volume control for the left channel weakness. You may just have to find another one.

 

[the_burmese]

I am pointing at the damaged volume control for the left channel weakness. You may just have to find another one.

That pot may be one of those unobtainium .... oh well
But you may be correct. I just upgraded all the caps and transistors on the flat amp board today. The unit still has the issue on the left channel.

 

[the_burmese]

BTW, found this from MTF - https://audiokarma.org/forums/index...hannel-any-suggestion-on-where-to-fix.772123/

Will try it tomorrow after I am done with control amp rebuild.

 

[markthefixer]

Control amp audio level problem is far more likely than a bad pot. Pot can only reduce what it is fed, not shift the whole shebang even lower.
ALSO, "Control amp" is a misnomer, there is a flat amp, a tone amp and a filter amp.

There is:
0. dump truck load of jacks and switches, ending with stereo/mono signal combiner
1. balance control
2. buffer amp after balance control (kind of part of flat amp, powered by +/-49v from the equalizer board!!!)
3. volume control
4. flat amp (powered by +/- 25v from power supply) also has 2sa725 problem child transistors onboard!!!
5. tone control (bypassable)
6. filter amp (ALSO bypassable)
7. pre-out jack
8. power amp in jack
9. power amp itself
10. protection relay
11. speaker switches
12. speaker terminals.


which are not yet done?

 

[the_burmese]

Just finished tone amp. I think it had some work done before. Some of the caps are a bit strange. Nonetheless,I replaced them all 2.2uF and 4.7uF with WIMA caps.
Now, what is a bit strange is that the weakness in the L channel might be gone. I am trying again and again to make it sure. Will report back.

I still have filter amp, equalizer amp and power amp in jack to be done.

 

[the_burmese]

Filter amp and equalizer amp are all done! The left channel issue is completely gone.
I have got delivery from Mouser with these parts for remaining transistors on Power amp board -

bias / vas
2sa858 -> 512-KSA1381es
2sc1439 -> 512-KSC3503estu

pre-driver
2sa899 -> 512-KSA1381es
2sc1904 -> 512-KSC3503estu

drivers
2sb630 -> 512-KSA940TU
2sd610 -> 512-KSC2073TU

Please confirm as they are preferred ones as @markthefixer mentioned.

Also, I'd like to get education the VR2 bias. I understand it increases with temperature but does it also increase with volume control? If it does, is it ok to go above 100mV for a few millisecond during audio playback? Or should it be stable below 100mV?

Thanks! The sound quality is amazing. I can't wait to try it on the SDA-1Cs!!!!

Control amp after recap with WIMA 5%. it is super tight!



Filter amp after rebuild

 

[merlynski]

Bias and temperature:
When a pn junction's temperature increases (in a diode or transistor) the voltage of the junction decreases (physics explains this). This allows more current to flow through the junction for a given driving source. If this is a transistor BE junction, then more base current flows, thus causing more collector current to flow. Then more power is dissipated by the transistor, causing more heat. If this continues the transistor goes into thermal runaway and magic smoke starts escaping. The current through VR2 and the STV is held constant by Q5 (current source) and its related components. As the temperature rises the voltage drop across the junctions inside STV D1 decreases. Because the current through Q5 also supplies current through the BE junctions of Q7, Q8, Q9, Q10 (the actual bias driving current) the reduction in D1 voltage decreases the drive of the bias spreader (Q5, D1, VR2, Q4), which decreases the current through BE junctions of Q7, Q8, Q9, Q10, and the output transistors, which decreases their collector currents, preventing them from going into thermal runaway. Because this is such a critical current flow, is has to be centered (adjusted, VR2) for a bias current that is sufficient to put the output transistors in the conduction state above the knee in their characteristic curve, but below the point where they go into thermal runaway. The feed back (thermal) is physically through the heatsink, from the transistors to the STV, and is relatively slow. The STV is chosen by the design engineer to decrease its junction voltages as temperature increases to match the temperature responses of the BE junctions of Q7, Q8, Q9, Q10, and the output transistors. This is not directly related to volume, except that higher volume causes higher power dissipation. When you are playing music, during the positive voltage excursions the output transistors Q3 & Q7 are supplying power to the speakers, and output transistors Q1 & Q5 are in cutoff, during negative voltage excursion that switches the other way. If there was no bias you would get 'crossover distortion' as the signal crossed over 0v (in either direction).
Bias current will always drift some, just breathing on the heatsink and/or STV can cause drift.
As I commonly do, I may have way overdone this explanation .

 

[markthefixer]

Also, I'd like to get education the VR2 bias. I understand it increases with temperature but does it also increase with volume control? If it does, is it ok to go above 100mV for a few millisecond during audio playback? Or should it be stable below 100mV?

merlynski's explanation is good, and tailored to the sx-1250.

Idle current is a DC measurement - NO audio in the circuit when it is being set.
It is a MINIMUM setting, and as AC is fed into the amp for driving a speaker, more current flows - but it is NOT idle current - the amp is NOT idle!!!


The stv diode is a voltage reference - a rubber reference with the stretching influence being temperature.

The STV diode's voltage is generally specified at 7 milliamps of current on the datasheet. Increasing or decreasing that 7mA WILL increase or decrease the diode's voltage.
Getting the current to run through the diode usually uses one of two methods:
1. Cheaper models use "swamping" by tapping a current from a regulated power rail through a large value resistor, where other effects in the circuit will be swamped by the large resistance.
2. regulated current sources (or current mirrors in differential amp stages) which are active circuits supplying a precise current.

The VR2 resistance added in series with the STV diode is used to gain a certain extra (and temperature constant!!) voltage under the affect of OHM'S LAW.

Voltage expectations:
Output transistors +0.600VDC and -0.600VDC
Driver transistors (Q9 & Q10) +0.600VDC and -0.600VDC
pre-Driver transistors (Q7 & Q8) +0.600VDC and -0.600VDC

So with a direct coupled current amplifier as found in the output section of a Pioneer SX-1250 power amp:

1.8v on base (-0.600v), 1.2v on emitter of pre-driver, then
1.2v on base (-0.600v) 0.6v on emitter of driver, then
0.6v on base (-0.600v) 0v on emitter of output.
thus
1.800v > 1.200v > 0.600v > 0.00v
don't forget, there are the npn and pnp halves of a push pull power amp circuit, so there is another
-1.800v > -1.200v > -0.600v > -0.00v

total voltage needed:
3.6v (+1.8v + -1.8v) into pre-drivers
2.4v (+1.2v + -1.2v) into drivers
1.2v (+0.600v + -0.600v) into outputs

Now, the three above stages get an additional VR2 voltage (say 0.02v)
which the DC offset control through the feedback system splits in half
+1.801v > +1.201v > +0.601v > +0.01v
-1.801v > -1.201v > -0.601v > -0.01v

Notice the final voltages of +0.01v and -0.01v across the power output transistor's emitter resistors? THAT'S your idle current set by VR2.

Hope your eyes aren't glazed - these are the KEYS to the KINGDOM!!!!!
There are 3 parts of a power amp
1. input differential amp (feedback system)
2. voltage amplifier section
3. current driver output section - usually "push - pull" or "class B".

Whenever I am asking for the base to base voltages, I am most likely asking for these numbers to be found in the push pull output amplifier.

 

[the_burmese]

Wow, gents! These are really good stuff. The EE301 I took in college doesn't go like that. I will have to digest this more and put the numbers from @markthefixer into schematics and follow the circuit to fully understand it. Sorry EE301 has been a while ...

I just finished upgrading L side power amp board with all new active components. Turned down VR2 all the way CCW and brought it back on DBT and it passed! Set the VR1 to 0.0v. It is around +/-0.5mV. It does move around but within that range so I believe it is ok. May be newer components are more sensitive?

Slowly brought up VR2 on DBT to about 50mV and it is stable. Will continue the work on the R channel power board tonight. However I hit a snag. I only ordered 10 pcs of UF4004. Apparently I am 2 short! I do have some 1N4004-G and 621-UF1004.
Would they be ok as subs?

Trying to get to-126 into to-92!


ALL shiny and new

 

[Watthour]

Yes. Small bits in a pin vise are great for that. They're also helpful for other small work - Like when some rocket-brain surgeon with an impact driver who nkows more than everyone else ruins a terminal on a brand new RVSS.

He was "too smart" to take advice and apparently knew more than anyone else. The $1600 hit in the paycheck could have been an attention-getter. It cost 17¢ and about two hours to repair the ignorance, and I didn't let on for a day (let them sweat).

 

[the_burmese]

Yes. Small bits in a pin vise are great for that. They're also helpful for other small work - Like when some rocket-brain surgeon with an impact driver who nkows more than everyone else ruins a terminal on a brand new RVSS.

He was "too smart" to take advice and apparently knew more than anyone else. The $1600 hit in the paycheck could have been an attention-getter. It cost 17¢ and about two hours to repair the ignorance, and I didn't let on for a day (let them sweat).

Impact drivers and circuit boards do not match ... Different line of professions. But the workbench could be shared so that is the dangerous "handyness" factor ....

 

[the_burmese]

Looking at 750-1N4004-G and 621-UF1004, i am thinking about going with 1N4004 since Vf 1.1v is closer to that of UF4004 (1v). Original 1s1886 has 1.2v i think. Even thought 1N4004 is not a fast recovery like UF4004, i dont think 1s1886 is either.

Planning to use them to replace D3 and D4.
Let me know if this is not the right choice.

 

[tsd71]

Looking at 750-1N4004-G and 621-UF1004, i am thinking about going with 1N4004 since Vf 1.1v is closer to that of UF4004 (1v). Original 1s1886 has 1.2v i think. Even thought 1N4004 is not a fast recovery like UF4004, i dont think 1s1886 is either.

Planning to use them to replace D3 and D4.
Let me know if this is not the right choice.

1N4004 will be fine. I’ve used them on 1250’s before with good results.

 

[markthefixer]

TO-126 into a TO-92 lead hole? The hole is clogged with flux residue, some heat or mechanical cleaning will get it in. A clean hole will not be a problem. BTDT a lot.....
Pin vise for cleaning, not cutting board material. 0.020 to 0.030 max.

I count 9 diodes in the power amp.
The UF (ultra fast) 4004's depend a bit on WHERE they are used - the usual 1n4004's power rectifiers don't generate (much) commutation noise at zero crossing, but a UF in a power supply should have a snubber cap across it for hum considerations. The UF's are for power in FAST supplies, switchers stc... that's where their low capacitance boosts efficiency and cuts down on heat.

If I had to "use" a few slow diodes, I would put them in positions where their speed would not be in constant play.
Like D6, D7, D10 & D11 - where they are used for their voltage drop and the current is "constant" (i.e. no periodic reversals )

as to UF1004 versus UF4004 - I haven't yet dug out the difference. I did try a bit....

 

[the_burmese]

Hi Mark, your counting is perfect. I just ordered wrong QTY. I used 1N4004 in D6 and D7.
My understanding with difference between UF1004 and UF4004 is Vf. The rest seems to be same.

I also want to replace the rectifiers on the power supply board AWR-107. They are listed as 1s1885. Can I use the UF1004?

Also, @markthefixer, you asked me to check for any bad rectifier connected to that leaking filter cap. I believe it is one of the S5151 diodes on the AWM-091. Is there any replacement for these funky looking ones?
I will just replace all, since I am already there ...

R channel power amp board is done. I have hooked it up on DBT and brought back up without issue.

It seems that the VR1 0mV bias on the L channel board is more "jumpy" vs the R side. It doesn't move much - only +/-5mV vs about +/-1mV on the R side. I hope this is ok.

I am pretty much done, except for checking on those rectifiers. Once they are checked, will re-adjust all biases. Plan to adjust the VR2 on house power and target about 80mV so keep it safe for now and keep it running for about a couple of hours.
After that, it will be reassembled to regain its stately beauty.

 

[the_burmese]

Something is not right or may be I am not doing something right ...
Set the VR1 to 0mV on DBT ( power amp jumpbers out with 5.1k ohm input).

However, VR2 on R side doesn't move from 0v! I could move it to different level on L side. It is already half turn and still zero. I do not think that should be expected.
I have checked all other components and they all match and correctly oriented.

The only thing consistent on R side is D6,D7 are 1N4004 but D10,D11 are UF4004.

I can try changing them all to 1N4004 but I will need to order UF4004. What else should I check?

Below is comparison between L and R power amp boards..

 

[markthefixer]

I also want to replace the rectifiers on the power supply board AWR-107. They are listed as 1s1885. Can I use the UF1004?

grrr... PRECISELY why I posted this:

The UF (ultra fast) 4004's depend a bit on WHERE they are used - the usual 1n4004's power rectifiers don't generate (much) commutation noise at zero crossing, but a UF in a power supply should have a snubber cap across it for hum considerations. The UF's are for power in FAST supplies, switchers stc... that's where their low capacitance boosts efficiency and cuts down on heat.

Read what I posted. it is best to use low speed 1N4004.
edit: "commutation noise" == HUM.

Go ahead and swap out the
"The only thing consistent on R side is D6,D7 are 1N4004 but D10,D11 are UF4004."
to the OTHER channel, and see if it makes a difference. It would surprise me if it did, but we gotta try...

after that - well VR2 without idle current - remember you said you have to:

I will have to digest this more and put the numbers from @markthefixer into schematics and follow the circuit to fully understand it.

THAT'S EXACTLY PRECISELY what you (or I) have to do next - to see where VR2's voltage gets lost or why it isn't lighting up the output transistors. BOTH output transistors have to be working for idle current to flow.

 

[the_burmese]

Changing D10 and D11 to 1N4004 on the R side doesn't help as you suspected.

Yeap, I expected that we will need to know transistor bias points. So I measured points that I can reach with the short micro grabber I have.

Q9 emitter 0.63v
Q4 collector -6.9v (neg 6.9v)

I couldn't reach base of the Q8 or other points. I am ordering longer grabber micro clamp. And a few more UF4004 to completely match between 2 boards.