Yamaha cr-2020 repair and restoration

rcaman

cornbread george
I bought a non working cr-2020 upon inspection someone had made an attempt at repairing the power supply
and made a mess. after a complete power supply recap and a correct relay and the Yamaha modifications resistor upgrade no relay click at power up. measuring the +25 volts was good but the -25 volt line was being pulled down to -2 volts. short on the -25 volt line was traced down to tr753 or tr754 i forget which one they are on the pre mod. e board i replaced both and she fired up and went to singing.as for the panel lights the Yamaha engineers should have been fired on this receiver. Pitiful to say the least i removed the green filters on the watt meters and tuning pointer use white nail polist to cover up the brown stains from the old lamps. As for the sound quality i am not at all impressed a Harman Kardon 330c sitting next to it puts it to shame. anyway another Yamaha cr-2020 lives again. i also added extra heat sinks to the two regulator transistors on the power supply.
 

Attachments

  • cr-2020-1 (Small).JPG
    cr-2020-1 (Small).JPG
    82.2 KB · Views: 275
  • cr-2020-2 (Small).JPG
    cr-2020-2 (Small).JPG
    140 KB · Views: 284
  • cr-2020-3 (Small).JPG
    cr-2020-3 (Small).JPG
    68.5 KB · Views: 276
  • cr-2020-4 (Small).JPG
    cr-2020-4 (Small).JPG
    72.9 KB · Views: 270
I have not seen one of these in 15 years that the preamp was not wore out on. If you don't recap it completely, you won't be impressed. Finish it! It is a very fine sounding receiver when all is right IMO...
 
Agreed, it won't sing unless you completely recap it. I don't say that about many vintage units, but if it follows the pattern of the CR-620, quite a few caps might be bad. You won't find them with an esr meter, only a proper value test. It might also need a tuner alignment.
 
Well, then something seems amiss; it should sound pretty decent. Are you judging it by FM, CD or something else?
 
I don't agree with the above statement. Have had several HK 430, 630 & 730 receivers and loved everything about them.
Also have 3 Yamaha CR receivers and also love them. Just because 2 things are different doesn't mean I can enjoy both.
It's hard to imagine not liking a fully functional 2020
 
Hi there.

I assume you meant TR573/574 on pre-module board E (not 753/754)

It is not surprising at all that these transistors failed. Since I am currently doing a deep and full rebuilt of a CR-2020, I stumbled some weeks ago on the strange fact that they are off-factory completely underrated power-wise.

Board D and E being identical by the way, the complementary transistors TR571/572 and TR573/574 do have to dissipate around 1.0~1.1 watt each (40~45mA * 24.7V).

Although the service manual schematic DOES list high power transistors 2SC1509/2SA777 combo or 2SD458/2SB560 combo (1.0W and 0.9W respectively), in reality Yamaha factory installed very low power devices: 2SA872(A) & 2SC1775(A), being 300mW rated only.

This is really surprising, and just a matter of time to have a naughty failure happening.

One reason for selecting 2SA872(A) & 2SC1775(A) could be their very high bandwidth (200 mHz), low capacitance (1.6pF) and very low noise properties; quite important properties for these sensitive pre-amp stages. The listed devices in the service manual are indeed much less quality parts in that sense (100~120 mHz, 10-15pF, not low noise).

Finding a decent sub which meets all criteria (high power, high bandwidth, low capacitance, low noise), resulted into using specific TO-126 devices: the fabulous 2SC3503/2SA1381 combo (no longer in production sadly). Also known as KSC3503/KSA1381 if made by Fairchild.

I actually used them in the output stages of all pre-module boards (A thru E), as they all suffer the same factory issue.

In the TO-92(L) format, the KSC1008C/KSA708C are the least subs (still available), though only 800mW, and no brilliant audio properties. TO-126 devices seem the only decent subs to put in from that point.

So.... what subs did you use?

Also, it seems you only replaced the PNP devices ? That would bad practice. In such complementary push-pull output stages, both NPN and PNP devices should be replaced to maintain identical circuit behaviour between the negative and positive signal half. Perhaps one of the reasons why the sound of your CR-2020 is not up to your expectations.
 
Last edited:
Hi there.

I assume you meant TR573/574 on pre-module board E (not 753/754)

It is not surprising at all that these transistors failed. Since I am currently doing a deep and full rebuilt of a CR-2020, I stumbled some weeks ago on the strange fact that they are off-factory completely underrated power-wise.

Board D and E being identical by the way, the complementary transistors TR571/572 and TR573/574 do have to dissipate around 1.0~1.1 watt each (40~45mA * 24.7V).

Although the service manual schematic DOES list high power transistors 2SC1509/2SA777 combo or 2SD458/2SB560 combo (1.0W and 0.9W respectively), in reality Yamaha factory installed very low power devices: 2SA872(A) & 2SC1775(A), being 300mW rated only.

This is really surprising, and just a matter of time to have a naughty failure happening.

One reason for selecting 2SA872(A) & 2SC1775(A) could be their very high bandwidth (200 mHz), low capacitance (1.6pF) and very low noise properties; quite important properties for these sensitive pre-amp stages. The listed devices in the service manual are indeed much less quality parts in that sense (100~120 mHz, 10-15pF, not low noise).

Finding a decent sub which meets all criteria (high power, high bandwidth, low capacitance, low noise), resulted into using specific TO-126 devices: the fabulous 2SC3503/2SA1381 combo (no longer in production sadly). Also known as KSC3503/KSA1381 if made by Fairchild.

I actually used them in the output stages of all pre-module boards (A thru E), as they all suffer the same factory issue.

In the TO-92(L) format, the KSC1008C/KSA708C are the least subs (still available), though only 800mW, and no brilliant audio properties. TO-126 devices seem the only decent subs to put in from that point.

So.... what subs did you use?

Also, it seems you only replaced the PNP devices ? That would bad practice. In such complementary push-pull output stages, both NPN and PNP devices should be replaced to maintain identical circuit behaviour between the negative and positive signal half. Perhaps one of the reasons why the sound of your CR-2020 is not up to your expectations.
Yes, strange but not unusual. I noted the Yamaha did a strange " arrangement" on the C-1 PSU, where that paired a T092 with a (edit ) T092 "giant" device as a PNP/NPN pair.
These are marked on the schematic as such yet, both have similar functions, regulating the - and + rails, respectively.
 
Last edited:
Oh, but those transistor in this CR-2020 are no regulators.
These pre-modules are buffer stages, close to unity gain (very similar to the output stages of a power amp, with a fixed bias setting the output stage in class A).
In modern gear those would have been ICs (as they cost nothing these days)
 
i did replace the pnp and npn i just didn't make myself clear on that. Don't get me wrong it does sound good but i have heard lots better. steve
 
Oh, but those transistor in this CR-2020 are no regulators.
These pre-modules are buffer stages, close to unity gain (very similar to the output stages of a power amp, with a fixed bias setting the output stage in class A).
In modern gear those would have been ICs (as they cost nothing these days)
ouch
 
Although the service manual schematic DOES list high power transistors 2SC1509/2SA777 combo or 2SD458/2SB560 combo (1.0W and 0.9W respectively), in reality Yamaha factory installed very low power devices: 2SA872(A) & 2SC1775(A), being 300mW rated only.

For information, I am just starting to make a full list of parts for a CR-2020 rebuild, and in my pre modules D and E, TR571, 572, 573, 574 I found 2SC1509 / 2SB560 combos. (and definitely nobody was there before me)

Another question : on the amp boards, the heat-tracking transistor TR608 is a 2SC458. Should I replace it ? I know they are known trouble-makers in the signal path, but here ? I have plenty of KSC1845 on hand...
 
For information, I am just starting to make a full list of parts for a CR-2020 rebuild, and in my pre modules D and E, TR571, 572, 573, 574 I found 2SC1509 / 2SB560 combos. (and definitely nobody was there before me)

Another question : on the amp boards, the heat-tracking transistor TR608 is a 2SC458. Should I replace it ? I know they are known trouble-makers in the signal path, but here ? I have plenty of KSC1845 on hand...

Well those 2SC1509 / 2SB560 are adequately rated, so they can stay.

Regarding TR608: I would (and I did).
The problem here: the leg orientation of the original 2SC458 is conveniently in reverse order (BCE seen from printed side) as opposed of all other regular TO-92 transistors (ECB seen from printed side).
Meaning that all subs would be facing the heatsink with their round side..... bummer !!!
The reason why they selected the 2SC458 is its flat pack design, to have a large heat transfer surface from heatsink to transistor

A decent large surface is important to ensure that the transistor is following the heatsink temperature without too much delay and too much offset.
Otherwise there is a risk of 'thermal runaway' whereby the self-increasing bias through the output transistors is not adequately compensated in the bias circuit (when output transistor temperature rises)

The only solution that I could define was a TO-126 shaped transistor (nice large surface!).
But, due to the normally higher power ratings of TO-126 transistors, they most often have much higher Cob capacitance, which I would not recommend in that place in the circuit.
Again, the fabulous 2SC3503 has a very low 2.6pF capacitance (as compared to 10~20pF for most other TO-126 devices)
Although that transistor has no amplifying function (just a steady bias voltage job), it is in the middle of the amplifying stage and any capacitance in it will be charged/discharged by the signal voltage swings across it
Also, the 2SC3503 has no naked metal backside, so no isolation pad required (another thing to take care off).
see attached photo

In other words: do take care of the leg orientation of this transistor, don't be mislead by the orientation of the original transistor.
In fact, the transistor symbol printed on the PCB is correct (bulged side towards to heatsink) for regular TO-92 transistors
Having a decent heat transfer surface is also important.

I powered up my unit last night and the bias stayed rock stable after initial warm up.
 

Attachments

  • DSC08949.jpg
    DSC08949.jpg
    69.8 KB · Views: 156
Last edited:
Back
Top Bottom