New project... Yamaha RX396 Upgrade

[Adrculda]

Yes, I'm at it again...
i got an older yamaha amp that i got from a friend and it sounds good, but it can use a bit more juice. So i pulled the cover and looked inside to find a quite small trafo, but will be back to this latter.

The amp is built around a complementary pair of 2STC4467 and 2STA1694 and was thinking of replacing them with 2STC4468 and 2STA1695 which is useless without upgrading the power supply itself.

The traffo has 7 output pins and 2 of them are sepparatelly with an output of 4.2VAC the other are supply for the amp and heres what i found...

All readings were taken at idle
pin 1 - pin 5 = 71.3VAC
pin 1 - pin 3 = 35.7VAC
pin 3 - pin 5 = 35.6VAC
pin 2 - pin 4 = 53.3VAC
pin 1 - pin 2 = 8.9VAC
pin 4 - pin 5 = 8.9VAC

seems to me its just nothing special than a multi-tap trafo that im guessing is no more that 200VA.
The impedance circuit however its a bit odd as all it does it switches the supply voltage from 56VAC to 71VAC and to boot it has 2 small 5600uF caps rated for 56V.

SO...
I was thinking up upgrading the complimentary output pairs and then get a 500va or 750va toroid with dual 35v secondaries and upgrade the caps to 8200uf 80V units and also upgrade the rectifier bridge which is tiny.

What do you guys think ??

[Adrculda]

Rail Voltages sag horrendously under a 4ohm load !!!!
under the one setting the rail voltage is 70VDC and under the other is 95VDC.
At the 70V setting rails sag to ~55V and at the 90V setting its even worse as the rails sag to ~63V. Both test were done using the same 80Hz note at the same volume setting ( 50% on the dial ).

As i was saying earlier the power supply needs some serious work before i can replace the output devices and should improve the sound quality as well.

Specs according you Yamaha:

Max Output Power / Ch 70 W
Dynamic Power / Ch (8 / 4 / 2 ohms) 70 / 89 / 100 W
Damping Factor (8 ohms, 20-20,000 Hz) 100

Yamaha RX-396

[ostripper]

All readings were taken at idle pin 1 - pin 5 = 71.3VAC pin 1 - pin 3 = 35.7VAC pin 3 - pin 5 = 35.6VAC pin 2 - pin 4 = 53.3VAC pin 1 - pin 2 = 8.9VAC pin 4 - pin 5 = 8.9VAC

You might of taken them readings wrong.
It looks like you have a 35-0-35vac center-tapped main secondary , a 8-0-8vac (5v supply)aux. secondary for your microprocessor /tuner and a higher supply tap for your florescent display.

This equates to @ 52-0-52vdc (rails) , hence.. the 56v caps.
My deceased sony 70w receiver had the same trafo. since your present trafo has custom secondaries running the aux. electronics it might be hard to replace. The only upgrade that makes sense is larger caps ... 8200 - 10,000uf @63v would give a little more headroom and a better soundstage. Replacing the outputs with different devices might create biasing problems or degrade the sound of the amp.

To measure the trafo right , either look for color (color -black-color = main amp secondary) or the thickest gauge winding (on 3 pins).
the auxiliary secondaries are usually (color-color) or smaller gauge windings close together(if out on pins).
OS

[Adrculda]

Thanks OS...
But already done that...
The voltage reading was done at the Collector of the output transistors.
Also I forgot to mention that there's a separate 12V source alongside the larger one.

I'm uploading pictures as we speak. As far as the biasing goes i believe its done by 2 smaller transistors that act as diodes "throttling" the bias voltage by the temperature of the heat sink. On a separate note the replacement output devices have less thermal resistance between the junction and the case and that would mean a faster heat exchange between the die and the sink itself which coupled with the auto-biasing that is in place would greatly benefit and improve sound quality.

[Adrculda]

OK...
Pictures are up :

Power Supply

Colors are as follows ( from left to right - from pin 1 to 7 ):
orange - yellow - black - yellow - orange - - - blue - blue




Impedance Selection Circuit ( i think i should call it voltage selector Circuit )


Rail voltage in position 1 is 95VDC and in position 2 is 70VDC measured at the Collector of the output devices !


Output Devices and Biasing Transistor


Pin 1 - Base
Pin 2 - Collector
Pin 3 - Emitter

[rabbitz]

I played around with one of these and changed a lot of the more important caps especially in the signal path and opamp power supply. The change to the sound was minimal but changing the power supply caps to 6800uF 63V made a very worthwhile difference adding more weight, refinement and control to the sound. It seemed to take out some of the bright character of this amp to make it a better listen.

I haven't changed any transformers or rectifiers etc.

[Adrculda]

Originally posted by rabbitz
I played around with one of these and changed a lot of the more important caps especially in the signal path and opamp power supply. The change to the sound was minimal but changing the power supply caps to 6800uF 63V made a very worthwhile difference adding more weight, refinement and control to the sound. It seemed to take out some of the bright character of this amp to make it a better listen.

I haven't changed any transformers or rectifiers etc.

I personally think that a good power supply should not sag more than a 5-10% sag in the rails under full load/volume since its the foundation of the amp.
I wont be running any impedance under 4 ohms or above 8 ohms at all so the upgrades mentioned before will definatelly help the amp sound smoother at higher volumes and lower impedance loads.

The opamp and its power supply will be a latter project as i will be running a separate board and run some decent opamps in there.

From the datasheet for the output devices seem to be a direct replacement with more muscle behind them , but the major limitation in this amp is the power supply itself, and that's the First priority, not the output stage of it.

[ostripper]

Pictures are cool... I would guess (starting at orange) 35-14-0-14-35 VAC (referenced to the blacks - centertap) ... , the blues are the standby voltage (6vac).

Same as the sony HT's , sherwoods (almost exact -OPstage and all) , etc. Looks like a 300-350VA unit.

Rail voltage in position 1 is 95VDC and in position 2 is 70VDC measured at the Collector of the output devices !

Measured from the centertap of the trafo or the chassis , you should have +52 and -52 volts , you would only see 70-90v rails on a "superamp" ...300-400w . PS , your Op board looks EXACTLY like the sherwood , 56v caps and all..!

Now that I had a good look , the little tranny on the HS is your VBE and there is no bias trim. You might luck out and have similar Vbe and beta with your replacement Op devices , but don't bet on it.

A more advanced upgrade would be to get the schema and ADD a bias trim to the circuit. Then you could adjust to the best bias (50-70ma) with any OP device. failure to do this would negate any improvement in device replacement , or even result in failure.

From the datasheet for the output devices seem to be a direct replacement with more muscle behind them , but the major limitation in this amp is the power supply itself, and that's the First priority, not the output stage of it.

The trafo is matched for a single pair of OP devices , it "sags" as to not exceed the limited SOA of a single pair of devices. You could add another pair if you increased the trafo size , but then you would need "beefier" drivers as well !

PS . I looked at your devices and the $1.20 NJW0281/0302's are a perfect replacement as well (that's what I put in the sherwood.)
OS

[Adrculda]

Originally posted by ostripper
I looked at your devices and the $1.20 NJW0281/0302's are a perfect replacement as well (that's what I put in the sherwood.)
OS

Your Evil !!!!
I just ordered 2 pairs and 2 10000uF 80V Nichicon caps.

New Trafo will be a Plitron 750VA with dual 35VAC secondaries.

Once i have everything at hand i will be posting pictures.

BTW...
Does Anyone have the Schematic for the RX-396 ???

[Adrculda]

Parts have ARRIVED !!!

Does anyone have a balanced to unbalanced converter schematic and a simple BIAS circuit i can implement on this unit ???

NJW0281G and NJW0302G went in quite simply and while i was in there decided to replace the S4VB20 rectifier that came inside with a GBPC5002 ( 200VRMS @ 50A - $1.95 ) and mounted it to the chassis with a silicone pad and grease and added the heat sink on the outside of the case with silicone compound.

Initial Impressions:
Cap Upgrade definatelly improved response and mid/high fade is gone. Transistors further improved overall sound quality and definition specially in the low end where its more controlled and the mid / highs are more clear and reached a new deafening level where i have to turn both the bass and the treble down to 0, where as before to get a clear sound i had to crank them up.

All in all well worth an upgrade and definatelly looking forward to the torroid upgrade !!