• 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

Realistic STA-90 Amp Circuit A-Z

videolady201

videolady201

AK Member-Design Engineer
I have posted two images of the Realistic STA-90 amp circuit. My goal is to get opinions on optimizing the circuit and to determine why this unit has such a unique sound. Very smooth and rich with very good transparency. I also wish to balance the differential stage and correct any DC Offset issues. Unless you have heard one of these units, it's hard to explain the sound quality.:scratch2:
 

Attachments

  • sta-90 amp.jpg
    sta-90 amp.jpg
    84.3 KB · Views: 199
  • sta-90 amp2.jpg
    sta-90 amp2.jpg
    58.2 KB · Views: 168
Register to hide this ad
I couldn`t agree more with your assessment of the sound quality of this unit. I believe it may have been made by Sansui or at least have some of their inherent design characteristics with the electronics. The rich, warm sound reminds me of a lower powered 881 (also have) and is much better overall than my STA-2000D.
 
It's a quasi-complementary (QC) power amp. I've seen at least one other discussion about the sonic qualities of asymetrical output stages, and the consensus seems to be that relatively high second harmonic distortion can make them sound a bit romantic.

There's some information missing in the schematics. I don't see the power supply to the diff-amp emitters, and component values on its input side are nearly illegible.

C605 probably has a substantial effect on sonic quality, but it's too big to be replaced with a plastic film type. High-performance 'lytics might yield worthwhile improvement. If the input coupling cap is 2.2uF, then it could be replaced with film.

Some folks like to disconnect SOA protection circuits (TR609~610), especially if they upgrade the output transistors. Do you have a spec for the output devices? Don't do this if you don't understand the risk.

QC output circuits can often be improved by adding a diode, with parallel capacitor and resistor, in series to the PNP driver emitter. The bias network needs a little alteration as well. There's more to know on this topic, and I'm no expert! Search on Baxandal Diode and Linsley Hood for references. One could easily revise the circuit for full complementary operation, but I suspect that its sonic qualities would be materially altered.

I see the Zobel network is downstream of the output inductor, which makes it ineffective at RF. That might be okay with this particular amp, but it makes me wonder if the designer wasn't thinking clearly.
 
Amp Circuit Info

BinaryMike said:
It's a quasi-complementary (QC) power amp. I've seen at least one other discussion about the sonic qualities of asymetrical output stages, and the consensus seems to be that relatively high second harmonic distortion can make them sound a bit romantic.

There's some information missing in the schematics. I don't see the power supply to the diff-amp emitters, and component values on its input side are nearly illegible.
Click to expand...

It will take me a little while but I will place the full circuit on my server and add a link so you can see all of it and it should be full size and hi-res this way!
 
Excellent viewing on the new schematic!

If D601 is a 16V zener, then each side of the diff-amp should operate at 939uA. R605 should therefore be 654R, assuming 0.6V for Vbe and TR603 current gain of 200.
 
Last edited:
Here we have much fun in perspective!

In the original circuit, the base resistance on TR601 is high, 120k. But on the feedback side, this is much lower at 22k (R609). In order to have low offset voltage, the collector currents in TR601 and TR602 must be different. And different, they are!

With the original value for R605, the collector current in TR601 is very small, at ~0.15mA. The rest of the tail current, which is ~1.9ma flows by TR602. So, the collector current in TR602 is ~1.75mA, or if you prefer ~12 times as large as the collector current in TR601. I neglected the base current in TR603.

The conflictual relationship between low distortion (equal collector current in the differential stage) and low output offset voltage is evident in this particular design.

Reducing R605, as BinaryMike suggested, is a good point for balancing the differential stage, but if the components on the feedback side are not changed you will have an offset high in the sky... You must also change R609 for a 120k, R601 for a 4k5 (to keep the same gain, but you can change it to another value if you want to change the gain), C605 can be lowered to 10uF. This is assuming a matched (hFE) pair for TR601/TR602.

The voltage on the collector of TR602 (-38V) is wrong, it is approximately the same as the emitter voltage of TR603, certainly not more negative.
 
Opinions on Sound

After seeing the full circuit of this amp, any suggestions on why it would sound as good as it does. DC offset is around 20mv on one channel and around 65mv on the other. It is very emotional and rich in detail. Low level decay is great. It has warmth and snap at the same time. Does anything you see make sense that it should sound so good?:scratch2:
 
I overlooked the imbalanced input bias resistances. That feature, as well as the placement of the Zobel network, looks an awful lot like the work of an apprentice designer. I'm suspicious that this amp's apparently prized sonic signature is more accident than intention. I generally prefer not to comment on sonics in any great detail, because they're so dependent on factors outside the amplifier. Lemme think about it...
 
I dont know a hell of alot about electronics, but i know this receiver is of amazing sound and construction. Ive been using mine non stop for a long time without any trouble whatsoever. It has an awesome phono section (not by specs but clearly on sound alone) and sounds amazing with a cd player hooked up. Its probably my favorite sounding receiver. And i own a Pioneer SX-980. I own 3 STA-90's, just in case one dies.
 
STA-90's Sound

Metalownz said:
I dont know a hell of alot about electronics, but i know this receiver is of amazing sound and construction. Ive been using mine non stop for a long time without any trouble whatsoever. It has an awesome phono section (not by specs but clearly on sound alone) and sounds amazing with a cd player hooked up. Its probably my favorite sounding receiver. And i own a Pioneer SX-980. I own 3 STA-90's, just in case one dies.
Click to expand...

I have the 2100, 2080, 225, 95 and the 90. The 90 blows them all away! There is something special about this receiver. That's why this thread is here, trying to figure it all out. :scratch2:
 
STA-90 Qvox Sound Option

I have read crazy stories on the web about Qvox and what it is. All have been out in left field. Hoping you guys can confirm it's operation. Best I can tell it is simply taking a discrete channel's signal and sending it to the B set of speakers on the same channel. I have sent a mono signal into the unit and never a sound from the Qvox rear speaker. I can send a mono signal to it with the left channel at higher output than the right channel and I get sound from the rear channel. The stronger the signal it to any one side the louder the rear channel is. So for example is a drum roll swings strong into the left channel it also swings to the left rear. If you set the rear channel speakers off to the sides but still in front of you it really deepens the depth of the sound stage. Can you guys confirm this looking through the circuit diagram? It would be nice to let people know just what it is!:scratch2:
 
Metalownz said:
It has an awesome phono section (not by specs but clearly on sound alone)
Click to expand...

The phono preamp has a simplistic feedback network that probably deviates from the RIAA spec enough to produce obvious coloration. My guess is that you could get similar effects from any good equalizer.
 
STA-90 Phono Pre

BinaryMike said:
The phono preamp has a simplistic feedback network that probably deviates from the RIAA spec enough to produce obvious coloration. My guess is that you could get similar effects from any good equalizer.
Click to expand...

I have measured the Marantz 2252B, STA-2100, Parasound and many others. The STA-90 is the only one that came real close to the standard RIAA curve. All the others missed by more than a few db and the low end rolloff started at around 60Hz. They must of included a sub filter in the RIAA in the other units. The STA-90 only starts to rolloff after 23Hz! It dones sound really good! They may also have a few different production models that have different parts than whats listed in the circuit diagram. I have seen this in many RS units.:yes:
 
Qvox STA-90

BinaryMike

Did you look to see what the circuit looks likes that's doing the Qvox mode? If you get a chance, I would love your opinion how you think it's working.:scratch2:
 
Last edited:
Metalownz said:
I dont know a hell of alot about electronics, but i know this receiver is of amazing sound and construction. ... I own 3 STA-90's, just in case one dies.
Click to expand...

You have an enviable opportunity to discover precisely which design elements are responsible for the sonic footprint. Change/upgrade one item at a time, and compare the modified unit with unmodified ones.
 
Let me see if I can do this an not confuse anyone (me included)

The QVox looks to be a basic implementation of the original Dolby Surround. The rear is playing the difference between the channels (the absolute value of L-R, aka |L-R| ) The QVox speakers are connected to the + side of the L and R channels. As you noted, a mono signal will produce no sound from the rear, as both channels are in phase and equal in volume- the difference will be equal to 0 [ |1-1| ] . If you have the sound 'balanced' to one or the other channel, there will be difference between the + sides equal to 1 assuming a mono signal [ |1-0| or |0-1| ]. If you have an out of phase signal playing through both channels at the same time, the difference between the + connections will be 2 [ |1-(-1)| or |-1-1| ]. That is, as one channels voltage/wattage approaches maximum (more positive voltage), the other channel will be approaching negative (more negative voltage).

Dolby simply did this math at line-level (before the amp) and then added an amplifier for the rear speakers. This allowed you to adjust the volume and balance in the rear to correct any channel skew in the original source.

Mike
 
Yeah, ill have to start swapping components out. But the STA-90 sounds awesome with alot of different equipment. The sound i get with Qvox is pretty spacious sound. Almost a live sound if you will. Some recordings dont sound much different though and others do. Overall though...i hold a very special place in my heart for this receiver. Its very underrated.
 
You must log in or register to reply here.
Back
Top Bottom