Anyone know what crosstalk on SX-1980 should be?

[markthefixer]

Haha! I can read a lot of skepticism into that first question; I'm sure it is justified.

<snip>

When it comes to audio equipment, I have no experience whatsoever actually fixing, upgrading, or otherwise building any of the components. I have always enjoyed good quality sound and I've always sought to reconcile the subjective human evaluation of sound (like warm, tight, crisp, involving, come to life, flat, etc.) with a more deterministic engineering based evaluation.

The EE in me says that what I want in any preamp or amplifier is absolute neutrality so that I get out exactly what I put in, just amplified. However, I'm not certain my ears would agree.

Furthermore, I'll admit that the traditional set of measurements (frequency response, THD) don't adequately cull out good sounding equipment from great sounding equipment.

I'm sure there must be a test yet developed, or a combination of tests that can do a better job of making this distinction.

You and me both...

I have come to rationalize that good sound is an individual phenomenon, because it is an interaction between the ears and the mind, and that it is even a moving target in each individual because the brain is constantly being trained to "hear better".

That said, it is also my opinion that "masking" occurs as energy at that frequency "band" and the cilia in the ear involved interact, and I am far less worried about phenomena at 60 plus db down.

Since the shorting input didn't cut the crosstalk further, I suspect it is in the inter-stage cables, and possibly the boards and the (operative) grounding.

With the exceptions of bypassing electrolytic caps with lower inductance caps (stacked film for upper audio and ultrasonic, and possibly ceramic caps for whatever RF), LED illumination source conversions and the current source replacements for unobtanium 50v jfets, doing recabling and other tweaks I relegate to the category of "modifications" from the original and don't spend time discussing or arguing them. I have enough worries already. The effects of bypassing electrolytics with stacked film caps in output capacitor coupled amplifiers is worry some (stability and "lost rolloff") enough to keep me distracted.

As for the original question, I was worried about (arguing with / or educating) someone with a great new toy, that they put on a golden pedestal and practically worshiped, without understanding the real word limitations involved. Even cabling can have it's own effects - sometimes I have to fight with my cabling and their connections to get repeatable, consistent results. Then I start to fight with the variable resistances inherent in old switches.... and connectors, and relays.... and so forth..... even AFTER I have "cleaned the daylights" out of them.

 

[capn]

As for the original question, I was worried about (arguing with / or educating) someone with a great new toy, that they put on a golden pedestal and practically worshiped, without understanding the real word limitations involved. [/I]

I do hold the 8903B in high esteem, but worship? No.

I save that for the fabled HP 8510 family of Microwave Network Analyzers

 

[Ed in Tx]

Out of curiosity I took a foot long piece of the stereo hookup coax from my scrap SX-950 rear panel, and measured the capacitance between the two conductors, about 15 pF. And if you shorted the input that would kill capacitive coupling from the input side, so happy hunting for the source of the crosstalk.

And I went to a HP "Signature Analysis" workshop sometime back in 1980! The national service manager who was ex-HP wanted us to go even though I had nothing to do with anything other than some HP test equipment on my bench. Still have a few of the blue HP pencils now with rotted erasers.

 

[rcs16]

Crosstalk discussion

Hello guys,

If you read Doug Self's book"Self on Audio", he explains in chapter 6, that the source/function selector switches, attribute in a large part to cross/inter- channel cross talk.
he sights, that he measured at 10KHz:
Simple rotary 71db
Morgan-Jones rotary 76dB ( I do not know exactly what this is)
2 pole change over switch 74 dB ( I take this as a regular interlock switch such as used in your case.)
4 c/o switch 95 dB ( special wiring arrangement to minimize cap coupling, which is the issue that you have discovered.
So your measurements are in line with what Doug has stated and make sense to me as well.
You have discovered the limitation of all these Pioneer designs. Of course Pioneer, do not want to spec something that they would not be otherwise proud of.
To get better performance, I use relays today in a pre-amp design, which offer much better performance/reliability/serviceability and route the traces on a pcb with ground planes, the discrete wiring as Pioneer has done, creates issues as well, probably more with regard to stray hum pickup. If you compare a SX-950 to a SX-1250, you can see the extra shielding that Pioneer used on the flagship SX-1250, even with a torroid xformer.
I do not think that -70db crosstalk is a bad thing or even noticeable.

On another note the HP8903, was good in its day, well if you compare it to a HP333. By today's standards it shows its age. I used to service these things back in the 80's when I was at Motorola. It was used in the automatic test systems that required HPIB/GPIB/IEEE488 control. Its distortion residuals are fair, a 339A has higher performance, as do many others.
I own a Amber 3501, I measure its own internal osc at -100db(0.001%), which is good enough for what I design/test. If I build a circuit and it does not add appreciably to this residual, I am pleased and look no further into THD.
If I remember correctly I do not think that a 8903 can even measure IMD, you need two osc's, and some LP/HF filtering, well that was a poor choice by HP, at the time. The Amber that I own, has this option and it was deigned in the same era.
If you want to learn some more about DA's, I suggest that you read this post, these guys are the expert EEs in this subject.
http://www.diyaudio.com/forums/equi...n-audio-range-oscillator-214.html#post3447194
Low distortion oscillator tests measurement circuits by
Vojtěch Janásek, JanasCard, Czech Republic- May 3, 2012
article in Test & Measurement World shows that he has designed a oscillator using standard OPAs that attains THD of -140dB, fantastic for such a simple ckt. I am sure Bill Hewlett would be amazed as well, having once met him, what can be done in this day.
Okay that's my input into this subject, have fun folks & take care.
Rick

 

[capn]

I want to thank all who've provided feedback to this thread; it is most appreciated.

I'll close it out by making one additional request, which is to ask any of you that find yourselves with an SX-1980 on the bench (and some extra time) to measure the crosstalk at 1 KHz and 20 KHz and to post the results back on this thread.

Thanks again