• 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

SX-780 Left Channel Oscillation

ccheath

ccheath

Got an Idea!
Subscriber
I purchased this receiver a couple years ago from Ebay. Blown right channel had been replaced with NTE1281 at that time. About 1 year later, I recapped this entire unit and had been working well ever since. I had been using headphone before without problem and it had been awhile since I used headphone. Finally I decided to connect the headphone then release the speaker button and noticed my left channel meter moved up to a few watts. Pushed the button back in then meter returns to zero. I pulled it out of service due to suspicious oscillation. Putting it on the scope with 1 microsecond setting, I found it oscillating approximately 3 1/4 mhz at around 3 volts output(see pic). I checked the feedback resistor and the disk capacitor, both appears ok. Checked DC offset and appears good. Bias checked at zero even though no adjustment for it. Upon researching and found a thread about SX-680 oscillating (not SX-780 with different topology), I have considered either swapping modules but don't like to do so unless really necessary. Also have considered pulling the module out of my other SX-780 (need repair anyways) to test any changes but wouldn't want to take apart my other SX-780 unless really necessary either. Any pointer for me to look at? Would it be the faulty modules?

EDIT: After writing the above, I decided to check with my audio generator @ 10khz approx 1 watt @ output:

1st pic: described from above as an oscillation waveform ("L" channel).
2nd pic: Sine on "R" channel (not bad for NTE device)
3rd pic: Square on "R" channel
4th pic: Sine on "L" channel original module (ringing crossover distortion)
5th pic: Square on "L" channel
6th pic: Sine on"L" channel (o-scope speed halved to see - to + swing)

Does that look like a faulty output module?

Thanks.
 

Attachments

  • 100_1211.jpg
    100_1211.jpg
    67.3 KB · Views: 160
  • 100_1215.jpg
    100_1215.jpg
    89.3 KB · Views: 58
  • 100_1216.jpg
    100_1216.jpg
    72.4 KB · Views: 47
  • 100_1217.jpg
    100_1217.jpg
    81.8 KB · Views: 77
  • 100_1218.jpg
    100_1218.jpg
    76.3 KB · Views: 155
  • 100_1220.jpg
    100_1220.jpg
    74.6 KB · Views: 158
Last edited:
Register to hide this ad
Frankly, I'd swap modules and see what changes on each channel. Starting with voltage across the emitter resistors before and after. If the speaker load kills the oscillation, I'd try shorting the amp input, and then removing the load and seeing if it oscillates.

Of course, since I keep a few modules setting around, I'd try new ones, but swapping from and to a good channel works....
 
markthefixer said:
Frankly, I'd swap modules and see what changes on each channel. Starting with voltage across the emitter resistors before and after. If the speaker load kills the oscillation, I'd try shorting the amp input, and then removing the load and seeing if it oscillates.

Of course, since I keep a few modules setting around, I'd try new ones, but swapping from and to a good channel works....
Click to expand...

You have convinced me to do it, so I did and I thank you for that. I feel better knowing my receiver is fine(it's my favorite of all receivers I have for my bedside). It turns out the exact symptoms have changed sides. NTE (good module) measured 0.2mv (yes mv). Probably explains a little gap in the crossover of the waveform (pic 2 of the 1st post). Original (bad module) measured 0.0mv. I did find the waveform came out flawless in the original module until I ran it 1 watt for a couple minutes to warm it up. After it warms up, that's when it started acting up.

I did try shorting the input at the amp side of R233. It won't start oscillating unless signal is applied even if I removed the jumper. Once the oscillation starts, cannot be stopped by shorting the input again (at R233). Shorting the input while oscillating only reduces the output by approx a watt out of a few watts oscillating. Reengaging the speaker switch stopped the oscillation. It only happens after the module 'warms up'.


EchoWars said:
That's fookin' hysterical. Not a hint of bias within a mile of that module (OK, maybe the other channel, if you want to get picky).

That module is a serious POS. I suppose it is possible that something else is going on, but I'm betting you have a crappy module.

http://www.bdent.com/search/part.jsp?partnum=STK0050
Click to expand...

You're right about the bad module. Looks like the bias circuit is pretty messed up inside since it only messes up after 'warming up' for a minute or two. While waiting for any reply, I was working on my other SX-780 (yes I have two) that's gonna need alot of work and found the amp working FLAWLESSLY with the original modules. I think I'm gonna buy two from that company you listed. My 1st SX-780 being clean and recapped, having crossover gap on one channel with NTE is not something I'd like to skimp on. Yeah, I'm picky. :D But I'll live with it if I have to.:o


Thank you both Mark and Glenn for your help. I am going to order them when my payday comes. I'll post back how it goes when I order and the parts arrive in a few weeks.
 
Last edited:
Yes, let's buy the cheapest possible part, because that must be the one that isn't a useless counterfeit and has no bias circuitry...just like the one he's replacing.
 
EchoWars said:
Yes, let's buy the cheapest possible part, because that must be the one that isn't a useless counterfeit and has no bias circuitry...just like the one he's replacing.
Click to expand...

I was under the impression that the modules from Electronix.com were good quality I think Mark directed me that way for mine. If this isn't so please let me know Glen as I just ordered 4 more for future use.
 
All I know is that there are lots of bad STK's out there, most with no biasing circuitry (because that would make them more expensive to make, and if you're making ****, why try to make **** that really works?)

The ONLY STK modules I have ever bought have been from B&D, and they have worked just fine...biasing and all.

It's a ****ing minefield out there, I'm sad to say.
 
EchoWars said:
All I know is that there are lots of bad STK's out there, most with no biasing circuitry (because that would make them more expensive to make, and if you're making ****, why try to make **** that really works?)

The ONLY STK modules I have ever bought have been from B&D, and they have worked just fine...biasing and all.

It's a ****ing minefield out there, I'm sad to say.
Click to expand...

I totally understand I have been burned before ... A friend of mine bought me 2 STK's after frying mine .... he purchased them from ebay and they were total garbage. I wasn't trying to direct anyone to counterfeit parts I was just relaying what I've read and apparently many others have had good success with Electronix.com. I meant no harm.
 
Use them. But once you have them installed, measure the idle current through the emitter resistors. It it's zero, then make a mental note not to buy any more from them.

Personally, I'd find it hard to believe I was buying anything other than a counterfeit at that price.
 
I have decided to get another ehm... NTE1281 at my local electronic store. I WANT that receiver BACK at my bedside. :D I still plan to buy Echo's recommended modules. Here's why (See pic):

First 3 pics of NTE's performance and last 2 is my 2nd SX-780 with original Sanyo modules

1st pic: @ approx 50mw output (scope repositioned to see the crossover) (NTE)
2nd pic: @ approx 1w output (scope repositioned to see the crossover) (NTE)
3rd pic: @ approx 1w output (scope slowed) (NTE)
4th pic: @ 1w output (Sanyo)
5th pic: @ 1w output (Sanyo)

The NTE shows the bias of 0.0-0.2mv (yes mv). The originals Sanyo shows in the left 7.5mv and right shows 3.5mv but both shows identical performance in the last 2 pics.
 

Attachments

  • 100_1239.jpg
    100_1239.jpg
    22.8 KB · Views: 47
  • 100_1233.jpg
    100_1233.jpg
    89.8 KB · Views: 46
  • 100_1232.jpg
    100_1232.jpg
    79.5 KB · Views: 51
  • 100_1231.jpg
    100_1231.jpg
    79.7 KB · Views: 54
  • 100_1241.jpg
    100_1241.jpg
    55.7 KB · Views: 62
I am resurrecting this thread regarding the module that I had been planning to buy from BDent. I finally got it here. I plan to install it sometime this week or this weekend and I'll post the outcome of this module. Here's the comparison:

attachment.php


Looks pretty similar except the body looks like the original. The numbering on the NTE's right top corner is 4325 and BDent's module right top corner is 3261.
 

Attachments

  • 000_0052.jpg
    000_0052.jpg
    133 KB · Views: 524
Last edited:
I just installed one module on the right channel. I checked the biasing and it looks really disappointing since it only shows 0.5mv almost as bad as NTE which is 0.2mv. BUT checking with 10khz signal, it looks promising. I could not see any crossover distortion. I'd use my distortion analyzer but it needs work. When I complete the work and do further testing, I'll keep this posted.
 
Last edited:
I had been doing extensive testing on this chip. I had been feeding it 100hz, 1khz, 10khz, and 50khz signal using mostly at 0.01 watt out. While 10khz does show difference, I can see drastic difference between NTE's and BDent's modules at 50khz in 0.01 watt. Since it's getting late, I'll post some pictures tomorrow to see for yourself if the module is worthy of consideration. Now, I'm doing the listening test.
 
Last edited:
Some results

Here are some results. On the last picture is about the BIAS. In that picture, I forgot to mention the first row is 1khz at 1watt. The second row is 10khz .01watt and third row is 50khz .01watt.
 

Attachments

  • 1khzsine01w.jpg
    1khzsine01w.jpg
    36 KB · Views: 32
  • 1khzsquare01w.jpg
    1khzsquare01w.jpg
    33.7 KB · Views: 30
  • 10khz01w.jpg
    10khz01w.jpg
    73.5 KB · Views: 161
  • 50khz01w.jpg
    50khz01w.jpg
    72.8 KB · Views: 34
  • BIAS.jpg
    BIAS.jpg
    61 KB · Views: 42
Nice flat tops with no overshoot. Rise and decay are good in the 1k trace but not so good in the 10k trace. You pointed out a notch. What does that mean?
 
That these crappy modules are operating in a nearly pure Class B mode (no bias current to speak of).
 
EchoWars said:
That these crappy modules are operating in a nearly pure Class B mode (no bias current to speak of).
Click to expand...

So even the 3261 modules (from BD) are not really any good? What are the other possibilities, if any, other than just not driving the originals too hard?

John
 
You must log in or register to reply here.
Back
Top Bottom