SX-3800 repair

[Way San]

I have checked every component on the power supply and audio output boards. Replaced numerous blown resistors, a few caps and the output transistors. If I run it without the output transistors connected, I hear the relay click and the DBT has a dim glow. When I connect the output transistors the DBT lights a little too bright. Something seems amiss when outputs are connected. Yet the output transistors check good in all 6 positions of my Sencore Super Cricket tester. What other components could be creating this problem?

Sorry for being rather vague in this description but just grasping for some ideas at this point. And I had to put it aside for a while so my memory is fading on the details.

I am running:
2SC2525 with 2SA1075 in one channel
2SC2526 with 2SA1076 in the other channel

Any questions or suggestions would be appreciated.

 

[Bob Chieldon]

Are the bias pots (VR 4 & 5 according to the manual) wound all the way counter-clockwise as per the biasing procedure?
I have a SX D5000 with the same type of power amplifier design and it can be a bit ferocious to bias unless you follow the manual procedure carefully.

 

[QSilver]

This is an NSA amplifier - DO NOT TOUCH THE IDLE CURRENT TRIMPOTS - until you have replaced them with new ones. The originals are open element style trimmers. One bad contact as you turn them will blow the amplifier sky high. Someone with several A-9's learnt that the hard way recently - a forcible explosion and plastic shrapnel.

 

[Way San]

Thanks for your response Bob. Yes, I believe I have the bias pots counter-clockwise. I will double-check though. As I recall, the offset trimmers start at center position and the two bias trimmers should be counter-clockwise. I did also check those trimmer values and worked them with Deoxit.

 

[grinx76]

One bad contact as you turn them will blow the amplifier sky high. Someone with several A-9's learnt that the hard way recently - a forcible explosion and plastic shrapnel.

newb to fixing things here......
why is this and what about the NSA aspect makes it dicey?
do you mean one bad turn (like a scratchy pot)?

 

[Way San]

Any suggestions as to good replaement trimmers?

 

[QSilver]

Is it worth the whole amp going up? It's your call but for the sake of less than $5 of components to replace them I'd go ahead and do it. If you're ever unsure about the idle current adjustments, with power off, check the resistance across the trimmer. For idle current adjustments minimum idle current will be 0 Ohms.

You replaced numerous bad resistors. Did the amp blow out its output transistors?

 

[Way San]

I know the biasing is crucial. Its a direct coupled amp. Very unforgiving if anything is off. Thats why just about every component in the amp and supply circuits need to be carefully checked.

Have to be careful even with new trimmers. I bought some recently that had a dead spot in one. Brand new multi-turn trimmer. That was a disaster waiting to happen. Glad I checked it before installing.

The only thing I could not check was the zener diodes, if they are clamping at the correct voltage. ?? Need to look at the circuit again to see if they are in path when the output transistors are connected.

 

[Way San]

You replaced numerous bad resistors. Did the amp blow out its output transistors?

At some point one side was blown. An odd combo of output transistors was tried by someone. It obviously did not solve the initial problem. I received it blown again. I replaced the outputs with a good leftover pair from a previous SX-3900 job. The .47 ohm dual power resistor was also open in this channel. Could not find a direct replacement so had to gang together two individual 5 watts.

 

[Way San]

I also replaced the original output component connectors with the Phoenix type.

 

[QSilver]

Ok, there is a doucment called the Pioneer Tuning Fork. It details the way these amps work and how they fail and what fails and in what order when an output goes boom. Speicifally, its The Pioneer Tuning Fork Issue 2 supplement.

Could not find a direct replacement so had to gang together two individual 5 watts.

That's a bit surprising. Did you try searching for something that was 470m Ohm? Thats a more popular category than 0.47 Ohm in vendors sites. Do you have a scope? Have you checked the the driver transistors are genuine and are operating properly when under full circuit voltage and not under test with a meter?

MTF usually says if the NSA has blown, all the silicon has to be replaced. All transistors and all diodes except the thermal bias diode.

Here is an excerpt from the supplement:

 

[QSilver]

Any suggestions as to good replaement trimmers?

Bourns Trimmers

newb to fixing things here......
why is this and what about the NSA aspect makes it dicey?
do you mean one bad turn (like a scratchy pot)?

Its a very highly strung amplifier. Rather than having an idle current it sticks to, it uses high speed bias servos - these adjust the idle current based on the input and output of the amplifier itself - making sure distortion is at a minimum. The servos run VERY fast. These transistors are all rated upwards of 100MHz for all stages except the outputs which are rated 30-60MHz. It requires this high speed to adjust the bias in real-time and not miss the peak.

Things that run that fast, only take milliseconds to turn both sides of the amplifier, positive and negative on causing an instant short circuit between +/- rails. So if the rails are +45V and -45V for example, thats 90V and a high current that suddenly has nowhere to go and its converted to heat in the outputs first. These will heat instantly overload and explode in a worse case scenario. The trimmer resistors are responsible for controlling the bias servos, if they go high resistance or open circuit for an instant, the bias servos are slammed on full.

do you mean one bad turn (like a scratchy pot)?

Exactly, scratchy pot. Sometimes moving them less than a degree will cause a "crackle".

 

[Way San]

Bourns Trimmers

Any particular wattage? multi-turn? What would you recomment that will fit physically? What have others used that worked well?

Thanks for all the helpful suggestions.

 

[grinx76]

Ok, there is a doucment called the Pioneer Tuning Fork. It details the way these amps work and how they fail and what fails and in what order when an output goes boom. Speicifally, its The Pioneer Tuning Fork Issue 2 supplement.


That's a bit surprising. Did you try searching for something that was 470m Ohm? Thats a more popular category than 0.47 Ohm in vendors sites. Do you have a scope? Have you checked the the driver transistors are genuine and are operating properly when under full circuit voltage and not under test with a meter?

MTF usually says if the NSA has blown, all the silicon has to be replaced. All transistors and all diodes except the thermal bias diode.

Here is an excerpt from the supplement:

View attachment 2538431

Thanks for the reading lesson for homework, doing it now.
Thanks for the high-level reply to my inquiry as well.

 

[mikiekimi]

Have you tried it with some of the outputs disconnected to see which circuit might be the issue?

My SX 3900 (a current repair thread) was having a similar issue. Not sure how much the 3800 and 3900 have in common, but it sounds like the issue I was working with MarkTheFixer on. Start at post #23 to follow my process along.

https://audiokarma.org/forums/index...nd-or-relay-click.990586/page-2#post-15556862

It's still very much a work in progress, but I did get over the glowing DBT hump.

 

[Way San]

Tried to download some of the Tuning Fork files but the downloads fail for some reason.

Thanks for all the helpful thoughts everybody. Need time to dig into this.