• Please note that there are a few updates and clarifications made in the Audiokarma Rules, mostly relating to advertising and the addition of the new "Paying it Forward" & "Giving back" forums in the AudioKarma Audio Marketplace section.

SX-3800 Repair/Restoration Dead End

C

ClintB

New Member
At the risk of starting another SX-3800 thread, I'm at my wits end and I hope someone here can help me. Here's the story:

When I first received this SX-3800, I powered it up through a DBT. The only thing that came on were the lights that illuminated the AM/FM tuning scale. Nothing else worked.

After several weeks of replacing various components, I ultimately ended up replacing all of the electrolytic capacitors and most of the transistors on both the Power Supply (GWR-122B) and Power Amp (GWH-139B) boards. In terms of the transistors I replaced what I could. I did not touch the two JFET's (Q202/206) on the power supply nor did I mess with the four 5-leg units on the power amp board. Otherwise, all transistors have been replaced.

Same with the diodes on both boards. I pulled and tested all (yes, all) of the resistors on both boards. Everything tested fine though I did replace one resistor on the power amp board because it looked like it had gotten a bit hot over time.

I re-flowed ALL of the solder joints that I didn't otherwise touch in the course of replacing components on both boards.

I also replaced the output transistors that are mounted to the large heat sink that the two boards are attached to. I also replaced the four plugs with the recommended Phoenix connectors.

What all of the above has resulted in is the unit will power on with the DBT light immediately going out, all of the various button lights, fluorescent tubes and digital display working. The protection relay closes. The catch is all of that stuff works so long as I have those four large output transistors unplugged. When I try to power it on (again, through the DBT) with those output transistors plugged in, the DBT bulb is brighter than bright and nothing else comes on nor does the protection relay close.

I have certainly reached the limits of my knowledge and probably the limit of my skills. I am humbly asking for some help here in this forum. Otherwise, this thing is probably going in the dumpster and I'll chalk it up as a learning experience.

Thanks in advance.
 
Register to hide this ad
Don't pitch it just yet. I threw away a few things decades ago that I now know could have been saved for less than $20 and a few hours poking around under the hood. I still consider myself a bit of a newb and that servo bias circuit is something I have not come across yet. I would probably have to read the circuit description at least a few times before understanding what's going on there.

I read through your original posts and looked through the SM & schematic. I think you still had some questions about the correct power supply voltages? You can check yours against the AF block diagram on the bottom of page 7 of the SM.

Did the original output transistors measure shorted? As you may have read, most being sold on eBay will be fakes.
 
Thanks for the reply and I have read that post from Oleg's Vintage Audio many times. In fact, most of the replacement components I've used were based on what he had listed.

My rationale for replacing the original output transistors was they weren't all original. Two of them appeared to be (2SA1075/2SC2575) but the other two were something completely different. I ended replacing all four with MJL1302AG/MJL3281AG pairs per another repair thread here on AK. And frankly, I'm not entirely sure how to test a transistor like that.

My gut feeling is my power board is pretty well sorted. In the course of my repair/replacements on that board, I got my pin voltages to match up pretty well with what others were seeing and what the schematic called for. I also base that on the fact that the protection relay is closing now where it didn't before.

I've been able to set the DC Balance per the SM procedure but I can't get anywhere close on the idle current. And part of my troubleshooting was replacing all of the VR 1-6 pots with new Bourns units that tested good before installation.

Thanks again for the reply and advice.
 
you don't say on the idle current what you do get. IS IT HIGHER THAN TARGETED? OR LOWER??
Changing the transistors can change the Vbe calculations the circuitry was designed for.
The replaced drivers and outputs have Vbe changes from the original transistors.
A few millivolts each, but in EXACTLY the wrong place.

At 0 ohms on R3 & R4 (the 100 ohm vr's used for setting the 120 millivoilts) what is the idle current reading?
This assumes VR1 & VR2 are at their 0.000v target.

The DBT throws things off badly. Yet this is where it is needed. What wattage are you using? It should be 150 watts (or more !)

Basically we have to come up with a graph showing the resistance range of VR3/4 and the idle current produced at that resistance.
WITHOUT endangering everything with excessive idle current. (tens of amperes)

Then there are a few threads where changes around VR3/4 tame things.
Remember, the 3800, 3900, d5000 & d7000 all have approximately the same amp circuit.
 
Double check the Phoenix connectors for proper wiring.
For some reason I mess that up all the time. No matter how many images I take. In fact. The wire colors from the factory will not always match each other from one set to another. It only takes one transistor miss wired to make that DBT bright.
Start from the board pins and inventory each wire to its correct output pin.
The driver transistor specs are very important on this amp (Marks post). If the originals are good, I would continue to use them.
 
Ok, confession time....

Once I was able to power on the unit (with the DBT) and have the light on the DBT go out (100W bulb) AND the protection relay close, I've since been doing various measurements and attempting the idle current procedure sans DBT. I can only do this with the output transistors unplugged. Rolling the dice, I know.

I am able to set 0v with VR1 and VR2. The readings bounce around a bit but I've seen that sort of thing before.

In regards to VR3 and VR4 idle current, I'm not seeing any voltage at all. Zero. And moving either VR3 or VR4 produces no reading change on my DMM.

Maybe I'm doing this wrong? Since I've installed the Phoenix connectors, my technique has been to unscrew the wires for pins 28 and 25 from the connector and attach the alligator clips of my DMM to those wires. Same with the other side (20 & 17) with the same result.

Again, all of the above is done without the DBT and with the output transistors disconnected.

Thanks....
 
zebulon1 said:
Double check the Phoenix connectors for proper wiring.
For some reason I mess that up all the time. No matter how many images I take. In fact. The wire colors from the factory will not always match each other from one set to another. It only takes one transistor miss wired to make that DBT bright.
Start from the board pins and inventory each wire to its correct output pin.
The driver transistor specs are very important on this amp (Marks post). If the originals are good, I would continue to use them.
Click to expand...
I had the exact same thought because it seems the orientation of those Phoenix connectors is different from the originals. And the wires cross over one another. I had a couple of wires get detached and I had to solder them to the pins. When that happened, I checked everything over but I will do it again and make sure I have it right.

Thank You!
 
Double check the outputs for shorts.
You could connect one output at a time and DBT for a dim lamp. I used the term blipping the DBT as in turning it on and off watching for the bright lamp.
Once you can tell is not going to dim, which is around a half a second, switch it off.
This amps feedback arrangement will have some give, letting you connect one at a time.
You get the picture?
 
Last edited:
zebulon1 said:
Double check the outputs for shorts.
You could connect one output at a time and DBT for a dim lamp. I used the term blipping the DBT as in turning it on and off watching for the bright lamp.
Once you can tell is not going to dim, which is around a half a second, switch it off.
This amps feedback arrangement will have some give, letting you connect one at a time.
You get the picture?
Click to expand...
I think so... I'm assuming that the purpose is to attempt to narrow down which output is giving me trouble.

I've tried powering it on (always through the DBT) with two outputs connected (one PNP and its matching NPN) at a time with the result always being a bright light. I'll definitely give your suggestion a try.

Thanks!
 
After double checking the output connectors wiring and outputs for condition look closely at Mark's post above.
 
Double check the trimmer part numbers (Ohms values) and if their in the correct positions, if you replaced them.
 
FWIW, I wanted to follow-up on this.

Over the last week, I had the Power Amp board out. I was double and triple-checking wiring, solder joints, etc. I pulled all of the trimmers and confirmed that they worked and that I had the correct components in the correct places.

I reinstalled the board with the same result as before. Powering on the unit with a DBT (150W bulb this time), the DBT will go out, the protection relay will close, the front panel display will come on, I can tune radio stations, and the florescent displays will work IF the output transistors (Q1-Q4) are disconnected. None of the previous will happen if Q1-Q4 are connected.

I did do some experimenting and found that I can get the unit to power on with the DBT lamp going out and the relay closing with each individual output transistor connected one at a time. It will do the same with either the two NPN's or two PNP's connected at the same time. I can also get it to power on normally with the opposite channel NPN/PNP connected. Either Q1 and Q4 connected or Q2 and Q3 connected. As soon as I have either Q2 and Q4 connected (the other two disconnected) or Q1 and Q3 connected, its a very bright light on the DBT and the protection relay is open.

I don't know what else to check. Clearly, there's something going on here that's way above my pay grade. I also think that I'm not ready to try and set the idle current until I get this sorted out. I can get the DC Balance to zero out.

Any additional ideas/help is appreciated.

Thanks!
 
it is as I first posted.

your inclusion of the output transistors allow them
to do something outrageous that the rest of the amp is calling for,
so we look at what they are being told to do.
There's NO indication of your trying to get a zero idle current with the 100 ohm pot at 0 ohms.

I also have deep concerns about your measurement method, these measurements for idle current are
in millivolts, read across pins 28 to pin 25 (120mv or 150mv targets) or pins 17 & 20
all of which are connected across (0.47 + 0.47 = 0.94 Ω) the emitter resistors.
Ohms law gives us the current, from the voltage across them.
It SOUNDED like you were BREAKING THE CIRCUIT:
Maybe I'm doing this wrong? Since I've installed the Phoenix connectors, my technique has been to unscrew the wires for pins 28 and 25 from the connector and attach the alligator clips of my DMM to those wires. Same with the other side (20 & 17) with the same result.
Click to expand...


So THIS time, with all four outputs disconnected, take voltage measurements as such:
(SINCE THE OUTPUTS ARE DSCONNECTED, THERE IS NO CIRCUIT TO BREAK)

for left channel:
red dmm probe to pin 29 base of npn output transistor
black dmm probe to pin 28 emitter of npn output transistor
take and post voltage reading

red dmm probe to pin 26 base of pnp output transistor
black dmm probe to pin 25 emitter of pnp output transistor
take and post voltage reading


for right channel:
red dmm probe to pin 16 base of npn output transistor
black dmm probe to pin 17 emitter of npn output transistor
take and post voltage reading

red dmm probe to pin 19 base of pnp output transistor
black dmm probe to pin 20 emitter of pnp output transistor
take and post voltage reading
 
Here are the readings as requested:

Pins 29 & 28= 2.25v DC
Pins 26 & 25= -2.22v DC

Pins 16 & 17= 2.26v DC
Pins 19 & 20= -2.26v DC

The unit was attached to a DBT with a 150W bulb. The bulb dimmed immediately and the protection relay closed.

The attached photo shows how I had my DMM set. Alligator clips directly on the pins. A tight fit...

Thanks for your help!
 

Attachments

  • IMG_1669.jpg
    IMG_1669.jpg
    138.3 KB · Views: 5
Last edited:
this is with the 100 ohm pots at about 0 ohms?

they should be about +0.6 volts and -0.6v, and adjustable around that voltage. 2.25v is wayyyy too high.

with the output transistors STILL removed, get off the DBT, plug straight into the wall.

Then repeat the measurements. Then there will be other measurements, later.
 
I repeated the measurements. I confirmed Vr3 and Vr4 were both at 0 ohms and had the unit plugged into the wall. Clips directly on the pins.

The readings were significantly different this time:

Pins 29 & 28= 0.67v DC
Pins 26 & 25= -0.66v DC

Pins 16 & 17= 0.67v DC
Pins 19 & 20= -0.67v DC

The difference was where Vr3 and Vr4 were set. Neither was at 0 ohms the first time.

Thanks.
 
VR3 & VR4, as you have seen, are profoundly important.

those 0.67 and -0.67v are exactly what we needed.

NOW, output transistors can be connected and idle current set.

although DBT can be connected for a TRIAL, TEMPORARY test setting of the idle current, that setting will be wayy too high when not on DBT.
So idle current VR3 & VR4 HAVE to be returned to zero for non-dbt operation.

ANY time power supply or amplifier work is done, the idle current HAS to be zero'd!!!! THEN the correct idle current adjusted in.
FROM BELOW (i.e. 0 current)
 
Last edited:
You must log in or register to reply here.
Back
Top Bottom