SX-3800 with dead display and no sound

Thespeakerdude8

Super Member
I have a SX-3800 on the bench with no flouroscan display other than the L and R lighting up. Also, it has no sound and the selector lamps do not light.

I checked the voltages at the power supply board and I am not getting 5.5v at pin 17 or 12.8v at pin 15. I believe these supply the digital devices and the flouroscan display. What components should I check, and is there any chance that the digital devices are beyond repair if the 5.5v rail blew?
 
if you have it, yes.. if not, test it thoroughly.

loss of a rail should not hurt the digital devices in the display. the drivers are open collector and isolated iiac.
 
if you have it, yes.. if not, test it thoroughly.

loss of a rail should not hurt the digital devices in the display. the drivers are open collector and isolated iiac.

I replaced the 2SD313 and I tested the 2SC2320 and it was bad. I replaced it with a 2SC1000 I scavenged since it is a aproximate sub. Voltages are back, flouro lights up and everything works! :music:
Thanks guys, sometimes even though I am riding on two wheels I like to know I have some training wheels to help me in case I do something stupid :D
 
3800 Question.

I replaced the 2SD313 and I tested the 2SC2320 and it was bad. I replaced it with a 2SC1000 I scavenged since it is a aproximate sub. Voltages are back, flouro lights up and everything works! :music:
Thanks guys, sometimes even though I am riding on two wheels I like to know I have some training wheels to help me in case I do something stupid :D

I know it's been a while, but I see 5 units of 2575 in a row on the flouro board, are these the ones you replaced? I have a 3800 on the bench now with low volume and no flouro display.

The 5 transistors I am wondering about are right under the tuning string in this picture. Thanks!!
 

Attachments

  • FullSizeRender 6.jpg
    FullSizeRender 6.jpg
    108.8 KB · Views: 81
  • FullSizeRender 5.jpg
    FullSizeRender 5.jpg
    92.5 KB · Views: 74
i believe the transistors in question are on the power supply board
q210 q211 q209
first check all power supply voltages to be sure you are on the right track .
 
Last edited:
Hi there, first off this forum has been fantastic in reviving my old sx-3800. It also suffered from the cracking power transistor bases causing the left channel to clip. Now that I know what that problem was and how to fix it I have moved on to the next problem.

I am getting near 0volts on the power supply board (pin 15 and 17) which from what I have read provides power to the tuner and two displays which do not work. I have re-soldered the 3 transistors mentioned (which also shows dark marks around them) and checked continuity to nearby pins. They seem to make contact so I am suspecting they are bad. I do not have an in circuit tester and need to find my solder wick!

I was hoping to get a little feedback on this, should I just go ahead and replace the three as the first step? Anything else I should be checking? I see several here have worked on this model specifically, any feedback would be greatly appreciated.

From another post these are the equivalent transistors (Thank you Mark!):

gwr-122 q209 2sd313 pwr ksc2073 512-KSC2073TU
gwr-122 q210 2sd313 pwr ksc2073 512-KSC2073TU
gwr-122 q211 2sc2575 darl ksc2383 512-KSC2383YBU

It looks like 512-KSC2383YBU is no longer available, anyone have any suggestions for an equivalent?

Lastly, its an old machine that still sounds great, anything else I should checkup on?

Thanks a bunch!
 
Welcome!! You REALLY should start your very own thread!!! One unit per thread is the etiquette. Otherwise it is a "thread hijack", even tho it may be a "thread resurrection" - but it's a different person and a different unit.

512-KSC2383YTA

pin 15 is +13v, and controls pin 17 in a way
pin 17 is +5.5v WHEN pin 16 is 12.8v telling it to turn on. pin 16 is a switched pin 17, possibly with diodes too.

I have to assume your +/- 47v is ok, and thus D207 is supplying +14v to Q211's base (and through it, to the other half of the high gain darlington pair : Q209)

q209, q211 have to have at least (or about) 20 volts on their collectors, for the transistors to work, with q211 base at 14 volts. Q211 emitter and Q209 base should be at (14 - 0.6v) 13.4v, and Q209 emitter at (14 - 0.6 - 0.6) 12.8v

Check those voltages before unsoldering anything.

Q210 could be OK too.

I'll let others talk you through "anything else".
 
Last edited:
Thanks Mark! I was also wondering if I should have revived this old post. I will remember that next time. Mostly in this case I was hoping to group like problems together, thought it might help others down the road not have to look in multiple locations.

I took a closer look and tested the additional voltages at the points described and WALLA!

the 47V appears to be pretty good (45.5V), I was getting the 14v at Q211's base, emitter voltage was at 32V, Q211 was passing through about 13.6v to Q209, q209 was passing through 0V and its emitter was also at 32V. I suspect the 2SD313 is bad.

I noticed Q201 and Q205 also have dark heat marks. I re-soldered them just for good measure, should I consider replacing them as well?

Unfortunately I have very few spare parts so its time for an order, would you recommend other common parts I should order to have on hand? I have a SX850 that I would use more often if I was brave enough (don't want it to break! :-D). I will be sure to add the new transistor bases as well (Mouser P/N: 651-1757022).

Thank you VERY much Mark! As I was reading through other threads it is apparent that you have great passion for this! Just think of how many electronics you have saved! :)
 
I did find some BD139 transistors. They are also 1.5A, NPN, but are only 80V, and .5v off on saturation voltage. They are also generics. Do you think the .5V increase on the output would matter? I could add a diode to the output to get a little closer and I don't think I would ever exceed the 80volts.
 

Attachments

  • upload_2017-6-10_16-21-28.png
    upload_2017-6-10_16-21-28.png
    166.4 KB · Views: 29
Last edited:
FIRST the major question - for WHICH circuit position. The same transistor can be recommended for several positions that have DIFFERENT requirements and that will impact the suitability of any replacement.

Then when I take some time to LOOK at the two you list, first off, the replacement and original are different cases to-220 versus to-126.
MAJOR difference in heat dissipation ability.

My time is limited, and balanced between the needs of various posters. I'm not killing time here - time is killing me.
 
Based on the testing results I suspect the Q209 (2SD313) is bad, I see the equivalent recommended is the Mouser 512-KSC2073TU, when I checked my spares box I did find some BD139's but was hesitant use it as a substitute based on the differences I found with the Mouser comparison. According to the service manual only 300mA runs through pin 15, it made me wonder why they cooked to begin with.

I hope you get the time you need, it does seem like the days are not long enough! Appreciate the help but no rush here.
 
I did find some BD139 transistors. They are also 1.5A, NPN, but are only 80V, and .5v off on saturation voltage. They are also generics. Do you think the .5V increase on the output would matter? I could add a diode to the output to get a little closer and I don't think I would ever exceed the 80volts.

By chance could someone take a look at my attached comparison on the transistors? I would like to try my BD139 for the position Q209 (2SD313) (512-KSC2073TU). I believe I can still attach the stock heat sink but only 300mA flowing (per the drawing) I would think the temps will be ok? I would have about 32volts on the collector, 14v base. Thanks!
 
Last edited:
You are trying to get someone to "bless" the substitution of a SMALLER to-126 cased transistor (12.5W) for a to-220 cased transistor(25W).

Go ahead, it's your unit, put it in like you seem hell bent to do.
It will be interesting to see how long it lasts, and how many other parts it takes along with it when it fries.

Sometimes they used to-220's without heat sinks in the regulators. IF the to-220 didn't have one (it does), and the to-126 you use gets a BETTER heat sink, it may not blow up instantly. 32v - 14v = 18 volts x 0.3amps = 5.4 watts. The to-126 is good for 1 watt bare, so it will quickly fry bare.
Plus this is a generic, so the internal heat resistance (die to outside of package) will probably be higher, meaning the transistor's die will get hotter under the same conditions as compared to one a manufacturer guarantees specs on.
 
Understood, I will get the correct parts order going now. I have added all the components from your power supply board parts list + the new plastic transistor sockets + Suggested replacements for Power amp (GWH 139) board:
GWH-139: Q107: 2SA750 (2SA726S): 512-KSA992FBU (common failure: causes crackling/popping noise through speakers)
GWH-139: Q108: 2SA750 (2SA726S): 512-KSA992FBU (common failure: causes crackling/popping noise through speakers)
Also found on a different thread that the JFET 512-J113 may sub out for the 2sk34 JFET.

I have also added in transistors from other threads in regards to my SX-850, I would like to have a small useful collection on hand for both units.

Here is what I gathered from other forums on the SX-850:
512-KSA992FBU Transistors
512-KSC2690AYS Transistors
512-2N4401TFR Transistors
470uF 80v Caps

Do you have any other recommendations of useful spare parts to have on hand?

Thanks a bunch.
 
2sk34 are left alone and RARELY fail.

2n4401 for what? use lots of ksa992 and ksc1845 (I buy them in the 1k quantity), less of ksc2383 (ksa1013) , ksc2690(ksa1220), ksc3503 (ksa1381) , ksa940,ksc2073, mje15032, mje15033
but these cover far more than a 3800.
 
Good to know about the 2sk34's, I thought the 2n4401 was a substitute for the 2sc1318 in the power board on the sx-850 but I may be getting things crossed by now. It is amazing how lost you can get sifting through the forums, truly a wealth of knowledge!

I will be sure to have several of the ksa992 and ksc1845 (not in the 1000's though, haha) and some of the others. Used to have a nice resistor collection, I think those are long gone though!

Do you think pin 15 (12.9vdc) could handle the extra load of a computer fan? Maybe even pin 17 5.5v for slower speeds? Given how warm the power supply board is I was thinking of adding a low speed fan to keep the temps down.
 
Last edited:
You may be getting tripped up by some of my ancient threads.

If you see anything for a ksc2310, ksd1616 those are either no longer available or a slightly bad choice that has been dropped. There are others.
Since you have a 3800 it is a NSA amp, finding TO-220 drivers for the power amp is becoming problematic because they need to be closer to 200 mhz Gain bandwidth product, pretty darn fast.
You also will find some to-92 transistors that have a THICKER center collector lead in the power amps - they are woefully under-heat sinked for their dissipation and are replaced with to-126 parts. they both are fast too.

Fans, even low speed, pack the dust and dirt in. I get the SX-D7000 power supply a lot cooler by judicious bundling of wires to allow much less restricted air flow. Given the chance, heat will rise and make it's own airflow.

Then a final gotcha for now: one that will become more problematic in the future since the merger of ON-Semi and Fairchild, sometimes there is a difference in the Vbe drop of different manufacturer's transistors, and that interferes with idle current range of adjustments, usually making idle current too high. I tried On-Semi in some circuit functions, and HAD to drop back to Fairchild.
 
Back
Top Bottom