Hey all, the title pretty much sums it up. All the higher wattage amps seem to get all the love but I am looking for the values of the internal components of the STK439. I have the datasheets which give an equivalent circuit but no values. I was thinking of using TIP31/TIP32 for output trannies and 2N3904 and 2N3906 for the driver section. I know TIP29 and TIP30 would be a closer match but I got a deal on the others, so was gonna give them a shot. I also have, what I am sure is a fake STK439, since I just bought it on ebay, though it seems to work ok. I suppose I could cut it open but I would hate to ruin what I paid money for. I would rather build my own as a more permanent solution. I have the original STK439 with one dead channel. I cut it open but I am not sure I can get all the right values with the connections all still in tact. Maybe I could cut the traces before and after each component?
My other option is to use the circuit shown on Fig 16 at the following link.
https://www.nutsvolts.com/magazine/article/bipolar_transistor_cookbook_part_7
Not the same output power but at 38V input it may be OK .
STK439 Datasheet: https://datasheetspdf.com/pdf/944755/Sanyo/STK439/1
I plan to release my final schematic, PCB gerber files, etc when I am done. Any help would be appreciated.
My other option is to use the circuit shown on Fig 16 at the following link.
https://www.nutsvolts.com/magazine/article/bipolar_transistor_cookbook_part_7
Not the same output power but at 38V input it may be OK .
STK439 Datasheet: https://datasheetspdf.com/pdf/944755/Sanyo/STK439/1
I plan to release my final schematic, PCB gerber files, etc when I am done. Any help would be appreciated.
The two compensation caps are easy: 22 pF.
You likely won't be able to find an exact match for the transistors used, but substitutes should be somewhat easy to find. TR1 and TR2 are responsible for the voltage gain of the circuit (mostly TR2). TR5-8 is a quasi-complementary output stage. TR3-4 form a two-transistor VBE multiplier.
The resistors look to be thick film types - so basically black conductive paint applied to the PCB. To measure them you'd have to take them out of the circuit either by cutting the traces or by cutting the bond wires to the transistors they connect to. If you're looking to recreate the circuit with substitute transistors you'll have to tweak all the resistors anyway. So why not just design something similar?
Fig. 16 from the Nuts & Volts article you link to seems like a better place to start. Personally I would limit the range of RV2 both so builders won't accidentally turn the bias to the max and fry the transistors and also so the circuit fails to minimum bias if the wiper of RV2 lets go.
Tom
You likely won't be able to find an exact match for the transistors used, but substitutes should be somewhat easy to find. TR1 and TR2 are responsible for the voltage gain of the circuit (mostly TR2). TR5-8 is a quasi-complementary output stage. TR3-4 form a two-transistor VBE multiplier.
The resistors look to be thick film types - so basically black conductive paint applied to the PCB. To measure them you'd have to take them out of the circuit either by cutting the traces or by cutting the bond wires to the transistors they connect to. If you're looking to recreate the circuit with substitute transistors you'll have to tweak all the resistors anyway. So why not just design something similar?
Fig. 16 from the Nuts & Volts article you link to seems like a better place to start. Personally I would limit the range of RV2 both so builders won't accidentally turn the bias to the max and fry the transistors and also so the circuit fails to minimum bias if the wiper of RV2 lets go.
Tom
Thanks for the reply Tom. Yeah, the caps were easy. I wanted to try and stay as true to the original as I could, for no other reason than for the experience. But I was leaning toward the article since it would be easier and seems close enough.
Don't change the volume too fast on the new STK, and use proper heat sink mounting procedures.
If it is working, just buy and keep another as a spare.
Or build a new set of mono chip amps, wire to fit.
STK modules were known to fail without any reason, so you are lucky your supply or speaker did not get damaged along with it.
Too risky to repair, really.
And it is within 15 Euros / USD.
Check the supply, it is supposed to run at 39 V, max. 56 V, so a 3886 might work, and I personally would strip out the STK and put a 3886 or ready transistor based amp board, they are quite cheap here for transistor boards.
Safer than blown transformer and speakers, happened to me.
I was able to get the transformer rebuilt, and a Chinese STK worked, speaker also repaired.
You may not have such facilities in your area, and may be expensive.
Think about that before starting on repairing a damaged STK.
Also, they went out of production about 1995, so ageing issues also are likely.
If it is working, just buy and keep another as a spare.
Or build a new set of mono chip amps, wire to fit.
STK modules were known to fail without any reason, so you are lucky your supply or speaker did not get damaged along with it.
Too risky to repair, really.
And it is within 15 Euros / USD.
Check the supply, it is supposed to run at 39 V, max. 56 V, so a 3886 might work, and I personally would strip out the STK and put a 3886 or ready transistor based amp board, they are quite cheap here for transistor boards.
Safer than blown transformer and speakers, happened to me.
I was able to get the transformer rebuilt, and a Chinese STK worked, speaker also repaired.
You may not have such facilities in your area, and may be expensive.
Think about that before starting on repairing a damaged STK.
Also, they went out of production about 1995, so ageing issues also are likely.
Last edited:
I bought this receiver already with a dead channel in hopes of repairing and restoring the unit. That's basically been done. I've already recapped it, replaced all the faceplate lights, and realligned the tuner. The new STK439 is mounted to the heat sink with new thermal grease. I'm only doing this because I can and like to learn. It keeps me busy and my mind sharp. It's been a long time since engineering school and I have only managed to find tech work in manufacturing and refurb around my area but I'm happy. I never really got to do any real engineering, so this helps fill a need for me. 🙂 I'm not looking to repair the original STK but make a replacement of my own for it and maybe help a few others out. I have plans to mill my own aluminum mounting plates to mount the new components and PCB. It will be shaped like the original so it can be a drop in replacement.
First glance at the datasheet shows a total lack of short circuit and SOAR protection, which is asking for trouble in the real world
I've read the thread on here already and have it bookmarked. It is helpful but doesn't have what I had wanted as far as the STK439.
Not looking to start a business or anything. I plan on releasing all the documents and gerbers on here and elsewhere for free. I might sell a few boards or something on ebay to make up for the stuff I buy like the aluminum and so forth for my own project, but that's about it.
Beware that it's an equivalent circuit schematic in the data sheet. The actual circuit could be different. A simple way to check would be to count the BJTs.
Tom
I would expect as a minimum four extra small signal BJTs and some Rs and Cs. They don't seem to be there
I have not done any measurements as of yet, time is limited these days. But I believe the drivers to be the 3 pointed components with two wires forming a partially triangular shaped pattern. Either way, I think I'm just gonna breadboard a few amplifier circuits that meet the voltage supply available in the receiver and see which I like better. The more I think about it, the more trying to make something as close to the original as possible seems counter productive to making a reliable replacement. These types of amplifiers were not known for reliability.
Due to the 2004 Christmas tsunami Sanyo had to shut down it's solid state plant and the complete department. So any STK thick film module marketed today most probably is a fake. Why not replace both modules by a pair of chip amps, adjusted to appropriate gain? LM3886 or TDA7293 come to mind...
Best regards!
Best regards!
Yep. The little blobs with two bond wires connecting to them are indeed individual transistor dice. The collector connection is through the die itself.
Gutting the chassis except for the power supply and heat sinks and installing a chip amp instead has a lot going for it.
Tom
I nearly bought some stk modules many many years ago when I built a mobile disco.
However instead I went for a discrete 225WRMS amplifier.
Since then the stk has become notorious for failures so I made the right choice.
However instead I went for a discrete 225WRMS amplifier.
Since then the stk has become notorious for failures so I made the right choice.
STK439 is a little small for a mobile disco. And no, I wouldn’t trust 100 watt STK modules either. Temperature cycling will eventually kill them.
If I was going to build an STK439 replacement I’d be using TIP41/42 and MPSA06/56. And add the two transistors and support components to add a simple current limiter. No need to worry about DC protect because the 435/7/9/441 were meant for single supply use with an output cap. An LM3886 is “better” and can run off a single supply if you want to badly enough - but try finding one these days. STK439’s are more common right now.
If I was going to build an STK439 replacement I’d be using TIP41/42 and MPSA06/56. And add the two transistors and support components to add a simple current limiter. No need to worry about DC protect because the 435/7/9/441 were meant for single supply use with an output cap. An LM3886 is “better” and can run off a single supply if you want to badly enough - but try finding one these days. STK439’s are more common right now.
I haven't ruled out the idea of a chip amp. The initial goal was to make a drop in push/pull transistor style amp on an aluminum plate that would be easily serviceable like others I have seen for the Pioneer SX-780, for example. I now have two different receivers that use STK modules. The other uses a STK-043. I haven't decided to make a separate circuit for both or just one. The The STK439 fake I have been using seems to be holding up well on the first receiver. I'm waiting on parts before I do some testing on various designs. I might try a chip amp; I have a couple different ones already laying around.