LM, TDA Chip Amps is **** Comparing with STK

Let me share my experience with small power amps.
Sometime in the early 90s, I assembled JLH 10W Class A amplifier.
Bench tested it and passed with flying colors.
Stable, no oscillations.
I was a proud owner of a Class A amp.

For auditioning, I plugged it to a pair of JBL4341 Studio Monitors.
The JLH fell flat.
Basically, couldn't drive the 15" and 10" woofers.

From then on, I always use a 100W or more power amplifier for bass.
Threw my JLH away after that.

Regards
Mike
AmpsLab.com
 
The OP may have a design problem but while a 10Hz to 20kHz amp is an easy option the speakers to handle it wont be. For sunsonic there are some posts around on ripoles. A speaker to reach and an F3 of 20Hz wont be easy, Same can be said for 30Hz really.

As JLH was a Wireless World man I suspect a KEF B139 would figure in his speakers so assume his amp will drive something similar and to a fairly low F3.
 
Point is, no STK available new except for couple or so released by the OnSemi company, which took over the Sanyo chips business.
There are 'good' Chinese versions, YMMV.

3886 and 7294 in higher output, and only the LM 1875 in the medium range (some ST models are sold as well)...choice is limited for new builds in the analog class.

It would improve the situation if the OP looked at the offerings sold in Class D by many makers...

And yes, good speakers with proper crossover selection are very important.
 
I don't know about that.
I remember one of the vintage EV speaker had a STK inside.
Onkyo also had a range of amplifiers using STK amps.

Mike
AmpsLab.com
STK failure rates were still damn high compared to discretes - and I'm not talking about thermal overload. Using open chip-and-wire die in a non hermetic, non overmolded air cavity package is a recipe for disaster. They have a nasty habit of just going bad without the user burning them out.

You could get discrete transistors with the SAME issue. The infamous 2SC458 (and others) and early TI “TIP“ types. The non-sealed packages let in moisture and eventually they just went bad.

The all-SMD modern STK replacements HAD thermal issues. And fake transistors inside didn’t help matters.
 
STK failure rates were still damn high compared to discretes - and I'm not talking about thermal overload. Using open chip-and-wire die in a non hermetic, non overmolded air cavity package is a recipe for disaster. They have a nasty habit of just going bad without the user burning them out.

You could get discrete transistors with the SAME issue. The infamous 2SC458 (and others) and early TI “TIP“ types. The non-sealed packages let in moisture and eventually they just went bad.
Well, maybe some did, but not all. Here's my real experience.
As mentioned above, I built an amp with the original Sanyo STK4048XII some 35 years ago. It was 150 Wrms stereo, dual Talema 200 VA torroids and 4 Nichicon 10K uF filter caps.
It was in storage in my basement when our neighborhood had a record flood from Hurricane Irene in 2011.
The amp, along with bunch of my other gear was completely submerged for 5 full days - not just wet, but six feet under.
All I did was remove the cover and power wash it to get the silt out, blow it out with compressed air and let it sit for about 6 months.
Cleaned the internal plug-in connectors and powered it up. Worked fine and didn't need to replace any parts. Still works and sounds great 12 years later.
Yes, I've repaired a bunch of other gear with failed STK devices over the years.
But every failed STK I pryed open for failure analysis was conformally coated with some thick clear goo that seemed to do a good job protecting the silicon from moisture.
And, as you may know, most board level electronic components are designed to be moisture resistant since they need to endure automated cleaning with aquaeous solutions in machines like the one listed below.
https://www.orion-industry.com/fabrication/washing_machine.html
 
My observation is that the environment had some effect, and Sanyo did not mark them with the name for OEM supplies, it is hard for repair people to decide genuine or fake in old sets.
Service parts sold by Sanyo did have the Sanyo mark, the volume was small in comparison to OEM supplies.

Also, the decision of supply was left to the builder.
For example, Pioneer ran STK 403-90 at +/- 40, and Sony ran them at +45/-46, on a device rated at +/- 50V max.
So the Sony would have a higher failure rate, and the STK was put to blame......

The genuine ones are said to have green epoxy boards, and some atmospheric sealant, but in any case they were much more prone to failure than other similar class devices, though in some cases they did not have competition except from transistor circuits, it is also to be borne in mind.

TDA 7293 and LM498xx are out of production, now I think the TDA 7294 is the highest output rated analog chip amp available...
 
Well, the STK thick film devices might be prone to fail, due to lack of personal experience I'm not about to judge, though. Fact is that genuine ones have vanished due to the 2004 Xmas tsunami that forced Sanyo to shut down their completely destroyed SS plants.

Best regards!
 
Sanyo was started by Mr. Matsushita's brother in law, and a few years back ran into financial problems, the semiconductor business was taken over by OnSemi, IIRC.
Much of the rest went to Mr. Matsushita's existing business, aka Panasonic.

OnSemi did release a couple of STK types, I have no idea of their availability...I get transistor kits with voltages from 15-0-15 to 80-0-80 in the local market here, the prices are competitive so I would rather buy a ready populated board if needed...I have excess amps in stock, and several await assembly.
 
Well, the STK thick film devices might be prone to fail, due to lack of personal experience I'm not about to judge, though. Fact is that genuine ones have vanished due to the 2004 Xmas tsunami that forced Sanyo to shut down their completely destroyed SS plants.

Best regards!
So is THAT what happened to my 2SC3902/A1507???? And I always thought it was one of those “business decisions”. No matter, I did a lifetime buy before the EOL.


The C2344/A1011 going away years ago was enough - didn’t need to go through that again.
 
Last edited:
There was a mention that a region specific launch had occured, the usual dealers like Digikey and element14 did not have in stock or not listed, for me in India.
No news after that, so it has not sold well, or at least no new variant has shown up.

Our prices have to be competitive, among the reason companies like GM and Ford quit India,they were unable to deal with our mindset.

We get D2030 and M1875 from China at 10 cents a pop, that is much less than TI price for 1875.
STK4141 (equivalent Chinese) was US$ 1.50....
 
Last edited:
https://www.onsemi.com/products/sig...parators/audio-power-amplifiers/stk404-070n-e

https://www.onsemi.com/products/sig...parators/audio-power-amplifiers/STK433-040N-E

Yes, they are detailed, so must be available.
And other versions for brake control and IGBT control are also listed.
They say original thick film technology is used.

Bear in mind that the same technology was used for SMPS controllers, electrical brakes, audio power and differential amplifiers, and other applications.

All were called STKxxxx, so STK was actually a description of the technology rather than a family of chips with similar applications.
 
Riiiiight.
 

Attachments

  • 7C70E106-0CE5-4D1F-841C-0620FF7ECEA3.jpeg
    7C70E106-0CE5-4D1F-841C-0620FF7ECEA3.jpeg
    130.6 KB · Views: 88
  • Like
Reactions: NareshBrd
So they stopped selling STK audio amps, for whatever reason....

As an aside, the TV techs here preferred to replace the entire PCB on a TV rather than repair a SMPS with an STK controller...those were not reliable, and a newer PCB had more functions, better remote, more sound output and cost Rupees 1000 (US$12.5 0) or so.
The STK SMPS repair would cost about Rs. 600 ($7.50) and be unreliable...and would have a lot of obsolete chips, which would fail sooner rather than later.

Changing the PCB took half an hour...hook up power, speakers, remote sensor, deflection coils, RF and RCA...there you were...
 
“For whatever reason” is because nobody uses them anymore (for new design). Same reason they change Mosfet offerings every 6 months. If they don’t get a major design win right away with a new product it goes EOL. Thats life today.
 
I feel that there is no point in making a product that has poor demand, and the Chinese will step in at times, their profit expectations are lower than US / European makers.

Volumes in OEM are far more than service / repair, and in a situation where a Class D amp costs $10, and tech time is above $50 an hour, there is no point in spending on repair.
People trash 5.1 home theaters because of broken transformer, don't want to spend money, even the independent service people are closing shop.

In any case, the production has changed over to Class D for the most part, because of many factors, and gradually Class AB devices will go EOL, as has happened with older ST products.