New transistor source

Hi everyone

I've been working professionally with solid state amplifiers for over 3 decades , and have been surprised by the sudden surge of devices going EOL. I believe that traditional class AB amplifiers are not going away, but are transitioning away from the mainstream to a boutique market. Valve amplifiers saw this happen a long time ago. With that said, there are a lot of high end solid state amps being built, and specialized parts sources are out there to supply them. I would like to list a number of sought after device type easily available, in some cases with a different part number.
They are available from Profusion plc in the UK. We have verified samples of everything listed here in pro audio gear made by LAR audio systems.
LAR audio is reason I had to locate some new sources.

KEC DEVICES (Korea Electronics Corp.):

KTA1381 / KTC3503 300V FAST VAS STAGE TO126 Replacement for the EOL ON devices really fast high voltage VAS
They should be in stock by Jan.

KTA1837 / KTC4793 230V VAS / DRIVER TO220 Replacement for toshiba 2SA1837 / 2SC4793 , good VAS and driver , in stock

KTA1700 / KTC2800 160V VAS / DRIVER TO126 Replacement for 2SA1220A / 2SC2690A , GOOD VAS and driver in TO-126
They should be in stock in Jan

KTB2640G / KTD1640G 150V 130W FAST DARLINGTON OUTPUT TO3P N Replacement for Sanken 2SB1647 / 2SD2560 improved SOA, in stock
This is a really nice output device for simple hifi amps

KTA1943A / KTC5200A 230V 150W WIDE SOA FAST OUTPUT TO3P L (TO264) This is an enhanced version of Toshiba devices, almost identical to 2SA1302 / 2SC3281 in every way , but higher voltage rating.

Unisonic This is not the most serious sounding name for semiconductors, but it is an established OEM supplier in Taiwan.

2SD669/A / 2SB649/A The TO126 versions are a good substitue for 2SA1220A / 2SC2690A There is some confusion about power ratings on the data sheets because they come various packages, including SMD, and the rated power is in free air We have verified that the SOA curves for the 2SD669A apply to all TO126 devices when on a heatsink.

2SA1837 / 2SC4793 The TO220 versions are also like the old favorite, but they are available in various case styles.

I may have post this in the wrong area, but searching threads, most appeared to discuss a specific need for a specific project. I hope I have been helpful to continue on with the construction of fine Transistor amplifiers.
 
I would love to buy KSC2690A / KSA1220A but they sound, um, questionable on your list.

I would love to buy 2SC5171 / 2SA1930 but they don't appear on your list.

I would love to buy 2SC4883A / 2SA1859A but they don't appear on your list.
 
Everthing listed has been verified with curve tracing, ft test, and an endurance test of a one second pulse at twice the rated SOA in the second breakdown region. Then, these parts went into LAR audio product for sale. We initially obtained factory samples of these transistors, then looked for a distributor to stock them. Profusion emerged as the best choice for us.
The KTA1700 / KTC2800 are great replacements for KSC2690A / KSA1220A, you would need to wait until Jan. The 2SD669/A / 2SB649/A TO126 are proven too. The only thing questionable is using other than TO126 packages, because the thermal resistance is higher with the alternative cases. Unisonic is a long standing OEM supplier for consumer products.

I have listed only those devices verified in our lab. If you explore the profusion plc site (it needs to searched with the "plc" because of similar names out there) , you will find many other types listed, including some current Sanken types.
For your 2SC5171 / 2SA1930 , please check the KTB1369 / KTD2061 These also will sub for 2SA968B / 2SC2238B. We have not verified these. The only spec difference is the ft. The Toshiba version could only do 200 mhz at higher than 10 mA collector current anyway. We used them in the past.

I totally understand that most engineers (and hobbyists) would like to have their much loved original parts back. With the analog amplification world going boutique, that will not be possible. However , very good functional replacements are out there. Some of these have actually improved end product test results for us. Have a look at the LAR 2600G amplifier. These are used in some amplifiers costing 5 figures of Dollars , Euros , or Pounds. Many parts are assigned different part #s because of intellectual property rights issues from the original manufacturer . The world of valves has experienced all this. And yet, once engineers embraced what was available and made the best choices, we have incredible valve amps of new build.
If anyone is interested in my qualifications in amplifier design, please message me. I don't want to come across as advertising anything here, beyond mentioning a couple of qualifying factors for sourcing parts.
BTW, those using legacy Fairchild from ON should stock up. Rumors are that this line will be EOL in phases, PNP parts first. This years EOL list seems to give that credibility.
 
Yes , all the huge semiconductor corporations are now only interested in electric vehicles , energy conversion, and networks. All very high margin. This comes right from sales reps when we asked for an explanation for disappearing parts. Especially when only the PNP part went EOL. We told them we need the small parts to use the big ones. Our 2600G has 36 power devices, but only 4 VAS transistors. They did not care. This is why I shared information on where to find continued sources. Just as with valves, there are a few semiconductor manufacturers content to supply a market which is stable and profitable, but not growing exponentially.
 
  • Like
Reactions: huggygood
I wonder when/if KEC and UTC are going to be stocked by Mouser, Digikey, Newark….. They already stock overpriced second sources for old obsolete industrial/military types, and those can’t possibly be very high volume (Microsemi, etc). I’m sure there would be enough turnover for the Korean Toshiba, Hitachi, and Sanken replacements to justify it - and if they were priced as high as two bucks a driver transistor they’d still sell.

Maybe they have some agreement with ON that prohibits it? But if they EXIT the market those agreements would go out the window, even if in place now.
 
Profusion has now added the expected new arrivals from KEC to the price list.
KTA1700 / KTC2800 160V VAS / DRIVER TO126 is a pair that should find fans here. These are exactly the same as 2SA1220A / 2SC2690A .

KTA1381 / KTC3503 300V FAST VAS STAGE TO126 is now on the price list as well.

I remember seeing some inquiries about Toshiba 2SA968A / 2SC2238A drivers here , regarding replacements for some hifi amps.
The specific curves for these were a bit unusual when compared to the types which were much copied between Japanese manufacturers. As a result , substitutions often had poor results in some designs. Designers liked these because of the near perfect gain linearity . Now there is KTB1369 / KTD2061 from KEC. These are an exact match electrically. The difference is that these are a full molded package.
 
Moderate Cob up to 25pf is fine even in VAS , if the design currents are correct for it. A current mirror front end is a good example of an application where higher Cob will still slew fast. We often used 2SA1837/2SC4793 in commercial amplifiers for VAS and pre driver in order to have some resistance to chain reaction failures. For more high end applications , we used the much missed MPSU10 / MPSU60 , which was available until recently from Central. Now KTA1381 / KTC3503 takes that spot (new board changes for pin outs) , with even faster slew rates.