faster transistors

[yugaaa]

Does sound improve when fast transistors are used ?Naim claims that the use of fast switching transistors improves the sound (their special output transistors) have a fast switching capability .

I have tried using some fast transistors ,and found remarkably good sound ,but I did also like to know views of others here ,on diyaudio .

 

[AndrewT]

I'm listening too.
info on predriver, driver and output transistors would be nice.

 

[Mr Evil]

Well, faster transistors in themselves generally aren't going to be better for audio, given that they tend to have lower gain and be less linear. However, there are indirect benefits. For instance you may be able to get away with smaller compensation capacitor(s);if the slew rate for the stage preceding the output stage is increased then you may find that crossover distortion is reduced; gain at higher frequencies may be higher, which could result in lower distortion at higher frequencies (but probably higher distortion at lower frequencies).

 

[cunningham]

Really fast transistors are not generally a good idea for the VAS. If it is too fast, RF occilations may occur and cause problems. Generally you want to use a high gain, not too fast device for this.

 

[djk]

One of the better sounding amps I own uses outputs rated at 800Khz.

It sounds way better than most newer stuff with 30Mhz outputs.

It will get too hot if doing full power sine wave testing above 10Khz, so what? I don't listen to 800W worth of 20Khz too often.

 

[Elso Kwak]

Fater is Better?

One of the better sounding amps I own uses outputs rated at 800Khz.

It sounds way better than most newer stuff with 30Mhz outputs.

It will get too hot if doing full power sine wave testing above 10Khz, so what? I don't listen to 800W worth of 20Khz too often.

I concur. Current trend is faster and faster also made possible by faster semiconductors being available. I feel we run into oscillation problems, RFI interference etc. resulting in glare and harsh sound.
The Metaxas Audio amplifiers come to mind.

 

[Elso Kwak]

fast as Light?

One of the better sounding amps I own uses outputs rated at 800Khz.

It sounds way better than most newer stuff with 30Mhz outputs.

It will get too hot if doing full power sine wave testing above 10Khz, so what? I don't listen to 800W worth of 20Khz too often.

Hi djk, I concur the better sounding amplifiers are not fast as light................
(did you get my email?)

 

[djk]

No phone for the last week, weather related. Trying to play catch up now.

 

[edl]

One of the better sounding amps I own uses outputs rated at 800Khz.

It sounds way better than most newer stuff with 30Mhz outputs.

It will get too hot if doing full power sine wave testing above 10Khz, so what? I don't listen to 800W worth of 20Khz too often.

2N3055?

 

[Elso Kwak]

2N3055?

Hi edl,
I am more inclined using MJ21193 and MJ21194 with a Ft of about 6MHz.

 

[PMA]

Re: Fater is Better?

Current trend is faster and faster also made possible by faster semiconductors being available. I feel we run into oscillation problems, RFI interference etc. resulting in glare and harsh sound.

Currently you have to be a better designer - that's all.

 

[djk]

"2N3055?"

It is an 800W amplifier, I suppose I could design one as an exercise around the 2N3055, but it would be tough.

The device in question is the 2N6259 which has even more safe area at 100V than the MJ21193/94 does and more current rating too.

 

[Upupa Epops]

To PMA : I agree - 2N3055 is only for heating.

 

[MikeB]

"2N3055?"

It is an 800W amplifier, I suppose I could design one as an exercise around the 2N3055, but it would be tough.

The device in question is the 2N6259 which has even more safe area at 100V than the MJ21193/94 does and more current rating too.

Why not choose a proven device like the 21193/4 ? And it seems to
have higher powerrating than the 2n6259.

 

[djk]

"Why not choose a proven device like the 21193/4 ? And it seems to have higher powerrating than the 2n6259."

Why don't you bother to look up the data before responding?

A 2N6259 is a 25A device, the MJ21193/94 are only 16A. The SOA of the 2N6259 is 25% higher at 100V than the MJ21193/94.

It is a 30 year old amplifier, if the MJ21193/94 had been available way back then they might have thrown a few more in the design and tried them. Now, of course, I have to suffer with the reduced SOA of the newer parts or pay through the nose for NOS parts.

 

[MikeB]

Why don't you bother to look up the data before responding?

I did, did you ?

mj21193/4: 16A, 250Volts, 250Watts
2N6259: 16A, 150Volts, 150Watts

All datasheets i found with google say this.
And i think powerdissipation is not too unimportant...

 

[djk]

I have no idea where you found that bogus data, probably from Mospec.

The RCA SSD204 handbook lists it as a 25A part and 250W SOA at 100V.

Video gaming people are familiar with this part too.

VDTs that use the 2N6259 will blow up lesser transistors in very short order.

 

[MikeB]

I've checked several sources and manufacturers including the mospec.
And yes, the SOA at 100 volts is slightly higher.
But, if one source says 25a, and others 16a, which is correct ?
What do you get if you order these ?

Mike

 

[djk]

I only accept NOS RCA or GE devices and even then they have to be selected for 170V for my amplifiers, I don't have much choice.

I guess with four in parallel for each leg I am only running between 7A~8A per device so the maximum current rating really doesn't matter, just the Beta at my peak current.

Looking at the MJ21195/96 I see that they have the same SOA at 80V that the 2N6259 does, I may have to try these.

 

[sajti]

I saw amplifier which use one(!) pair of MJL21193/94 and gives almost 200W on 4ohm load. The owner use it since 2 years, and it still running. I was totally frustrated, when I opened the top. I tried to find more transistors inside, but there was only two.

sajti