Looking for a power amp upgrade - currently using the Toshiba 2SA1943 / 2SC5200 what's new and available that would be an easy upgrade for those parts.
I have looked at the Sanken 2SA1216 / 2SC2922 and Sanken 2SC3858 and 2SA1494 but they don't seem that available from trusted sources.
Any other recommendations that could more or less drop in and are available?
I have looked at the Sanken 2SA1216 / 2SC2922 and Sanken 2SC3858 and 2SA1494 but they don't seem that available from trusted sources.
Any other recommendations that could more or less drop in and are available?
Last edited:
Oh, I see it now - they are using the more modern Toshiba TTC004BQ TTA004BQ transistors.
I will look into them then - thanks!
I will look into them then - thanks!
Looks like the the TTA1943 and TTC5200 are designed to be similar to the original 2SA1943 and 2SC5200.
So would they improve the sound quality ?
So would they improve the sound quality ?
Last edited:
Sound quality depends on the circuit design. Type of stages, amount of feedback. You should read the thread about wolverine design first.
I checked and the TTA1943 and 2SA1943 differ primarily in DC current gain and collector output capacitance.
"TTA1943 and 2SA1943 have minimal hFEs of 80 and 55, respectively. TTA1943 and 2SA1943 have COBs of 240 pF and 360 pF, respectively"
That may translate to better slew rates which is what I am looking for.
Slew Rate Estimation Formula
The slew rate (SR) can be estimated using a simplified formula:
SR ≈ Ib / C
Where:
- SR is the slew rate (in volts per microsecond, V/µs)
- Ib is the available base current (in amperes)
- C is the capacitance (in farads), which in this case is COB
1. 2SA1943 / 2SC5200
- COB (Ccb) = 360 pF = 360 x 10^-12 F
- Assume Ib = 20mA = 0.020 A
2. TTA1943 / TTA5200
- COB (Ccb) = 240 pF = 240 x 10^-12 F
- Assume Ib = 20mA = 0.020 A
Results
Based on these calculations, we get the following estimated slew rates:
- 2SA1943/2SC5200: Approximately 55 V/µs
- TTA1943/TTA5200: Approximately 83 V/µs
I doubt seriously that the drivers will start running out of gain at only 20 mA. More like 200 in EF 2, over an amp in EF3. The limitation on slew rate is almost never the output stage. IF the driver stage can cope with the higher capacitance, the larger devices will be better. If you are running multiple parallel outputs the smaller ones may again be better if the SOA is sufficient, as all that capacitance starts to add up.
Swapping in other power transistors without changing anything else in the circuit would in my opinion not result in a better sound quality as you put it. You would be very hard pressed to hear any difference,
There is a dreaded NE5534 driving the small/ medium transistor stage, so I plan to change those out for one that matches the amps frequency function. OPA627 for bass and OPA828 for the amps doing the upper mid and high = there are 4 amps in total.
Dreaded 5534? Really?
What the heck are you reading anyway? The performance of the 5534 is beyond what you need, a 5532 is as well. If you were talking about a 1458 or 4558, I could see it. The 4136 is really bad, as is the TL072 (TL071 single).
Also, maybe changing the output transistors might help at higher power levels, but up that high the distortion produced by your loudspeakers swamps any improvements you are talking about. Take a breath, sit back and think about where the bottlenecks are in your amplifier/speaker/source chain. Now for the most sobering thought. Have you got any idea how many op amps a signal passes through during recording? Some aren't even close to the 5534 in performance, discrete designs are even worse. If the music ever went to tape, well that's another world of hurt for low distortion compared to what you are talking about.
Perspective. I'm all for real improvements, but that isn't the path you are on.
What the heck are you reading anyway? The performance of the 5534 is beyond what you need, a 5532 is as well. If you were talking about a 1458 or 4558, I could see it. The 4136 is really bad, as is the TL072 (TL071 single).
Also, maybe changing the output transistors might help at higher power levels, but up that high the distortion produced by your loudspeakers swamps any improvements you are talking about. Take a breath, sit back and think about where the bottlenecks are in your amplifier/speaker/source chain. Now for the most sobering thought. Have you got any idea how many op amps a signal passes through during recording? Some aren't even close to the 5534 in performance, discrete designs are even worse. If the music ever went to tape, well that's another world of hurt for low distortion compared to what you are talking about.
Perspective. I'm all for real improvements, but that isn't the path you are on.
My speakers are ALcons Pro-ribbons which go to 130db and have a flat response all the way up. Historically that was always always the problem with PA ribbons no audible warning of impending failure till they smoked. My whole chain has been redesigned the amp is the last part. My completely rebuilt 4-way mono analog crossovers are here: https://www.diyaudio.com/community/threads/dbx-234xl-opamp-crossover-upgrade.416571
I am aware of the production chain of music so I only use a few older tracks if testing - most of my music is very recent EDM straight from the DAW stuff and custom 32 bit recordings.
The old 1970's 5532 /4 has a slew of only 12v/uS - swapping them out is like night and day on my system.
But I get it, how people process sound in the 8 different areas of the brain does vary - some people can resolve detail other simply can't.
Heck 30% of people don't have an inner dialog and another % can't even visualize anything so of course we don't all process sound in the same way!
And for those people a 5532 and an 828 will sound the same and a CD will be just fine - that's why those audiophiles with 100K system always seem to play such dire straits type music, which as you say, has been mixed through a ton of 1970's opamps.
I am aware of the production chain of music so I only use a few older tracks if testing - most of my music is very recent EDM straight from the DAW stuff and custom 32 bit recordings.
The old 1970's 5532 /4 has a slew of only 12v/uS - swapping them out is like night and day on my system.
But I get it, how people process sound in the 8 different areas of the brain does vary - some people can resolve detail other simply can't.
Heck 30% of people don't have an inner dialog and another % can't even visualize anything so of course we don't all process sound in the same way!
And for those people a 5532 and an 828 will sound the same and a CD will be just fine - that's why those audiophiles with 100K system always seem to play such dire straits type music, which as you say, has been mixed through a ton of 1970's opamps.
Well, it isn't higher slew rate that is what you are hearing. If it misbehaves ultrasonically, the effects are heard lower in frequency. Various interactions are more likely. Faster op amps may well oscillate well above hearing too. Sometimes old op amps are just old and degraded depending on run time and temperature. Same for transistors. So replacing the old ones with current manufacture new ones might solve your problems. That's as long as you buy them through authorized distribution from known manufacturers of those products. It's too easy to change something and make a generalization as to why something was "fixed" or improved.
Yes, ribbons can fail without warning when you're having way too much fun. For that you want a tractrix horn. Just installed a pair of Klipsch Jubilees for a friend recently. Shipping ribbon speakers is a real nail biter. We saw many shipped in that didn't make it (Carver Amazing with magnets, they can explode). Canadian winters are the worst time to ship, like any cold climate.
As for people subjectively processing sound, sure. Some don't care, some can't tell while others can really assess things well. Those people tend to agree with measured results as long as you have the right equipment and know what you are doing. Expectation bias throws a wrench into subjective opinion.
Yes, ribbons can fail without warning when you're having way too much fun. For that you want a tractrix horn. Just installed a pair of Klipsch Jubilees for a friend recently. Shipping ribbon speakers is a real nail biter. We saw many shipped in that didn't make it (Carver Amazing with magnets, they can explode). Canadian winters are the worst time to ship, like any cold climate.
As for people subjectively processing sound, sure. Some don't care, some can't tell while others can really assess things well. Those people tend to agree with measured results as long as you have the right equipment and know what you are doing. Expectation bias throws a wrench into subjective opinion.
The high slew rate does have a beneficial effect on odd order distortion anomalies.
These are PA Pro-ribbons intended for the rigorous of touring - these are small 8-12" and in the range of 1-3Kw - not to be confused with large magneplanner HiFi types.
These are PA Pro-ribbons intended for the rigorous of touring - these are small 8-12" and in the range of 1-3Kw - not to be confused with large magneplanner HiFi types.
I beg to disagree. Completely in fact.
I don't know where you are getting your information, but it is incorrect. Slew rate distortion doesn't just pick on odd or even harmonics. If you are nowhere near the slew rate limit, it simply does not factor in.
I don't know where you are getting your information, but it is incorrect. Slew rate distortion doesn't just pick on odd or even harmonics. If you are nowhere near the slew rate limit, it simply does not factor in.