how can you be so sure?...without looking at the datasheet?
the current figures i give are from sanken, the manufacturer..
seems that for you, what you don t know don t exist..
these devices are 50 mHZ Ft....
MP1620 Datasheet pdf - Transistor De Puissance De Silicium De Type De Moule - Sanken
MN2488 Datasheet pdf - Power Transistor - Sanken
for those interested in good devices...
MN2488 Datasheet pdf - Power Transistor - Sanken
for those interested in good devices...
This is right.
But on the other hand Mr. Nelson Pass don't use such 40MHz-80MHz BjT devices neither in the Threshold aera nor in the passlabs aera. He use only Motorola BjT parts in the Threshold area and in the passlabs aera only power MOSFETs (instead expensive high Ft BjT's) from IRF. Also by the currently models, where are inside PP instead single ended output stages.
And as I know, nevertheless there are nothing other Class-A solid state amp model with 80 MHz BjT devices inside, that clearly do outperform the models from "X" and "XA" series, even not the legendary 400A and 800A with slow Motorola BjT's.
He is right. At 50MHz transition freq, the gain is down to one. Starting at min maybe 5000 at DC. So you can easily see that at higher freqs the gain starts to drop very quickly. Discrete darlingtons can have much flatter gain versus freq curves. Only some specialized integrated darlingtons come close but don't make it.
Edit: the data sheet doesn't show gain vs freq curves. That tells me enough ;-)
jd
yes, i know it...the thing is that the -3db point is not at dc....
if you had read the datasheet, you would have see that there s
3 FLAVOURS of the transistors, gain given for 7 A collector current..
mini 5000
mini 6500
mini 15 000
the higher grade should do it..
i think that at 50khz, it should be enough..
these value of current gain are sustained in a range
largely exceding the audio band...
2AndrewT
It can, in measurements I posted in this thread ( post #14) You can find results for 200W/4ohms from two stage EF output with doubled output devices, all ONsemi´s (drivers MJE15032/33, output 2xMJL21193/21194). VAS current about 8mA. Dual slope protection in output included, most of distortion is originated from this protection (close copying SOAR).
Last edited:
yes, lumba, i was stunned, they really do the things more than well..
far from the dreadfull TIP14XX and other motorola...
sanken is not onsemi, for sure..
by the way, those who advocate that the FT is not enough are simply
missing an important point : if 30 A peak are to be supplied to the speakers,
it s ONLY the WOOFER which will see this current, so at a low frequency, where the beta is at his higher value...
Well it's up to you to try it, by all means. I'm just trying to warn you based on trying this same thing several times myself and finding out the hard way.
In the smaller Ic devices (8A) you can see that the 80MHz bw is down to 20MHz at 8A Ie. The Hfe - Ic is a DC spec. At 8A and 20MHz the gain is down to one. This is the same as a single BJT with an 20MHz GBW product.
jd
In the smaller Ic devices (8A) you can see that the 80MHz bw is down to 20MHz at 8A Ie. The Hfe - Ic is a DC spec. At 8A and 20MHz the gain is down to one. This is the same as a single BJT with an 20MHz GBW product.
jd
i have no trouble trying, since you miss one more case..
since a darlington is two transistors in serial , each add his -6 dBfrequency roll off, ending with a 12db /octave total roll off....it s easy to make the calculus and see that from the UNITY gain at 20 Mhz , the gain increase rapidly when the frequency decrease and reach his full value as soon as 300 khz;; a single transistors is 6 db/octave...not the same thing...
of course no one will use a 8 A rated bjt to his full current, paraleling
is mandatory, thus increasing the Ft...
you took the most unfavourable figure, not realistic in real use, and
even then, it s not enough..
it s true that darlingtons are heavily vilified by people who lacks
some basical technical understandings...
wahab,
or take a look at this superbly matched pair for voltage amplification, 2SC3600/2SA1406, conveniently reaching 300MHz@10mA , Cob=1.8/2.3pF, with exceptional linearity.
In the other thread, tiefbassuebertr has been fooling around for months, seriously investigating the sloth MJE 340 / MJE 350, giving immense distortion in the demanding VAS, I don`t get it...
or take a look at this superbly matched pair for voltage amplification, 2SC3600/2SA1406, conveniently reaching 300MHz@10mA , Cob=1.8/2.3pF, with exceptional linearity.
In the other thread, tiefbassuebertr has been fooling around for months, seriously investigating the sloth MJE 340 / MJE 350, giving immense distortion in the demanding VAS, I don`t get it...
yes, i have them in my PDF datasheets, but as they are video type,
i refrained to llok further...it does remember the old BF471/472 ; also
video, with low Cob, but unpleasant results..
i will make a try..
currently, i use 2SA1145/2SC2705 , fairly old devices, but i got many of
them for free, and after all, at 200 mhz/2.5 and 1.8 pf respectively,they are good enough..
2SA1360/2SC3423 have the same specification, frequency and Cob,
only the casing is changed...
i use mainly the latter and of in low level, 2SA872A/2SC1775A,
2SA970/2SC2240, 2SA992/2SC1845....also get them for free..
crazy what the old japanese amps that are scrapped here
in france by the hundreds can have as transistors !!!
about tiefbassuebertr, i just don t understand why he goes after
thise antic bjts...perhaps the price, that allow to burn them by
the tens...i would understand for experimentation purpose,
but for a serious amplifier, it wouldn t be cautious...i just remember all
the horrible amps i did repair that were using them along with
motorola MJ15XXX...really, ugly sounds...
I have briefly reviewed this thread.
Gentlemen, I would like to ask a simple question. How many of you have build and compared output stages with A1943/C5200, MJL1302/3281 and MJL21194/93? If you did, how was the 'linearity'? I know that BV has done such comparisons. Others, would you expect the 1302/3281 to give better results than 21194/93? Would you expect there are only advantages of faster output devices in a real world amplifier?
Gentlemen, I would like to ask a simple question. How many of you have build and compared output stages with A1943/C5200, MJL1302/3281 and MJL21194/93? If you did, how was the 'linearity'? I know that BV has done such comparisons. Others, would you expect the 1302/3281 to give better results than 21194/93? Would you expect there are only advantages of faster output devices in a real world amplifier?
hi, pavel..
in my experiences, the toshiba and sanken devices with higher Ft give almost
a little less distorsion than the onsemi....
the main difference is in the nature of this distorsion :
the 2SA1943/2SC5200 and to a lesser extent 2SA1302/2SC3281 give
significantly less high order harmonics residuals...
as soon as the seventh harmonics is reached, the onsemi are outperformed..
surely that high levels of NFB reduce the difference, but personnaly,
i use relatively small amounts of NFB, and in this case, the difference
is not to be forgoten when designing such amps...
as to tell from where the difference originate, tough question...
i think the parasistic capacitances, and surely a more refined process..
in my experiences, the toshiba and sanken devices with higher Ft give almost
a little less distorsion than the onsemi....
the main difference is in the nature of this distorsion :
the 2SA1943/2SC5200 and to a lesser extent 2SA1302/2SC3281 give
significantly less high order harmonics residuals...
as soon as the seventh harmonics is reached, the onsemi are outperformed..
surely that high levels of NFB reduce the difference, but personnaly,
i use relatively small amounts of NFB, and in this case, the difference
is not to be forgoten when designing such amps...
as to tell from where the difference originate, tough question...
i think the parasistic capacitances, and surely a more refined process..