Number of 2SA1934/2SC5200 OD?
I see (if I'm right) that some variations exist on the number of output devices. In several number of cases, it's stated that there are 3 pairs as on the sch. I originally saw, but some state more. I think (do not know yet) that 3 pairs would suit my needs, and that would basically mean that I need 6pcs of 2SA1934 and 6 2SC5200 for a stereo amp.
The ESL I have are fairly sensitive just have low impedance most of the time so I suppose I don't really need to go over 50W/8Ohm pch in A class (made the ESLs work with my 8W Class A SE tube amp)
That leads me to conclusion that I can go with the basic design and get away with it?
There's a lot and lot of reading to do, I'd just need a confirmation so I can place an order for the transistors because I expect a month or so to get them.
Since I'll run the amp (if I get to that point at all) at mainly 4R and lower, I presume the same rules apply as with other Solid State designs, same or even lower voltage across rails, more Amperes for lower impedance loads. So, I would probably have to keep the voltage at moderate 37VDC, right? Of course, I'm aware that two 500VA toroids and 4 pairs of 25.000/75V will have to be sacrificed for the project but I have these already, just will have to remove a few Volts from the secondary windings (now they are at 2x35VAC) and that shouldn't be a big thing, I already did manual rewinding some toroids from 24 to 270V and this will be a piece of cake compared to that...
I see (if I'm right) that some variations exist on the number of output devices. In several number of cases, it's stated that there are 3 pairs as on the sch. I originally saw, but some state more. I think (do not know yet) that 3 pairs would suit my needs, and that would basically mean that I need 6pcs of 2SA1934 and 6 2SC5200 for a stereo amp.
The ESL I have are fairly sensitive just have low impedance most of the time so I suppose I don't really need to go over 50W/8Ohm pch in A class (made the ESLs work with my 8W Class A SE tube amp)
That leads me to conclusion that I can go with the basic design and get away with it?
There's a lot and lot of reading to do, I'd just need a confirmation so I can place an order for the transistors because I expect a month or so to get them.
Since I'll run the amp (if I get to that point at all) at mainly 4R and lower, I presume the same rules apply as with other Solid State designs, same or even lower voltage across rails, more Amperes for lower impedance loads. So, I would probably have to keep the voltage at moderate 37VDC, right? Of course, I'm aware that two 500VA toroids and 4 pairs of 25.000/75V will have to be sacrificed for the project but I have these already, just will have to remove a few Volts from the secondary windings (now they are at 2x35VAC) and that shouldn't be a big thing, I already did manual rewinding some toroids from 24 to 270V and this will be a piece of cake compared to that...
Hi,
a 3pair 2sa1943/c5200 will drive a 4 ohm 60degree phase angle load to 110W with Tc<=80degC using +-50mF and 500VA (5% regulation) and stay within the DC SOAR.
2ohm 60degree phase angle and Tc<=90degC is just inside the 100mS SOAR.
BTW,
2pair of MJ15003/4 have about 36% more current (SOAR) ability @ 80degC than 3pair 1943/5200.
If you intend (and should) match your output devices then I suggest you buy a few more than you need and keep the spare for another project. Buy 16off each and choose the best 4 sets of 3matches for the Klone.
But, removing turns from your 35Vac transformer will reduce the VA rating. The current output remains fixed @ 7.1Aac
a 3pair 2sa1943/c5200 will drive a 4 ohm 60degree phase angle load to 110W with Tc<=80degC using +-50mF and 500VA (5% regulation) and stay within the DC SOAR.
2ohm 60degree phase angle and Tc<=90degC is just inside the 100mS SOAR.
BTW,
2pair of MJ15003/4 have about 36% more current (SOAR) ability @ 80degC than 3pair 1943/5200.
If you intend (and should) match your output devices then I suggest you buy a few more than you need and keep the spare for another project. Buy 16off each and choose the best 4 sets of 3matches for the Klone.
But, removing turns from your 35Vac transformer will reduce the VA rating. The current output remains fixed @ 7.1Aac
john65b said:
Putting a cap in parallel with a resistor creates a low pass RC filter.
Placed in the NFB network of a power amp, such a filter rolls off the high freq. response at 6dB per octave beyond the RC cornering frequency [f= 1/(Pi*R*C) ]
Means that the NFB is attenuated at high frequencies, resulting in a gain boost to counterbalance the Hfe behaviour of the output devices.
Only trouble is that the resistor and capacitor in the parallel RC filter have a phase difference that can be as high as 90 degrees.
Use sustained beta output devices and the combination of phase shift and an effective gain level which is higher than nominal leads to oscillation, reducing the C-value of the parallel RC filter raises the roll off frequency to a higher level where it does no harm.
The KSA50 only has a DC blocking cap in the NFB network, not a compensation C.
(you probably mean pF instead of uF)
Rewinding a 35Vac toroidal to 27Vac (37Vdc +1.2V rectifier loss) reduces the nominal rating of a 500VA to less than 400VA.
Still means the transformer can deliver more than 500VA during peaks, but throwing away so much continuous power is a bit of a shame imo.
An alternative is to wind some extra primary on top instead of taking secondary wiring off.
Why not rewind it to 30Vac and add some extra heatsinking ?
I'm currently rewinding R-cores, which is a whole different ball game compared to the rewinding of toroidals.
R-cores have multiple secondaries on top of eachother and on both sides of the core, a rewiring job that'll drive you nuts.
My rule of thumb is not to rewind beyond 10% of the nominal rating, otherwise look for different transformers and flog off the ones you can't use.
I was thinking a bit and came to conclusion that it will be simpler to order two new 500VA rated toroids at 2x26VAC - @~50€ per transformer might be a better solution and leave these I already have as they are, after all, have prepared them for quad replicas which run at +/-50VDC...
So, new transformers will be better...
So, new transformers will be better...
You just confirmed it now. The ones available are obviously brand new and labeled with "M" as in "Motorola" and I remember back some fifteen years ago how the original ones looked like. And these are definitely not(!) Onsemi! Maybe they are of excellent quality but I'll let someone else see if they are or not. I'll go for 2SA1943/2SC5200 and it suits me better for the coolers I have because I'll go with passives (w/o fan) or fit the fan only for summer+rock concerts combination.
Here is the image of the coolers I have at 205mm height each, two (as on image) should be sufficient per channel.
Here is the image of the coolers I have at 205mm height each, two (as on image) should be sufficient per channel.
Attachments
jacco vermeulen said:
Wowza for sure. Two of these babies and four Spragues pull down to mother Earth by 10 kilos. Meaning, that four of these, eight Spragues and two 500VA (4kg/each) will generate 28 kgs alone, plus the box/case/housing and misc stuff ... I think that will come very close to the weight/power ratio of a decent tube amplifier...
Forgot to ask about additional voltages for protection circuitry... I see some do use it, some don't and should I use an additional transformer for that (because that seems to be the way to go) and is there something recommended? Nothing too fancy, though, I usually play my music at moderate levels and in the last twenty years or so, burned down only one 5" Tannoy driver with an oscillating Gainclone (my mistake) ...
Nordic said:
Thanks for the effort, I will get one PCB and assistance from Zen Mod who owes me for the diodes for his friend's MIG (not the CCCP made aeroplane, of course
)@ Jozua: Thanks but I will stick to lower (original) voltage because my ESLs are fairly sensitive (90dB) and run from around 100Hz, they only have a bit demanding impedance curve... To support this idea I can say that they work with a decent (AB) 30W amp, but those all have a hard time with low impedance loads (and mainly behave very well at nominal 8 Ohm loads)