Hi Andrew,
Original schematic has two pairs of 2sa968/2sc2238. Clone uses mpsa42/92 for the first pair and mje15032/33 for the second (predrivers). After that is another pair of mje15032/33 (original mje 15030/31) for drivers. MPSA 42/92 have different pinout, but mje 15032/33 have the same as 2sa968/2sc2238. So, can I just put 2sa/sc instead of mje's in the predriver section?
Original schematic has two pairs of 2sa968/2sc2238. Clone uses mpsa42/92 for the first pair and mje15032/33 for the second (predrivers). After that is another pair of mje15032/33 (original mje 15030/31) for drivers. MPSA 42/92 have different pinout, but mje 15032/33 have the same as 2sa968/2sc2238. So, can I just put 2sa/sc instead of mje's in the predriver section?
Hi,
the VAS can be 2sa970/c2238. Use the "y" version rather than the "o" if you can find a supplier.
I think a low capacitance device here would perform better. What do others think about using 2sa1360/c3423 or 2sa1540/c3955 or similar for the VAS devices?
Note, the To126 devices must be reversed due to the different pin out from the To220 devices.
practical measurements
I've read modest amount of this thread but could not spot consolidated measurements that have posted for a built design. Could someone point me to the right post having measurements/graphs like THD, IMD (SMPTE), Signal to noise ratio, Damping factor, Frequency response etc.
I'm wondering how this performs compare to Leach, DPA386SE, modified AEM6005 ... etc.
cheers,
-Navin
I've read modest amount of this thread but could not spot consolidated measurements that have posted for a built design. Could someone point me to the right post having measurements/graphs like THD, IMD (SMPTE), Signal to noise ratio, Damping factor, Frequency response etc.
I'm wondering how this performs compare to Leach, DPA386SE, modified AEM6005 ... etc.
cheers,
-Navin
Re: practical measurements
We are relying on word of mouth (recommendation) about how good/bad this amp is.
you haven't a hope.nsaidor said:I've read modest amount of this thread but could not spot consolidated measurements that have posted for a built design. Could someone point me to the right post having measurements/graphs like THD, IMD (SMPTE), Signal to noise ratio, Damping factor, Frequency response etc.
I'm wondering how this performs compare to Leach, DPA386SE, modified AEM6005 ... etc.
cheers,
-Navin
We are relying on word of mouth (recommendation) about how good/bad this amp is.
Hello all,
I am quite familiar with this thread, but new to post. I got the bug up my *** to build a Krell clone and purchased a set of Pinkmouse boards and BrianGT Power supply board (don't think I will use the PS board since I have all the separate components)...I should have read a little deeper into the WIKI before purchasing the enclosure (wasn't cheap with shipping to the states)
Anyway, I wanted to put the amp in a nice chassis and purchased a 3U Pesante Dissapante chassis from hifi-2000.
Anyway, I just received the attached data (with pic of nice looking chassis) regarding the heatsinks and their documented cooling rating and think i may need a beefer heatsink.
Before I change direction and buy a new chassis/heatsink, I thought I would post here for a final opinion...The data sheet does say the ability of dissipating 280 watts both heatsinks with six T03-P on each heatsink - Each heatsink is .41C/W -
Planning on dissipating 140W per channel to be within the 280W total.
When I work backward from the 25C ambient, the .41C/W cooling rating, and the planned 140W dissipated, I get 57C change, which is way hotter than I have been reading here - am I calculating this correctly?
Why would they post a possibility of 280W if the six TO3-P transistors had to run at 82C? Is this an error on their assumption of a transistor reliably running this hot?
BTW - I am leaning toward the 2SA1943/2SC5200 as transistors, single 25-0-25 500VA - 700VA Tranny and (hopefully) driving 4 ohm Magnapan MGIIIa - maybe 50W class A not enough power, but maybe planning a different set of speakers if the maggies don't work out)
Sorry for the tyro-ish question, but just want an opinion before I render this beautiful chassis to a 6 channel Gainclone amp.
I am quite familiar with this thread, but new to post. I got the bug up my *** to build a Krell clone and purchased a set of Pinkmouse boards and BrianGT Power supply board (don't think I will use the PS board since I have all the separate components)...I should have read a little deeper into the WIKI before purchasing the enclosure (wasn't cheap with shipping to the states)
Anyway, I wanted to put the amp in a nice chassis and purchased a 3U Pesante Dissapante chassis from hifi-2000.
Anyway, I just received the attached data (with pic of nice looking chassis) regarding the heatsinks and their documented cooling rating and think i may need a beefer heatsink.
Before I change direction and buy a new chassis/heatsink, I thought I would post here for a final opinion...The data sheet does say the ability of dissipating 280 watts both heatsinks with six T03-P on each heatsink - Each heatsink is .41C/W -
Planning on dissipating 140W per channel to be within the 280W total.
When I work backward from the 25C ambient, the .41C/W cooling rating, and the planned 140W dissipated, I get 57C change, which is way hotter than I have been reading here - am I calculating this correctly?
Why would they post a possibility of 280W if the six TO3-P transistors had to run at 82C? Is this an error on their assumption of a transistor reliably running this hot?
BTW - I am leaning toward the 2SA1943/2SC5200 as transistors, single 25-0-25 500VA - 700VA Tranny and (hopefully) driving 4 ohm Magnapan MGIIIa - maybe 50W class A not enough power, but maybe planning a different set of speakers if the maggies don't work out)
Sorry for the tyro-ish question, but just want an opinion before I render this beautiful chassis to a 6 channel Gainclone amp.
Attachments
Hi Pink,
I was not being negative.
I was telling him he hadn't a hope of finding
I have the 50 Klone PCBs and if I can't get the four stage 100 Klone PCBs to perform I will revert to using the three stage 50 PCBs to drive the 6pair output stage. There is nothing inherent in the 50 Klone PCBs that stops it from using the same PSU and output stage of the 100 Klone.
I think both these amps are good. It's just I don't have the skills necessary to debug the 100K yet.
I was not being negative.
I was telling him he hadn't a hope of finding
.
I have the 50 Klone PCBs and if I can't get the four stage 100 Klone PCBs to perform I will revert to using the three stage 50 PCBs to drive the 6pair output stage. There is nothing inherent in the 50 Klone PCBs that stops it from using the same PSU and output stage of the 100 Klone.
I think both these amps are good. It's just I don't have the skills necessary to debug the 100K yet.
You don't need much to do some of the mentioned measurements, the DF for instance can easily be determined.
S/N, THD, DF and freq/-3db are equal or better than the published ones of the original, parts+powersupply are the main reasons.
Why someone desires them is beyond me, the original KSA50 has DF>100, Freq>100KHz, S/N>90dBV, a SR higher than 25V/uS, a risetime within 2uS and a goodlooking decreasing harmonic spectrum, but what gives ?
Actually, if you go back about 16months you'll find some spectra and measurements I took with a 50 clone, but when nobody showed interest I didn't bother with more testing. I'll probably take some more in the near future with a) the original clones b) Pinkmouse's clones with MJL3281/1302 outputs and c) commercial KSA50. I have extensive measurements of the KSA100, both Mk1 and Mk2, the latter before and after transistor upgrades. The 100 Mk1 is a very similar design to the KSA50 (actually identical with a few small changes for the higher power) so the same applies. IIRC a quick summary of all the KSA's are so-so THD and noise, but these figures remaining quite consistent with load power, i.e. the distortion doesn't increase as much as the load power increases as with most other amps, which also explains the typical "Krell bass grip".