Hi msamkl,
Interesting question, I would consider mine to be in a different class (or I probably would not have bothered making it). Quite a few have been delivered in the last 3 weeks so we should have some independent subjective assessments soon.
SKA is capable of 150W though. I would happily audition it against a Leach or Krell, for some surprises. Better put it in a box.
Thanks Jorge for -
..."SKA - a new proposition, unusual circuit (the only really new in decades)."
Indeed it is.
Hi mikelm,
the SKA has a closed loop bandwidth of about 7MHz tailored back to 1MHz with a 5pF FB cap. ( the only comp) providing +45deg at 1MHz. It doesn't need an output R//L. A passive input filter works at 160KHz.
How's it sound - very open and transparent with or without the AC coupling FB C. But I have only soft dome tweeters and it's fast but not 'tizzy'. No hum or noise. I haven't heard the others...
Full complementary and high bandwidth Jess. And super PSRR so no need for huge PS.
Hi Lumanauw,
I would expect your 2 cascaded diff'l stages to improve PSRR by maybe 20-30 dB, but you must now tailor the gains of 2 stages symmetrically as well as your Vas for stability. HF unbalance due to Ccomp could lose it.
Cheers,
Greg.
Interesting question, I would consider mine to be in a different class (or I probably would not have bothered making it). Quite a few have been delivered in the last 3 weeks so we should have some independent subjective assessments soon.
SKA is capable of 150W though. I would happily audition it against a Leach or Krell, for some surprises. Better put it in a box.
Thanks Jorge for -
..."SKA - a new proposition, unusual circuit (the only really new in decades)."
Indeed it is.
Hi mikelm,
the SKA has a closed loop bandwidth of about 7MHz tailored back to 1MHz with a 5pF FB cap. ( the only comp) providing +45deg at 1MHz. It doesn't need an output R//L. A passive input filter works at 160KHz.
How's it sound - very open and transparent with or without the AC coupling FB C. But I have only soft dome tweeters and it's fast but not 'tizzy'. No hum or noise. I haven't heard the others...
Full complementary and high bandwidth Jess. And super PSRR so no need for huge PS.
Hi Lumanauw,
I would expect your 2 cascaded diff'l stages to improve PSRR by maybe 20-30 dB, but you must now tailor the gains of 2 stages symmetrically as well as your Vas for stability. HF unbalance due to Ccomp could lose it.
Cheers,
Greg.
Sorry if this has been covered, how many watts of thermal output would such an amp have worst case scenario?
also, what are the dimensions of the MOSFET's?
and around what temperature would one like them to run?
also, what are the dimensions of the MOSFET's?
and around what temperature would one like them to run?
Hi Zenon,
You have asked a very difficult question, given that the amplifier is designed around music as a signal, not sine waves or square waves.
Around 1/3 power is meant to generate the most Pd in a Class B power amplifier. But this is impractical resulting in gross overbuilding of output stages and heatsinks for an unrealistic scenerio, pushing up price out of the reach of some, for no reason.
The IHF came up with a better test signal more reflective of the typical peak to mean of music. A pulsed toneburst offerring 20 dB peak to mean.
So for a 150W amplifier running at an average power of 15W, my rough calculation for this amp puts Pd around 50W for a 37C rise on a recommended 0.75C/W heatsink. Junction temps will be some 13C higher due to junction to case 1C/W (150W IRFP's) so the junctions run up to 75C from 25C ambient.
However much music has a higher peak to mean than this and/or would not be run so hard with breaks, so typical temp rise could be halved.
That help?
Cheers,
Greg
You have asked a very difficult question, given that the amplifier is designed around music as a signal, not sine waves or square waves.
Around 1/3 power is meant to generate the most Pd in a Class B power amplifier. But this is impractical resulting in gross overbuilding of output stages and heatsinks for an unrealistic scenerio, pushing up price out of the reach of some, for no reason.
The IHF came up with a better test signal more reflective of the typical peak to mean of music. A pulsed toneburst offerring 20 dB peak to mean.
So for a 150W amplifier running at an average power of 15W, my rough calculation for this amp puts Pd around 50W for a 37C rise on a recommended 0.75C/W heatsink. Junction temps will be some 13C higher due to junction to case 1C/W (150W IRFP's) so the junctions run up to 75C from 25C ambient.
However much music has a higher peak to mean than this and/or would not be run so hard with breaks, so typical temp rise could be halved.
That help?
Cheers,
Greg
Sorry my question was so complex!
I'm a mechanical vs. electrical so all this amp talk makes my head spin, but I'm looking into making either your amp or a gainclone of some incarnation within the next 12mths (after I make a nebula witin the next 6 mths to play it on) and I will liquid cool it. So I was just wanting a typical power output and operating temperature.
so 50W and 75C target temperature? pfft fegettaboutit!
Looks like this so far, in terms of delta-T from ambient:
1-channel: 3.2C
2-channel: 3.9C
5-channel: 16.8C
can you measure the MOSFET's (that's what we're cooling correct?) dimensions please? Everything will be super cheap here, this would only use a tiny $5 fountain pump used in those desktop fountains.
What kind of pricepoint would I have for a typical heat exchanger?
I'm a mechanical vs. electrical so all this amp talk makes my head spin, but I'm looking into making either your amp or a gainclone of some incarnation within the next 12mths (after I make a nebula witin the next 6 mths to play it on) and I will liquid cool it. So I was just wanting a typical power output and operating temperature.
so 50W and 75C target temperature? pfft fegettaboutit!
Looks like this so far, in terms of delta-T from ambient:
1-channel: 3.2C
2-channel: 3.9C
5-channel: 16.8C
can you measure the MOSFET's (that's what we're cooling correct?) dimensions please? Everything will be super cheap here, this would only use a tiny $5 fountain pump used in those desktop fountains.
What kind of pricepoint would I have for a typical heat exchanger?
Ah Zenon,
Now I understand what you're asking.
With the minimum recommended heatsink, maximum supply voltages for 150W power on music I expect roughly 25C rise above ambient, for a minimum size monoblock.
However, if multi channel is desired money can be saved by using larger shared heatsinks and liquid cooling as you state. Expect to need to sink 30W per amplifier. The SKA is still a Class AB amplifier (not Class D) for sound quality but is effectively Rail2Rail so more efficient than others. And Full Complementary which characteristically has less (objectionable) high order harmonics.
The MOSFETS pkg is 20 x 16 mm each x 4 but mounted on the module they are spaced to need a flat contact area of 100 x 20 mm.
Heatsinks actually work more efficiently the larger the dT so 25C is a fair compromise.
Cheers,
Greg
Now I understand what you're asking.
With the minimum recommended heatsink, maximum supply voltages for 150W power on music I expect roughly 25C rise above ambient, for a minimum size monoblock.
However, if multi channel is desired money can be saved by using larger shared heatsinks and liquid cooling as you state. Expect to need to sink 30W per amplifier. The SKA is still a Class AB amplifier (not Class D) for sound quality but is effectively Rail2Rail so more efficient than others. And Full Complementary which characteristically has less (objectionable) high order harmonics.
The MOSFETS pkg is 20 x 16 mm each x 4 but mounted on the module they are spaced to need a flat contact area of 100 x 20 mm.
Heatsinks actually work more efficiently the larger the dT so 25C is a fair compromise.
Cheers,
Greg
So assuming a two channel amplifier how much would a typical heat sink cost?
Thanks, I can hardly wait go get all this stuff built!
Thanks, I can hardly wait go get all this stuff built!
Hi Zenon,
Standard CAST heatsinks are made here in Australia by Conrad -
www.conradheatsinks.com
and for a 2 channel 150W amplifier with convection cooling I would use an MF25-151.5 ( 0.4C/W corrected) or for monoblocks 2 pieces of MF20-75 (0.75C/W corrected). The smaller one actually works out as slightly better value if you ignore the shared reserve sinking ability of the larger. Around US$27.
Cheers,
Greg
Standard CAST heatsinks are made here in Australia by Conrad -
www.conradheatsinks.com
and for a 2 channel 150W amplifier with convection cooling I would use an MF25-151.5 ( 0.4C/W corrected) or for monoblocks 2 pieces of MF20-75 (0.75C/W corrected). The smaller one actually works out as slightly better value if you ignore the shared reserve sinking ability of the larger. Around US$27.
Cheers,
Greg
Hello, amplifierguru
🙂
A question.
Is there any chance you will publish more details,
like components and values, on SKA Mk1
when a new version is released?
Nelson Pass is using this method with his First Watt series:
Each design is a different exploration into high quality at low power.
At the end of the sales cycle of new product, the schematics and technical details of each design
will be posted for the benefit of the "DIY" community.
🙂
A question.
Is there any chance you will publish more details,
like components and values, on SKA Mk1
when a new version is released?
Nelson Pass is using this method with his First Watt series:
Each design is a different exploration into high quality at low power.
At the end of the sales cycle of new product, the schematics and technical details of each design
will be posted for the benefit of the "DIY" community.
Hi Greg,
Just a quick post to let you know I sent you an email this morning with an order for two amplifier and two power supply PCBs. I'm looking forward to giving your design a try as I have done with many other new topologies presented here on DIYaudio.com. I'll be happy to post some performance figures, THD results, etc. in this thread if anyone is interested. I have the needed equipment to do such measurements so just let me know. Due to the holiday season I don't expect to have these amplifiers up and running before the new years though.
Best regards,
Sander Sassen
http://www.hardwareanalysis.com
Just a quick post to let you know I sent you an email this morning with an order for two amplifier and two power supply PCBs. I'm looking forward to giving your design a try as I have done with many other new topologies presented here on DIYaudio.com. I'll be happy to post some performance figures, THD results, etc. in this thread if anyone is interested. I have the needed equipment to do such measurements so just let me know. Due to the holiday season I don't expect to have these amplifiers up and running before the new years though.
Best regards,
Sander Sassen
http://www.hardwareanalysis.com
Hi Sander,
Yes I check my emails 3 times a day. You are sourcing your components, that's fine. The only 'special' parts required are matched BC546C's and 556C's and 10mVgs matched IRFP240's and IRFP9240's. If your supply rails are less than +/-50V you could use IRFP9140's instead of the 9240's as they are a closer complement to the 240's with higher transconductance.
Cheers,
Greg
Yes I check my emails 3 times a day. You are sourcing your components, that's fine. The only 'special' parts required are matched BC546C's and 556C's and 10mVgs matched IRFP240's and IRFP9240's. If your supply rails are less than +/-50V you could use IRFP9140's instead of the 9240's as they are a closer complement to the 240's with higher transconductance.
Cheers,
Greg
Excuse my ignorance, Greg. Do the connected base and emitter pads on D5 and D6 connect to the + or - holes at Q8 and Q6?
Thanks,
Nathan
Thanks,
Nathan
Hi Nathan,
I think if you have a look at the 'stuffing' guide picture on the instructions, I have marked + and - beside the symbol of D5 and D6. They should connect to the corresponding + and - pads (or holes) in front of Q8 and Q6 respectively. Use the colour coded wires supplied to ensure correct polarity.
If your copy is corrupt, I'll send another by email.
Good luck.
Cheers,
greg
I think if you have a look at the 'stuffing' guide picture on the instructions, I have marked + and - beside the symbol of D5 and D6. They should connect to the corresponding + and - pads (or holes) in front of Q8 and Q6 respectively. Use the colour coded wires supplied to ensure correct polarity.
If your copy is corrupt, I'll send another by email.
Good luck.
Cheers,
greg
amplifierguru
can you post complete schematic of this amp
just want to check out, you havent made some less good things
if so, we will help you improve
and this is good for you
😎
can you post complete schematic of this amp
just want to check out, you havent made some less good things
if so, we will help you improve
and this is good for you
😎
Hi lineup,
Thanks for the offer but I'll take my chances at present and keep the detail proprietary. Maybe after the first 100 pairs I'll reconsider. I trust you'll understand.
I still have to work out where I'm taking this.
Cheers,
Greg
Thanks for the offer but I'll take my chances at present and keep the detail proprietary. Maybe after the first 100 pairs I'll reconsider. I trust you'll understand.
I still have to work out where I'm taking this.
Cheers,
Greg
power supply
What would be a suitable transformer? Can I use a single torridial with dual secondaries? What va? I thought 300-400va. And a voltage output /secondary expressed in ac and rectified to what voltage vdc/channel ?Most likely plan to use plitron.
thank you
doggy
🙂
What would be a suitable transformer? Can I use a single torridial with dual secondaries? What va? I thought 300-400va. And a voltage output /secondary expressed in ac and rectified to what voltage vdc/channel ?Most likely plan to use plitron.
thank you
doggy
🙂
- Status
- Not open for further replies.