Aah, yes, I see that. However, there is a later circuit ... and BoM.
Irrespective, on X's advice, I will buy some more KSA992 and match them.
I have no idea whether it will make a difference or not - I simply build the circuit the experts have decided upon.
Aah, OK - thanks, X.
Can you direct me to a pic which shows how V111 & V112 are (thermally) bonded together?
Andy
btw, I can buy 100 from RS Components here:
https://au.rs-online.com/web/p/digital-transistors/1868558/?relevancy-data=636F3D3126696E3D4931384E53656172636847656E65726963266C753D656E266D6D3D6D61746368616C6C7061727469616C26706D3D5E2E2A2426706F3D31313326736E3D592673723D2673743D43415443485F414C4C5F44454641554C542673633D592677633D4E4F4E45267573743D466169726368696C64204B5341393932465441267374613D466169726368696C64204B534139393246544126&searchHistory=%7B%22enabled%22%3Atrue%7D
... for A$11.20 ... with free delivery!
Andy
Hi Andy,
Look at the transistors on the amp on the left. They are next to the black heat sink. They are flat to flat and covered with (yellow) heat shrink tube. You can squirt some thermal grease in between if you wish. I used yellow heat shrink tube instead of black because I knew someone would ask so it’s easier to point out.
Look at the transistors on the amp on the left. They are next to the black heat sink. They are flat to flat and covered with (yellow) heat shrink tube. You can squirt some thermal grease in between if you wish. I used yellow heat shrink tube instead of black because I knew someone would ask so it’s easier to point out.
Hi Andy,
Look at the transistors on the amp on the left. They are next to the black heat sink. They are flat to flat and covered with (yellow) heat shrink tube. You can squirt some thermal grease in between if you wish. I used yellow heat shrink tube instead of black because I knew someone would ask so it’s easier to point out.
Thank you, X - very cool!
btw, the PCBs that I have (Version 001 12/19/2019) use the pin sequence for KSA992 (ECB!) not 2N5401 (EBC). So I have just ordered 100x KSA992s.
Andy
Last edited:
I ran the test using a 10ohm 300w fan cooled EBG resistor dummy load, I connected an O-scope to show the trace shape to verify that the power delivered was just before clipping (clipping is at about 18.05Vrms on this amp) so I ran it at 17.8Vrms (about 40w at 8ohm but only 32w at 10ohms) with a 1kHz sine wave. Also had a Fluke 115 True RMS meter hooked up to verify the voltage setting. Before starting the test, I let the amp reach steady state temp by listening to soft music for about 45 min. I checked the temp of the middle of the heatsink with an IR thermometer. I got the usual 44.7C. Ambient temp is 20.1C. After 5 minutes, the temp was 41.6C. I ran it another 5 minutes for 10 minutes total and it was 40.1C. The whole time the Oscope trace did not flinch, the amp did not smoke or smell funny, or acted stressed in any way. I think I could run it here indefinitely. In case you have not followed my build, the PSU consists of a 600VA Antek AS-6224 24v transformer. This feeds an SLB PSU with an LT4320 active bridge and CRC followed by a cap multiplier that provides +/-28.5V rails. The amp is running at about 1.7A bias current. That is 200w of total dissipation in the MOSFETs alone and another 10w in the cap multiplier, close to about 1/3rd capacity of the transformer's 600VA rating - not much stress to the system because this is where it runs all the time. Music playing loud or not.
Here is the setup:
Here is a closeup of the Oscope showing the 53.6Vpp signal going into 10ohms:
Here is the IR thermometer after 5 min:
Here is the IR thermometer of the amp at idle and warmed up:
I suppose you were trying to show that if this was a push-pull amp going out of Class A and into Class AB, the temperature would not have decreased at max power?
This is consistent with what 6L6 observed with the Aleph J, another amp using the Pass active CCS.
Attachments
Last edited:
Nice measurements X !
Here are simulations for the damping factor,
I've compared the Alpha20 with the Nirvana, both with 8R load.
Top graph is Nirvana.
The damping factor for the Nirvana is 10x higher in simulation.
Probably the Nirvana DF simulation is not correct, Hugh calculated it at as 160.
Still there's no need for higher damping with the Nirvana.
Here are simulations for the damping factor,
I've compared the Alpha20 with the Nirvana, both with 8R load.
Top graph is Nirvana.
The damping factor for the Nirvana is 10x higher in simulation.
Probably the Nirvana DF simulation is not correct, Hugh calculated it at as 160.
Still there's no need for higher damping with the Nirvana.
Attachments
Last edited:
Use your famous 10 ohm dummy load and easily calculate damping factor by measuring the difference between the unloaded and loaded output voltage
DF = V no load / ( V no load - V loaded ).
You can repeat at different frequencies as DF may vary.
I ran the test using a 10ohm 300w fan cooled EBG resistor dummy load, I connected an O-scope to show the trace shape to verify that the power delivered was just before clipping (clipping is at about 18.05Vrms on this amp) so I ran it at 17.8Vrms (about 40w at 8ohm but only 32w at 10ohms) with a 1kHz sine wave. Also had a Fluke 115 True RMS meter hooked up to verify the voltage setting. Before starting the test, I let the amp reach steady state temp by listening to soft music for about 45 min. I checked the temp of the middle of the heatsink with an IR thermometer. I got the usual 44.7C. Ambient temp is 20.1C. After 5 minutes, the temp was 41.6C. I ran it another 5 minutes for 10 minutes total and it was 40.1C. The whole time the Oscope trace did not flinch, the amp did not smoke or smell funny, or acted stressed in any way. I think I could run it here indefinitely. In case you have not followed my build, the PSU consists of a 600VA Antek AS-6224 24v transformer. This feeds an SLB PSU with an LT4320 active bridge and CRC followed by a cap multiplier that provides +/-28.5V rails. The amp is running at about 1.7A bias current. That is 200w of total dissipation in the MOSFETs alone and another 10w in the cap multiplier, close to about 1/3rd capacity of the transformer's 600VA rating - not much stress to the system because this is where it runs all the time. Music playing loud or not.
Here is the setup:
Here is a closeup of the Oscope showing the 53.6Vpp signal going into 10ohms:
Here is the IR thermometer after 5 min:
Here is the IR thermometer of the amp at idle and warmed up:
I suppose you were trying to show that if this was a push-pull amp going out of Class A and into Class AB, the temperature would not have decreased at max power?
This is consistent with what 6L6 observed with the Aleph J, another amp using the Pass active CCS.
>96 watts for an output of 40 watts into 8 ohms is around 42% class A is therefore verified... single ended even...any pp amp will have efficiency much greater than 50%
> ambient of 20 *C and sink temp of 40 *C means a temp rise of 20*C so that in my country where ambient can reach 30*C sink temp can rise 50*C, this is better than what we are getting with the NP ACA where sink temps go as high as 70*C...but the good news is that when you train a fan to the heatsinks, the sinks get colder....
With my simulator I get output impedance of 0.03 ohm DF 250 for 8 ohm . Nearly double of it at 30hz.
Attachments
Last edited:
