Member
Joined 2009
Paid Member
I never had trouble saying AKSA, it's rather close to 'Laksa' which happens to be a very delicious curry from Malaysia 
my 2c: don't forget the power supply, I keep hearing from those more experienced than I that this is a crucial element to any 'amplifier' design and impacts those subjective descriptions of sound quality.
p.s. I never heard a Class D - they don't seem very DIY friendly - no fun there.
my 2c: don't forget the power supply, I keep hearing from those more experienced than I that this is a crucial element to any 'amplifier' design and impacts those subjective descriptions of sound quality.
p.s. I never heard a Class D - they don't seem very DIY friendly - no fun there.
Last edited:
When I christened the amp I was mindful of the catchiness of Kodak, a GREAT name. AKSA seemed like something that held attention, even though the acronym was quite prosaic (Aspen Kit Set Amplifier) but I found to my surprise that AKSA was the largest textile company in Turkey, and of course a deeply significant landmark in old Jerusalem, the Al Aqsa mosque.
Hugh
Hugh
Well thanks to Greg and Hughs generosity, I now am set to build one of these (once I procure a few more parts and have some time!)
Quick question about the VAS transistors, I note discussions earlier in the thread about the BD139's being inadequate. I have a four BF469's matched for Hfe which I had left over from an upgrade I did to my Playmaster series 200.
I've run up the schematic in LTSpice and it certainly seems to work ok with these transistors. Anyone have any reasons I shouldn't go with these for the VAS? Is there even any advantage of matching the VAS transistors?
Gaetan8888 recommended these
Tony.
Quick question about the VAS transistors, I note discussions earlier in the thread about the BD139's being inadequate. I have a four BF469's matched for Hfe which I had left over from an upgrade I did to my Playmaster series 200.
I've run up the schematic in LTSpice and it certainly seems to work ok with these transistors. Anyone have any reasons I shouldn't go with these for the VAS? Is there even any advantage of matching the VAS transistors?
Gaetan8888 recommended these
But since I already had the matched BF469's thought I'd ask
Tony.
Tony,
Yes, BF469, one of the early Philips video transistors with a 250V rating, will do just fine.
If you want to get really picky about it, I would suggest the Toshiba 2SC3423, but it's tricky to find. Bdent.com in the US is a good contact for this one. They have a 120V rating, 200MHz speed, and 50mA collector current rating with low Cob of 2pF. Recommended.
Hugh
Yes, BF469, one of the early Philips video transistors with a 250V rating, will do just fine.
If you want to get really picky about it, I would suggest the Toshiba 2SC3423, but it's tricky to find. Bdent.com in the US is a good contact for this one. They have a 120V rating, 200MHz speed, and 50mA collector current rating with low Cob of 2pF. Recommended.
Hugh
Well thanks to Greg and Hughs generosity, I now am set to build one of these (once I procure a few more parts and have some time!)
Quick question about the VAS transistors, I note discussions earlier in the thread about the BD139's being inadequate. I have a four BF469's matched for Hfe which I had left over from an upgrade I did to my Playmaster series 200.
I've run up the schematic in LTSpice and it certainly seems to work ok with these transistors. Anyone have any reasons I shouldn't go with these for the VAS? Is there even any advantage of matching the VAS transistors?
Gaetan8888 recommended these But since I already had the matched BF469's thought I'd ask
Tony.
Hello Tony
My list was just a suggestion, there is other transistors who can be use, most video transistors can do the job.
BF469 are a video transistor and do have a Ft between 60 and 125 mhz and a very low COB of 1.8 pF max.
For a low power amp I did even use a 2N5551 for the VAS, it was the only one I have at the time.
Now I use KSC3503 because I can have them at Mouser.
Bye
Gaetan
Last edited:
Thanks Hugh, Greg, Rabbitz and Gaetan, I'll probably run with the BF469's for now.
Hugh there was another surprise as well. The driver pair you snuck in is a matched pair (and I mean exactly) was that luck or planned? also I can get two pairs fairly closely matched from the outputs, at hfe of 80 each for two of the 1943's and 79 and 82 for two of the 5200's... again I don't really know how much of a difference this makes but figure it can't hurt
I think I originally got the 469's from WES Rabbitz, good to know that they have the more exotic stuff too, I hadn't even thought to check their catalog, and from memory they were quite a bit cheaper than farnell or RS.
Greg I wouldn't hold your breath, but I would love to come over once I have finished the amp. Would be nice to see (and hear) another diyer's work
Gaetan I had looked at the BF469 datasheet but wasn't sure if the capactitance figure listed (as Cre) was the same as the COB parameter you had mentioned earlier, now I know it is
Cheers,
Tony.
Hugh there was another surprise as well. The driver pair you snuck in is a matched pair (and I mean exactly) was that luck or planned?
also I can get two pairs fairly closely matched from the outputs, at hfe of 80 each for two of the 1943's and 79 and 82 for two of the 5200's... again I don't really know how much of a difference this makes but figure it can't hurt I think I originally got the 469's from WES Rabbitz, good to know that they have the more exotic stuff too, I hadn't even thought to check their catalog, and from memory they were quite a bit cheaper than farnell or RS.
Greg I wouldn't hold your breath, but I would love to come over once I have finished the amp. Would be nice to see (and hear) another diyer's work
Gaetan I had looked at the BF469 datasheet but wasn't sure if the capactitance figure listed (as Cre) was the same as the COB parameter you had mentioned earlier, now I know it is
Cheers,
Tony.
Member
Joined 2009
Paid Member
Oh please !! You'all make my head spin with this claptrap about low Cob, etc...
Yes, if you are designing VIDEO amplifiers with bandwidth compensation transistor capacitance effects are VERY big issues. You will NEVER hear the difference at audio frequencies - if you do, then your design needs to be redone
Yes, if you are designing VIDEO amplifiers with bandwidth compensation transistor capacitance effects are VERY big issues. You will NEVER hear the difference at audio frequencies - if you do, then your design needs to be redone
Hi balaboo,
Thanks for your input. It is a timely reminder that there are 2 sides to every story (maybe 3 in audio)
In this relative simple design I have changed the VAS transistor and found it makes a difference. So in this case, adding a 20c transistor improved the sound, but I can not say it s because of the Cob or not. This is a lot easier than redoing the circuit which is not in the scope of this project.regards
Hi Greg,
I wish to comment on what you say above. I would argue partly based on my own experiences that there are many variables which can change the sound.
What you've implied above is that there exists a unique part, in this case a transistor, which gives you the 'best sound'. Furthermore if I understand correctly you seek to select the best part for each component which gives you the best sound.
Let's consider the variables. Do you agree that the room, the listener, the speakers, the room temperature, just to name a few in no specific order are variables? If so, suppose your amplifier contained 10 parts which could be specially selected to give you the best sound, how many possible combinations ought to be considered before we form any conclusions about the sound quality?
Lastly may I ask how many cases you have actually considered before coming to the conlusion stated above?
Hi Professor smith,
All I can say is part A was better than part B in my opinion under my test conditions. I can't say part A is unique or the best part.
No, the room, the listener, the speakers, the room temperature are fixed in my tests. Lots. The main objective is please yourself. Do as much testing and comparisons as you like. Come to your own conclusions. You can see in this thread already there are conflicting opinions on the VAS transistor. We have use 2SC1819, 2SC3423, 2SC3503, KSC3503, 2SC3953, 2SC2705, 2SC2910, BF469 or BD139 and redo the design to make it tolerant to any transistor.
12.
All I can say is part A was better than part B in my opinion under my test conditions. I can't say part A is unique or the best part.
No, the room, the listener, the speakers, the room temperature are fixed in my tests. Lots. The main objective is please yourself. Do as much testing and comparisons as you like. Come to your own conclusions. You can see in this thread already there are conflicting opinions on the VAS transistor. We have use 2SC1819, 2SC3423, 2SC3503, KSC3503, 2SC3953, 2SC2705, 2SC2910, BF469 or BD139 and redo the design to make it tolerant to any transistor.
12.
Well I think I can add some input at this point (especially since I raised the question about the VAS transistors).
When I did a simple AC analysis in spice it showed everything as being fine, howerver when I ran a transient analysis and an FFT things were VERY different. I can see from having tried three different transistors in this position (admittedly only simulations) that they do indeed make a difference. In the case of the BF469's a very negative one (this could be a dodgy model I don't know). The distortion increased by an order of magnitude with the BF469's... I tried adjusting the pot but it didn't help. Unfortunately I don't know enough to try to postulate what it is about these transistors that makes them perform so badly in this application (in the simulation). Real world results may be perfectly fine.
THD is regarded by a lot of people as a meaningless number, but the other thing that was apparent was that the distortion spectra changed as well, (ie the relative heights of the individual harmonics) with the different transistors. Strangely the BF169's apart from the overall higher distortion, seemed to perform better than the others in that the higher order harmonics were relatively lower.
Dr Smith, different transistors have different characteristics, do you not believe that using a device with different characteristics can change the sound? Yes the room is going to have a massive effect (as are the speakers) but in this case I would propose that any improvement in odd order higher harmonics due to changing a part could well be subjectively better.
edit: this bit was me remembering what balaboo had said but thinking it was Dr Smith. Or are you saying that the circuit should be designed to cope with a very wide range of possible devices with little change? That might be ok for mass produced products, but surely if one is trying to ring every last drop of performance out of a design, then the opposite must be done?
Tony.
When I did a simple AC analysis in spice it showed everything as being fine, howerver when I ran a transient analysis and an FFT things were VERY different. I can see from having tried three different transistors in this position (admittedly only simulations) that they do indeed make a difference. In the case of the BF469's a very negative one (this could be a dodgy model I don't know). The distortion increased by an order of magnitude with the BF469's... I tried adjusting the pot but it didn't help. Unfortunately I don't know enough to try to postulate what it is about these transistors that makes them perform so badly in this application (in the simulation). Real world results may be perfectly fine.
THD is regarded by a lot of people as a meaningless number, but the other thing that was apparent was that the distortion spectra changed as well, (ie the relative heights of the individual harmonics) with the different transistors. Strangely the BF169's apart from the overall higher distortion, seemed to perform better than the others in that the higher order harmonics were relatively lower.
Dr Smith, different transistors have different characteristics, do you not believe that using a device with different characteristics can change the sound? Yes the room is going to have a massive effect (as are the speakers) but in this case I would propose that any improvement in odd order higher harmonics due to changing a part could well be subjectively better.
edit: this bit was me remembering what balaboo had said but thinking it was Dr Smith. Or are you saying that the circuit should be designed to cope with a very wide range of possible devices with little change? That might be ok for mass produced products, but surely if one is trying to ring every last drop of performance out of a design, then the opposite must be done?
Tony.
Last edited: