Aikido 24V headamp
Nov 28, 2008 at 3:49 AM Post #286 of 570
Quote:

Originally Posted by digger945 /img/forum/go_quote.gif
OK, one more ultra critical question,
Were you thinking of using this 24V aikido to do a 12SX7 amp in the future, or did you have plans for a totally new amp?..or did you just want to try the 6GM8 at 48V?

I just thought of it, the 6GM8 and the 12SX7 are completely different tubes, one in 9pin and one 8, they need completely different pcb's to work.
Just wondering Rod.



The 6GM8 cannot run at 48V, JB has confirmed that with me. My intentions in if any, were to leave it open to 12SX7 in the future. JB himself said he was wondering what the 12SX7 would sound like at 48V?
 
Nov 28, 2008 at 3:56 AM Post #287 of 570
OK, rodent, so this will be just a 24V 6GM8 amp then. So we don't really need to build it with all the 50V caps and such. So are we ordering the higher voltage parts for the future build of the 12SX7 amp? ...and if so I have no idea what values the Octal pcb needs to build that amp.
But I do have some suggestions for the 24V 6GM8 amp.
 
Nov 28, 2008 at 4:01 AM Post #288 of 570
Thank you Renato for the help to understand how this all works. I have learned a great deal today. I am bidding on some 6H30 tubes on ebay, if I lose, and I usually do, I will just order some from tubedepot or one of the other familiar ones.
I have the owners manual for the mono noval pcb's in front of me. I think that I will just order the vishay resistors when I finish up the Dynamite order from Mouser tomorrow or Saturday. I will take your suggestions on parts and such and add them to my list.
Later guys, I am tired and must get to bed now.
Thanks, it's been fun today.
 
Nov 28, 2008 at 4:11 AM Post #289 of 570
Quote:

Originally Posted by looser101 /img/forum/go_quote.gif
The output of the Aikido like all single ended amps has an output that is (for sake of illustration) say B+/2, so 12Volts in the case of the 24V Aikido. In this case the headphone driver would see 12v on one side and ground (0volts) on the other (which would be bad since a current would flow across the driver when no music was playing). So what can be done about this problem? We use a coupling capacitor. In this case it will block DC in the output stage and pass AC (ie. music). The DC voltage at the output of the coupling cap is effectively 0 and since ground is also 0 then no current will flow. Hope that is what you are asking.




I don't know, that is why I am asking! LOL! Right, so as I see it, unless this idea is fataly flawed, if we make one board B+ +24V and one B+ -24V and leave all caps for the tubes in place except remove the buffer lytics which are replaced by resistors tied to the other board... I am just guessing but I know I read it somewhere.
 
Nov 28, 2008 at 4:15 AM Post #290 of 570
Quote:

Originally Posted by digger945 /img/forum/go_quote.gif
OK, rodent, so this will be just a 24V 6GM8 amp then. So we don't really need to build it with all the 50V caps and such. So are we ordering the higher voltage parts for the future build of the 12SX7 amp? ...and if so I have no idea what values the Octal pcb needs to build that amp.
But I do have some suggestions for the 24V 6GM8 amp.



I was just thinking that I would use adaptors and change some resistors and voila! 12SX7 Aikido, ya know, something like that. The Pan FM 1000uF/50V will do fine for the heater string and the lytics for now. If I am happy with the FMs in there I may avoid Black Gates. Does anyone know how the Rubycon stack up to the Black Gates?
 
Nov 28, 2008 at 3:08 PM Post #291 of 570
Do they make such an adapter to put an Octal tube into a Noval socket?
This might be more hassle and could, if it exists, be way more expensive than just building two seperate amps.
I gotta do some googling first.
On C7,8,9,10= I have 100uF 25V in mine. I'm not sure that the 1000uF 50V you have listed will fit next to the tube sockets. JB's BOM says 10-1000uF and rated at B+. I think 100uF would be fine here, it is not directly in the signal path.
On C5,6= I have the same Wima, but 0.1uF.
C3= Fierce kept the .47 Multicap. I now have an Auricap 2.2uF. JB calls for 1-3uF. I think I priced the Multicap .47, IIRC they ain't cheap.
C4= I think the Panasonic FM should do fine here, get the BG to play with, I would buy both if it were me. This is the main coupling cap from buffer to headphone.
C11= Same as C4, only this one is B+, this is the main power supply cap. I have a BG 1000uF 25V in mine now but I have used many different ones. I think I liked the Panasonic HA that I robbed from the Dynahi parts stash best. Needs to be a minimum 1000uF and rated at B+ and no more than about 28mm diameter.
If you really want to build this for a 48V PS and 50V stuff aint available, then 63V will work fine, maybe 80 or 100V if they don't get too big to fit on the board.
C1 is only necessary if you plan to use the line out, it does not affect the headphone out, it is just the coupling cap to the next amp in your system.
 
Nov 28, 2008 at 3:38 PM Post #292 of 570
On the resistor side...
Does your amp have an R13? Mine does not and the revised schematic does not but it is listed in the revised BOM on tubecad.

R1,5,6,7,12,14= 1M. This is what I have.
R2,4= 180R is what I have. JB says 100-240 ohm, and recommends 180.
R3,9,10= I have 330R, JB recommends 300. Try to get 300 if you can.
R8,11= I have 180R, JB recommends 180.
R15= I used to have 18R when I ran the 317's, JB recommends 20. I would order 18, 20, and 24 or 25 to play with.
R16= I had 5R when I got it. JB recommends 2R. I would order 2 & 5 to play with.
R18= 100k.
R19= 75k.
R20/21= Your pcb should be revised to show just one part number. By the schematic this should be 10k, and that's what I had also.
I have read a lot of John's articles that have 20 ohm or 24 ohm for both R15 & 16.
If I were ordering I would get some extra 10,20,10k,and 4 100k's to play with later, should you decide to upgrade the output to mosfet.
I still think you would be more satisfied with two amps, one for the 6GM8 and one for the 12SX7.
 
Nov 28, 2008 at 9:39 PM Post #293 of 570
If I use the FM 1000uF 50V for the coupling caps that I have and the HV LM reg I could be covered in both scenarios. The only difference will be a couple of resistors and the octal adapters which can be made from an octal socket soldered to the base of a broken noval tube and then glued in place. One could even to a vibration mount such as the $45 sockets available to reduce microphonics. I could use a chassis mount socket with leads to the board. Many options for octal conversion. I think what has to be done first is someone needs to buy 4 12SX7GT(native 26.5V-28V) tubes and try them at 24V and see how they sound. If they are better than the 6GM8 at 24V then any voltage up to 48V will likely be better. The stock regs can be used with a B+ of double their rating since they only see half the voltage. If you surpass that voltage then using the HV would be required. The heatsink will have to be bigger though.

We need to see what a quad of 12SX7GT in the set-up at 25.2V is like. I guess I will have to build one board with octals and one with novals. They will the same caps and I will use the FM's 50V for coupling and I can always go 25V BG NX after the comparison. I will try a 48V PS as well. So I need a 48V PS and some way to convert the signal from balanced to SE. I can probably get a couple of Sowter 1:1 transformers to do the SE conversion and I will need suggestion for a simple linear PS for 48V.
If we were to run it at 48V, are there any better options for mosfet output devices?
 
Nov 28, 2008 at 9:57 PM Post #294 of 570
I don't think you gotta worry about the heatsinks if you run the 317's. They run cool even at 28V and biased with 2 ohms(I tried it, they don't even get warm).
Mosfets? There are probably few that wouldn't work, pins installed correctly.
 
Nov 28, 2008 at 10:09 PM Post #295 of 570
OK, so all I need to do to run 12SX7GT is adjust the resistor values from 180 to 220, the 75K to 82K and make sure all of the caps are 50V or higher. The PS is the next thing. The one that John offers is max 25.2V, but what is limiting it? If the device cannot handle the 48V, could I just change it to a HV LM317 and use the same board? I am also looking for a shunt reg PS for my D1 I/V with +/-30V rails if you have any ideas. Twisted Pear is coming out with a shunt reg but it won't do 30V.
 
Nov 28, 2008 at 10:24 PM Post #296 of 570
+/- 30V? I dunno. Do you need a lot of current? I could sell you an extra Dynahi PS pcb, but you may have better success just ordering an amb alpha kit from Glass Jar.
I am not familiar with the D1.
 
Nov 29, 2008 at 4:50 AM Post #298 of 570
What about other output devices than the LM317? Could you use some JFETs like the 2SK170? It is good to about 30v I think so it would be fine with the 24V/2 and even 48/2 for 24V at the JFET. I don't know what else would have to be done?
 
Nov 30, 2008 at 12:27 AM Post #299 of 570
I have a question about tube choices, I have noticed a lot of auctions on ebay with 6N1P tubes that are supposed to be "close" to the same specs as a 6dj8. Does anyone know what the difference would be and if it would be possible to use those tubes with the component values recommended by JB for using 6DJ8/6922 tubes? Would be nice if it was an option because they are cheap and plentiful on ebay.
 

Users who are viewing this thread

Back
Top