|
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
205.238.183.48
Has anyone here ever built a transformer-coupled, cathode follower power output stage? Simple designs are limited to low power levels, but I'm considering this for a headphone amplifier, so that's OK. I'm just wondering how well they perform in practice. Very little seems to have been written on this subject, other than how to drive it. Any real-world experiences here?
--------------------------
Buy Chinese. Bury freedom.
Follow Ups:
My CF build uses LC coupling with a nice grid choke employed on each output. I have it on the bench...just need to complete. I got some promising results from the prototype. I loaded both sections of a 6SN7 for the voltage amplifier section with CCS to accommodate the swing of the output valves.
Is there enough HT for the driver to swing the EL34 grid to 0V ?
Al
I have done one with a 12b4 headphone amp. It was a matter of lacking enough L, bottom end so I swapped it over and it did work well for a few years for me. I have done a 6v6 cf se speaker amp too ideal if you can get the dcr to set the bias right. Both application you need about the same load I have read of others experiencing similar results. cascade for a driver is a good application . What about stc?
I am building a dht headphone amp next. Its one place I don't see the need to have to use fb trading bass impact verses midrange openness I guess.
Al,
Not there yet. Built the prototype totally by ear. This is my rabbit hole. I'll keep you posted...I could have ignored all of the Tubecad articles and not had this much fun...as I refined the voltage section and opened up the output section with options like a good high impedance grid choke, I heard the differences and reacted to them...I will make a good assessment and report back to the forum regardless of numbers...
Stuben
The change that kept me going was the grid chokes for the voltage amp sections. This made the drive easier....I'll take measurements after I fire up the latest version. Been fun...If I'm not satisfied, I'll probably convert the amp to a 2a3 build. I just like the thought of getting a Pentode to perform well as a Pentode...
Cheers
Stuben
Al,
I need to clarify...PS calcs, valve bias and normal operating parameters were not established by ear...
The topology was...
Stuben
I've done a CF amp with triode connected PL504 . Worked well , good bass , overall good sound . This used a 2C22--A2134 DC coupled driver with it's own 450V HT supply
Suggest using a lower ratio OPTX , I used a 600 ohm unit .
Al
Thanks Al, glad to hear the overall results were favorable. The PL504 appears to have a much lower Rp than the tubes I'm considering, but the concept is the same. FWIW, I did a quick load line for the PL504 in triode, and it came out to about 1K. A 600 ohm load is close enough to not cause significant issues. However, the tubes I have in mind need about 5K, so 600 won't work.
The impedance issue is something I've thought about in depth over the last few weeks. Some tubes that I wouldn't normally use in common cathode service due to poor linearity have a very low Rp and would suit CF use well. None of the "strange" shapes I have are like this, they all exhibit fairly high Rp. However, I do have some NOS ST-bottle 6AS7s, and those could definitely use a 500-600 ohm transformer.
Another thing... when the load Z is reduced, it becomes easier to make power because less voltage swing is required at the grid. Where a triode 6L6 at 5k requires almost 200V p-p to produce one watt as a follower, the 6AS7 at 500 ohms could do it with 65V p-p. Raise that to 125-150V and the 6AS7 will make a respectable amount of power. The biggest hurdle is finding a transformer suited to this use (I asked about that in a thread here not too long ago).
Anyway, I'm anxious to try this out. It's quickly moving up my long list of priorities. :)
--------------------------
Buy Chinese. Bury freedom.
6AS7 requires crippling swing fromn the driver due to it's low mu . E130L is the one to find for CF duty . I have a few but none tested anywhere near to new spec hence using PL504 (which I had a load of) . E130L have Mu 8 Ra 280 ohms as a triode . Pentode CF also interesting but I have never tried in this mode
Al
"6AS7 requires crippling swing fromn the driver due to it's low mu ."
When the 6AS7 is used as a cathode follower, isn't the voltage swing at the grid determined primarily by the transformation ratio (OPT) and the desired output power?
--------------------------
Buy Chinese. Bury freedom.
Gain of a CF is approximately mu/mu+1 . Work that one out with the 6AS7 , then recalculate with a mu> 5 triode
Al
"mu/mu+1"
OK, I understand now. I usually think about CF gain as 0.9 or so, and I don't bother to calculate it during early planning. The 6AS7 is much different than other tubes I'm considering though. I'll definitely need to weigh it against the anticipated driver swing before going to the trouble to prototype. Thanks for bringing this up.
--------------------------
Buy Chinese. Bury freedom.
Try a search for Mapletree Audio. Their headphone amps use cathode followers into the Hammond 119DA. There are schematics available.
TK
Been there done that ...
KT90 Type II with MagneQuest Iron
Sounded very detailed and articulate.
Not involving or endearing at all.
Dynamics were accentuated.
WOW without the YUM.
DanL
Dan, that's an interesting circuit. One thing though... shouldn't the OPT be 4K? The optimum Z for the tube doesn't change when the load is moved from the anode to the cathode. Maybe that's why you weren't thrilled with the sound. Did you have an opportunity to measure harmonics or other performance?
--------------------------
Buy Chinese. Bury freedom.
This was the first tube project for me.
So I didn't know alot like I do now.
I have no way of measuring just listening.
I kept waiting for it to breakin but
after 100+ hours it never changed.
It seemed to have enough power but
the sound just wasn't for me.
DanL
Well, it's a beautiful amplifier, do you still have it? 4K-5K OPTs in place of the 800 ohm parafeeds might really open up the sound.
--------------------------
Buy Chinese. Bury freedom.
Yes and No
I modded it to become my SE KT66/6L6 Amp.
Very happy with it.
DanL
Good, I'm glad to see it didn't go to waste. Nice work!
--------------------------
Buy Chinese. Bury freedom.
Edits: 09/13/14
I have dim recall of a power amp that used a 300B to drive a 6L6 cathode follower, which was transformer coupled to the speaker. Claims for good sound were made.
Eli D.
That's good to hear. My goal in pursuing this is to make use of more unusual triodes like the 829B that normally wouldn't perform well in SE, or that would be cumbersome to use for one reason or another. Using them as followers resolves linearity issues, but driving them to their usual output power would require an extremely high voltage swing on the grid. So, I'm thinking about using them at the 0.5-1W level, strictly as headphone amps. Even at that, the required swing with a 5K OPT is pretty large, but it should be manageable with a 6SN7 driver and +400 driver supply. To spice up the cosmetics, I'm also thinking about attaching nice finned anode radiators, and I want to simply ground the anodes to the chassis directly behind the tubes in order to avoid the usual shock hazard. The output tubes (and maybe the drivers) would be powered with a negative supply at the cathodes, rather than the typical B+ supply. There are other tubes with unusual shapes and styles that would also look good in this sort of arrangement. I'm concerned as to why no one has done this before (?), so I'm hoping to get a better handle on all the pros and cons before I start the actual work. Thanks for your input!
--------------------------
Buy Chinese. Bury freedom.
Perhaps a step up interstage trafo could be employed between the driver circuitry and the "final". A CF presents the lightest of loads to the driving circuitry. You should be able to get enough O/P power to use mid 90s sensitive speakers.
The 829B is interesting. Is that a built in cap. between the common cathode and the internally connected screen grids? For stereo, you will have to run PSE and 2X tubes.
Eli D.
Yes, that's an internal bypass capacitor. It's intended to decouple the screen at UHF. If you scroll down the datasheet, total G2-K capacitance (including the bypass capacitor) is shown as 65pF. I think that's small enough not to cause problems. Incidentally, this tube isn't necessarily my first choice for this use, mainly due to the somewhat hefty filament requirements to run a pair. I just happen to have a number of NOS Mullard 829Bs that I'd like to use someday. There are other "alien/space" tubes that would look great sitting out in the open on top of a modern chassis. There's the 829B's little brother, the 832, and a few ST shapes with dual external anode caps.
The IT concept is appealing for several reasons, but I'm not thrilled by the additional cost. I think this can be accomplished as a low-power amplifier without the need to add another transformer. I did that in my larger SETs, and I'm happy with the results.
--------------------------
Buy Chinese. Bury freedom.
Post a Followup:
FAQ |
Post a Message! |
Forgot Password? |
|
||||||||||||||
|
This post is made possible by the generous support of people like you and our sponsors: