Here's what I tried in PSPICE as an alternative to interstage coupling... never tried it on a bench, though.
An externally hosted image should be here but it was not working when we last tested it.
Marcus
Clean DC is a must for 811 filaments. I looked into using linear regulators and chokes but regulating 4 amps will generate a good bit of heat. The solution was to modify two 5v switching supplies to output 6.3v. Adding a resistor in series with the trim pot is all thats needed. Be careful with this because some supplies use filter caps rated for only 6v.
AC on the filaments is not a problem with the 811 in push-pull. I have an A2 PP 811 amp running AC filaments with no hum what-so-ever. I can't comment on using AC in SE mode though where it would be more of a problem.
As stated by Tubelab in another post, High frequency AC (50khz and above) on the filaments only is an improvement if a CLEAN HIGH FREQUENCY SINE-WAVE is used to power the filaments. This is EXTREMELY HARD TO IMPLEMENT. High frequency square wave filament power causes WAY MORE noise problems that standard 50/60 hz AC filament power will give.
CLEAN HIGH FREQUENCY SINE-WAVE power supplies are HUGE are quite complicated and very inefficient.
Good luck, Daniel
As stated by Tubelab in another post, High frequency AC (50khz and above) on the filaments only is an improvement if a CLEAN HIGH FREQUENCY SINE-WAVE is used to power the filaments. This is EXTREMELY HARD TO IMPLEMENT. High frequency square wave filament power causes WAY MORE noise problems that standard 50/60 hz AC filament power will give.
CLEAN HIGH FREQUENCY SINE-WAVE power supplies are HUGE are quite complicated and very inefficient.
Good luck, Daniel
AC on 811A filaments
One problem with heating the 811A filaments is that it is a very high-mu tube, so that hum is a lot harder to eliminate than with low mu tubes like the 2A3 and 300B. DC or RF is essential when run in SE mode.
Another effect of the 811A being high-mu is that its output plate resistance is fairly high, leading to poor damping factor. That is why Ganti put in feedback from the transformer secondary to the 811A filaments - it lowered the output impedance enough to tame his speakers.
- John
One problem with heating the 811A filaments is that it is a very high-mu tube, so that hum is a lot harder to eliminate than with low mu tubes like the 2A3 and 300B. DC or RF is essential when run in SE mode.
Another effect of the 811A being high-mu is that its output plate resistance is fairly high, leading to poor damping factor. That is why Ganti put in feedback from the transformer secondary to the 811A filaments - it lowered the output impedance enough to tame his speakers.
- John
I´ve seen several transmitter triode SE A2 schematics that employs cathode feedback through the OPT secondary, so obviously it works well (I use the same thing in some of my more conventional SE amps).
I´m planning to go the same way in my future 808 SE amp, but what happens when the grid voltage dives below 0V?
I mean, there must be some kind of glitch when the cathode current suddenly decreases by 20mA or so?
My (theoretical) conclusion is that CFB only works as long as the tube operates in pure A1 with no grid current at any time or pure A2 with grid current all the time.
I will investigate this as soon as my OPT´s arrive from Hong Kong so that I can build a prototype.
BTW I´m going to use Mosfet followers to feed the grids, I grew tired of direct coupled cathode followers when I messed with 811A a couple of years ago. I also believe Mosfets will do a better job supplying current to the very nonlinear grid impedance.
PS: Has anyone actually tried driving this kind of tubes in grounded grid mode with success?
It would be kinda cool to build an hybrid amp with a big direct heated output tube and the rest of the circuit all SS...
I´m planning to go the same way in my future 808 SE amp, but what happens when the grid voltage dives below 0V?
I mean, there must be some kind of glitch when the cathode current suddenly decreases by 20mA or so?
My (theoretical) conclusion is that CFB only works as long as the tube operates in pure A1 with no grid current at any time or pure A2 with grid current all the time.
I will investigate this as soon as my OPT´s arrive from Hong Kong so that I can build a prototype.
BTW I´m going to use Mosfet followers to feed the grids, I grew tired of direct coupled cathode followers when I messed with 811A a couple of years ago. I also believe Mosfets will do a better job supplying current to the very nonlinear grid impedance.
PS: Has anyone actually tried driving this kind of tubes in grounded grid mode with success?
It would be kinda cool to build an hybrid amp with a big direct heated output tube and the rest of the circuit all SS...
Fuling said:
There is that problem whenever you stray from the *1 classes of operation. The grid impedance will be nonlinear, and the traditional fix was xfmr coupled drivers and low u triodes to reduce the driver imedance as much as possible. Using MOSFET source followers is a much better solution since the MOSFET, being a much higher gain device whose g(m)'s are measured in amps per volt, have very low Zo's. That should pretty much take care of that little problem.
Feedback continues to work so long as the VT isn't actually cutoff. If there's signal. there's NFB.
It's really your best choice.
That's more of an RF topology since it doesn't need neutralizing, and the likely drivers already are loZ devices (i.e. 20W ham transciever driving big linear).
Nothing stopping you from trying it. If you do, let us know how it worked out.
Just as an additional possiblilty, the 815 tube that i've been playing around with seems to be intended for being driven with grid current gives AB2 operating specs in the data sheets. It's two beam tetrodes in one envelope, with two plate caps. It's rated for 54 watts class B. It was originally meant as a VHF transmitter tube. It looks like a little robot.
When I was developing the 845SE, I tried grounding the grid of an 845 and puting a mosfet where the cathode resistor would go. I could never get the bias stable, and the amp tended to run away leading to glowing 845's and blown mosfets. I think that this combination just amplified the nonlinearities already present in the mosfet. I did not try any type of servo bias correction circuits.
Then I tried stacking the (GG) 845 on top of a tube. I tried a few different tubes, with the best being a 6AS7 (sounded pretty good). You need a tube that is linear with the plate voltage equal to the bias voltage of the output tube, which is a lot for the 845. I was thinking of using a diff pair made with the 6AS7. I abandoned this idea when I realized that any type of grounded grid circuitry was costing me about 140 volts of B+ (compared to grounded cathode fixed bias), which I could make up by raising the B+.
This would not be an issue with the 811A since the cathode voltage would actually be negative compared to the grid.
If you really want to build a hybrid amp using some of the best circuitry from the SS world, think about this. Build a killer low power SE amp using the best tube that you can find (45 comes to mind) and a reasonable OPT ($18 Edcor) resistively loaded. In the cathode circuit of the 45 you put the sensing leg of a SS current mirror circuit. The current mirror output is scaled up by a factor of 3 to 5 and wired into the cathode of a big power tube whose grid is grounded. Now the current through the big tube is an exact copy of the 45 tube's current only multiplied by the scaling factor. A lot cheaper and more stable than 5 X 45's in parallel.
The little Edcors are amazing transformers for $18. Like any small transformer, they will saturate if you try to pump a lot of bass through them. If your speakers are very efficient (100+db) then your amp can operate at 1 or 2 watts, and these transformers would work well. If your speakers are not that efficient, but don't have much bass response, they may still work well. My studio monitors are 86 db and have no response below 70 hZ. with these transformers I get some saturtion at high volume levels. These would work better if my speakers were 90db. If you have big speakers and want loud bass, you will not be happy with these.
It also depends on what music you play, and how loud you play it.
I got a pair for experimentation. They have been used in several experiments. I have always used better transormers in the final design.
It also depends on what music you play, and how loud you play it.
I got a pair for experimentation. They have been used in several experiments. I have always used better transormers in the final design.
Class A2 grid behavior
When the grid voltage falls below zero, the plate current will continue to decrease, although with a high-mu triode like the 811, it will quickly get cut-off. However, the grid current will abruptly fall to zero, lightening the load on the driver stage. This is where the non-linear behavior occurs. To avoid this, always keep the grid positive! It seems kind of strange to bias an amp according to the instructions: "adjust the bias to about +22V to reach proper plate current", but this is how is works.
Back in the early 1930s there was a development of class A2 audio output stages where a cathode follower would directly drive a high mu power triode. called the "Dynamic-Coupled Amplifier". The two triodes were often combined in the same envelope, such as the 6B5, 6N6G, etc. The 6AC5GT was a stand-alone high-mu power triode designed to be driven by a type 76 cathode follower. See this article for how this works:http://www.clarisonus.com/Archives/Amp_Design/New2B6.pdf I think this concept originated as a way to get around the pentode patents.
- John
When the grid voltage falls below zero, the plate current will continue to decrease, although with a high-mu triode like the 811, it will quickly get cut-off. However, the grid current will abruptly fall to zero, lightening the load on the driver stage. This is where the non-linear behavior occurs. To avoid this, always keep the grid positive! It seems kind of strange to bias an amp according to the instructions: "adjust the bias to about +22V to reach proper plate current", but this is how is works.
Back in the early 1930s there was a development of class A2 audio output stages where a cathode follower would directly drive a high mu power triode. called the "Dynamic-Coupled Amplifier". The two triodes were often combined in the same envelope, such as the 6B5, 6N6G, etc. The 6AC5GT was a stand-alone high-mu power triode designed to be driven by a type 76 cathode follower. See this article for how this works:http://www.clarisonus.com/Archives/Amp_Design/New2B6.pdf I think this concept originated as a way to get around the pentode patents.
- John
I have a pair of UBT3's. They do sound good, but there are other transformers in the same price range that sound similar. I live is a warm humid location, and my UBT3's have developed rust on the laminations. They have never been outside of the air conditioned house. I haven't been to Singapore since 1988, but I think your humidity is similar to ours. No other transformers have rusted except for some surplus guitar amp transformers.
I use Transcendars (Ebay) in my current amps, but they are not available outside the US. They are very similar in sound to the UBT's but half the cost.
I have only heard good things about James transformers but have not tried them yet. Their US distributor is not currently selling transformers. I plan to get a pair of them when they become available again, and I get caught up on everything else.
Transformers are heavy and international shipping can be really expensive. If James are available locally, try them.
John:
Thanks for the info on the 2B6. 4 watts out, driver built in, sounds like a unique spud amp to me. I know that I have some of those in my warehouse. It is going to take some serious digging to find them.
I use Transcendars (Ebay) in my current amps, but they are not available outside the US. They are very similar in sound to the UBT's but half the cost.
I have only heard good things about James transformers but have not tried them yet. Their US distributor is not currently selling transformers. I plan to get a pair of them when they become available again, and I get caught up on everything else.
Transformers are heavy and international shipping can be really expensive. If James are available locally, try them.
John:
Thanks for the info on the 2B6. 4 watts out, driver built in, sounds like a unique spud amp to me. I know that I have some of those in my warehouse. It is going to take some serious digging to find them.
Tubelab: Regarding grounded grid, I was thinking about something simpler like a cascoded mosfet CCS at the cathode and the grid connected to +80V or so. The input signal would modulate the CCS by the gate of the lower Mosfet..
Perhaps some DC feedback could be taken from the cathode to balance things up a bit?
The whole thing would have some similarities to the Zen amps.
Perhaps some DC feedback could be taken from the cathode to balance things up a bit?
The whole thing would have some similarities to the Zen amps.
yeah, the humidity here is pretty high. in the nineties...
in that case, i will go with my james 6123hs plans.. i've picked up a massive mains trans,
its 330-310-0 @ 1.5A
and there are 2 heater taps. 6.3-3.15-0 and 8-0, both rated 10A.
i was wondering if i could rectify the 8 volt tapping for use as a 811 heater. if i were running DC, could i put them in parallel?
in that case, i will go with my james 6123hs plans.. i've picked up a massive mains trans,
its 330-310-0 @ 1.5A
and there are 2 heater taps. 6.3-3.15-0 and 8-0, both rated 10A.
i was wondering if i could rectify the 8 volt tapping for use as a 811 heater. if i were running DC, could i put them in parallel?
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.