Yes, DC balance. But as you increase the signal on the triode/UL side it will draw more DC due to 2nd harmonic distortion. So one will need some slight adjustment to allow for average signal level, or use a slow acting DC servo to keep things in order. Its possible that an E-I core may have enough effective gap slack in it to tolerate the DC shift.
----------------------------------------------
Trying to run the P-P amp in these unbalanced drive SE modes will certainly stress the power supply, so that would need regulation like for a SE amp design. It will improve the P-P amp anyway.
Triode on one side and driven pentode on the other side (class A) would come close to constant current draw on the power supply.
----------------------------------------------
Trying to run the P-P amp in these unbalanced drive SE modes will certainly stress the power supply, so that would need regulation like for a SE amp design. It will improve the P-P amp anyway.
Triode on one side and driven pentode on the other side (class A) would come close to constant current draw on the power supply.
Last edited:
Is SE a lot more than just having more 2nd harmonics. How about imbalance the LTP to get some 2nd harmonics? Just a stupid question.
But ground pentode only provide DC current, it cannot compensate signal on the triode/UL side also.
DC servo. An Op Amp integrator (input from current sense) with slow response changes the bias on the pentode (or triode/UL) side. This is often used for normal P-P stages into a toroid core that has no tolerance for DC imbalance. (tube drift, aging)
One could get some SE 2nd harmonic effect from an unbalanced LTP. Would have to do a listening test to see if it passes muster. Just changing the gain on one phase will produce some discontinuity at crossover and higher harmonics, it needs to be something smoothly varying across the full +/- range.
Other ways to get SE effect from P-P would be to send the global feedback back to the input stage (single input device) screen grid (internal Mu) instead of the cathode.
Another way is to make the output stage a differential stage with a Mosfet on one side. It's high gm just mirrors the complementary current developed by the other side while holding the cathode tail near constant voltage.
One could also just put a reversed triode in for the feedback attenuator of a SS amp. Grounded grid mode, with a CCS on the cathode. (output signal into the plate, attenuated FDBK from the cathode, 1/Mu attenuator giving Mu gain for the SS amp)
I wouldn't claim that these all produce the "same" SE sound effect. The classical SE amplifier has permeability variation in its core due to the unbalanced DC+AC, and high magnetizing current due to the air gap in the core. I can't imagine that 90 degree shifted magnetizing current really adds to the sound quality though. But who knows what these audiophiles are listening to.
One could get some SE 2nd harmonic effect from an unbalanced LTP. Would have to do a listening test to see if it passes muster. Just changing the gain on one phase will produce some discontinuity at crossover and higher harmonics, it needs to be something smoothly varying across the full +/- range.
Other ways to get SE effect from P-P would be to send the global feedback back to the input stage (single input device) screen grid (internal Mu) instead of the cathode.
Another way is to make the output stage a differential stage with a Mosfet on one side. It's high gm just mirrors the complementary current developed by the other side while holding the cathode tail near constant voltage.
One could also just put a reversed triode in for the feedback attenuator of a SS amp. Grounded grid mode, with a CCS on the cathode. (output signal into the plate, attenuated FDBK from the cathode, 1/Mu attenuator giving Mu gain for the SS amp)
I wouldn't claim that these all produce the "same" SE sound effect. The classical SE amplifier has permeability variation in its core due to the unbalanced DC+AC, and high magnetizing current due to the air gap in the core. I can't imagine that 90 degree shifted magnetizing current really adds to the sound quality though. But who knows what these audiophiles are listening to.
Last edited:
This is getting a little too fancy!!! I think I am inclined to buy a good OPT that can take some imbalance DC current. Then I just simply ground the grid on one side and switch it to pentode. I know Lundahl have OPT that can work with PP and SE. But I have to pay import tax that can end up to be super expensive. Do you know of any comparable quality OPT that is made in USA that I can order?
How about just use a simple common cathode with resistor plate resistor as the first stage. That will give some 2nd harmonic and avoid imbalancing the LTP. I see people use SRPP or even mu-follower for the first stage, just jumper the top triode with a switch with a resistor!!!
Last edited:
Someone else will have to chime in on that. I have cheapo Edcor and Hammond E-I OTs here. They do make SE OTs as well. But not cut core type. I know there are other high end winders around.
---------------
"How about just use a simple common cathode with resistor plate resistor as the first stage. That will give some 2nd harmonic and avoid imbalancing the LTP. I see people use SRPP or even mu-follower for the first stage, just jumper the top triode with a switch with a resistor!!!"
That might work too. Put a cap in series with the resistor to avoid a sudden DC shift (or loud pop from speaker)
Cat nap time here. ZZzzzzz
---------------
"How about just use a simple common cathode with resistor plate resistor as the first stage. That will give some 2nd harmonic and avoid imbalancing the LTP. I see people use SRPP or even mu-follower for the first stage, just jumper the top triode with a switch with a resistor!!!"
That might work too. Put a cap in series with the resistor to avoid a sudden DC shift (or loud pop from speaker)
Cat nap time here. ZZzzzzz
Last edited:
most positive reviews have real vintage tubes, 10y,45/46,71a,801a...
these have box plates, w filament, plate lines in datasheets are very evenly spaced.
beam tubes are far away from this (its ok in pp with tons of fb)
these have box plates, w filament, plate lines in datasheets are very evenly spaced.
beam tubes are far away from this (its ok in pp with tons of fb)
Thanks, you give me a lot to think about. I was wondering when are you going to go to bed as you are 3 hours ahead of me ( Calif).
Truly a touchy subject
Short answer... it depends. Basically, it depends on a lot of variables. System as a whole is critical to whether or not a single-ended design will make you happy. You need to consider the speakers you're using and determine if a SE amp will be able to drive them without difficulty.
I also lean towards true triodes as output tubes in a SE design. Beam power tubes don't really cut it here and the design needs to compensate, usually requiring feedback to make it perform reasonably well.
As most single-ended amps go, power is on the low side (sans some of the folks that have used transmitting triodes) so amplifier performance is key. Specifications do mean something and getting the design, engineering, parts selection and build are essential. My view for minimal acceptable specifications are:
Frequency response: 25Hz to 25KHz (10+ octaves preferred) within 1dB.
Measured THD at 1-watt: less than 0.5%
Signal-to-noise: 80dB minimum referenced to 1-watt output.
Square-wave response: 10KHz w/clean edges, sans any overshoot or ringing.
The above are required (IMNSHO) for any amplifier actually... but when you're dealing with single-digit power, higher efficiency speakers are common, so any level of noise becomes more apparent/annoying.
If you can manage the build and get the specs right, then yes... a well done single-ended design can be excellent, preferably a zero feedback design, hence true triodes, the 45 as an example, which is a gem, albeit only 2-watts.
Regards, KM
Short answer... it depends. Basically, it depends on a lot of variables. System as a whole is critical to whether or not a single-ended design will make you happy. You need to consider the speakers you're using and determine if a SE amp will be able to drive them without difficulty.
I also lean towards true triodes as output tubes in a SE design. Beam power tubes don't really cut it here and the design needs to compensate, usually requiring feedback to make it perform reasonably well.
As most single-ended amps go, power is on the low side (sans some of the folks that have used transmitting triodes) so amplifier performance is key. Specifications do mean something and getting the design, engineering, parts selection and build are essential. My view for minimal acceptable specifications are:
Frequency response: 25Hz to 25KHz (10+ octaves preferred) within 1dB.
Measured THD at 1-watt: less than 0.5%
Signal-to-noise: 80dB minimum referenced to 1-watt output.
Square-wave response: 10KHz w/clean edges, sans any overshoot or ringing.
The above are required (IMNSHO) for any amplifier actually... but when you're dealing with single-digit power, higher efficiency speakers are common, so any level of noise becomes more apparent/annoying.
If you can manage the build and get the specs right, then yes... a well done single-ended design can be excellent, preferably a zero feedback design, hence true triodes, the 45 as an example, which is a gem, albeit only 2-watts.
Regards, KM
Yes - it is all about the iron running somewhere away from the zero flux point. This linearizes the OPT.
There's an article by Eddie Vaughn c. 2006 "Single Ended vs. Push Pull (Continued): The Deep, Dark Secrets of Output Transformers" which covers this rather well. In particular:
Eddie Vaughn said:
Whereas, at low level our PP OPT is stuck in hysteresis hell.
Repeat after me: SE is all about being linear.
That depends a great deal on the choice of PP OT. It maybe true of most PP iron - but not all. I would point you to Piltron as an example.
The Bigger the introduced gap the worse the problem - so avoid gapped PP iron.
Eddie Vaughn goes into some detail in the article you mentioned to highlight why toroidal transformers avoid most of the issues normally associated with EI PP transformers, at a slight penalty in increased circuit complexity (zero DC offset been the requirement - which incidentally is easily achieved with a Servo, CCS or Garter Bias).
So I think a toroidal transformer need not suffer from the low level detail loss you mention, and that has been my experience.
Shoog
Last edited:
Was getting late last night.
My comment about the unbalanced load LTP was wrong. (was thinking class B for some reason, with a sudden gain transition) The gain should smoothly change from low to high in a class A LTP with unbalanced load resistors. This might actually give one very pure 2nd harmonic addition. The usual loaded down triode approach is also making some 3rd and 4th harmonic, although in a benign tapering sequence. So these won't sound quite the same.
You may still run into some DC imbalance in the OT with these pre-distortion approaches if one output tube is handling higher power (so making some extra 2nd harmonic itself). (the earlier inserted 2nd harmonic approach just looks like normal signal, and the cap coupling removes any DC component there anyway.) But lets keep this DC imbalance issue in perspective. The SE case has maximum signal, 100%, DC imbalance in the core right from the start, so needs a big gap. The 2nd harmonic imbalance issue with rigged P-P is only like a couple of %, many E-I OTs can handle this, but if not, this only requires a small re-adjustment of the tube biasing to offset it.
The triodes used for SET are usually low Mu types (and low Rp), and old low gm types (often DHT) So KT-88 is not filling the specs here well when trioded. The TV sweep type tubes are generally low internal Mu (like 3 or 4), and low Rp in triode, but have higher gm (just degenerate it some to linearize, some have even put a damper diode below to degenerate ). There have been some well received 6CB5 sweep tube "SETs" for example.
On this SE OT idea of less hysteresis distortion, I'm not buying that. From knowing some magnetics, when you put DC on a core, the best high permeability domains are the 1st ones to saturate, you are left with the low permeability domains which have higher anisotropy (stickiness). The hysteresis is just not so obvious for the gapped core since the linear (air gap) magnetizing current is so much higher as to hide it. But then even the linear magnetizing current is 90 degrees out of phase with the signal, so technically it's still distortion. A grain oriented toroid clearly is the way to go for low OT distortion. Local feedback around the outputs to lower Zout is another way to handle any hysteresis current, and should work fine for an E-I core. Many like the E-I sound, its classic as well. Less sensitive to DC imbalance.
My comment about the unbalanced load LTP was wrong. (was thinking class B for some reason, with a sudden gain transition) The gain should smoothly change from low to high in a class A LTP with unbalanced load resistors. This might actually give one very pure 2nd harmonic addition. The usual loaded down triode approach is also making some 3rd and 4th harmonic, although in a benign tapering sequence. So these won't sound quite the same.
You may still run into some DC imbalance in the OT with these pre-distortion approaches if one output tube is handling higher power (so making some extra 2nd harmonic itself). (the earlier inserted 2nd harmonic approach just looks like normal signal, and the cap coupling removes any DC component there anyway.) But lets keep this DC imbalance issue in perspective. The SE case has maximum signal, 100%, DC imbalance in the core right from the start, so needs a big gap. The 2nd harmonic imbalance issue with rigged P-P is only like a couple of %, many E-I OTs can handle this, but if not, this only requires a small re-adjustment of the tube biasing to offset it.
The triodes used for SET are usually low Mu types (and low Rp), and old low gm types (often DHT) So KT-88 is not filling the specs here well when trioded. The TV sweep type tubes are generally low internal Mu (like 3 or 4), and low Rp in triode, but have higher gm (just degenerate it some to linearize, some have even put a damper diode below to degenerate ). There have been some well received 6CB5 sweep tube "SETs" for example.
On this SE OT idea of less hysteresis distortion, I'm not buying that. From knowing some magnetics, when you put DC on a core, the best high permeability domains are the 1st ones to saturate, you are left with the low permeability domains which have higher anisotropy (stickiness). The hysteresis is just not so obvious for the gapped core since the linear (air gap) magnetizing current is so much higher as to hide it. But then even the linear magnetizing current is 90 degrees out of phase with the signal, so technically it's still distortion. A grain oriented toroid clearly is the way to go for low OT distortion. Local feedback around the outputs to lower Zout is another way to handle any hysteresis current, and should work fine for an E-I core. Many like the E-I sound, its classic as well. Less sensitive to DC imbalance.
Last edited:
One could put the E-I lams all one way, or in blocks, (but no intentional gap) to increase the OT DC handling for a rigged P-P design.
-------------------------------------
Using some local feedback around the outputs (low Zout) also overcomes the low permeability near zero flux in a typical OT, so should preserve detail and overcome hysteresis current as well.
-------------------------------------
Using some local feedback around the outputs (low Zout) also overcomes the low permeability near zero flux in a typical OT, so should preserve detail and overcome hysteresis current as well.
Last edited:
If you are after SET signature sound and have the speakers that can allow it to shine (high efficiency), then yes. Otherwise keep your SS amps and add equalizer to play around with the sound.
And don't skimp on the room acoustics.
It is worth it ONLY if you have the correct speakers to use with this kind of amplifier. If you are driving the typical modern commercial stereo speaker you find in most stores today then No 8-10 WPC is not enough
Just as a data point I have a guitar amp that uses one 6V6 tube in a single ended design. The 6V6 is much smaller than your KT88. I use a 105 db/W at 1M speaker. If I turn up the volume the sound can be heard INSIDE the house across the street. (Don't ask how I know this.)
Typically for practice I install a 9db "L-Pad" attenuator (made with some 25W ceramic power resisters) and play through a MUCH smaller speaker with the volume at about 3 (out of 10) and it still annoys others in the house who are trying to watch TV or something.
Think of it this way, to make your 8W amp 4X as powerful your need 32W. With tubes this is getting expensive. But you get the same effect if you buy a speaker with 12db greater efficiency.
In my opinion, 80% of that "tube sound" comes from the speakers that tube amp owners have to use. These speakers are typically in quite large cabinets and have large drivers and operate at about maybe close to 95 db/w An 8WPC stereo would be "way loud" if you have those speakers. It's these speakers that are so easy to drive that have a sound of their own. The amp is at most only 20% of this sound. So the reason you build the SE amp is so these really nice speakers can "shine."
Last edited:
Whatever the rationale behind SE amps, it is NOT linearity. All SE amps have a higher amount of even harmonics compared to PP. This makes SE sound sweet and "musical". If SE really were linear, the amount of harmonic distortion would be lower compared to PP, OK?
Worth it?
Well, if you've got one, nobody can say you haven't. I have. So whatever I say about them, nobody can say I don't speak from experience.
That's worth something to me.
Well, if you've got one, nobody can say you haven't. I have. So whatever I say about them, nobody can say I don't speak from experience.
That's worth something to me.
I am not against using negative feedback in designs just to clear that up right away. BUT, I am very interested building an amplifier without the use of feedback. I was originally thinking about low power SET for driving some headphones, but nothing is set in stone. Anyway looking for triodes with low internal resistance I found the 6C33B-C🙂 I am now very interested in this tube.
SET doesn't have to only be 2 watts output power, I think it's the output impedance when zero feedback is used which gives these amps their characteristics, along with the added 2H.
SET doesn't have to only be 2 watts output power, I think it's the output impedance when zero feedback is used which gives these amps their characteristics, along with the added 2H.
- Status
- Not open for further replies.