Hi All,
How well do KT88 and 6550's perform in Single end triode strapped?
I have read that the have less distortion than 6l6gc's and EL34's...but they don't seem to simulate much lower distortion.
Any comments are appreciated.
Thanks
How well do KT88 and 6550's perform in Single end triode strapped?
I have read that the have less distortion than 6l6gc's and EL34's...but they don't seem to simulate much lower distortion.
Any comments are appreciated.
Thanks
look here: 6GU7+ KT120 SET, better than 6550 and KT88
som test data from my amp gives good results... - Balle Clorin - Tubes Asylum
som test data from my amp gives good results... - Balle Clorin - Tubes Asylum
A pentode will behave like a pentode as will a tetrode and so will a pentode wired as a triode.
Try reading the manufacturers specification sheet. That will tell you all that is needed. Otherwise all you get is opinion, which sometimes is meaningless.
Try reading the manufacturers specification sheet. That will tell you all that is needed. Otherwise all you get is opinion, which sometimes is meaningless.
True in practice as well, frankly I prefer the triode connected EL34 to triode connected 6550.. The EL34 has pretty good curves triode connected, not I think so much the case with the 6550 at least imvl experience.
Thank you for your replies...
I have spent many hours using Trioda load line simulator. Here are my findings using this simulator.
6l6GC is generally low in 4th order distortion. Higher in the 2nd order. (I love their blue glow btw...even though it has no impact on sound)
EL34 seems lower in 2nd and 3rd but higher in 4th.
The KT88 in triode just seems generally higher in all orders.
This differs from other information I found where it shows the KT88 lower.
Am I doing something wrong with the simulator?
I asked in another thread if the anode voltage was the same as the plate voltage...meaning the voltage after the OT. I believe it is the same.
I have spent many hours using Trioda load line simulator. Here are my findings using this simulator.
6l6GC is generally low in 4th order distortion. Higher in the 2nd order. (I love their blue glow btw...even though it has no impact on sound)
EL34 seems lower in 2nd and 3rd but higher in 4th.
The KT88 in triode just seems generally higher in all orders.
This differs from other information I found where it shows the KT88 lower.
Am I doing something wrong with the simulator?
I asked in another thread if the anode voltage was the same as the plate voltage...meaning the voltage after the OT. I believe it is the same.
Since you did not specify what the operating conditions were used for the various tubes, no way to tell if you are making apples to apples comparison or not...
So I guess a triode connected KT88 Williamson style amp is also out of the question? - I wouldn't want to "cheat" by using UL taps.........
Last edited:
The original Williamson used triode connected KT66 so there is your answer. Differences in output power are significant but I would go for the KT66 regardless.
Hi yes, I am aware that the original used KT66s in triode giving 15W theoretical output. But then according to KT88 sheets from back in the day state you can get up to 30Wp-p in triode mode...
I've also just realised this thread is about SE mode LOL I think you did very well to get 20W ish, in single-ended, no.....???
I've also just realised this thread is about SE mode LOL I think you did very well to get 20W ish, in single-ended, no.....???
I've never gotten 20W out of an SE KT88, not even sure that is possible.. Something like 15W is attainable in pentode and a bit less in UL connection, about 12W?
I am not sure what the problem with triode connected 6550/KT88 comes from, but I just found them very "uninvolving" and slow sounding. Possibly the greatly increased miller capacitance was just too much for the wimpy driver stages I used at the time. (True for both PP and SE designs I worked on then.) No longer have slew rate measurements from that time, but they were substantially slower; again this might have been nothing more than a reflection of the limited current capability in the driver stage
I no longer use any of these types for anything other than series pass elements in my regulated supply so I have not revisited this issue in decades. Any comment I make should be taken with a grain of salt as a result.
Maybe use something like 6N6P, 6H30, or ECC99 in the second stage of the Williamson running at quite high currents or tubelab's powerdrive - no major problem with mr. miller then.
I am not sure what the problem with triode connected 6550/KT88 comes from, but I just found them very "uninvolving" and slow sounding. Possibly the greatly increased miller capacitance was just too much for the wimpy driver stages I used at the time. (True for both PP and SE designs I worked on then.) No longer have slew rate measurements from that time, but they were substantially slower; again this might have been nothing more than a reflection of the limited current capability in the driver stage
I no longer use any of these types for anything other than series pass elements in my regulated supply so I have not revisited this issue in decades. Any comment I make should be taken with a grain of salt as a result.
Maybe use something like 6N6P, 6H30, or ECC99 in the second stage of the Williamson running at quite high currents or tubelab's powerdrive - no major problem with mr. miller then.
I meant over a broad range of operating points.
Anywhere between 25-500V, 50-100+ mA. 7.7K- 4K loads....2 - 4watts
I understand that different tubes sound best at differing paramterers.
I'm doing another home brew and have enough stuff to get any of the parameters above.
I'm just noticed that the KT88, 6550's are said to be lower in distortion...however I generally get different results using the simulator at trioda.com
Simulation with just the output tube by itself isn't that meaningful, afterall you have to drive it with another stage, so the actual distortion of the the amplifier may be quite different. If you want to rely on the simulation, at least try it with two stages.
You might be absolutely right there. I've a SE power amp running either KT88 or 6550 in triode mode (triode only) and it certainly doesn't sound uninvolving nor slow. But I use a cathode follower after the input stage so there's at least some current feeding that power tubes grid capacitance.
I've developed this mindset, that if anything looks like a triode, drive it with at least some current capability. No ECC83's driving power tubes directly.
yes but I was considering using the KT88 to get more than the original 15W. 30W in triode mode p-p sounds very good indeed. Maybe I take into account the driver valve as stated above......
High-gm grids tend to be not so linear. I think it has something to do with the grid having to be so close to the cathode to get that high gm. So you will notice that with the high gm tubes like the KT88 there will be a lot of curve slope change as the voltage goes up.
You will get much better linearity out of a low gm tube like a 300B but they are mechanically very complicated and are expensive to make.
That's the bad news. The good news is that you can use simple plate-grid feedback on a high-gm pentode and come out with better linearity, more gain, and more output power (without grid current) than the 300B if you operate it as a pentode.
I built a simple plate-grid feedback KT88 amp and described it here: Tube Amps with a Twist: A Push-Pull Amplifier with Simple Plate-Grid Voltage Feedback Driven by a P-Channel FET This amp gave about 50W. I made it without a global feedback loop but you could easily choose a higher gain input tube and add feedback if you want distortion to be lower.
You will get much better linearity out of a low gm tube like a 300B but they are mechanically very complicated and are expensive to make.
That's the bad news. The good news is that you can use simple plate-grid feedback on a high-gm pentode and come out with better linearity, more gain, and more output power (without grid current) than the 300B if you operate it as a pentode.
I built a simple plate-grid feedback KT88 amp and described it here: Tube Amps with a Twist: A Push-Pull Amplifier with Simple Plate-Grid Voltage Feedback Driven by a P-Channel FET This amp gave about 50W. I made it without a global feedback loop but you could easily choose a higher gain input tube and add feedback if you want distortion to be lower.
KT88 triode strapped in pp Dynaco MKIII
I am experimentng with strapping the KT88s in triode in my Dynaco MKIIIs with the Poseidon boards. External linear power supplies are providing filament current. Initial impressions are pretty good. Speakers are fairly high efficiency Zu Druid IV 08, 14 ohms, so I am using the 16 ohm speaker taps from the mono-blocks. I am getting pretty good results. First album tried in Gary Burton's "Times Square" (ECM). Roy Haynes' cymbals breath out into the room and the kick drum is qick and impactful. Strikes are likewise articulate and the harmonies have a nice interlacing sustain. Tiger Okoshi's trumpet is super smooth and Steve Swallow provides a nice bass foundation. There is less damping with the triode strapped KT88s, so there is a pleasing openness to the lower frequencies. I may tighten things using the 8 ohm taps just to experiment. I did not measure the power output, but guess it's 20-25 watts?
Fortunately, the existing bias pot was sufficient to get enough juice to the power tubes.
Here, as with everything else we do, it's all about the synergy.
I am experimentng with strapping the KT88s in triode in my Dynaco MKIIIs with the Poseidon boards. External linear power supplies are providing filament current. Initial impressions are pretty good. Speakers are fairly high efficiency Zu Druid IV 08, 14 ohms, so I am using the 16 ohm speaker taps from the mono-blocks. I am getting pretty good results. First album tried in Gary Burton's "Times Square" (ECM). Roy Haynes' cymbals breath out into the room and the kick drum is qick and impactful. Strikes are likewise articulate and the harmonies have a nice interlacing sustain. Tiger Okoshi's trumpet is super smooth and Steve Swallow provides a nice bass foundation. There is less damping with the triode strapped KT88s, so there is a pleasing openness to the lower frequencies. I may tighten things using the 8 ohm taps just to experiment. I did not measure the power output, but guess it's 20-25 watts?
Fortunately, the existing bias pot was sufficient to get enough juice to the power tubes.
Here, as with everything else we do, it's all about the synergy.
- Status
- Not open for further replies.