Another view of distortion and negative feedback.
http://www.passdiy.com/pdf/distortion_feedback.pdf
Same bottom line as DF96 says.
DT
http://www.passdiy.com/pdf/distortion_feedback.pdf
Same bottom line as DF96 says.
DT
UL with NFB is a good combination for hi-fi. It is the combination I use.
Triode mode gives lower gain and less output power. Without NFB it will probably give higher output impedance. Useful for those who wish to avoid NFB.
Pentode mode gives higher gain and slightly higher output power, and higher distortion. It can't do hi-fi without significant NFB.
Sorry, that is just repeating what I said earlier.
Triode mode gives lower gain and less output power. Without NFB it will probably give higher output impedance. Useful for those who wish to avoid NFB.
Pentode mode gives higher gain and slightly higher output power, and higher distortion. It can't do hi-fi without significant NFB.
Sorry, that is just repeating what I said earlier.
The never ending discussion of neg. feedback effects on distortion harmonics has another shark hiding in the dark, which mostly invalidates even carefully derived conclusions. Take a 2.0 power device generating just 2nd harmonic, then a 1.9 power device generating a tapering tail of higher harmonics above the 2nd H. No discernible difference in sound though. One has to take the phase of the harmonics into consideration as well. The 1.9 power device generates a tail of alternating phase harmonics, it's just the mathematics modeling the less curved transfer function. Sonically, the sum of the higher harmonics cancel.
Similarly, neg. feedback straightens out the transfer function of distorting devices. A tapering tail of harmonics may well indicate improved linearity. But the typical FFT magnitude displays give you no clue whether it is good or bad. FFTs need a post analysis program to calculate the transfer curvature. Neg. feedback should always improve the "real" linearity (assuming it is covering the bandwidth), even at low amount. (it may increase some higher harmonics slightly, in a beneficial way, to do that) Whether it sounds better is another matter, since 2nd harmonic is the easiest to remove. People pay good money for 2nd H. dist. No 2nd H. gets called "clinical", "sterile", "flat"......
Similarly, neg. feedback straightens out the transfer function of distorting devices. A tapering tail of harmonics may well indicate improved linearity. But the typical FFT magnitude displays give you no clue whether it is good or bad. FFTs need a post analysis program to calculate the transfer curvature. Neg. feedback should always improve the "real" linearity (assuming it is covering the bandwidth), even at low amount. (it may increase some higher harmonics slightly, in a beneficial way, to do that) Whether it sounds better is another matter, since 2nd harmonic is the easiest to remove. People pay good money for 2nd H. dist. No 2nd H. gets called "clinical", "sterile", "flat"......
Last edited:
+1
I wanted to switch between quickly many years ago so I fitted a b+ switch that I could turn off the B+ to the power tubes change over and power up again.
However you have to make sure you don't kill the Bias supply when you do this and that the reduced load on the PSU creating voltage rise that may exceed the PSU cap working voltage.
Nearly every amp I have built has a change over switch even single ended I can't live with just one output configuration and after disconnecting and changing over so many times about 40 years ago I fitted switches. I find it to limiting..I'm not a fan of U/L many are YMMV. I prefer triode..or ES U/L.
I like to be able to use different tubes as well so I make the bias so I can change tubes..that way you don't end up with a pile of tubes you can't use.
A bit like a diesel/petrol car in case of shortage
Max Max tube amp can use anything with octal base..NB watch the bias setting when you switch..the current can be different at idle
Its interesting to be able to run most octal tubes..and switch between triode etc..I have switches that have Pentode/U/L/triode/ES U/L. I even did it with an amp running PPP..and on a friends amp Vellman and it worked well. Its not until you have switched back and forth about 100 times while listening that you will find a preference but even then its nice to have a change occasionally.
I guess if you build a PP and a SE with switchable configuration for different tubes / feedback/ input tube/ bias setting etc you only need two amps with possibility of specialist Triode and OTL..but I don't see the point in building different amps for octal tubes etc..(unless you like building amps
)Universal PP and Universal SE..
Regards
M. Gregg
Last edited:
Interesting post Gregg.
I love the ability to have it all, with as many switches and knobs as possible.
But I also appreciate a single trick pony that does absolutely the best for what it's meant to.
It is nice to have endless options on a guitar amp but with Hi-Fi?
Here is a quick photo of the back of my modded guitar amp.
It has so much unused holes front and back, children with pokey fingers should be kept in sight...
On top is my DIY load box which mimics exactly a guitar speaker impedance (200W), for recording, measuring, re-amping with SS amp for cranked tone in bedroom levels, etc..
I love the ability to have it all, with as many switches and knobs as possible.
But I also appreciate a single trick pony that does absolutely the best for what it's meant to.
It is nice to have endless options on a guitar amp but with Hi-Fi?
Here is a quick photo of the back of my modded guitar amp.
It has so much unused holes front and back, children with pokey fingers should be kept in sight...
On top is my DIY load box which mimics exactly a guitar speaker impedance (200W), for recording, measuring, re-amping with SS amp for cranked tone in bedroom levels, etc..
An externally hosted image should be here but it was not working when we last tested it.
Afaik, GU50 has no published limit on UL operation. Most material is published on the LS50 or its sibling EL152. Some pentode operating points specify up to 300V of g2. IIRC I've also seen some triode operating points of up to 800V.
Apparently, the g2 voltage on the GU50 is limited by screen dissipation, so you could run a little higher than 300V for UL, where Ug2 swings down in sync with the plate voltage. I believe there was a guy on a german forum who ran the GU50 in UL with Ua 400V, Ik 100mA on full power without any damage. So for your typical hifi amp, which spends most of the time way below max power, it's probably ok.
Apparently, the g2 voltage on the GU50 is limited by screen dissipation, so you could run a little higher than 300V for UL, where Ug2 swings down in sync with the plate voltage. I believe there was a guy on a german forum who ran the GU50 in UL with Ua 400V, Ik 100mA on full power without any damage. So for your typical hifi amp, which spends most of the time way below max power, it's probably ok.
Yes, I'm sure the GU50 could go to higher voltage in Triode, g2 current is very small at 400V so screen dissipation is minimal.
I haven't tried it in UL so can't really comment but it is g2 dissipation limited rather than g1-g2 insulation and I know people have pushed the limits an awful long way. It really is a very tough tube and a hell of a lot cheaper than 6550/KT88s. They could withstand extended red-plating in transmitter use, hard glass envelope.
I haven't tried it in UL so can't really comment but it is g2 dissipation limited rather than g1-g2 insulation and I know people have pushed the limits an awful long way. It really is a very tough tube and a hell of a lot cheaper than 6550/KT88s. They could withstand extended red-plating in transmitter use, hard glass envelope.
Think about it..what is different...bias, maybe loading..we build with all types of tubes so how is it a one trick pony? You can buy it all again and stick a different tube in..what have you gained? Build the same amp 5 times..put el34 /6L6/KT66/KT88/6550 in each one sounds nuts to me..YMMV
Build a different amp 5 times all the same and do the same again..you could build the output transformers on a different chassis and have plug and play amp circuits..
In the end there will be something you like better than the rest..however how many times are you going to build the same circuit?If a particular tube goes out of fashion (hard to get) you are still good to go! KT66 comes to mind..I have 8 Chinese KT66 so what do you do build for el34 and say oh no you cant use the KT66 they are not meant for this amp? Sorry but if it works and sounds good this isn't about effects boxes its just being practical, you can still use the different tubes to the best ability of the circuit. I have 8 6L6 8 EL34 it goes on..I remember talking to a guy in a HIFI demo room and mentioned 6L6 his immediate reaction was "isn't that a guitar tube designed for distortion"?
(but a KT66 is for audio..)Interestingly I found I preferred different tubes in the same circuit with different output configuration..Eg EL34 in triode KT66 in U/L..etc YMMV
Regards
M. Gregg
Last edited:
Modular design is useful and has been done with guitar pre-amps by Randall.
Throughout history, guitar pre-amps got various famous designs with different amount of gain stages from clean to super distorted, but the power section remained virtually the same.
What you are suggesting is make EVERYTHING modular but only keep the chassis with the output transformers?
I would say it is impractical.
If one is a fan of ultimate versatility, just build the chassis bigger and slap on it every Pot & Switch imaginable.
After the hassle/fun of modding my guitar tube amp (a million times) I find myself sticking with one good setting the matches my speaker and music style.
If I correctly understand the goal of Hi-Fi amplifiers, they should be FAR LESS personal, they should be a flat and predictable with no bells and whistles.
The best Hi-Fi amplifier is basically a "black box" that outputs exactly what has been put in but larger.
Or is it....
+1, that's my experience as well, you end up with a few favorite settings/tones that is about it. That is the reason why modular designs like Randall and amp sims are not that popular (relatively speaking), because they offer too many options where most players want just a few good ones...
Interesting question,
So what's the point of switching for triode pentode U/L etc..the reason is everyone wants "better" its hard to define better..perhaps it should say different?
If the amp and everything is flat then that's it correct? Not many people stick with equipment..interestingly an amp that gets a right pasting for being inferior is one I have seen stay in systems for years and years its the Vellman. I have seen triode amps and every other kind swapped and changed perhaps to follow fashion..How many new designs stay the course...and get left for ever. That's why I still think a universal is the way to go..particularly with octal tubes.
Most audio people set up a system listen to a few records and then think even though they don't say..what can I do to make it better..or the treble is a bit off or the image could be a bit clearer etc etc..Those that don't or can't build normally change tubes until the sockets wear out
or change cables then speakers..I'm supprised the U/L triode switches aren't worn out in a couple of months..
So much for the system is flat..more like the system is boring..or doesn't rock the boat any more..
Most people will stick with a setting they find most enjoyable..but if you then build an amp with that setting (no switches) they miss the opportunity to change over just to see if they are right..I think a universal amp should have switches for output configuration because I think different tube types sound better in different configurations.
But all amps sound the same don't they<<well maybe, so its the interaction of the equipment that's the issue..and perhaps the lack of versatility.
So your amp should be versatile ..just like an egg
Regards
M. Gregg
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.