keithgreenhalgh said:Two different people had the same amp as me.they Both were sure that changing the kt88 tubes in one case was a great improvement were the second person tried it and said the origional kt88 was the best.They can't both be right so it must either be that they think they hear a difference or perhaps one of them perfers a certain sound but whats to say the sound he hears(if he can hear a difference) is better?
1 thing is for sure. They felt a strong difference. All the other factors are highly unknown and debatable.
P.S. Not long ago I recommended Genalex KT88 reissue to a member in another thread along with 12AX7 Tungsol reissue. He got em and replied positively.
regal said:I guess I would rather tube roll than apply NFB, if I wanted NFB I would be working with opamps.
Don't follow myths-rolling, follow Engineering!
NFB is bad only and only when used to correct very dirty amplifiers. I.e. applying deep NFB by voltage to a typical opamp you don't get much better linearity of current amplification, and output/input resistances will be frequency dependent though overall voltage gain measurements will be nice, so as the result good according to measurements amp sounds horrible. But when the amp is already very good NFB may be used to make it perfect.
I.e. all superstitious myths are made from real facts, but in order to make a bad myth the fact must be wrongly interpreted, like generalized from one case to many other cases when it is not applicable at all.
SY said:Not unless you're doing something very strange. In fact, it often has the opposite effect.
One exception is single VAS (voltage amplifier stage) based circuits where the feedback is applied to the grid circuit. (This can include a single VAS driving a CF.) Input impedance is simply Rin as would be the case in a simple inverting op-amp stage. Rin would normally tend to be a lowish resistance value particularly if more than unity gain was required. Input impedances much above 100K aren't such a great idea with this topology particularly if gain is needed as well.
Heck, manufacturers themselves did "cherry picking" of tubes... in essence, a type of tube-rolling.
Whenever you see a reference to a "red dot" or "red mark" Telefunken or RCA, et al, tube, those were tubes that were tested and picked off the assembly line for close-tolerance use (such as electronic test equipment, military radar, etc). The remaining "rejects" were sold as consumer-use tubes.
Many 6DJ8/6922s, EF804s, and 12AX7s (and others) went through this process... and these were for some of the most critical-function circuits possible (i.e, they HAD to work). So, what to make of it, if even the tube factories and OEM manufacturers of the most critical equipment, did exactly what's being discussed here, in essence?
Just want to put a bit of historical perspective on things...
Regards,
Gordon.
Whenever you see a reference to a "red dot" or "red mark" Telefunken or RCA, et al, tube, those were tubes that were tested and picked off the assembly line for close-tolerance use (such as electronic test equipment, military radar, etc). The remaining "rejects" were sold as consumer-use tubes.
Many 6DJ8/6922s, EF804s, and 12AX7s (and others) went through this process... and these were for some of the most critical-function circuits possible (i.e, they HAD to work). So, what to make of it, if even the tube factories and OEM manufacturers of the most critical equipment, did exactly what's being discussed here, in essence?
Just want to put a bit of historical perspective on things...
Regards,
Gordon.
If a negative feedback is applied in parallel with input signal (i.e. to the same grid of the tube) it decreases input resistance, but when it is applied in series with input signal (i.e. to cathode) it increases it.regal said:With a tube amp, applying NFB reduces the input impedance to unacceptable levels doesn't it?
Here is the picture I drew long time ago for my students to explain basics:
Tubes4e4 said:Hi Gordon,
I wouldn´t call this classical tube rolling, because I understand what you are talking about is just factory-side testing and selecting to certain electrical specs or tolerance ranges, etc.
Regards,
Tom
The point I'm making is, though, that there was a claim that the only way to good engineering, was through making a device tube-independent, at least for different OEM's for the same type tube.
I'm only pointing out, that some of the MOST SOPHISTICATED, most carefully-engineered devices OF ALL, required a level of device selection BEYOND tube rolling... to the point where even some tubes OFF THE SAME ASSEMBLY LINE wouldn't work.
So, there's obviously some problems, with using device-interchangeability as a prime evaluator of what constitutes "good engineering"...
Regards,
Gordon.
I've been looking through my data sheets to try to find the EL821 audio reference, but haven't found it yet.
It's in the STC data sheet:
An externally hosted image should be here but it was not working when we last tested it.
Could it be that the STC 6CH6 is not the same design as the Philips EL821, although close enough to be an equivalent for the purposes of video output? I had not heard before that this was intended to be an 'EL84'.
This thread contains the usual mixture of wisdom and nonsense which we love here on DIYaudio!
This thread contains the usual mixture of wisdom and nonsense which we love here on DIYaudio!
What about the same tube type that uses different internal materials and construction?
I built a simple 6SN7 preamp similar to Tubecad CCDA. Operating at a safe 12ma 150V,nice a linear on the curves. Tying an RCA which has black plates vs. a GE with gray plates vs. a Sylvania with black box plates yield different sound. They look different and sound at least slightly different.
A very nice preamp.
Scott
I built a simple 6SN7 preamp similar to Tubecad CCDA. Operating at a safe 12ma 150V,nice a linear on the curves. Tying an RCA which has black plates vs. a GE with gray plates vs. a Sylvania with black box plates yield different sound. They look different and sound at least slightly different.
A very nice preamp.
Scott
What about the same tube type that uses different internal materials and construction?
I built a simple 6SN7 preamp similar to Tubecad CCDA. Operating at a safe 12ma 150V,nice a linear on the curves. Tying an RCA which has black plates vs. a GE with gray plates vs. a Sylvania with black box plates yield different sound. They look different and sound at least slightly different.
A very nice preamp.
Scott
replacing the resistor loads with mosfet CCS and you have an aikido on a pair of tubes....
That's a nice way of putting it...Thread necromancy is alive and well!
Keep in mind that Manufacturers back in the day and even today SCREEN their products... and sell the screened products for a premium...
Tubes, transistors, zeners, resistors ....ect..ect.. can be purchased in a tolerance band, matched pairs..ect... It's always been like that... nothing new..
Consumer products need to be designed such that these variances, characteristics are no longer dominant in the circuit... I think the first class I took in freshman college with Solid State circuit design taught us to design out the Beta from the circuit, ie Beta becomes a 3rd order effect...this way circuits will behave themselves over usable life of product...
Tubes, transistors, zeners, resistors ....ect..ect.. can be purchased in a tolerance band, matched pairs..ect... It's always been like that... nothing new..
Consumer products need to be designed such that these variances, characteristics are no longer dominant in the circuit... I think the first class I took in freshman college with Solid State circuit design taught us to design out the Beta from the circuit, ie Beta becomes a 3rd order effect...this way circuits will behave themselves over usable life of product...
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
- Home
- Member Areas
- The Lounge
- Engineering vs tube rolling