Home |
Search |
Today's Posts |
#1
|
|||
|
|||
Linear gain stage
I've been doing a bit of searching regarding linearization of individual
stages before feedback is applied. One interesting old techinque was discussed at diyaudio.com, where a diode tube inversely predistorts the input to the triode (see "The long lost linear gain stage" thread in the tubes forum). Another thing I came across is this: http://www.wolcottaudio.com/WA_whitepapers.htm The problem is that a) there's no schematic, and b) it's hard to get through all the marketing BS. Nonetheless, I'm interested in figuring out what exactly is being done, specifically with the 2nd stage. I'm amazed Wolcott didn't actually list the patents being referred to. USPTO's website doesn't search properly pre-1975 patents, but, ironically, the German patent site does (for US patents at that). From the 1960s, I found the following amplifier patents by the guy: US3111630, US3328711, US3361981, and US3428912. Of the two being referred to on the website, the second one from that list (US3328711) is clearly one of them (fig. 6), but the other one I'm not sure. By the way, the best way to view patents is by downloading them as PDF files from free.patentfetcher.com. My goal in posting here is to get comments on a) how this design has been upgraded to fit the description on the website, as I failed to get a clear idea; b) the merit of the idea overall; and c) how well it stands on its own, in terms of me being able to use a similar stage in an OTL amplifier, as he's got feedbacks from various other stages, including the output transformer (my tube amplifier plans relate to driving high voltage plasma and electrostatic panels directly, as well as Berning's high frequency switcher-based impedance matching output that he calls ZOTL). |
#2
|
|||
|
|||
Hi,
I've fought my way through the description and am of the opinion that there's nothing outrageously new about the Woolcott circuit. If I understand it correctly, it (1) has a dedicated gain/splitter circuit for a positive feedback loop sourced from a tertiary winding on the output transformer, (2) has a second phase-splitting stage per push-pull half to drive (3) a set of McIntosh-inspired final driver stages, and (4) a conventional pentode output stage. So, marketing hype aside, it's defintely quite an involved circuit for any amplifier. However, even assuming that the designer has come up with a listenable product (I suspect it'll have a rather solid-state like quality to it), getting such a complicated circuit stable will require very, very careful attention to power supplies, grounding, and precise component values. There's an awful lot happening to the audio signal already, and an amplifier of this complexity runs a serious risk of instability. At any rate, this is an interesting idea - but I'd really need to be convinced that it's practical and desirable! Best regards, Russ Sadd Prune wrote: Another thing I came across is this: http://www.wolcottaudio.com/WA_whitepapers.htm The problem is that a) there's no schematic, and b) it's hard to get through all the marketing BS. Nonetheless, I'm interested in figuring out what exactly is being done, specifically with the 2nd stage. I'm amazed Wolcott didn't actually list the patents being referred to. |
#3
|
|||
|
|||
Referring to fig. 6 in the patent (US3328711), I was interested in the
triode/tetrode second/third stage as driven by the cathode follower input. I quoted this particular patent as the second and third stages essentially correspond to the website description (in the webpage, the first stage is replaced with White cathode followers, and there is only feedback signal and no input at e1, and the connection to the output stage is DC coupled with current mirror, and there is stuff on the secondary 119). Now, there's probably more to it, given a comment at audioasylum (http://www.audioasylum.com/scripts/t...ubediy&m=51536), but... Now I have four specific questions, please: a) About the cross-coupled phase splitter -- the way I understand it, a mere CSS loading of the triode also has the benefit of appearing like infinite plate impedance, so what's the point of doing it the way in the patent? The only thing I see is that the positive feedback from the third stage cathodes makes for very high gain, allowing very high feedback (and very high frequency problems from this are dealt with by the feedback from the output through capacitors 122). b) Capacitors 104 can be replaced by DC coupling through level shifting such as by a zener(as in other figs in the patent); what about capacitors 123 -- given it's in a local feedback loop, it's still in the signal path; any way to do away with it? c) What about using this configuration in an OTL amp? Adding a negative rail, the load could be connected between the output devices, like the balanced single ended Aleph-X amps (see diyaudio forums/Pass Labs). But what if the output was to be unbalanced? How would the third stage outputs be properly coupled to a single output device? I'm asking because I have specific applications in mind. d) In an OTL, since there's no tetriary winding 120, where would the main NFB be sourced? And in a way so that balanced input can be done without input transformers -- where would the main NFB be injected? |
#4
|
|||
|
|||
Prune wrote: I've been doing a bit of searching regarding linearization of individual stages before feedback is applied. One interesting old techinque was discussed at diyaudio.com, where a diode tube inversely predistorts the input to the triode (see "The long lost linear gain stage" thread in the tubes forum). Another thing I came across is this: http://www.wolcottaudio.com/WA_whitepapers.htm The problem is that a) there's no schematic, and b) it's hard to get through all the marketing BS. Nonetheless, I'm interested in figuring out what exactly is being done, specifically with the 2nd stage. I'm amazed Wolcott didn't actually list the patents being referred to. USPTO's website doesn't search properly pre-1975 patents, but, ironically, the German patent site does (for US patents at that). From the 1960s, I found the following amplifier patents by the guy: US3111630, US3328711, US3361981, and US3428912. Of the two being referred to on the website, the second one from that list (US3328711) is clearly one of them (fig. 6), but the other one I'm not sure. By the way, the best way to view patents is by downloading them as PDF files from free.patentfetcher.com. My goal in posting here is to get comments on a) how this design has been upgraded to fit the description on the website, as I failed to get a clear idea; b) the merit of the idea overall; and c) how well it stands on its own, in terms of me being able to use a similar stage in an OTL amplifier, as he's got feedbacks from various other stages, including the output transformer (my tube amplifier plans relate to driving high voltage plasma and electrostatic panels directly, as well as Berning's high frequency switcher-based impedance matching output that he calls ZOTL). The comments here won't help you as much as using a few diodes and triodes on a breadboard set up, and then measure and observe all scientifically with meters and CRO. Patrick Turner. |
#5
|
|||
|
|||
Agreed.
"Patrick Turner" wrote in message ... Prune wrote: I've been doing a bit of searching regarding linearization of individual stages before feedback is applied. One interesting old techinque was discussed at diyaudio.com, where a diode tube inversely predistorts the input to the triode (see "The long lost linear gain stage" thread in the tubes forum). Another thing I came across is this: http://www.wolcottaudio.com/WA_whitepapers.htm The problem is that a) there's no schematic, and b) it's hard to get through all the marketing BS. Nonetheless, I'm interested in figuring out what exactly is being done, specifically with the 2nd stage. I'm amazed Wolcott didn't actually list the patents being referred to. USPTO's website doesn't search properly pre-1975 patents, but, ironically, the German patent site does (for US patents at that). From the 1960s, I found the following amplifier patents by the guy: US3111630, US3328711, US3361981, and US3428912. Of the two being referred to on the website, the second one from that list (US3328711) is clearly one of them (fig. 6), but the other one I'm not sure. By the way, the best way to view patents is by downloading them as PDF files from free.patentfetcher.com. My goal in posting here is to get comments on a) how this design has been upgraded to fit the description on the website, as I failed to get a clear idea; b) the merit of the idea overall; and c) how well it stands on its own, in terms of me being able to use a similar stage in an OTL amplifier, as he's got feedbacks from various other stages, including the output transformer (my tube amplifier plans relate to driving high voltage plasma and electrostatic panels directly, as well as Berning's high frequency switcher-based impedance matching output that he calls ZOTL). The comments here won't help you as much as using a few diodes and triodes on a breadboard set up, and then measure and observe all scientifically with meters and CRO. Patrick Turner. |
Reply |
Thread Tools | |
Display Modes | |
|
|
Similar Threads | ||||
Thread | Forum | |||
Books & tools for transformer design and building | Vacuum Tubes | |||
Re Once more into the breach, dear friends/NFB in 300B | Vacuum Tubes | |||
KISS amp.Andre Jute.Stewart Pinkerton | Vacuum Tubes | |||
Who needs NFB when there is error correction? | Vacuum Tubes | |||
Turner the Ostrich ?? | Vacuum Tubes |