Konnichiwa,
Okay. I'm not in for that either. But unless your Powersupply is solid you might as well not bother. But as a hint, I normally build single supplies for both channels, with the final Filter stages split. But that does not mean I reduce the number of iron pieces.
The PSU for my Phonostage has 6 pcs of 120uF Film Capaictors and 8pcs 10H Chokes, for some serious filtering and is followed by low resitance RC filtering per channel after that. The Supply with it's GZ34 Rectifier would be happy feediong a 45 SE Stereo Amp or something a little bigger...
Wisdom is an interesting thing. I prefer experience. I rarely other peoplkes powersupplies, usually neither the step response nor the noise levels are much too my liking.
You need to unserstand the context.... It is not as such the silver but rather what it allows. BTW, silver transformers are on my list. Silver is a sufficiently industrial metal as not to impact on the priuce of a really good transformer too much. I intend to use bade silver foil secondary, paper insulated and enameled silver wire for the primary. Permalloy core. Cost per transformer, scary.
Now I'll comment once more before I leave to make your own choices.
If I wanted an effective and affordable SE Amp I would use a High Gm Triode/Triode wired pentode driver (preferred Siemens/Valvo EC8020, E810F & WE437A or STC 3A/167M, the others after that, E180F/6688 being a good cheap option) with solid state active load like the DN2540 FET (10...20mA/150...200V) and a suitable Film Bypass Cap for the Cathode (rubycon SWR?), coupled via gridchoke and silver mica coupling cap (still a lot less than a decent IT) to a self bias 300B (Reflector/Sovtek/EHX or TJ, depending on wallet size) with a 100uF electrolytic cathode cap and 22...33uF Film "Ultrapath" (+B to Cathode) Cap.
The whole PSU based on Electrolytic Cap's and Hammond Iron, well filtered (LCLC) using heavily oversized electrolytic cap's and RCA 5R4 or WE274A/B. The 300B Heater Supply DC would be schottky rectified and regulated with an LT1085 post filtered by Ferrite Chokes. You can find 1mH/1.5A and 47uF/63V cap's quite easily - makes a 500Hz/2nd Order lowpass. I'd also throw some Common Mode Chokes plus a further of Mylar Film Caps in and use Sanyo Os-Con's around the LT1085.
The cathode resistors would be split between the two sides of the filament and adjusted to ensure identical current flow from either side of the filament to ground (evens the wear of the cathode DC heated), cathode bypass cap & ultrapath cap returned to the positive terminal of the heater.
I'd probably shell out big on the output transformer, I'm having something developed (very slow work), but otherwise I'd probably just order Tango X20S or FC-30-3.5S and be done, at around $ 700 shipped (with current dollar exchange rate not too bad). Suitable nickel core grid chokes from S&B and Magnequest (the S&B has more inductance, allowing smaller value coupling cap's) and probably other sources too.
The resulting amp (build well) will outperform many (almost any?) commercial SE Amp's, will be dead quiet (< 0.1mV noise), with a better dynamic range than most and will sound very transparent. It would certainly worth it to put "premium" Valves (NOS WE437A, WE300B, WE274A/B) into it, for the "real WE sound" (yes, there is such) at a cool 3-4k american depending on market conditions....
For something really uncompromised I have commented before, it would likely require an industrial lifting device or serious body building lessons in the gym to move about....
Sayonara
DVDHack said:
Okay. I'm not in for that either. But unless your Powersupply is solid you might as well not bother. But as a hint, I normally build single supplies for both channels, with the final Filter stages split. But that does not mean I reduce the number of iron pieces.
The PSU for my Phonostage has 6 pcs of 120uF Film Capaictors and 8pcs 10H Chokes, for some serious filtering and is followed by low resitance RC filtering per channel after that. The Supply with it's GZ34 Rectifier would be happy feediong a 45 SE Stereo Amp or something a little bigger...
DVDHack said:
Wisdom is an interesting thing. I prefer experience. I rarely other peoplkes powersupplies, usually neither the step response nor the noise levels are much too my liking.
DVDHack said:
You need to unserstand the context.... It is not as such the silver but rather what it allows. BTW, silver transformers are on my list. Silver is a sufficiently industrial metal as not to impact on the priuce of a really good transformer too much. I intend to use bade silver foil secondary, paper insulated and enameled silver wire for the primary. Permalloy core. Cost per transformer, scary.
Now I'll comment once more before I leave to make your own choices.
If I wanted an effective and affordable SE Amp I would use a High Gm Triode/Triode wired pentode driver (preferred Siemens/Valvo EC8020, E810F & WE437A or STC 3A/167M, the others after that, E180F/6688 being a good cheap option) with solid state active load like the DN2540 FET (10...20mA/150...200V) and a suitable Film Bypass Cap for the Cathode (rubycon SWR?), coupled via gridchoke and silver mica coupling cap (still a lot less than a decent IT) to a self bias 300B (Reflector/Sovtek/EHX or TJ, depending on wallet size) with a 100uF electrolytic cathode cap and 22...33uF Film "Ultrapath" (+B to Cathode) Cap.
The whole PSU based on Electrolytic Cap's and Hammond Iron, well filtered (LCLC) using heavily oversized electrolytic cap's and RCA 5R4 or WE274A/B. The 300B Heater Supply DC would be schottky rectified and regulated with an LT1085 post filtered by Ferrite Chokes. You can find 1mH/1.5A and 47uF/63V cap's quite easily - makes a 500Hz/2nd Order lowpass. I'd also throw some Common Mode Chokes plus a further of Mylar Film Caps in and use Sanyo Os-Con's around the LT1085.
The cathode resistors would be split between the two sides of the filament and adjusted to ensure identical current flow from either side of the filament to ground (evens the wear of the cathode DC heated), cathode bypass cap & ultrapath cap returned to the positive terminal of the heater.
I'd probably shell out big on the output transformer, I'm having something developed (very slow work), but otherwise I'd probably just order Tango X20S or FC-30-3.5S and be done, at around $ 700 shipped (with current dollar exchange rate not too bad). Suitable nickel core grid chokes from S&B and Magnequest (the S&B has more inductance, allowing smaller value coupling cap's) and probably other sources too.
The resulting amp (build well) will outperform many (almost any?) commercial SE Amp's, will be dead quiet (< 0.1mV noise), with a better dynamic range than most and will sound very transparent. It would certainly worth it to put "premium" Valves (NOS WE437A, WE300B, WE274A/B) into it, for the "real WE sound" (yes, there is such) at a cool 3-4k american depending on market conditions....
For something really uncompromised I have commented before, it would likely require an industrial lifting device or serious body building lessons in the gym to move about....
Sayonara
If I wanted to achieve a higher power output from this design would it be just a matter of adding a tube in parallel connections to the existing ones and an additional heater circuit or is more engineering required.
I've checked a few designs and there seems to be very lttle in making it a PSE.
Thanks
Ralf
I've checked a few designs and there seems to be very lttle in making it a PSE.
Thanks
Ralf
Konnichiwa,
In theory, yes. In parctice the result is not always as expected. I found significant sonic improvements in simply "pulling the second 300B" out of a PSE Amp and re-connecting the 4 Ohm Speakers to the 2 Ohm Output tap.
I observed similar things with Kondo Amp's (Baransu vs. Kegon) where the only difference was the use of PSE vs SE and the Output transformer Impedance/Size, but otherwise identical circuits, parts quality etc. The difference between PSE and SE is only subtle on some recordings, on others it is literally day/night.
Similar observations apply to the linestage of the Arthur Loesch Preamp.
If you want to use more than one output Valve I feel that you should use a complete Output transformer & Cathode resistor/cap combo (where applicable) and connect the output transformer secondaries in series (eg the two 4 Ohm sections in series give an 8 Ohm Speaker match).
This way each output valve operates completely unaffected by the other. If you simply parallel the valves they need to share the current equally at all levels of signal, something that real valves rarely do well....
So, basically use either a suitable valve (any of the "Super 300B's") or use seperate Transformers with series secondaries.
Yes, so it appears. As always, apparency is not neccesarily reality.
Sayonara
DVDHack said:
In theory, yes. In parctice the result is not always as expected. I found significant sonic improvements in simply "pulling the second 300B" out of a PSE Amp and re-connecting the 4 Ohm Speakers to the 2 Ohm Output tap.
I observed similar things with Kondo Amp's (Baransu vs. Kegon) where the only difference was the use of PSE vs SE and the Output transformer Impedance/Size, but otherwise identical circuits, parts quality etc. The difference between PSE and SE is only subtle on some recordings, on others it is literally day/night.
Similar observations apply to the linestage of the Arthur Loesch Preamp.
If you want to use more than one output Valve I feel that you should use a complete Output transformer & Cathode resistor/cap combo (where applicable) and connect the output transformer secondaries in series (eg the two 4 Ohm sections in series give an 8 Ohm Speaker match).
This way each output valve operates completely unaffected by the other. If you simply parallel the valves they need to share the current equally at all levels of signal, something that real valves rarely do well....
So, basically use either a suitable valve (any of the "Super 300B's") or use seperate Transformers with series secondaries.
DVDHack said:
Yes, so it appears. As always, apparency is not neccesarily reality.
Sayonara
Konnichiwa,
Not sure, I have little exposure to them. You get them from KR, Vaic/Emission Labs and EAT (is EAT a continuation of Vaic?) as well as Tianjin (Full Music).
Yes, they usally draw more heater current and need more +B Voltage AND current to reach their higher output power. You often need different output transformers as well.
Well, you're going to have to ask the "pee-pull" why "they" don't use them. I see little need for more than the 6 - 10W you get from a normal 300B, with high performance speakers such power levels invariably can raise lease breaking levels.
And Electrostatic Speakers really don't mix with SET's (ESL are the only "legit" high performance speakers having low sensitivity) due to their problematic impedance and the fact that they require "voltage" drive for low distortion.
Even more. If you really max out the normal 300B (like done in the Electraprint DRD Amp from VTV) you can get around 12W, using any of the 300B's on steroids can get into the 18 - 20W region with 30W claimed for the KR & Vaic Behemoth Valves.
Sayonara
protos said:
Not sure, I have little exposure to them. You get them from KR, Vaic/Emission Labs and EAT (is EAT a continuation of Vaic?) as well as Tianjin (Full Music).
protos said:
Yes, they usally draw more heater current and need more +B Voltage AND current to reach their higher output power. You often need different output transformers as well.
protos said:
Well, you're going to have to ask the "pee-pull" why "they" don't use them. I see little need for more than the 6 - 10W you get from a normal 300B, with high performance speakers such power levels invariably can raise lease breaking levels.
And Electrostatic Speakers really don't mix with SET's (ESL are the only "legit" high performance speakers having low sensitivity) due to their problematic impedance and the fact that they require "voltage" drive for low distortion.
protos said:
Even more. If you really max out the normal 300B (like done in the Electraprint DRD Amp from VTV) you can get around 12W, using any of the 300B's on steroids can get into the 18 - 20W region with 30W claimed for the KR & Vaic Behemoth Valves.
Sayonara
The output TFR realy drive up the cost so that defeats the purpose for me. What I'll do is build the circuit for PSE and I can always leave the Valve out if it is not up to scratch.
I assume that I can share the same heater power supply for these two tubes? I'll just make a cover plate for the second valve and it will be "PSE Ready"
Interesting note on the linearity of the tubes - the "need to share current equally at all levels". Many DACs are using DAC chips in parallel to reduce errors in linearity - in many cases 8 chips in parallel. I would have thought a similar efect with valves, parallel valves should reduce error as the average tends toward the norm.
Thanks for the input.
I assume that I can share the same heater power supply for these two tubes? I'll just make a cover plate for the second valve and it will be "PSE Ready"
Interesting note on the linearity of the tubes - the "need to share current equally at all levels". Many DACs are using DAC chips in parallel to reduce errors in linearity - in many cases 8 chips in parallel. I would have thought a similar efect with valves, parallel valves should reduce error as the average tends toward the norm.
Thanks for the input.
Konnichiwa,
For SE instead of PSE to be optimised you need a different Output transformer. Do one or the other and well, or simply build something else.
Ideally not, but if you must. Beware that the valves need seperate bias adjustments as well.
The reverse is true. DAC's are Current Sources, Valves are current sinks. One valve at any given point in the curve will "hog" the majority of the current sunk, often shifting between valves over the signal cycle.
If you wish to study the sonic effects do it with a cheap linestage and decide if you like the result or not, instead of doing it with an expensive Amp.
Sayonara
DVDHack said:
For SE instead of PSE to be optimised you need a different Output transformer. Do one or the other and well, or simply build something else.
DVDHack said:
Ideally not, but if you must. Beware that the valves need seperate bias adjustments as well.
DVDHack said:
The reverse is true. DAC's are Current Sources, Valves are current sinks. One valve at any given point in the curve will "hog" the majority of the current sunk, often shifting between valves over the signal cycle.
If you wish to study the sonic effects do it with a cheap linestage and decide if you like the result or not, instead of doing it with an expensive Amp.
Sayonara
As Thorsten has pointed out:
Emission Labs have a couple here
Euro Audio Team have a couple here
KR Audio have two here
TJ Full Music has their 300+ Premium here
If you look at the data sheets for all these tubes the filament current will range from 1.5A to over 2A.
The EML 300BLXS is the most reasonable at 1.5A. Which is close enough to standard 300B ops that it can be sub-ed with slight modification. The others require a dedicated designed circuit. The TJ Full Music is nothing more than 2 plates stacked on top of eachother, allowing for twice the power, at twice the filament current.
And as Thorsten remarked, although you can use them at standard 350Vp -72Vg 60mA you can reap more power at higher ops.
I personally use the EML 300BXLS in my amp at 400Vp -60Vg 90mA or thereabouts. Sounds great, maybe 12W.
One thing though is cost. The EML 300BXLS is ~$500US, TJ is $520US, KR is $560US and the EAT is $700US! a pair.
This is probably the main reason you see so few of them in use.
Hello,
The thread as a bit old but I'm discovering it now.
I have a question that bothers me: how can one calculate the load of the plate tube when connected to an interstage transformer?
Some IT transformer are given for 3K, 5K... but does it mean anything? When drawing the load line, what load is to choose. It make the things very different, if I take a 5K or a 1K load line to design my input/driver stage!!
And another question: in those condition: how is it possible to take in account an E810F as a driver for IT as it has a high ra of 42KOhms?
If the IT is done for 3K, how can it works???
Thank you for enlightening me...
Wadek
The thread as a bit old but I'm discovering it now.
I have a question that bothers me: how can one calculate the load of the plate tube when connected to an interstage transformer?
Some IT transformer are given for 3K, 5K... but does it mean anything? When drawing the load line, what load is to choose. It make the things very different, if I take a 5K or a 1K load line to design my input/driver stage!!
And another question: in those condition: how is it possible to take in account an E810F as a driver for IT as it has a high ra of 42KOhms?
If the IT is done for 3K, how can it works???
Thank you for enlightening me...
Wadek
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.