Euro21, I have an 801 in the queue. What is you catjode resistor and Vref for the gyrator? I am using j310 in lower spot.
A bit Offtopic, but the environment is similar, than 4P1L preamp, except higher B+, and higher filament current.
p.s.: I also have quite a few 4P1L (for some years ago), but apart from measurements they are waiting for use them.
Attachments
Hi Bela,
I use the 2SK170GR for lower currents. There are BL (IDSS=12mA) and V (IDSS=20mA) versions, if you can find them.
The BF862 can do 20mA and is a fantastic device still in production, albeit is SMD. I have a pair of PCB of my gyrator as prototype for testing. I can send you a pair to try them with some 2SK170. Ping me a PM.
Looking at your diagram, you don't need C3 to be 1uF in my view. With a 100nF cap you have a low enough LF pole.
Cheers
Ale
Hi Ale,
I made similar cGyrators, one with DN2540 (TO92), the other with J310 as "lower" device. The upper is IXTP01N100D.
The first is capable up to -about- 100mA, second is can only 37mA when saturated (J310 is the limiting factor).
I love using my 10/10Y/801/801a tubes about 30mA.
I adjusted both with same settings in my test (breadboard) system:
B+ is 400V (270VA R-core, Tungsram PV 600/200 rectifier, cLCLC, Tomchr's V 1.0 regulator).
Sylvania 10 tube, R.C. regulated, 10R filament biased to -12.3V, 285-286V plate voltage, 29mA plate current.
Measuring from "Low Z" point (my AC voltmeter has 10MOhm imp.), reference is output at 1kHz (0dB).
DN2540 gyrator J310-DN2540 gyrator
100kHz -0.3dB -0.5dB
-1dB point 240kHz 170kHz
-3dB point 550kHz 340kHz
The THD in both cases very low, 0.047/0.043 % (1V input from Victors 1kHz generator).
With J310 circuit the "upper" FET dissipation in a bit larger (about 3W), so heatsink (Fischer SK68 50mm: 5K/W) is lukewarm (about 40 C).
I made similar cGyrators, one with DN2540 (TO92), the other with J310 as "lower" device. The upper is IXTP01N100D.
The first is capable up to -about- 100mA, second is can only 37mA when saturated (J310 is the limiting factor).
I love using my 10/10Y/801/801a tubes about 30mA.
I adjusted both with same settings in my test (breadboard) system:
B+ is 400V (270VA R-core, Tungsram PV 600/200 rectifier, cLCLC, Tomchr's V 1.0 regulator).
Sylvania 10 tube, R.C. regulated, 10R filament biased to -12.3V, 285-286V plate voltage, 29mA plate current.
Measuring from "Low Z" point (my AC voltmeter has 10MOhm imp.), reference is output at 1kHz (0dB).
DN2540 gyrator J310-DN2540 gyrator
100kHz -0.3dB -0.5dB
-1dB point 240kHz 170kHz
-3dB point 550kHz 340kHz
The THD in both cases very low, 0.047/0.043 % (1V input from Victors 1kHz generator).
With J310 circuit the "upper" FET dissipation in a bit larger (about 3W), so heatsink (Fischer SK68 50mm: 5K/W) is lukewarm (about 40 C).
Last edited:
I just upgraded my filament supply to my 4P1L line stage. I use a 150v voltage reg. tube feeding a Hammond 126B as plate choke, and at around 25mA for the 4P1L that gives a cathode resistor of 13 ohms and a bias of 8.5v. Add 2v for the filament and you have 10.5v, and add 6.5v for Rod's reg (Ver 2) and you get 17v supply.
Previously I achieved that with an 18v toroid in cap input followed by a Hammond 159ZC 60mH choke. The new supply is a 25v EI transformer and a 280mH choke in choke input, LCRC. With a 6.8R dropper resistor this gives me the required 17v.
Sound is clearer and more detailed - an audible improvement. So the supply does matter. Thomas Mayer is a big believer in chokes in the filament supply, and I can only agree with him - it's worth it. For a single 4P1L at 640mA this is overkill - a 60VA transformer and a 2.7A choke. But hey, it works.....
Photo shows the supply for 2 channels - the 2 big chokes do actually fit in! The connector is a 4-pole XLR.
Previously I achieved that with an 18v toroid in cap input followed by a Hammond 159ZC 60mH choke. The new supply is a 25v EI transformer and a 280mH choke in choke input, LCRC. With a 6.8R dropper resistor this gives me the required 17v.
Sound is clearer and more detailed - an audible improvement. So the supply does matter. Thomas Mayer is a big believer in chokes in the filament supply, and I can only agree with him - it's worth it. For a single 4P1L at 640mA this is overkill - a 60VA transformer and a 2.7A choke. But hey, it works.....
Photo shows the supply for 2 channels - the 2 big chokes do actually fit in! The connector is a 4-pole XLR.
Attachments
Last edited:
I am using 80VA EI core mains for filament supply!!! I agree it is essential to quality of the sound! Also filtering is best whit inductor...
Hi Andy,
Yes, with Filament Bias, there is even greater sensitivity to noise leaking in from the mains supply.
The toroidal trafo has as much as 1nF leakage capacitance from primary to secondary. In contrast, all the EIs I could find measured 50-70pF - a major improvement.
Any leakage-current entering the dc supply through the trafo appears as common-mode noise to the raw dc +/- nodes (because the raw dc must float, and not be connected to ground). This common-mode noise returns to ground via the filament bias resistor - so it is directly mixed with the music signal.
If you only have a toroidal to use, it may be worth trying a common-mode choke (in the mH range) in the raw dc supply, and a 220nF stacked-film cap from each raw dc node (+ and -) to the anode supply's supply negative (at the filter cap -, via short wires). This may bypass and absorb some HF noise.
But it is usually better not to let the noise enter in the first place.
The problem only gets worse when your neighbour installs a Solar PV up-converter.
Yes, with Filament Bias, there is even greater sensitivity to noise leaking in from the mains supply.
The toroidal trafo has as much as 1nF leakage capacitance from primary to secondary. In contrast, all the EIs I could find measured 50-70pF - a major improvement.
Any leakage-current entering the dc supply through the trafo appears as common-mode noise to the raw dc +/- nodes (because the raw dc must float, and not be connected to ground). This common-mode noise returns to ground via the filament bias resistor - so it is directly mixed with the music signal.
If you only have a toroidal to use, it may be worth trying a common-mode choke (in the mH range) in the raw dc supply, and a 220nF stacked-film cap from each raw dc node (+ and -) to the anode supply's supply negative (at the filter cap -, via short wires). This may bypass and absorb some HF noise.
But it is usually better not to let the noise enter in the first place.
The problem only gets worse when your neighbour installs a Solar PV up-converter.
Given that the system ground is only in the main chassis, and my filament supply is floating via 4-pole XLRs (no ground), would the common mode choke then be in the main chassis, as shown? Should there be a capacitor both before and after the CMC? I usually use 1,000uF in the chassis itself.
Attachments
Hi Andy, In this case, yes, place the CMC in the signal chassis, but keep it away from signal wiring. The toroidal ferrite-core CMC will be best here.
Could try caps before the CMC, but they might make it worse, by causing current to flow in the supply wires. (When caps are connected after the CMC, the HF current is somewhat dissipated in the losses of the CMC).
This kind of CMC might be a good starting point:
B82721A2152N001 - EPCOS - CHOKE, RING CORE, DOUBLE, 3.3MH, 1.5A | Farnell element14
B82721A2152N001 - EPCOS - CHOKE, RING CORE, DOUBLE, 3.3MH, 1.5A | Farnell element14
With the L value so much higher, please check the dc resistance, as it may drop too much voltage.
Otherwise, so long as it is really a common-mode choke, it will be OK.
Hello
Hope someone can answer this.
with very low volume,when the volume pot is nearly turn down,,the music is a bit distorted
only a bit more up with the volume ,,all is right- (not a problem but like to know why)
What can that be ?
I have before talk about a high freqent noise (around 5 to 7khz),,,the same noise on 2 different 4P1L preamp
maybe i have found out why,,,when i use a filter on my main 230v net ,,it remove a good deal off the noise....so i think the 230v net is a bit unclean..next thing to try must be a regenerator for the main net
Best Bjarne
Hope someone can answer this.
with very low volume,when the volume pot is nearly turn down,,the music is a bit distorted
only a bit more up with the volume ,,all is right- (not a problem but like to know why)
What can that be ?
I have before talk about a high freqent noise (around 5 to 7khz),,,the same noise on 2 different 4P1L preamp
maybe i have found out why,,,when i use a filter on my main 230v net ,,it remove a good deal off the noise....so i think the 230v net is a bit unclean..next thing to try must be a regenerator for the main net
Best Bjarne
Some things to check:
Please check the resistance value of the volume poti's track - and compare to the input resistance of the preamp.
If you have a 100K Pot and a 100K grid leak resistor, then the Pot may be loaded too much by the input - at low levels.
Or, maybe the pot needs cleaning.
Or, maybe the 4P1L has some gas - this will cause grid current to flow, and if the grid is dc-coupled it may leak into the Pot, and cause noise & distortion.
Using industrial mains filters can cause problems for B+ supplies. If you have noise on the B+, we will need to see the full schematic, including parts types and values - and we may be able to help, and improve it!
Hello Rod
On one pre it is 100k pot and 68k leak
the other 15k pot ang 86k leak
they both have the fun with the low volume
i do not think it is a noisy B+....because the noise is gone when the pre comes to another town---and sometime during the day it is totally gone
my pre is build with the B+ in this way
from rectifire to a 470uf ..then a resistor ..then a 470uf..then a resistor..then a 470uf
no choke
then to Gyrator
no hum only those high freqence noise...as i said lesser with the filter--
why kan the filter give problem with the B+ ??there are no loss in 230v on output filter
what is the optimal value off pot and grid leak resistor ??
Best Bjarne
On one pre it is 100k pot and 68k leak
the other 15k pot ang 86k leak
they both have the fun with the low volume
i do not think it is a noisy B+....because the noise is gone when the pre comes to another town---and sometime during the day it is totally gone
my pre is build with the B+ in this way
from rectifire to a 470uf ..then a resistor ..then a 470uf..then a resistor..then a 470uf
no choke
then to Gyrator
no hum only those high freqence noise...as i said lesser with the filter--
why kan the filter give problem with the B+ ??there are no loss in 230v on output filter
what is the optimal value off pot and grid leak resistor ??
Best Bjarne
hello Bjarne,
If you have filament bias, you can use a grid leak value somewhere near to the 500K allowed in the specification. Maybe 330KΩ to allow for the gas (grid leakage) to be expected from 40 year old tubes.
Maybe you neighbour has some Wind-power or solar power feed scheme, making this noise on your mains supply?
470µF is a big capacitor, and the rectifier current-pulses will be large. The pulses can cause trouble with the inductance of any primary-side filter (depends on its type and design).
If you have noisy mains, a shunt regulator for the dc will probably help a lot (I should be able to help with these, soon).
If you have filament bias, you can use a grid leak value somewhere near to the 500K allowed in the specification. Maybe 330KΩ to allow for the gas (grid leakage) to be expected from 40 year old tubes.
Maybe you neighbour has some Wind-power or solar power feed scheme, making this noise on your mains supply?
470µF is a big capacitor, and the rectifier current-pulses will be large. The pulses can cause trouble with the inductance of any primary-side filter (depends on its type and design).
If you have noisy mains, a shunt regulator for the dc will probably help a lot (I should be able to help with these, soon).
Hello Rod
My neighbour is big farms...lots of freqence regulater..
i could try my Salas shunt regulater...to see if this help...but maybe my B+ will be at bit low to the gyrator
my trafo give max 186 v ac out
my main filter is a big 20amp version,,,made for RFI
will try the bigger grid leak...must hear if it give negative effekt to the sound in the pre
will you say that it is better to use smaller value HV cap--
i can use film caps,,size op to 47uf ..3.4.5 pieces
or 220uf 400v caps
or my choke 5,6 mh 8,7ohm
Best Bjarne
My neighbour is big farms...lots of freqence regulater..
i could try my Salas shunt regulater...to see if this help...but maybe my B+ will be at bit low to the gyrator
my trafo give max 186 v ac out
my main filter is a big 20amp version,,,made for RFI
will try the bigger grid leak...must hear if it give negative effekt to the sound in the pre
will you say that it is better to use smaller value HV cap--
i can use film caps,,size op to 47uf ..3.4.5 pieces
or 220uf 400v caps
or my choke 5,6 mh 8,7ohm
Best Bjarne
Hi Kevin.
Thanks for your tips with the IXYS IXTP08N50D2 as a replacement for DN2540 upper.
I just had solvered them into my Gyrator, and christ they are good.
For me it seems that they have a lower noise level, and a much cleaner sound.
Now it's time to check my favorite stuff once more.
Thanks Michael.
Thanks for your tips with the IXYS IXTP08N50D2 as a replacement for DN2540 upper.
I just had solvered them into my Gyrator, and christ they are good.
For me it seems that they have a lower noise level, and a much cleaner sound.
Now it's time to check my favorite stuff once more.
Thanks Michael.
Indeed they are. Hope that it will result in a more involving music experience...
Hello
just made some update for my 4P1L pre
first thing was New socket from K&K Audio,,they are cool,,togetter with some other things i have made..nearly no microfonic...absolute no problem anymore
today i get my Ver 7 from Rod Coleman
They give lower noise,,easy to get the DC right,,,and a more clear sound,,more easy to hear what happened in the music
all in all ,,this 4P1L pre goes better and better
for output cap i use a 4,7uf Mundorf silver in oil,,,on top of this i have a Rike Qcaps 0.47uf
those to caps playes super great together,,,where the very great sound from the Qcaps gives nearly all of the middle and top in the music,,,
it is nessesary for me to use a big caps ,,because my amp have only 10k ohm input impedance
next thing will be gyrator with Ixus
Best bjarne
just made some update for my 4P1L pre
first thing was New socket from K&K Audio,,they are cool,,togetter with some other things i have made..nearly no microfonic...absolute no problem anymore
today i get my Ver 7 from Rod Coleman
They give lower noise,,easy to get the DC right,,,and a more clear sound,,more easy to hear what happened in the music
all in all ,,this 4P1L pre goes better and better
for output cap i use a 4,7uf Mundorf silver in oil,,,on top of this i have a Rike Qcaps 0.47uf
those to caps playes super great together,,,where the very great sound from the Qcaps gives nearly all of the middle and top in the music,,,
it is nessesary for me to use a big caps ,,because my amp have only 10k ohm input impedance
next thing will be gyrator with Ixus
Best bjarne