Felipe, If you use 9V battery, you can operate at about 145V (135 to 155V maybe). The anode current will be about 6 to 8mA. Salas HV shunt will need some extra current, so you can set the current-source to 20mA or more (please check the instructions for the SSHV).
.
Thanks John, I'm glad to hear you like them. Unlike most dc regulators, there's no noisy reference IC anywhere, so no noise is added at all, except the transistor junctions & resistors, which are minimal.
And, so far as the cathode-current is concerned, there's no feedback at all. Meanwhile, all the mains-noise is screened out by the pass-transistors, which even AC-heat can't do!
This allows the regulators to work even with RIAA stage DHTs.
Andy runs the 26 with the filament current (1.05A) running through a 9-10 ohm cathode resistor to provide the bias. It's this low value that allows the omission of the capacitor without degrading the gain.
With a normal cathode bias setup (~1K resistor) the value is too high, and must be bypassed.
The price is the need for a big trafo to power 16V at 1.05A......
With a normal cathode bias setup (~1K resistor) the value is too high, and must be bypassed.
The price is the need for a big trafo to power 16V at 1.05A......
What do you suggest Rod? A std 5A/12V (60VA) trafo should be fine IMO
8A (or 16A) bridge (four discrete TO220s) followed by a couple of 10-15mF/25V caps and a suitable low ohm resistor in between.
What about the mod to your #26 std kit to get 11V out?
Last edited:
Hi Massimo, yes you can try filament bias, when you have the Filament Regulators supplied by me:
2x 10000uF/35v; 1N5822 rectifiers. resistors adjusted to give 16,0V at the regulator input terminals.
The Regulators are modded:
C1= 47uF/35V Panasonic AM (From Farnell. this has been tested.... Nichicon VR should be OK too, from Mouser)
R8 change to 470 (if you have 220 there now); or change to 1,8K (if you have 1K there now,)
The down side to filament bias is a big increase in rectifier pulse current. The trafo/rectifier/caps must be wired with VERY short leads to prevent magnetic coupling from these pulses into your amp. Put them in a shielded box, away from the tubes!
Alternatively, find a big choke 10-22mH or more (at 2A) and use PSUD2 to work out a choke-input supply. This also should be kept away from the tubes, but potentially offers the ultimate in performance.
2x 10000uF/35v; 1N5822 rectifiers. resistors adjusted to give 16,0V at the regulator input terminals.
The Regulators are modded:
C1= 47uF/35V Panasonic AM (From Farnell. this has been tested.... Nichicon VR should be OK too, from Mouser)
R8 change to 470 (if you have 220 there now); or change to 1,8K (if you have 1K there now,)
The down side to filament bias is a big increase in rectifier pulse current. The trafo/rectifier/caps must be wired with VERY short leads to prevent magnetic coupling from these pulses into your amp. Put them in a shielded box, away from the tubes!
Alternatively, find a big choke 10-22mH or more (at 2A) and use PSUD2 to work out a choke-input supply. This also should be kept away from the tubes, but potentially offers the ultimate in performance.
Wow, faster than light, Rod!
R8 is now 220 ohms, thus will become 470 1/4W, C1 from 100/16 to 47/35 pana FM (this should be the right band, not AM I guess
) I have some 47/35 pana FC, but they are 6.3 mm dia. instead of 5.... they might not fit between trim and BJT.
What I have in mind is a two chassis build, thus all trafos and choke will stay far from the tubes, while your boards just under them.
Is 16V at the reg input the very minimum? With 12V AC and dropping something in the Pi filter it will be difficut to get more than 15 V. OTOH a 15V AC trafo means drop nearly 5V with a big power resistor in the Pi filter and a lot of heat.
Or I can slightly reduce the Rk value changing the #26 operating point accordingly....
R8 is now 220 ohms, thus will become 470 1/4W, C1 from 100/16 to 47/35 pana FM (this should be the right band, not AM I guess
) I have some 47/35 pana FC, but they are 6.3 mm dia. instead of 5.... they might not fit between trim and BJT.What I have in mind is a two chassis build, thus all trafos and choke will stay far from the tubes, while your boards just under them.
Is 16V at the reg input the very minimum? With 12V AC and dropping something in the Pi filter it will be difficut to get more than 15 V. OTOH a 15V AC trafo means drop nearly 5V with a big power resistor in the Pi filter and a lot of heat.
Or I can slightly reduce the Rk value changing the #26 operating point accordingly....
Last edited:
Thank you Rod.
I'll stay 5V higher at the reg input.
I'm planning to start with a std cathode bias arrangement and later change to filament bias if I can find the right parts (dale flat resistors, new bigger filament trafos, new caps, etc.) but is always better to know in advance what look for.
I'll stay 5V higher at the reg input.
I'm planning to start with a std cathode bias arrangement and later change to filament bias if I can find the right parts (dale flat resistors, new bigger filament trafos, new caps, etc.) but is always better to know in advance what look for.
This can be benefit also for cathode bias changing R1 660 mohms for a choke of same 660 mohms (I saw Andy uses a choke of 280mH 2.7A but I don't know the DRC value), sorry my ignorance but why higher inductive value is more benetial & wich is maximum inductive value also I wanna know what is limiting the maximum inductive value?
There's 2 places for the choke, when you try Filament Bias:
1. right after the rectifier - make a CHOKE-INPUT supply. This is ultimate solution, because pulse-current in the rectifier is the lowest possible (pulses can transmit across to you tube-circuit). This must be a BIG iron core choke, more than 10mH @2A.
2. Choke between C1 and C2 (See my Filament Reg manual: ANdht01.PDF). This gives extra HF filtering, if your mains is noisy. This can be small ferrite choke (or CMC)
Andy uses 280mH, because he found them! High L gives more filtering, but 20-50mH is probably high enough.
1. right after the rectifier - make a CHOKE-INPUT supply. This is ultimate solution, because pulse-current in the rectifier is the lowest possible (pulses can transmit across to you tube-circuit). This must be a BIG iron core choke, more than 10mH @2A.
2. Choke between C1 and C2 (See my Filament Reg manual: ANdht01.PDF). This gives extra HF filtering, if your mains is noisy. This can be small ferrite choke (or CMC)
Andy uses 280mH, because he found them! High L gives more filtering, but 20-50mH is probably high enough.
Can be good this 3A 2mH 55 mOhms? or have to be the same RDC like the resistor 660 mohms? or just add in series a resistor 600 mOhms 5W?
http://es.rs-online.com/web/p/inductores-para-pcb/4656629/
http://es.rs-online.com/web/p/inductores-para-pcb/4656629/
All,
nice discussion going on and I fully agree to the remarks with regard to chokes. The more inductance the better imho. If you are on a budget you can go with chokes from Hammond. Lundahl has some nice chokes for filament supplies, the LL1694 and LL2733.
Also look for good quality power transformers for the filament. A screen between primary and secondary helps to keep a lot of the noise from the mains out.
This is what I use for power transformers:
VinylSavor: Power Transformers
Best regards
Thomas
nice discussion going on and I fully agree to the remarks with regard to chokes. The more inductance the better imho. If you are on a budget you can go with chokes from Hammond. Lundahl has some nice chokes for filament supplies, the LL1694 and LL2733.
Also look for good quality power transformers for the filament. A screen between primary and secondary helps to keep a lot of the noise from the mains out.
This is what I use for power transformers:
VinylSavor: Power Transformers
Best regards
Thomas
I got the preamp power transformer model from Thomas and can just say that the quality is superb at a very reasonable price. Highly recommended.