It sounds like you're onto the clipping problem with the OPT primary impedance. This could impact the soundstage also I guess.
While you're in the investigative mode, I think it would be interesting to try LED or other diode bias on the 5842, replacing the resistor and bypass cap. It would be interesting to see in particular what impact this has on the soundstage definition.
Cheers,
Michael
While you're in the investigative mode, I think it would be interesting to try LED or other diode bias on the 5842, replacing the resistor and bypass cap. It would be interesting to see in particular what impact this has on the soundstage definition.
Cheers,
Michael
2.3 kOhm sounds a tad on the low side for a 300B. I think most people use about 3k for higher power (at the expense of higher THD) and 5k for lower THD (at the expense of power).
I use transformers with a primary that measures out to 4.2 kOhm, 375 V B+, 60 mA cathode current. I get about 6.6 W @ 3 % THD+N (midband).
I'd try running off of the 16-ohm winding. I seem to recall that you tried that already. Did that improve the sound stage?
I notice that you're using 7 kOhm as the plate load on the 5842. That sounds like a rather heavy load... I suggest running the 5842 'cooler' (I run mine at 8~12 mA). If you want any hope of getting any appreciable gain out of the 5842, I suggest using a current source (IXYS 10M45 is a popular one) or gyrator.
If you don't like sand in your glass, you could also dial the current way down (a few mA) and increase the plate load.
From what I've heard (I read it on the internet so it must be true!), the perception of sound stage is mostly related to the power response of the amp. I.e. does the output amplitude increase linearly when the input amplitude is increased linearly. My guess is that your input stage runs out of gas rather early.
~Tom
Last edited:
TubeLab has no grid stopper resistors at the 5842 in his TubelabSE design. Are there any drawbacks using grid stoppers?
I use a Lundahl LL1623 as output transformer in my TubelabSE amplifier. The LL1623 is strapped for 6 kohm primary load impedance. The 300b runs att B+ 425 V 80mA and the 5842 at 180V 10 mA. I then get 8 W over 16 ohm before clipping. I really like the Lundahl OT, I think they sounds great
I have the same oppinion on the Lundahl transformers. I used quite a few interstages and some output transformers and I think they are by far the best quality for the price, and sometimes better than some "top-of-the-range trafos independently of the price. Thay are also very flexible, and you can try various versions with the same trafo. Shame the casing is not very DIY friendly.
I replied but I have a message 'delivery...bla, bla,failed...'
This is my e-mail: jp.*****@****.com
A bientôt
Has been seen, now removed. Salas.
Hi Jeepy ,
I see another item in the schematic which could be tweaked :
the filament transformer for the 300B is in the signal-path , I've the experience that using two low-value resistors in series over the filament and connect the cathode-resistor parallel with decoupling capacitor to the connection of both resistors is sonically better. This removes the filament transformer and it's noise from the signal-path.
Another small capacitor ( 1 to 2,2 uF ) from top of OPT to pos. side of the decoupling capacitor will too enhance resolution.
grtz
Bouquet
I see another item in the schematic which could be tweaked :
the filament transformer for the 300B is in the signal-path , I've the experience that using two low-value resistors in series over the filament and connect the cathode-resistor parallel with decoupling capacitor to the connection of both resistors is sonically better. This removes the filament transformer and it's noise from the signal-path.
Another small capacitor ( 1 to 2,2 uF ) from top of OPT to pos. side of the decoupling capacitor will too enhance resolution.
grtz
Bouquet
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.