I'm cleaning up the wiring in my "Brute Force" linestage to get rid of some hum (built with 6SN7 instead of 12SN7). In the process I started looking at the heater elevation - I have separate heater windings, one for the 6SN7s and one for the 6BL7s, is it necessary to elevate the heaters for the 6SN7s in this case? Or can I just make a "virtual ground" for this winding with a pair of 180R resistors? Right now all heaters are referenced to 100V.
Attachments
Last edited:
The bottom 6sn7 will live fine with filaments close to zero. A 250-500 ohm pot
with the viper grounded will allow some hum-cancelling adjustments.
The top ones ought to be close to the cathodes of the top 6bl7's
Thanks! That's what I suspected, it will simplify the wiring a little bit. Or at least get rid of one wire 😀
The best result in my experience, is to take a second heater winding in the traffo, and refer it near the cathode voltage, but not couple directly to it, otherwise all traffo intra-capacitances will be in parallel to the output. So, taking from the cathode a 100KΩ resistor, decoupling it with, say, 1µF @ 250V to ground will place the wright DC voltage but no AC signal in the filaments.
I’m just taking B+ through a 60K/27K voltage divider, bypassing the 27K with a 10uF cap, and referencing the heaters through a virtual ground.
Those resistor values will take unnecessary high currents, spending watts and dollars. A 20 or 50 times greater resistances will do.
Showdown, given that you are 'cleaning up the wiring', have you confirmed the hum is related to the heater powering? Eg. have you powered either or both from 6V battery supplies, and are you able to measure the hum adequately (eg. with a spectrum analyser plot) ?
Do you have spare valves to tube-roll and check if hum varies with tube samples? That is likely to indicate if the heater-cathode resistance is a hum contributor, and if so then you may get some advantage by using a 20-40Vdc difference between cathode and elevated heater voltage.
Do you have spare valves to tube-roll and check if hum varies with tube samples? That is likely to indicate if the heater-cathode resistance is a hum contributor, and if so then you may get some advantage by using a 20-40Vdc difference between cathode and elevated heater voltage.
It turns out that most of the hum came from one of the 6SN7s. When I switched them over the hum appeared in the other channel. Also, the heater wiring was too close to the signal input wiring and I guess that contributed a bit. I’ve rerouted the heaters and changed the inputs to shielded cable instead of twisted wires and it got a lot better now. I have another set of 6SN7s I’m going to try out.
For the heater elevation I ended up using a 220k/100k divider from B+, bypassed with a 33uF cap to ground.
For the heater elevation I ended up using a 220k/100k divider from B+, bypassed with a 33uF cap to ground.
If a 6SN7 exhibits hum, compared to an ok 6SN7, then it is most likely related to Rhk being low for that particular valve. Rhk can usually be increased by applying a difference in heater voltage with respect to the cathode of at least 15-20Vdc (ie. elevating the heater with respect to 0V by at least 15-20V for the bottom triodes), so that may allow you to use the not-so-good 6SN7, or provide a little better hum with the good 6SN7's (if you can measure a difference).
Using a humdinger pot may allow any grid related pickup from nearby heaters to be further minimised.
Using a humdinger pot may allow any grid related pickup from nearby heaters to be further minimised.
I had elevated heaters on the 6SN7s before I rewired, and the hum was there. I suspect the tube is just bad. With the other set of 6SN7s it’s pretty quiet but they’re very microphonic. I guess it’s time to look for some tubes...
Also take a look at the input circuit for hum.
I would use screened wire from socket input to amplifier.
Make sure transformer isn't too close to the pcb/valves/circuit or you will pick up stray magnetic fields.
I would use screened wire from socket input to amplifier.
Make sure transformer isn't too close to the pcb/valves/circuit or you will pick up stray magnetic fields.
I got rid of the hum so that's OK, now there's only a little hiss when I turn up to way beyond normal listening levels so I'm pretty happy with that. I'm going to order another pair of tubes since one of the JAN 6SN7s hummed badly and my spare pair of Sovteks are very microphonic, and they also make some low klicking noises for some reason. For now I have the best one of each pair installed.
I should probably test the circuit more thoroughly, but I'm not sure how to go about it... I have an old oscilloscope, but it's been over 25 years since I used one in school so I can't remember much 🙂
I should probably test the circuit more thoroughly, but I'm not sure how to go about it... I have an old oscilloscope, but it's been over 25 years since I used one in school so I can't remember much 🙂
Old tubes are just - old tubes.
A lot of the so called "old stock" is tubes that someone has tried as sparepart
but rejected by some reason and put back in the box, saving for future. Now
a lot of these tubes surfaces and is sold for lots of money :-(
Get brand new tubes from someone that has a decent warranty scheme.
Showdown, how did you come up with those resistor values? 100 ohm in the cathodes and
10k in the anode of the bottom triodes? Also, why the 10k resistor at the ac-output when it
could be put to good use feeding some additional current to the 6BL7 cathodes?
10k in the anode of the bottom triodes? Also, why the 10k resistor at the ac-output when it
could be put to good use feeding some additional current to the 6BL7 cathodes?
Last edited:
I didn't come up with them at all 😀 This isn't my design, I just built it from the schematics I found. It was designed by Eric Barbour, and the article said he just used resistors he had lying around.
ok, lying around explains a lot 😉
I was just wondering why on earth one would want 100ohm in the cathodes, wich would give just 1V Ug at 10mA and squander 100V on the anode when having only 310V Ub.
I was just wondering why on earth one would want 100ohm in the cathodes, wich would give just 1V Ug at 10mA and squander 100V on the anode when having only 310V Ub.
Last edited:
I have a basic understanding of simple tube circuits and can read schematics and solder things together. I haven't read much up on Mu-followers/SRPPs and the like, so if you have any suggestions for modification of this circuit I'm only glad to hear them 🙂
My suggestion:
to be able to feed the heaters of both tubes from the same transformer winding
make sure that the upper cathode stays well below 200Vdc and elevate the transformer winding halfway.
6SN7(per section):
Double Ugk by increasing Rk from 200ohm to 680ohm.
Reduce the Ra voltage-loss and increase Ua/k of both tubes by reducing R4 from 20k to 10k.
6BL7(per section):
Increase Ug by increasing Rk from 200ohm to 330ohm
that should increase Ua/k to at least 120V to be safe regarding h/k voltage.
Increase Ia to 12-14mA by connecting a 22-27k resistor between grid/cathode-resistor junction and neg. supply.
Take the output from this same junction (or from the cathode with a 100-220ohm serial resistor to the cap).
Reduce the waste in the 10kohm resistor connected to the output (increase to 100k or more)
Reduce Rg to 1000k and reduce Cg to 0,047-0,1uF.
Unless you want to drive loads lower than 10kohm 1 section per tube will do.
Last edited:
I'm not completely sure what you mean, I have the same resistor values as in the schematic. Are you talking about values if I use only one section per tube? The sections are paralleled at the socket so there's only one resistor per 2 sections.
I think I'll need the current capabilities of the parallel triodes since I'm driving headphones with this amp.
I've done some measurements now:
I have 430V on the last reservoir cap and B+ is 339V after the regulator.
V1 - heaters referenced to ground:
plate/anode: 120V
cathode: 1.6V
That should mean that each section is drawing 8mA?
V2 - heaters referenced to 100V (should probably be more):
plate/anode: 339V
cathode: 287V
I think I'll need the current capabilities of the parallel triodes since I'm driving headphones with this amp.
I've done some measurements now:
I have 430V on the last reservoir cap and B+ is 339V after the regulator.
V1 - heaters referenced to ground:
plate/anode: 120V
cathode: 1.6V
That should mean that each section is drawing 8mA?
V2 - heaters referenced to 100V (should probably be more):
plate/anode: 339V
cathode: 287V
Last edited:
Ok, my suggestion wont work for you any better than what you allready have (except for h/k voltage stress).
A related, but different circuit would be better.
What is the impedance of your phones?
What is the current limit your ps can provide?
How much voltage amplification do you really need (needed input sensitivity)?
Do you have to use a 6SN7 and 6BL7 or can you choose any tube you want?