Almost every "improvement" causes a problem...
As Frank says, you no longer need to worry about hum due to the temperature (and emission) of the heater tracking the AC waveform. (That's why you chose 2A3 rather than 6B4.)
However. One end of your filament is 2.5V higher voltage than the other. Given that the filament is at 45V, this is only 10% error, but have a look at the anode curves. That 2.5V difference means that one end of the filament emits significantly more current than the other, reducing the life of the valve.
Perhaps you could arrange a big relay to swap the polarity applied to the filament each time the amplifier is switched on.
As Frank says, you no longer need to worry about hum due to the temperature (and emission) of the heater tracking the AC waveform. (That's why you chose 2A3 rather than 6B4.)
However. One end of your filament is 2.5V higher voltage than the other. Given that the filament is at 45V, this is only 10% error, but have a look at the anode curves. That 2.5V difference means that one end of the filament emits significantly more current than the other, reducing the life of the valve.
Perhaps you could arrange a big relay to swap the polarity applied to the filament each time the amplifier is switched on.
Hi,
I've other, cheaper tricks up my sleeve to tackle that...
Cheers,
Perhaps you could arrange a big relay to swap the polarity applied to the filament each time the amplifier is switched on.
I've other, cheaper tricks up my sleeve to tackle that...
Cheers,
Or I could wire the two sockets differently and swap tubes every now and then. Getting rid of the hum pots will help, because that's a little more real estate, plus I can use those pots for some of the regulator trimming, especially since my hum pots are 2W.
Grataku, my amp is the JE Labs/Angela 2A3 SET (the one with the SRPP 6SL7 driver stage). When you say "300bxls SE", that's a modified version of a JE Labs design too, right? I built the amp pretty much according to the original design (though I changed the power supply), and now I'm trying out some changes. The regulated DC filaments are first, and I might try direct coupling the driver stage to the 2A3s as well (haven't decided on that one yet, since I'll need a capacitor somewhere in my signal path to act as a high pass filter, so maybe I'll make the coupling cap do the filtering as well). I'm also getting better OPTs than the Hammond 125ESEs I built the amp with.
Grataku, my amp is the JE Labs/Angela 2A3 SET (the one with the SRPP 6SL7 driver stage). When you say "300bxls SE", that's a modified version of a JE Labs design too, right? I built the amp pretty much according to the original design (though I changed the power supply), and now I'm trying out some changes. The regulated DC filaments are first, and I might try direct coupling the driver stage to the 2A3s as well (haven't decided on that one yet, since I'll need a capacitor somewhere in my signal path to act as a high pass filter, so maybe I'll make the coupling cap do the filtering as well). I'm also getting better OPTs than the Hammond 125ESEs I built the amp with.
fdegrove said:I've other, cheaper tricks up my sleeve to tackle that...
Hello Frank, as you well know, the Imperial English plundered the world in search of good words (and anything else valuable), so you will undoubtedly understand our use of the word, "coquette".
Saurav said:Or I could wire the two sockets differently and swap tubes every now and then.
You could indeed, and at far less expense.
Hi,
As any well educated Belgian will tell you cockilly...English and French are historically closely related.
An awful lot of posh English expressions and words are etymologically very close to their French origin, no big surprise when you know your country's history.
Very interesting to know, no doubt...
Frank is being cocky, which isn't quit the the same as "cocquet"
but close nonetheless...
Ciao,
Hello Frank, as you well know, the Imperial English plundered the world in search of good words (and anything else valuable), so you will undoubtedly understand our use of the word, "coquette".
As any well educated Belgian will tell you cockilly...English and French are historically closely related.
An awful lot of posh English expressions and words are etymologically very close to their French origin, no big surprise when you know your country's history.
Very interesting to know, no doubt...
Frank is being cocky, which isn't quit the the same as "cocquet"
but close nonetheless...
Ciao,
Frank,
can you please pull that ace out of your sleeve and show us how it's done? I was initially thinking to make a dual +/- 2.5V PS to overcome the uneven heating but I couldn't get past that.
Saurav,
I am familiar with the angela schematics, what I am doing is a mix two of Ciuffoli's designs with the interstage LL1660 trafo and a LL1623/120mA output with the KR 300bxls.
can you please pull that ace out of your sleeve and show us how it's done? I was initially thinking to make a dual +/- 2.5V PS to overcome the uneven heating but I couldn't get past that.
Saurav,
I am familiar with the angela schematics, what I am doing is a mix two of Ciuffoli's designs with the interstage LL1660 trafo and a LL1623/120mA output with the KR 300bxls.
Hi,
Unfortunately that doesn't help.
When I said I had another cheap trick up my sleeve, Saurav must have read my mind.
He posted the idea in # 43.
Cheers,
I was initially thinking to make a dual +/- 2.5V PS to overcome the uneven heating but I couldn't get past that.
Unfortunately that doesn't help.
When I said I had another cheap trick up my sleeve, Saurav must have read my mind.
He posted the idea in # 43.
Cheers,
Saurav
You mean these links?
http://www.audiodesignguide.com/my/300bxls.html
http://www.audiodesignguide.com/my/pse2.html
http://www.audiodesignguide.com/my/se1.html
How are you going to wire it?
You mean these links?
http://www.audiodesignguide.com/my/300bxls.html
http://www.audiodesignguide.com/my/pse2.html
http://www.audiodesignguide.com/my/se1.html
How are you going to wire it?
I didn't understand what you meant when you said "Ciuffoli has a way of running the HV to ground using the transformer as a voltage divider."
> How are you going to wire it?
I think I'll wire the DC filament supply direct to the filament pins, and move the cathode bias resistor to one of the filament pins (probably the end that's 2.5V below the other end). So, no trimpot across the filament. I'll re-use the trimpot for the current regulator. And I'll wire the two sockets differently, so that I can swap tubes around as discussed above.
> How are you going to wire it?
I think I'll wire the DC filament supply direct to the filament pins, and move the cathode bias resistor to one of the filament pins (probably the end that's 2.5V below the other end). So, no trimpot across the filament. I'll re-use the trimpot for the current regulator. And I'll wire the two sockets differently, so that I can swap tubes around as discussed above.
I think a simple switch on the top panel of your amp to reverse the filament is the simplest idea. Every time you turn it on, you swap the switch position (ideally just before). Just get into the habit of doing it, and is preferable to me than unplugging the tubes all the time which I don't think will help reliability.
A decent switch shouldn't cost ant more than another socket.
A decent switch shouldn't cost ant more than another socket.
Dang it. That's a better idea than mine
How does one hook up a switch to reverse polarity? I guess a 4-pole dual-throw would do it, right? That gives me 2 poles per channel, wire up + and - from the regulator board to one set of switch contacts, reverse + and - to the other set, and the switch's output goes to the tube socket? That should do it. Now all I need is a 4PDT switch that can handle 2.5A.
Thanks.
How does one hook up a switch to reverse polarity? I guess a 4-pole dual-throw would do it, right? That gives me 2 poles per channel, wire up + and - from the regulator board to one set of switch contacts, reverse + and - to the other set, and the switch's output goes to the tube socket? That should do it. Now all I need is a 4PDT switch that can handle 2.5A.
Thanks.
Saurav,
a switch is so uncool... why not use a flip-flop relay for a polarity switch at every turn on? It easy to do.
About the ciuffoli design... in one case the filament winding is center tapped and the high voltage bias is set between the center tap and ground via the usual parallel RC network.
a switch is so uncool... why not use a flip-flop relay for a polarity switch at every turn on? It easy to do.
About the ciuffoli design... in one case the filament winding is center tapped and the high voltage bias is set between the center tap and ground via the usual parallel RC network.
Hi,
That's probably done because otherwise the cathode to heater insulation voltage would be exceeded.
Indirectly heated tubes usually have a stated insulation voltage in the 100 to 150 V range, by injecting a portion of the B+ into the heater this potential difference is lowered.
What you saw probably looked like this:
Cheers,
About the ciuffoli design... in one case the filament winding is center tapped and the high voltage bias is set between the center tap and ground via the usual parallel RC network.
That's probably done because otherwise the cathode to heater insulation voltage would be exceeded.
Indirectly heated tubes usually have a stated insulation voltage in the 100 to 150 V range, by injecting a portion of the B+ into the heater this potential difference is lowered.
What you saw probably looked like this:
Cheers,
Attachments
a switch is so uncool
Agreed, it certainly doesn't have the elegance of a fully automated solution
Saurav said:
Agreed, it certainly doesn't have the elegance of a fully automated solution
It has the practicality of reliability, and because of it's simplicity is extremely elegant. Switch one way odd days of the month, the other even.
Beware finding solutions that are more complex than they need to be.
grataku said:What's so unreliable about a flip-flop relay?
It's the same number of components as a relay and a switch.
With kids running around asking 'what does this do, daddy?' I would never know what position it was on last.
A switch is one component.
As for position, I said in the previous post "Switch one way odd days of the month, the other even."
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
- Home
- Amplifiers
- Tubes / Valves
- PSUD modelling of a filament supply