Hi!
Could someone explain what the term headroom exactly refers to when talking about a tube driver stage and how much is good?
From what I've understood it has something to do with the voltage swing the driver can produce... but in db's...
or maybe not 
When designing a two-stage se amp aimed to work without a linestage, the minimum required voltage swing is the amount needed to drive the output stage to full output. Is there also a maximum that shouldn't be exceeded for sonic reasons (rule of thumb?)? I've read that too much gain increases distortion + makes the amp sound nervous...
While we're at it: any recommendations for a very good sounding driver tube that can swing atleast 70V peak/peak to drive a 7236 single-ended? RC-coupled to keep things affordable. I was thinking of 6c45 or 5842 with a Gary Pimm CSS (would be my first ccs); for a 10-20mA drawing driver/channel I have about 200V ps voltage available... that rules out a lot of tubes that like higher voltages.
Thanks!!😀
Could someone explain what the term headroom exactly refers to when talking about a tube driver stage and how much is good?
From what I've understood it has something to do with the voltage swing the driver can produce... but in db's...
or maybe not When designing a two-stage se amp aimed to work without a linestage, the minimum required voltage swing is the amount needed to drive the output stage to full output. Is there also a maximum that shouldn't be exceeded for sonic reasons (rule of thumb?)? I've read that too much gain increases distortion + makes the amp sound nervous...
While we're at it: any recommendations for a very good sounding driver tube that can swing atleast 70V peak/peak to drive a 7236 single-ended? RC-coupled to keep things affordable. I was thinking of 6c45 or 5842 with a Gary Pimm CSS (would be my first ccs); for a 10-20mA drawing driver/channel I have about 200V ps voltage available... that rules out a lot of tubes that like higher voltages.
Thanks!!😀
Suppose you have an output stage that requires 10VRMS to drive it to clipping but your driver stage is capable of supplying 20VRMS before it clips. Your driver is capable of supplying twice as much signal as is necessary, and that's headroom. In dB (which makes it easier to compare), it's 20log(20/10) = 6dB.
Headroom just means that the driver can supply more than necessary, not that it will. Gain is a different matter entirely.
With only 200V of HT available, you ought to think seriously about a choke or transformer load in your driver. Your CCS idea might be a bit marginal.
Headroom just means that the driver can supply more than necessary, not that it will. Gain is a different matter entirely.
With only 200V of HT available, you ought to think seriously about a choke or transformer load in your driver. Your CCS idea might be a bit marginal.
Hi Klimon,
following up to EC8010´s explanations, you also should be aware that headroom (swing reserves) will be eroded in circuits using NFB of any kind.
Tom
following up to EC8010´s explanations, you also should be aware that headroom (swing reserves) will be eroded in circuits using NFB of any kind.
Tom
yup, I have a 5842 with diyAudio CCS and 200V won't be enough... I have another 5842 choke loaded from 220V B+ to drive my 45.
I believe this is perfectly true as EC8010 intended it. To take it further...EC8010 said:
What is necessary? I believe that SET amps are often overdriven. This is not usually a problem, the distortion from a SET output stage is fairly benign in moderation.
The problem begins when the driver stage distorts too, apart from it being unable to supply output stage grid current. Now the distortion is much less benign, putting a kind of upper limit on how much we can really push the amp.
I built a version of this amp using the 7236 instead of the 5998. The 7236 is a not well-known tube with very low distortion. Admittedly, the driver does run out of gas before the output tube does. I would follow the others recommendations and use a choke- or transformer loaded tube like a D3a, C3g, or E810F triode-connected. You will NOT be unhappy with the 7236 as an output tube.
John
http://www.diyaudio.com/forums/showthread.php?postid=348606#post348606
http://www.diyaudio.com/forums/showthread.php?postid=464286#post464286
John
http://www.diyaudio.com/forums/showthread.php?postid=348606#post348606
http://www.diyaudio.com/forums/showthread.php?postid=464286#post464286
Thanks for all the replies 
From what I understand the main target of having extra headroom is that the output tube clips before the driver, giving a more benign clipping behaviour...
About the planned amp:
I've looked into the circuit again and could go up to 70mA, 220V biased at -40V (with high 4k load using the opts I have = a small 3W output power) which would leave around 220 volts for the driver. That's exactly what arnoldc uses with his plate chokes 😎... I think I'll leave the CCS and go the iron way --- got some 5842, c3g/m and 6c45 (mil spec but possibly unmatched) that can be auditioned for this amp.
Cheers,
SImon
From what I understand the main target of having extra headroom is that the output tube clips before the driver, giving a more benign clipping behaviour...
About the planned amp:
I've looked into the circuit again and could go up to 70mA, 220V biased at -40V (with high 4k load using the opts I have = a small 3W output power) which would leave around 220 volts for the driver. That's exactly what arnoldc uses with his plate chokes 😎... I think I'll leave the CCS and go the iron way --- got some 5842, c3g/m and 6c45 (mil spec but possibly unmatched) that can be auditioned for this amp.
I didn't think it was a bad tube but this doesn't hurt as motivation (certainly when people start using capitals to point out how GOOD something is😀😀)! I breadboarded a 6080 se before and was particularly amazed by it's low-end, the driver stage was too compromised to give a real evaluation though... if other ears heard right the 7236 and 5998 should be another step up the ladder😉
Cheers,
SImon
- Status
- Not open for further replies.