This is a bit counterintuitive, but adding a resistor on the top/cold side of the inductor screws up the frequency response in the low end -- at least according to my spice sim. Basically, when the load impedance for the input tube transitions from being resistive to being inductive dominant, there's a hump in the frequency response. I think it's better if the plate load is as close to an ideal inductor as possible. One could add a cap across the resistor (R1 in my schematic below) but that gets us back to having a cap in the signal path. Though, it could be a smaller, hence, better quality cap than that used in the cathode on the 300B in my schematic earlier.
Note: I made the input cap 100 uF so it wouldn't interfere with the LF response of the circuit. In reality, it'll be smaller of course. Or completely eliminated as I do plan to use an input transformer.
~Tom
Note: I made the input cap 100 uF so it wouldn't interfere with the LF response of the circuit. In reality, it'll be smaller of course. Or completely eliminated as I do plan to use an input transformer.
~Tom
Attachments
Last edited:
ChrisB, I actually did not contact them. I picked out some 3k single ended Transcendar transformers that were up on Ebay just before Christmas a year ago wrote down the auction number and gave it to my wife and said I would like this for Christmas. The next thing I knew was Santa Clause delivered them on Christmas day! As far as buy from them directly, I do not know that they do that. I am only familiar with seeing the products on occasion up on Ebay such as a current listing for a pair of 5k SE ended transformers listed under item number 280618272554. You could probably go to this auction and send the lister an email asking about the availability. For the money, they make a pretty darn good transformer. I am not disapointed. Mickeystan
Latest contemplations...
I noticed that in many designs the cathode cap on the 300B goes to the B+ supply. Does anybody know what the advantage of this is compared to letting it go to ground? Better transient response?
In the final version I'll probably use an input tube with a bit higher mu. ECC81/82/83, 6AN4, or D3A come to mind. I just haven't bothered creating a spice model for those yet.
~Tom
I noticed that in many designs the cathode cap on the 300B goes to the B+ supply. Does anybody know what the advantage of this is compared to letting it go to ground? Better transient response?
In the final version I'll probably use an input tube with a bit higher mu. ECC81/82/83, 6AN4, or D3A come to mind. I just haven't bothered creating a spice model for those yet.
~Tom
Attachments
Last edited:
Turns out Lynn Olson has answered my question about the cathode cap to B+: ETF Presentation, Nutshell HiFi
~Tom
~Tom
Tom, The Cathode cap of the 300B returning to B+ rather than ground topology is what is known as "Loftin White" circuit I believe. By doing this it keeps the 300B AC primary (primary amplified current) returning without going back through the power supply picking up extra noise. The incorporation of this along with the DRD in Jack's design is what attracted me to it when I decided to built my amplifiers. Mickeystan
Nope, what you describe is an Ultrapath or Western Electric type connection. Loftin White is a dc coupled SE topology originally depending on a tapped resistor to derive all of the different operating voltages.
See post #8 for a true Loftin-White design. Ignore the bootstrapped follower, if that makes it easier to follow the current paths.
Sheldon
This thread has more explanation: http://www.diyaudio.com/forums/tubes-valves/125488-loftin-white-801-amp.html
Kevin and Sheldon, Thanks for correcting my error. I will take another look at the posts recommendation that Sheldon supplied as I want to better understand my error. Glad you guys corrected me! Mickeystan
No worries. Actually, it's pretty common to see pretty much any direct coupled SE amp called Loftin White. A true Loftin white uses a specific type of ripple canceling. This was novel and important at the time, because high value capacitors were not available at reasonable cost or size. Their approach takes a bit to get your head around, but is quite elegant.
Sheldon
If you can slog through this thread, you will get the full picture: http://www.diyaudio.com/forums/tubes-valves/123211-darius-loftin-white-explained.html
Sheldon
If you can slog through this thread, you will get the full picture: http://www.diyaudio.com/forums/tubes-valves/123211-darius-loftin-white-explained.html
Last edited:
I can haz caps
4 x 47 uF, 630 V metalized polypropylene fresh from AES. /me like... Hockey puck included as a scale indicator. These caps are *BIG*.
I'm settling on the Loftin-White circuit with the mods showed earlier. I'll control the negative B+ (C- ?) for the input tube with a Maida regulator. The supply for the 300B will be about 470 V. This leaves about 400 V anode-cathode on the 300B (there's a 15-ish V drop across the plate choke on the input tube).
Currently I'm using a triode-strapped D3a as the input tube. I like it! But there are other high-mu triodes out there to try.
~Tom
4 x 47 uF, 630 V metalized polypropylene fresh from AES. /me like... Hockey puck included as a scale indicator. These caps are *BIG*.
I'm settling on the Loftin-White circuit with the mods showed earlier. I'll control the negative B+ (C- ?) for the input tube with a Maida regulator. The supply for the 300B will be about 470 V. This leaves about 400 V anode-cathode on the 300B (there's a 15-ish V drop across the plate choke on the input tube).
Currently I'm using a triode-strapped D3a as the input tube. I like it! But there are other high-mu triodes out there to try.
~Tom
Attachments
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.