BHD: I totally agree with you, and I do have homebrew tube mono-block amps with OPTs and preamp for my listening pleasure. OTLs can sound spectacular; all I am saying is that they are quite inefficient, since driving a low impedance load from a high impedance source just clobbers efficient power transfer.
Well for my part the OTL has so much magic that ineffiency dosent matter. Use PWM (class D) amp if that is a concern...IMHO
This is the most power hungry beast I have ever constructed, but also the best sounding, so it remains in the chain...
Beautifull work!
I have a pair of Amphion Kripton 2 that are a hard load on bass region for any amplifier, but my DIY Atma sphere monos can control the Amphions like no other amplifeir i tried before, tube or transistor, and i don't even use any feedback on these amps. I never expected these level of performance with these speakers. I made these amps for my electrostatics.
Excatly!! Contrary to common belifies these amps are champs in the bass region. No feedback in mine either
I am curious about this statement "the pre-conditioning must be...etc" as well as "there must be sufficient cathode current limiting resistors for each cathode". Could you explain these statements, please?
I have also attached a schematic for a 7241 as v2 where the author has put in resistors in the 3 filaments. Does it make sense to you?
Attachments
Last edited:
It does. The Wattage of the resistors is too low! Essentially we are talking about 10Watt 2.5 Ohm resistors required for each cathode. If they are no present a cathode can hog current, causing premature failure. That is why Tung Sol designed this tube with three cathode connections. If you look at the tube you'll see its actually three sections from the 6336 dual power triode.
'Preconditioning' is the act of lighting up the filament on a previously unused (new, or NOS) tube for a predetermined amount of time. This is known to improve the life of the tube (perhaps doubling it) and reduces arcing due to flaking of the cathode coating. This works on any power tube. When B+ is applied the preconditioning process is permanently ended.
'Preconditioning' is the act of lighting up the filament on a previously unused (new, or NOS) tube for a predetermined amount of time. This is known to improve the life of the tube (perhaps doubling it) and reduces arcing due to flaking of the cathode coating. This works on any power tube. When B+ is applied the preconditioning process is permanently ended.
Thank you for your quick response. So you recommend using 10w 2.5 ohm resistors instead. Thanks. Very, very helpful
One last question: what difference in the functioning of the tube would it make to use the values as originally designed? It seems to me that with 3 cathodes in the tube and 6 resistors spec'd (2 each of 51 ohms in parallel), there would be 25 ohms for each cathode. Or was it a typo since 25 looks a lot like 2.5 ohms? What difference would that difference make?
The original drawing says that too: two 5 Ohm resistors in parallel (hence 6 resistors instead of 3) is 2.5 Ohms.
But now that you point that out, I had assumed the '51 Ohms' was really '5.1 Ohms'.
But that is in the context of a Circlotron output. If its used in a different way, then 25 Ohms would be OK, but not if only 2 Watts! Is there more to the circuit than just the drawing you showed?
Yes, I saw that discrepancy and I wanted clarification. In fact, I verified it with the original tube data sheet and it too recommends that the cathode resistor should be around 25 ohms, so I figured that someone misread the schematic.
More to the point, you helped me identify the fact that there are 3 cathodes and that they should have a higher wattage of resistor on them. So thanks for that.
That being said, yes, there is more to the schematic but I have confidence in its accuracy. So again, thanks for your help so far.
More to the point, you helped me identify the fact that there are 3 cathodes and that they should have a higher wattage of resistor on them. So thanks for that.
That being said, yes, there is more to the schematic but I have confidence in its accuracy. So again, thanks for your help so far.
The question of Wattage is dependent upon the current being drawn: if the tube is drawing 350 mA, you'll be dropping 2.975 v across these 6 paralleled resistors, which implies 1.04125 Watts.
Six 1 Watt resistors should be more than adequate in this scenario.
Six 1 Watt resistors should be more than adequate in this scenario.
What is a good amount of time? I just bought a few 5C8S rectifiers for a possible new build.
Rectifiers are a different beast- not tried it with them but if they are indirectly heated its worth a try.
Member
Joined 2009
Paid Member
Phase shift is a function of bandwidth. Its filter theory. A 6dB rolloff will cause phase shift to 10th or 10x the cutoff frequency, inside the passband. higher folloffs, like 12dB or 18, will have less phase shift over such a wide range but will have more and even some peakiness just prior to the -3dB point.
If you are not running feedback keeping phase shift down is a good idea. But this requires bandwidth, which is a lot easier to obtain if there are no transformers present.
If you are not running feedback keeping phase shift down is a good idea. But this requires bandwidth, which is a lot easier to obtain if there are no transformers present.
@valvemaniacuk
sorry to bother you, I saw in a post that you built a Lite Audio LS26
I'm also building Lite Audio LS26 and I have to buy the transformer for the power supply. What transformer did you buy? Can you tell me the features?
Ciao Francesco.
sorry to bother you, I saw in a post that you built a Lite Audio LS26
I'm also building Lite Audio LS26 and I have to buy the transformer for the power supply. What transformer did you buy? Can you tell me the features?
Ciao Francesco.
Yes. My U-OTL6080 delivers 10WRMS on a 16 Ohms impedance, having a measured output impedance of 20 Ohms. They solely use 3x6080WA + 1x12AX7 + 1x12AU7 per channel :
T
Actually Julius wasn't the first to create an OTL, although he was the first to actually manufacture one.