Guess it's time to order those novar sockets . . . . . .
It is interesting to know Bruse comment that EL519 gives lower power than EL509!
http://www.websitetoolbox.com/tool/post?id=163&action=vpost&username=tubehifi&forum=0&keywords=519
http://www.websitetoolbox.com/tool/post?id=163&action=vpost&username=tubehifi&forum=0&keywords=519
OTL, EL509, 519 6lf6 and other Stuff
Hi fellow Gizmologists. I've been building OTL of the Futterman variety since 1984 - 20 years when I was with New York Audio Labs as chief engineer with the inimitable Dr. Gizmo, Harvey Rosenberg.
Here is the straight poop on some of the recent questions.
EL509II's:- I asked Morgan to do the testing since I couldn't get any samples from SED or Techzotic. This tube is not a 6KG6/EL509 or 19. His tests were correct. Techzotic is threatening to sue him which is ridiculous. He should sue them for false advertising. The best thing to do is to correct the tube so that it matches the EI EL519. Now you'd have something. You could make otl's with octal tube sockets. What a lovely concept that would be!
Test parameters: I test the tubes at Vp=90v, Vg2=190V, Vg3=0 When Vg1=-35 Cathode I should be about 200 ma. When Vg1=0V, Ik should be >1.3 Amps. These are typical waveform peak conditions in an OTL.
Subbing the EL509 for the 6LF6: Outside of the plate ratings it is the same tube, different bottle. An element for element sub may be made. I was working on a socket adapter a while back. Contact me and if there is enough interest I'll make them. They will raise the tube about a 3/4 inch so make sure the lid fits on your amp.
Importance of plate ratings: The OTL shouldn't be run continuously at full power unless you have enough tube to spread the heat losses around. I have found that a 10 tube otl will pump 200 watts into 8 ohms for a long time without cherrying the plates if they are biased at about 16 watts or less/ tube. Music power is not continuous power. Average power is what is important which includes heat generated brom idle current. If I remember correctly, and I could be off, the max dissipation in an output stage is at 1/3 power. 10 tubes will do the job into 8 ohms for most speakers. I'm doing a 12 tub job on a Fourier Panther chassis. Looking for 250 +watts into 8. Should get 280+. The results are up to the power supply.
Here is a page out of the Bible, (RCA Receiving Tube Manual, Circa 1973).
Hi fellow Gizmologists. I've been building OTL of the Futterman variety since 1984 - 20 years when I was with New York Audio Labs as chief engineer with the inimitable Dr. Gizmo, Harvey Rosenberg.
Here is the straight poop on some of the recent questions.
EL509II's:- I asked Morgan to do the testing since I couldn't get any samples from SED or Techzotic. This tube is not a 6KG6/EL509 or 19. His tests were correct. Techzotic is threatening to sue him which is ridiculous. He should sue them for false advertising. The best thing to do is to correct the tube so that it matches the EI EL519. Now you'd have something. You could make otl's with octal tube sockets. What a lovely concept that would be!
Test parameters: I test the tubes at Vp=90v, Vg2=190V, Vg3=0 When Vg1=-35 Cathode I should be about 200 ma. When Vg1=0V, Ik should be >1.3 Amps. These are typical waveform peak conditions in an OTL.
Subbing the EL509 for the 6LF6: Outside of the plate ratings it is the same tube, different bottle. An element for element sub may be made. I was working on a socket adapter a while back. Contact me and if there is enough interest I'll make them. They will raise the tube about a 3/4 inch so make sure the lid fits on your amp.
Importance of plate ratings: The OTL shouldn't be run continuously at full power unless you have enough tube to spread the heat losses around. I have found that a 10 tube otl will pump 200 watts into 8 ohms for a long time without cherrying the plates if they are biased at about 16 watts or less/ tube. Music power is not continuous power. Average power is what is important which includes heat generated brom idle current. If I remember correctly, and I could be off, the max dissipation in an output stage is at 1/3 power. 10 tubes will do the job into 8 ohms for most speakers. I'm doing a 12 tub job on a Fourier Panther chassis. Looking for 250 +watts into 8. Should get 280+. The results are up to the power supply.
Here is a page out of the Bible, (RCA Receiving Tube Manual, Circa 1973).
Attachments
Hi,
Hmmm....Morgan himself says his tests were NOT correct.
DISREGARD ALL MY PREVIOUS STATEMENTS....
While I feel that interelectrode capacitances would be slightly different one from the other, I fail to see how simply resocketing a 6P45S would turn it into a totally different animal.
While the 6P45S may not be an exact copy of the 6KG6/EL509, IMO it's close enough to not cause any trouble as a remplacement.
Hands on experience has shown this to be the case, it works just fine.
Cheers,
Hmmm....Morgan himself says his tests were NOT correct.
DISREGARD ALL MY PREVIOUS STATEMENTS....
While I feel that interelectrode capacitances would be slightly different one from the other, I fail to see how simply resocketing a 6P45S would turn it into a totally different animal.
While the 6P45S may not be an exact copy of the 6KG6/EL509, IMO it's close enough to not cause any trouble as a remplacement.
Hands on experience has shown this to be the case, it works just fine.
Cheers,
Gentlemen, I know that some others have decided to not use the Svetlana EL509s for their OTL projects. I am still waiting to make a decision.
I have asked one of the US supplies of the EL509 MK IIs (which are exactly the same internally as the normal EL509s) to clarify this issue because I am buying (or was buying) a large set of MK IIs from him. So far, I don't have a response.
In fact, looking at all of the posts, I still am not able to piece together a definitive answer.
The Svetlana webpage shows plate curves that indicate that what moscode is saying is accurate. Others have said differently. moscode you've said what the Ik is at Vg=-35. What is it at Vg=0?
Is there anyone who can say, from direct experience using Svetlana EL509s or EL509 MK IIs whether or not they can deliver the necessary >1A currents that are shown in the standard EL509 spec sheets shown above and that are required for high powered OTLs? I sure would appreciate a definitive yes or no on this question. Thanks much.
I have asked one of the US supplies of the EL509 MK IIs (which are exactly the same internally as the normal EL509s) to clarify this issue because I am buying (or was buying) a large set of MK IIs from him. So far, I don't have a response.
In fact, looking at all of the posts, I still am not able to piece together a definitive answer.
The Svetlana webpage shows plate curves that indicate that what moscode is saying is accurate. Others have said differently. moscode you've said what the Ik is at Vg=-35. What is it at Vg=0?
Is there anyone who can say, from direct experience using Svetlana EL509s or EL509 MK IIs whether or not they can deliver the necessary >1A currents that are shown in the standard EL509 spec sheets shown above and that are required for high powered OTLs? I sure would appreciate a definitive yes or no on this question. Thanks much.
"While the 6P45S may not be an exact copy of the 6KG6/EL509, IMO it's close enough to not cause any trouble as a remplacement."
In the OTL setup I want as much current as possible. I would not use a tube that only passes 1/2 amp. You need many more tubes.You might as well go to a 6c33 and live with 48 watts of heat to light one up for 1.2 amps of cathode current. The 12 watts of heater power for the 6KG6 family make it a much more efficient choice. 10 tubes at 48 watts ea = 480 watts. That's too much heat for me.
Your 6p45 may be fine in a hig voltage transformer coupled circ. but it is IMO NG for OTL.
In the OTL setup I want as much current as possible. I would not use a tube that only passes 1/2 amp. You need many more tubes.You might as well go to a 6c33 and live with 48 watts of heat to light one up for 1.2 amps of cathode current. The 12 watts of heater power for the 6KG6 family make it a much more efficient choice. 10 tubes at 48 watts ea = 480 watts. That's too much heat for me.
Your 6p45 may be fine in a hig voltage transformer coupled circ. but it is IMO NG for OTL.
Where does this figure 1.2A comes from? I have measured peak 2.5A at low frequencies in 6C33C but I dont know if this is the saturation current or not, (I can't measure with higher current).
The saturation current and also max available current is dependant on cathode area, heater power and the cathode efficiency. Given the higher heater power and much larger cathode area and assuming that cathode efficiency is more or less the same it can be assumed that safe peak operating current for a 6C33C tube is much higher then for a EL509.
Please note that the specified peak current in PL519 of 1500mA is for pulse conditions in a TV line output stage, it is not given that this is valid for the peak of a low frequency sinus type signal, (oxide coated cathodes can sustain much higher currents during peak conditions with short pulses then for steady state)
Regards Hans
I current & efficiency
I'd like to think this post can put this issue to rest for me. I'm not open to intellectual debate on these specification. I've been building these amps since 1984, I've built tube testers specifically for these tubes (for myself and commercially) and my amplifiers work and give the outputs I state, I get paid and my customers are very happy. If you actually do the measurements and can't get the same results for the conditions I've stated in my first post then send me a private email.
I can only state what I've measured and I've measured 1.3 to up to 1.9 A continuous on 509's, 519, & 6LF6's . I limit the test to 3 seconds so I don't cherry the tube. I stated the test conditions in my original post and included the RCA spec sheet. I know the the 1.5 a rating is for the repeated horiz pulse but a good tube will sustain the full current until it starts to destruct. I built a tube tester for these tubes that test using real world otl conditions (my stated specs) and burn in the tubes using pulses so you can hit peak current without exceeding the average current. The average current times the voltage accross the tube can't exceed the power rating for the tube. We are talking long tem averaging.
If you pulse the tube at 1.5 amps for 1 second x 90 volts we have a 135 watt plate dissipation for that 1 second. If we do it every 5 seconds we only have a 27 watt average plate dissipation. That is why we can get away with using these tubes in an OTL. The average Pd is lower than the maximum ratings. Also using more tubes effectively increases the size of your heat sink. If your load impedance goes down use more tubes. I am trying to keep the heat dissipation to practical levels in my amps. That is why I favor the real EL509. Current / heat overhead ratio is higher than the 6c33.
I tested the 6C33's under the same conditions less the G2 voltage since there ain't one. You may be able to pull more current thru a 6c33 if you increase the plate voltage or drive the grid positive. If you raise the Vp then the heat dissipation goes up for the same idle current. If you want to drive the grid positive then you need a lotta current in the drive circuit which has a down side of bigger tubes or mosfets and bigger power supply. Here is a photo of one of my tube testers.
Here's the bottom line- I want octal based 509's that act like 509's.
I'd like to think this post can put this issue to rest for me. I'm not open to intellectual debate on these specification. I've been building these amps since 1984, I've built tube testers specifically for these tubes (for myself and commercially) and my amplifiers work and give the outputs I state, I get paid and my customers are very happy. If you actually do the measurements and can't get the same results for the conditions I've stated in my first post then send me a private email.
I can only state what I've measured and I've measured 1.3 to up to 1.9 A continuous on 509's, 519, & 6LF6's . I limit the test to 3 seconds so I don't cherry the tube. I stated the test conditions in my original post and included the RCA spec sheet. I know the the 1.5 a rating is for the repeated horiz pulse but a good tube will sustain the full current until it starts to destruct. I built a tube tester for these tubes that test using real world otl conditions (my stated specs) and burn in the tubes using pulses so you can hit peak current without exceeding the average current. The average current times the voltage accross the tube can't exceed the power rating for the tube. We are talking long tem averaging.
If you pulse the tube at 1.5 amps for 1 second x 90 volts we have a 135 watt plate dissipation for that 1 second. If we do it every 5 seconds we only have a 27 watt average plate dissipation. That is why we can get away with using these tubes in an OTL. The average Pd is lower than the maximum ratings. Also using more tubes effectively increases the size of your heat sink. If your load impedance goes down use more tubes. I am trying to keep the heat dissipation to practical levels in my amps. That is why I favor the real EL509. Current / heat overhead ratio is higher than the 6c33.
I tested the 6C33's under the same conditions less the G2 voltage since there ain't one. You may be able to pull more current thru a 6c33 if you increase the plate voltage or drive the grid positive. If you raise the Vp then the heat dissipation goes up for the same idle current. If you want to drive the grid positive then you need a lotta current in the drive circuit which has a down side of bigger tubes or mosfets and bigger power supply. Here is a photo of one of my tube testers.
Here's the bottom line- I want octal based 509's that act like 509's.
Attachments
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.