The Hammond 1650T is 1900 ohms CT, and the 1650R is 5000 ohms,CT, so one would naturally assume that the impedance for a primary half would be 950 ohms and 2500 ohms, respectively. Using a straighforward impedance calculation, I got a turns ratio of 10.9:1 for the 1650T and 17.7:1 for the 1650R, assuming I'm using the 8 ohm output tap. I calculated back the peak output voltages and currents for the three output power operating points I mentioned to determine the points on the tube operating curve. The 1650R intersects 100W peak at +50V drive right before the tube runs out of gas, with about 100V left on the plate.
I have an inquiry in to Hammond as to whther the 1650R can actually support this kind of voltage at 20-30Hz without saturating. If not, I'll squeeze back the drive a bit to keep from inadvertantly saturating the transformer. Better clipping than saturation...
I have an inquiry in to Hammond as to whther the 1650R can actually support this kind of voltage at 20-30Hz without saturating. If not, I'll squeeze back the drive a bit to keep from inadvertantly saturating the transformer. Better clipping than saturation...
Owie, I see where I made my mistake. I can't just whack the impedance in half for a primary half of course, as it goes as turns squared. I just re-ran the numbers for the 1650R. If you look at the whole primary to secondary turns ratio, it's 25:1, 1/2 of this is 12.5 :1. This means that I don't really even want to bother with the 1650T, as the peak currents and required drive will be right up in the stratosphere. I just replotted the curves with the new turns ratio, and assuming again a starting point of 10W, 800V on the plates, you hit 70W peak at just under +50V drive, so you need a total of about 70V swing to get there, as you're starting from -20V bias. The transformer sees about 400V across a primary 1/2, and the tube sees about 330ma peak. The numbers are still very reasonable, and one is way within the linear operating region of the tube, with plenty of meat left if one really wants to get crazy with the drive. The small triodes on the 6GF7 are a bit wimpy, I'll probably have to work pretty hard to squeeze 70-100V drive out of them.
Actually, it's worse than that, as I need a linear swing of twice that voltage. Sigh... Well, I wanted challenges, and I got them.
Wrenchone,
I built a push pull EL-509 amp and got all the transformers from Jack at Elecrto-Print. They outputs were 2k primary and were huge monsters. The circuit is Rozenblits 150w design. The amps sound fabulous with great bass. Jack will work with on the transformer design.
Andy
I built a push pull EL-509 amp and got all the transformers from Jack at Elecrto-Print. They outputs were 2k primary and were huge monsters. The circuit is Rozenblits 150w design. The amps sound fabulous with great bass. Jack will work with on the transformer design.
Andy
Does the Rozenblit 150W amp use the EL509s as pentodes or triodes? One of the attractions of using the 509s for me was to use them in triode mode and get some significant power for my efforts.
Tony,
I don't think there is anything on the web. I purchased Rozenblit's book that had the schematic of the 150w amplifier in it as projects. I built it up from the schematic. You can contact Bruce and he may be able to provide you witha copy of the schematic.
I don't think there is anything on the web. I purchased Rozenblit's book that had the schematic of the 150w amplifier in it as projects. I built it up from the schematic. You can contact Bruce and he may be able to provide you witha copy of the schematic.
Wrenchone,
This amp is set as pentode with regulated screens. I'm sure it can be set up as triode but the amp is optimized for pentode. So some messaging will be needed to get the best sound for triode operation.
This amp is set as pentode with regulated screens. I'm sure it can be set up as triode but the amp is optimized for pentode. So some messaging will be needed to get the best sound for triode operation.
From the tube curves for the EL509, a screen driven triode amp will be a very different beast as compared to a pentode implementation, as a lot more voltage drive is required for significant output power, along with some significant screen current. I'm currently wrestling with how to do that while at the same time keeping the number of stages to a minimum.
Wrenchone,
No argument, I was just recommending the output transformer from Electra-Print. For a driver you could consider a mu stage. El-84 over a 6N1P or something like that. Svetlana had a tech tip showing this circuit. They said it could swing something like 350v which would be enough.
No argument, I was just recommending the output transformer from Electra-Print. For a driver you could consider a mu stage. El-84 over a 6N1P or something like that. Svetlana had a tech tip showing this circuit. They said it could swing something like 350v which would be enough.
Andy,
thaks a lot.
Wrenchome,
have you seen david bernings' ut used 2x 12at7's and a 6cg7 tube complement.
thaks a lot.
Wrenchome,
have you seen david bernings' ut used 2x 12at7's and a 6cg7 tube complement.
I want to try and do my own circuit for this amp and not lean too heavily on other people's results.
I am very suspicious of using a mu-follwer in this application given the amount of drive voltage needed (at least 150V p-p). I don't see how someone could keep within the peak filament to cathode ratings of the tubes without using separate tubes for top and bottom and isolated filament supplies. At this point I'm thinking of using a pair of 6GF7 vertical oscillator/amplifier dissimilar triodes for amplification and follower duty. I may back this up with an 8CG7 for extra gain if needed. I I'm looking at using a differential pair at the input, maybe with cascode to get more gain, with a 400V regulated supply and about 250V at the plates.
I'll be using Hammond iron at the output, as is reasonably inexpensive and readily available - the 1650R, if I remember correctly. Anyway, it's the 5k, 100W transformer. This strikes a compromise between output power ( more than enough for my purposes) and required drive voltage and output tube peak current.
I am very suspicious of using a mu-follwer in this application given the amount of drive voltage needed (at least 150V p-p). I don't see how someone could keep within the peak filament to cathode ratings of the tubes without using separate tubes for top and bottom and isolated filament supplies. At this point I'm thinking of using a pair of 6GF7 vertical oscillator/amplifier dissimilar triodes for amplification and follower duty. I may back this up with an 8CG7 for extra gain if needed. I I'm looking at using a differential pair at the input, maybe with cascode to get more gain, with a 400V regulated supply and about 250V at the plates.
I'll be using Hammond iron at the output, as is reasonably inexpensive and readily available - the 1650R, if I remember correctly. Anyway, it's the 5k, 100W transformer. This strikes a compromise between output power ( more than enough for my purposes) and required drive voltage and output tube peak current.
have you read Allan Kimmel's article regarding mu followers? they can really swing a lot of voltage. for instance a 12by7 pentode on top, and a 6cg7 triode at bottom and a supply of 600volts can give you a lot. then you can buffer this with a 6cg7 to set output tube operating point.
downside is you have to have a separate floating supply for the top tube, 12by7 heaters.
downside is you have to have a separate floating supply for the top tube, 12by7 heaters.
- Status
- Not open for further replies.