Tweet
Change to 12AT7-based LTPI circuit?
I’ve been reading thru notes on Randall Aiken’s website on the design of the LTPI circuit White Papers along with the notes from Merlin B’s website for the 12AT7 The Valve Wizard and then finding an email thread from Ken Gilbert here http://www.blueguitar.org/new/text/t...n%20Advice.pdf where he laid out the procedure designing from scratch a LTPI circuit using the 12AT7.
I’m led to believe what I need to do on this Traynor YBA-1A to get the power output up from 50W is increase the drive capability of the LTPI circuit, since I only have the one pair of tubes….KT88’s or 6550’s, capable of 100W or close to it with the power supply potential I have. I’ve been unable to get that output level using the stock LTPI circuit based around the 12AX7. That circuit driving either a pair of power tubes for 50W or a quad of power tubes for 100W. works, since there, you’re doubling the output current by adding an additional pair of tubes.
Leaving the LTPI circuit as is with the 12AX7, or plugging in a 12AT7 with the circuit set up for the 12AX7 doesn’t yield anything but lower gain overall with a single pair of tubes, regardless of the type (EL34, 6L6GC, KT88). The output power level hasn’t changed
Working from the curves on the 12AT7 from Philips ECC81 data sheet while looking back and forth from the data KG presented, where he selected an operating point of 200V/10mA, and that of Merlin’s of 150V/4.5mA, there’s quite a difference in power dissipation (2W vs 0.68W), and also the symmetry in voltage drive (Merlin’s looks pretty symmetrical, and KG’s quite asymmetrical on the load line), though what the KT88/6550’s are looking for, as far as voltage swing goes is about 65V p-p for 100W output.
Recalling a good starting point for the plate resistors, stated by RA as being twice the driver tube’s rp value. With the 12AT7 at 4.5mA, the rp is close to 15.5k, and at 10mA, rp is close to 10.5k. For 4.5mA Ip, 30-33k would seem to be the right ballpark, though Merlin’s example used 18k for the Plate resistors. In KG’s exercise, this worked out to about 22k, which seemed to be more in line with RA’s comment.
I ran a set of numbers using an operating point of 170V at 8mA Ip, a tail voltage of 80V with a HT of 450V. Selected 24k as the plate resistors, with a dissipation of 1.36W (200V @ 8mA), bias voltage of -1.2V, cathode resistor of 75 ohm and tail resistor of 5k total. The drive voltage still looks quite asymmetrical on the load line
I haven’t ever gone thru this exercise before, so perhaps if the asymmetry is outside the required voltage drive, it doesn’t matter? And, am I on the right track in increasing the LTPI’s current drive capability to get the higher power output from a pair of KT88/6550’s?
I’ve been reading thru notes on Randall Aiken’s website on the design of the LTPI circuit White Papers along with the notes from Merlin B’s website for the 12AT7 The Valve Wizard and then finding an email thread from Ken Gilbert here http://www.blueguitar.org/new/text/t...n%20Advice.pdf where he laid out the procedure designing from scratch a LTPI circuit using the 12AT7.
I’m led to believe what I need to do on this Traynor YBA-1A to get the power output up from 50W is increase the drive capability of the LTPI circuit, since I only have the one pair of tubes….KT88’s or 6550’s, capable of 100W or close to it with the power supply potential I have. I’ve been unable to get that output level using the stock LTPI circuit based around the 12AX7. That circuit driving either a pair of power tubes for 50W or a quad of power tubes for 100W. works, since there, you’re doubling the output current by adding an additional pair of tubes.
Leaving the LTPI circuit as is with the 12AX7, or plugging in a 12AT7 with the circuit set up for the 12AX7 doesn’t yield anything but lower gain overall with a single pair of tubes, regardless of the type (EL34, 6L6GC, KT88). The output power level hasn’t changed
Working from the curves on the 12AT7 from Philips ECC81 data sheet while looking back and forth from the data KG presented, where he selected an operating point of 200V/10mA, and that of Merlin’s of 150V/4.5mA, there’s quite a difference in power dissipation (2W vs 0.68W), and also the symmetry in voltage drive (Merlin’s looks pretty symmetrical, and KG’s quite asymmetrical on the load line), though what the KT88/6550’s are looking for, as far as voltage swing goes is about 65V p-p for 100W output.
Recalling a good starting point for the plate resistors, stated by RA as being twice the driver tube’s rp value. With the 12AT7 at 4.5mA, the rp is close to 15.5k, and at 10mA, rp is close to 10.5k. For 4.5mA Ip, 30-33k would seem to be the right ballpark, though Merlin’s example used 18k for the Plate resistors. In KG’s exercise, this worked out to about 22k, which seemed to be more in line with RA’s comment.
I ran a set of numbers using an operating point of 170V at 8mA Ip, a tail voltage of 80V with a HT of 450V. Selected 24k as the plate resistors, with a dissipation of 1.36W (200V @ 8mA), bias voltage of -1.2V, cathode resistor of 75 ohm and tail resistor of 5k total. The drive voltage still looks quite asymmetrical on the load line
I haven’t ever gone thru this exercise before, so perhaps if the asymmetry is outside the required voltage drive, it doesn’t matter? And, am I on the right track in increasing the LTPI’s current drive capability to get the higher power output from a pair of KT88/6550’s?
Comment