Can any one direct me to a good simple KT88/6550 design? Looking for push pull with ultralinear, probably operating in class AB1. Want around 100 watts, which the KT88 should easily do.
Has anyone got a design for some output transformers that could suit such a design? I want to get some prices on having some OPT wound, otherwise my only option would be one of the Hamonds such as the 1650P.
Has anyone got a design for some output transformers that could suit such a design? I want to get some prices on having some OPT wound, otherwise my only option would be one of the Hamonds such as the 1650P.
While 100W is quite achievable with a single pair of KT88, with current production tubes that is really pushing it. For good long term reliability I would recommend a quartet of them in UL PP with plate voltages not significantly higher than 500V.
About 60Wrms is the most I would ask from a pair.
I designed and built a few 120Wrms per channel stereo amps commercially using 6550 and KT88 and achieved good reliability.
Incidentally the current JJ KT88 seems to have build quality issues, and I have had reliability problems with them at >525V. (Catastrophic failure is an understatement.)
I would check with Jim McShane for his take on the current best bets sonically and reliability wise.
About 60Wrms is the most I would ask from a pair.
I designed and built a few 120Wrms per channel stereo amps commercially using 6550 and KT88 and achieved good reliability.
Incidentally the current JJ KT88 seems to have build quality issues, and I have had reliability problems with them at >525V. (Catastrophic failure is an understatement.)
I would check with Jim McShane for his take on the current best bets sonically and reliability wise.
I built a prototype mono-block several years ago using a quad of Ultralinear connected KT88 driving a Hammond 1650T. It was very good but other projects took priority and the second channel never got built.
I used a CAE Dynaco MK4 front-end board to drive 6SN7 cathode followers to drive 2 pairs of KT88. I would probably do it a bit differently if designing it today BUT it may give you some ideas.
Schematics for the Amp and Power Supply and a "brag" document can be found on the Projects pages of Evatco website. Its the 4th project down at this link.
http://www.evatco.com.au/project.htm
I used a CAE Dynaco MK4 front-end board to drive 6SN7 cathode followers to drive 2 pairs of KT88. I would probably do it a bit differently if designing it today BUT it may give you some ideas.
Schematics for the Amp and Power Supply and a "brag" document can be found on the Projects pages of Evatco website. Its the 4th project down at this link.
http://www.evatco.com.au/project.htm
Get your tube vendor to tap test'em on burn in. I've had Svet 88's and JJ 88'S making rude noises when pushed at 500V+. Worse with KT90 when internal shorts occurred so regulary that I had to fit a fuse in the centre tap o/p tranny.
Running 6550C at 460V max in parallel pairs UL p-p should get one close to 100W around 2K a-a. Don't go and ruin the performance by using a crappy driver stage. The 6550C new version quotes 100K grid leak in fixed bias. I use half this but demands a slam driver stage capable of 50+50V rms at low Z at low thd. Once I get the software to run I will post suitable circuits.
The drawback is power supply issues...For "full pull" in a stereo pair in AB1/2 one is looking at a psu capability of 8x 0.15Ax470= 565 Watts with some reserve headroom for transients. I use a 1kW beast for min B+ volt drop between min and max sig. Amp weight becomes a measurable issue. A skimpy psu will ruin performance.
richj
Running 6550C at 460V max in parallel pairs UL p-p should get one close to 100W around 2K a-a. Don't go and ruin the performance by using a crappy driver stage. The 6550C new version quotes 100K grid leak in fixed bias. I use half this but demands a slam driver stage capable of 50+50V rms at low Z at low thd. Once I get the software to run I will post suitable circuits.
The drawback is power supply issues...For "full pull" in a stereo pair in AB1/2 one is looking at a psu capability of 8x 0.15Ax470= 565 Watts with some reserve headroom for transients. I use a 1kW beast for min B+ volt drop between min and max sig. Amp weight becomes a measurable issue. A skimpy psu will ruin performance.
richj
I just completed modifying an ELLA kt88 amp to run 100watts rms per channel.
Running the plates at 600vdc @ 140ma each and a pair of hammond 1650R output transformers.
B+ using Cree schottky diodes (bridge) and a power transformer from ebay China that supplies 550vac @ 750ma (and 3x6.3@5a and a 70v bias tapping)several Rubycon 450uf @400vdc in series parallel.
I need a choke that can handle 500ma and still fit under a 2" tall cassis if anyone can help.
Quad of re-issue gold lion kt88's.
Not expecting the tubes to last any where as long but it's holding up so far and makes a lot of smooth noise.
Just to see if it can be done easily.
Running the plates at 600vdc @ 140ma each and a pair of hammond 1650R output transformers.
B+ using Cree schottky diodes (bridge) and a power transformer from ebay China that supplies 550vac @ 750ma (and 3x6.3@5a and a 70v bias tapping)several Rubycon 450uf @400vdc in series parallel.
I need a choke that can handle 500ma and still fit under a 2" tall cassis if anyone can help.
Quad of re-issue gold lion kt88's.
Not expecting the tubes to last any where as long but it's holding up so far and makes a lot of smooth noise.
Just to see if it can be done easily.
Just to chime in on power available from a pair of tubes, and the need for good quality tubes.
I fixed up a bass guitar amp that put out over 80 watts from a pair of EL34 tubes !! It achieved this with a voltage doubled power supply mof 700+ volts. It blew fuses regularly, and when I looked, it had a pair of el-cheapo chinese noname EL34s in it. I put in a pair of Sovteks that were said to be rugged, and no more problems were had.
Now if it can get 80 watts reliably from 2 EL34s, then I am sure 100 watts can be done with KT88s as well, but as others have said, you might need to choose good quality tubes.
I fixed up a bass guitar amp that put out over 80 watts from a pair of EL34 tubes !! It achieved this with a voltage doubled power supply mof 700+ volts. It blew fuses regularly, and when I looked, it had a pair of el-cheapo chinese noname EL34s in it. I put in a pair of Sovteks that were said to be rugged, and no more problems were had.
Now if it can get 80 watts reliably from 2 EL34s, then I am sure 100 watts can be done with KT88s as well, but as others have said, you might need to choose good quality tubes.
MI watts are usually rated at much higher THD than hifi watts, even for the same basic design. I'm not confident of a good sounding 100W from a pair of KT88 in AB1 fixed bias. Last one I tried was awful. At 70W (lower B=, higher Ia, different OPT config) it was quite nice, but was never worked on further; ran out of clipleads.hotbottle said:
Brett said:
Not to mention that with most currently available output tubes life will be short under the conditions required for 100Wrms output. There is a good reason why most 100Wrms+ hifi amplifiers use 2 or more pairs of output tubes..
Brett is right about the sound quality issue as well, my findings as well.
I built a few 120Wrms per channel amplifiers for clients using quartets of Svetlana 6550C - 8 yrs later the originals are still running fine. (B+ 500V and idle current of 50mA per tube.)
just an odd idea I'll throw out there - check out the 50W plate EL156 tube. The Shuguang copies aren't bad - I've abused one in a SE amplifier test setup and it appeared to be quite tough. Eventually planning to make a pair of monoblock SE amplifiers (once I'm done moving to my next house).
http://www.mif.pg.gda.pl/homepages/frank/sheets/128/e/EL156.pdf
http://www.mif.pg.gda.pl/homepages/frank/sheets/128/e/EL156.pdf
The EL156 does require a higher filament current. 1.9A (EL156) vs 1.6 (6550). It also biases a little lower - eg, when I tested them on a a pair of Dynaco Mark III refurbed amps, I replaced the 10K bias pot with a 25K. I may have increased the resistor before the bias pot, but I can't remember off hand.
Also, in my SE experiment, I never went over 500V, so I'm not sure how the Shuguang version hold up after that.
Also, in my SE experiment, I never went over 500V, so I'm not sure how the Shuguang version hold up after that.
Hi,
Looks like the reliability of KT88 and lately K120 have become MUCH more reliable, hence I am thinking of restarting this project, ie 100W from a pair of KT88 and more power with K120. Any comments of the following are welcomed:
1. What B+ to use 500 or 550Vdc, I think 550 is better?
2. Which transformer to use, Hammond 1650R or Edcor CXPP100-MS-5K both have
about 5K anode to anode impedance?
3. Which driver tube to use, Musical Reference RM200 uses 6BQ7, any other tube to use? what current to bias at?
I believe an amp using 2 output tubes, a reliable 100W amp is achievable, e.g RM 200Mk2
Looks like the reliability of KT88 and lately K120 have become MUCH more reliable, hence I am thinking of restarting this project, ie 100W from a pair of KT88 and more power with K120. Any comments of the following are welcomed:
1. What B+ to use 500 or 550Vdc, I think 550 is better?
2. Which transformer to use, Hammond 1650R or Edcor CXPP100-MS-5K both have
about 5K anode to anode impedance?
3. Which driver tube to use, Musical Reference RM200 uses 6BQ7, any other tube to use? what current to bias at?
I believe an amp using 2 output tubes, a reliable 100W amp is achievable, e.g RM 200Mk2
You could easily get 100w with a pair of kt-120's. I wanted to do an amp using four kt-88's but an issue I was running into was the filament supply for four kt-88's which would be over 6 amps, most PT's are only rated to 6 amps. So one would have to purchase an additional filament transformer.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.