captainbeefheart

Yes I encourage anyone with the means to join in of the fun and build one themselves. Will provide schematic, bill of materials, layout etc...

I started this thread for as much feedback from people as possible. If folks want a little 'sizzle' and up the target price to say $500 that's fine with me. I am trying to figure out what people consider inexpensive. Originally this idea came from a mix of two threads, the tubecube 7 and the Crimson 275. I mean the more I look at similar offerings (quality 15 watt single ended tube stereo amp) $500 is peanuts compared to what I am seeing for commercial offerings. The reason I am leaning towards SE is because I believe the majority here using tubes have push pull amps so why not make something everyone wants to try but doesn't want to spend $4,000 for 15 watts. Honestly I also am leaning that way because I prefer SE and feel it can work very well with Klipsch speakers. I have heard and owned so many amplifiers and I also have the ability to build myself anything I want to try if it isn't offered commercially so I like to think if my ears have landed on SE amps as sounding the best it's for a reason. Clearly it isn't just vanishingly low distortion as SE typically will have more vs push pull so something else is at play here. I surmise it's just more natural to how we hear in nature, sound propagates through the air in a single ended fashion. Fluid dynamic equations clearly show the peaks of the waveform travel faster than the low energy troughs making whole process non-linear and rich with even harmonics dominant with the second harmonic. Processing this type of information is natural to us, our brains say 'normal' and we relax and feel we can just listen endlessly without fatigue. Sadly if one owns an amplifier that can't be dynamic and clean enough we don't get to enjoy the benefits of the topology. Wave equation that accounts for non-linearity up to the second harmonic where p {\displaystyle p} is the sound pressure, c 0 {\displaystyle c_{0}} is the small signal sound speed, δ {\displaystyle \delta } is the sound diffusivity, β {\displaystyle \beta } is the nonlinearity coefficient and ρ 0 {\displaystyle \rho _{0}} is the ambient density.

I agree, the proper component selection for best results I also consider the meat and not the sizzle. The sizzle for me is eye appeal, meters, fancy chassis, and sometimes boutique name capacitors. If I choose a paper capacitor what costs me $5 for our intensive purposes it will perform the same as a $100 Jensen paper capacitor. Same dielectric will have the same linearity and hysteresis properties, there may be a slightly different resonant frequency from the parasitic properties (series inductance) but that's true for any capacitor you purchase, a radial will have slightly different series inductance so resonant frequency will somewhere different but so far above the audio band it's moot, especially tube amps. Resistors, most have no idea what to look for, name doesn't matter. Look at the datasheet and look at the coefficients, typically voltage and temperature. The lower the number, mind you this is measured in parts per millionth (10-6) the better. This means the hotter the resistor gets or the more voltage across it will change the resistance value making it slightly non-linear. Well lets look at the problem and find a better solution. For the temperature issue size the resistor appropriately so it doesn't get even warm and it will stay linear. So even a carbon composition resistor with terribly high temperature coefficient will not drift as long as it doesn't get hot, so over size it to keep it's value within tolerance. That's easy enough right? The voltage issue isn't avoidable, any resistor with a large signal swing across it you choose a type that has a low voltage coefficient that way it's linear with the large signal swings. It really takes a lot of swing to start to see the distortion increase but it is a real issue and not just 'mojo'. Carbon composition resistors will have higher voltage coefficients so using them as say plate load resistors for a phase inverter where signals are larger will most likely show an increase in distortion. Luckily the voltage coefficient also is reduced with increase in resistor size/dissipation rating so again selecting a wattage rating very conservatively improves performance. I try and follow a 10x rule where the rating of the component is 10 times what the working value is. So for a resistor that sees 100mW I'll go ahead and use a 1 watt resistor. Typically 5x is adequate but if I can go ahead with 10x for not much money or loss in physical space it takes up I will. It's simple physics, there is no magic inside these parts, knowing how the data will relate to sound it's easy to see if you want low distortion there are solutions without ever mentioning "Allen Bradley" or "Jensen paper/wax" etc..... Choose things via the data and smart engineering practices and you will achieve your goal of good performance and sound.

That's fine, if someone asks for specific brand capacitors and I feel it will not diminish performance just pay the difference and I will build it with them. My point is the actual difference between these parts are so small, probably 1% or less where actual circuit improvements gets you a 20% increase in performance/sound it's obvious which gets priority to me. I choose parts by their application and how different component types will effect the total outcome. I'll most likely go paper capacitors for coupling so I don't know what anyone else prefers but if they want film types that's ok also. Connections are important and I never skimp on them, it's not expensive to get good In/outs, sockets, pots etc... I usually install lots of filters on the primary side of the power transformer so it sort of has internal 'power conditioning' if you will, common mode chokes, with X and Y capacitors for common and differential mode noises. This goes both ways, filtering both AC coming into the amp but also not polluting your house power either. Yes your amplifiers are awful non-linear loads polluting your power with distortion and harmonics. The average linear power supply has a power factor of .3 which is awful compared to the ideal 1 or .8 Federal standard requirement. Hence to meet compliance you see so many switch mode power supplies going into products these days.

First circuit I am drawing up is a KT88/6550 SEP amp class A2 operation. Gain - 20db 15 watts at 1.88% THD second harmonic dominant Fixed bias with external tip jacks and control to easily adjust bias. With a nice looking chassis I can build this for around $400. It blows my mind to see some of these builders asking $3,000 for a simple single ended amplifier (Aric Audio) that doesn't even drive grid current A2 operation. I get there needs to be profit but there should be a healthy profit margin with a retail of $1,200-$1,500 dollars. The fact they do not even show distortion output at any given power and frequency tells me they probably don't even measure anything once completed. How are they calculating compensation networks? Do they even have them? I would like to see some square waves from these amps. I am not saying the amp sounds bad, just not much info for a pretty healthy price tag. This is what is driving me to do this project. I have seen far too many amplifiers hit the market without much research and development. I see way too many guys that can solder and have a somewhat decent understanding jump into the market and start building and selling amplifiers for big money. The refinement in the product is missing on a technical measure, upgrades consist of brand name capacitors instead of putting effort into actual circuit improvements that will yield much more. For example the lack of SE smaller powered amps with no grid current drive capabilities. I truly feel that it makes a big difference with lower powered amplifiers to not run into blocking distortion bias shifting the amp until it recovers over time back to normal operation. With A2 operation the driver can push current into the grid maintaining the signal integrity with no recovery time smear. If you are going to ask a huge price tag at least put the money into something that will greatly improve performance instead of extremely minor esoteric boutique capacitors and silver wire. Those gains are diminished returns for the investment, you get FAR more bang for your buck implementing circuit changes to mitigate some of the pitfalls of the specific design. I really want people to have the opportunity to get good performance and sound for a reasonable price. I don't think the majority know enough about control networks to properly implement feedback so they stray away from it instead of fine tuning it. I use Lissajou formula to look at phase relationship input to output, mark the breakpoints of the amplifier on a Bode plot and then calculate the compensation networks to increase phase margin to where it's unconditionally stable. This is best done with the built amp and not in a simulation as stray capacitance will just change your simulation results. Once calculated and installed I do square wave tests which may lead to some empirical adjustments and fine tuning. This only needs to be done once if you continue to use the same output transformers. The majority of time is spent on the very first prototype, after that it's just rinse and repeat and I can probably knock a build out in one weekend. I hope this can come to fruition, I don't expect many here to jump in and purchase an amp so I am sure will be able to keep up with the demand if there is one. Basically once I am done with the first amplifier I will sell it, if anyone else wants one after that just let me know and I will build per order.

I was intrigued so I went and took a look, it appears his designs are ultralinear/triode which in my book is much different compared to a true pentode amplifier. I have built and listened to many UL amps and just feel pure pentode when done well is better driving real loads like speakers. Plus he is asking $4,000 for one!! Wow.

The vast majority of tube audio amplifiers are going to be push pull, and for the single ended fans I would say the majority of them tend to go the full Monty to single ended triode amplifiers. I could be wrong of course, this is just my observation and certainly there are not a lot of SEP amplifiers offered from major commercial tube amplifier companies which strengthens the notion that not many people have heard one. The few I have heard from commercial offerings were not terrible but also not great. Personally I jump back and forth between SEP and SET as my favorites, I just haven't cared for push pull amps for the past decade at least. Well not really, I am lying I have been listening to a Sherwood 700a SS push pull receiver with RB-81 mkII speakers and it sounds quite good. That system mainly plays FM radio all day for the wife, she's easy. BUT for when I get into serious listening mode it's La Scala room and tubes I have found that one major aspect to getting a SEP amplifier to sound great is the feedback. I still have global feedback but some nested loops at the output stage ala plate to grid yields excellent results. I like variable damping and no matter how the design of this amplifier turns out I will most likely include one ala current feedback from speaker. From being a member for a little while now and reading what people are mainly running I don't think many have amps that can do this and they might have fun with an amp and a new feature to tweak the amp/speaker relationship by ear for their preferred listening pleasure. So many circuits...........so little time ⏰

Thank you for your response, and all valid points. George has sent me two of his boards, the TSE and the SSE, I preferred the TSE better and ended up with 45's in it but that was a birthday gift to my bass player along with a pair of Heresy's. He desperately needed a good sound system for the level musician he is. I have been building other peoples designs and my own for decades and have made some really nice things for not much scratch. I am not looking to profit from these, so cost to forum members will be parts and shipping. Not necessarily, I can do a 15w single ended amplifier, I have some 6550, KT88's here I can breadboard with, owners can swap to KT120 also. That would be a decent idea, I do not have any of those to breadboard with. I will have to crunch numbers but they may push the price limit too high and have not much tube rolling options, I know folks love to try different tubes out as part of the fun.

To do so one would need to increase the size of the fuses which will void any warranty in order to reach the rated continuous output power. The actual physical swapping of the output transformer isn't very difficult, but whoever decides to take on this venture I hope they remove the existing compensation networks and plot the phase shift between input and output. Then re-calculate the compensation networks for best stability with the new iron. As for safety, after grounding the chassis and removing the resistor on the neutral noise would need to be checked and possible grounding changes may need to be done.

This would be an assembled amp sold to only forum members. For the members that like to build there own I will provide whatever necessary information for them to make one on their own. If anything has been learned from the Crimson 275 drama is that a 15 watt amplifier can work for many forum members. The goal now is to put that in peoples hands for a very low price point. I have made a few one 'tuber' or 'spud' amps before and 2-3 watts might be enough for some it's not enough for the majority. It is looking like we are aiming for 15 watts here. As soon as I settle on one I don't mind letting you in on what it will be. I needed to get some design goals first which is why I asked about power needs first. I figured a 15 watt amp might be much more appealing to many. I agree that iron is the most expensive parts, I have some tricks that worked well for me in the past I need to revisit to get the what we need and still fit the budget and performance parameters. It's panning out a 15 watt tube amp is what we are going with.

Thank you!! Pretty much what I expected accept it does have feedback, I will calculate how much they are using. I don't like the current source at the output tube, you just cannot beat a simple RC network, the CCS at the cathode will have awful overload characteristics and with such a low powered amp you do not want that. I know they are en vogue at the moment but it's just not a place I would use a CCS. Other than that it's a pretty cool little amp, I like it.

Will anyone be willing to open one up and take some pictures for us? Has anyone drawn up a schematic from one? I can probably do it from pictures if posted. My guess is it's a simple 12AX7 common cathode gain stage AC coupled to the EL84 wired for pentode operation. I don't think there is any global feedback as the measurements are that great, at 1 watt there is not much above 10kHz. If you stay under 1 watt performance is ok but at 3.5 watts I wouldn't consider it 'hifi' but I am sure it sounds good measurements be damned. There seems to be a huge market for inexpensive tube equipment for people to play around with. I have been thinking of designing a small amplifier like this for forum members except with much better performance. For $200 I am thinking I can do at least 5 watts per channel, I'll shoot for 10 watts but size will need to be larger with less than 1% THD at full output power 20-20kHz.

We have one and I agree they are very easy to repair. Usually the big main function/timer switch goes bad, I have two spares I pulled from units because new ones are like $150 each. I actually have a whole box labeled "dryer parts" that is filled with every single part needed for the unit. Most people don't know the schematic is hidden inside the top panel where the switches are, it's dead simple. Our daughter is at least on her third set of washer and dryer because they keep purchasing the more modern types, their latest is only a year or two old and they already have had to call for service on it. They purchase name brands, I believe they have a GE one now and it's total junk. The Armana brand is the same as Whirlpool, all the same parts and same design. I found a working one on the side of the road this summer and gutted all the good parts from it and scrapped the sheet metal at our towns recycling center. More parts to keep my dryer running forever.

Oh I forgot, the 100 watt model is really easy to install as there are only two primary wires which go to the 70v amplifier side and there are 3 wires for the secondary, use the two marked '0' and '8' to connect to the speakers and you will be all set. Enjoy

Ok then I suggest purchasing the 100 watt transformers, it will sound very full and detailed with those transformers as they are rated for full power between 40-20kHz which is really good, you will be very happy. God Bless you and your whole congregation and enjoy the music!

If your amplifier is a 70v constant voltage output your speakers wouldn't be happy with a direct connection, you cannot just directly hook it up to a normal 8 ohm speaker or damage will most likely ensue. If it's for just vocal PA uses and very light music then you can purchase the one I linked above. It sounds like you may be pushing music though these with the same system? If so and you plan to want to be able to crank the speakers to louder music volumes then you might want to get a higher powered one like this: https://www.parts-express.com/70V-100W-High-Power-Line-Matching-Transformer-300-225

You need to wire the constant voltage PA system through one of these: Secondary: Red to Black for 8 ohms wire to the Speaker Primary: White to Black wired to the 70v line Tape off all other connections and don't allow them to touch each other. https://www.parts-express.com/70V-15W-Line-Matching-Transformer-300-039

Yes they go in series with the tweeter, shouldn't matter + or - leg As for size, If memory serves correctly I don't think the K77 want anything over 2-2.5 watts maximum power, we'll say at 6 ohms, that's 645mA for 2.5 watts. So you look for a PTC that has a It (current trip) of 600mA. This littlefuse brand says Tt (trip time) of 2 seconds. So if current reaches 600mA or higher for longer than 2 seconds it will raise from a resistance of .8 ohms to 4 ohms, once current is reduced it will reset back to Rmin (resistance minimum) of .8 ohms. At $.82 they are cheap insurance for K77 protection. https://www.digikey.com/en/products/detail/littelfuse-inc/TRF250-180-2/1114120

No not a stock part

It will increase resistance with increased current. A PTC Thermistor is much better, it will have a much sharper knee at a specific current you choose, at this current and where you pass the knee increases resistance very quickly to a very high level. The light bulb will have much more gradual transfer between current and resistance and not offer as good protection.

I wonder if anyone has measured more distortion with these in the circuit. The non-linear load of the rectifier may introduce unwanted harmonics into the signal.

The vast majority of tubes being produced today are made in 2 plants, one in Russia and the other in China. The names like Mullard and Tung Sol have been sold and to be used for advertising, they are not the same company that originally made them. Some of the current production tubes are very good but many of your regular audio receiving tubes like EL84 and 12AX7 are not as good as some of the old ones. Since they are no longer in production supply and demand has skyrocketed the price. If your circuit is sensitive to tube quality, usually open loop designs with little to no feedback globally or locally then you will hear differences with all tubes. It's considered good engineering to have the overall circuit performance not critically rely on one specific device's linearity, negative feedback will make differences between tubes a non-issue.

Probably one of the best books out there is "The Art of Electronics" by Horowitz and Hill. You do not necessarily have to read the whole book but I strongly recommend the first two chapters on "Fundamentals" and "Transistors" before learning about amplifier circuits. Once you learn at least the first chapter you can pick up some of the old Radiotron handbooks about tubes but I advise learning fundamentals first before getting into active devices like transistors and tubes.

It may have been something that was only touching and shorting out from the vibration of the speakers at higher output levels. At lower levels it wasn't causing a short and played normal.

Well said. I 100% agree they new exactly what they were doing when they marketed this amplifier with 275 in the name. People are all over something they think performs like a McIntosh 275 with 75 real continuous watts per channel but for $2,000 dollars less. Power sells, I am certain it was a marketing and engineering decision to make the B+ over 600v and use a 2.5k plate to plate load, that way there they can technically make 75 watts, just not continuous and not over the entire audio bandwidth. Carver is innovative, and that is the first time I have seen a tube amp designed this way to provide a very short duration of higher transient power well above it's rated continuous power. They very well could have chosen a 350v plate supply for the KT120 but it wouldn't have reached it's transient peak power of 75 watts. The non-linear distortion created by the output transformer at low frequencies and higher power levels creates a lot of distortion but it's even order distortion and isn't necessarily heard as 'bad'. Between the bass boost they put into it's frequency response and the high amount of even order harmonic distortion at low frequencies it can be easily heard as 'full' bass as the distortion is going to mostly be an octave above the fundamental adding in more low frequency information. So for a 27Hz fundamental (low A on a piano) the output will also contain a 54Hz signal from the added distortion reinforcing the 27Hz fundamental. Now this can sound good as the amplifier reviews have proved but in it is not regarded as high fidelity as it's not an accurate reproduction of the original information. This is how many of those amplifiers with wimpy output transformers still have good bass sound, it's not that bass is attenuated, they are distorted adding in information that isn't originally in the program material making it not exactly hifi but acceptable in that it doesn't sound 'bad'. This would not be tolerated in the auto industry. If you your truck is rated at 400 ft/lbs of torque and you go to pull a boat and it really only has 150 ft/lbs of torque you will be pretty disappointed going up hills and accelerating from a complete stop. It doesn't surprise me how Bob finally addressed the issue, he isn't going to admit any malicious intent and deflect the real issue by just talking about how the amp sound good. I really would love one of these on my bench to test myself. I would love to see if the DC restorer circuit even does anything. I would also love the answer of how much voltage is across C36 and C37 during power up as they are only rated at 450v and without the preamp tubes conducting (they need time to warm up) I am certain the voltage rating of those capacitors will be grossly exceeded!! They should have at the very least installed a shunt resistor to make sure there is always load and a drop across the dropping resistors in the filter network during power on. Or better, I make sure the voltage rating of all the capacitors in the power supply can handle full unloaded B+ in the chance someone may power it on without any tubes installed. It would not have cost much more money at all to use two 350v caps in series for a combined rating of 700v, instead they chose to rate the capacitors for working voltages which is poor engineering with a SS rectifier. I am willing to bet C36 and C37 will start to fail soon since these have been out for a little while now and they can only take so much over voltage abuse in their lifespan.