tube fanatic Posted March 30, 2014 Share Posted March 30, 2014 Just when I thought that tube amp discussions were becoming a regular part of forum life, it all fizzled! What's going on? In the interest of fueling things (hopefully!), attached are pix of the latest prototype which came off of the bench this morning. It's a variant of the parallel SEP which I posted a while ago, and uses a single output tube per channel. It's a 6CM6, another 6V6 clone, and is much better sounding than the latter imo. Like the other amp, it is a totally dual mono unit sharing one chassis, and output power is approx. 3.75 watts/channel at 1 kHz.I'm often asked how I always seem to "know" exactly what kind of amp is needed for a particular situation, and the answer is that I don't. It's based on a very careful history and discussion, much as a medical doctor can often diagnose a problem in the same fashion. It starts with talking about listening tastes and habits- loudness, type of music, type of sound desired (since SETs and SEPs can differ greatly- I'm not doing push-pull amps currently), and where the listener is seated in relation to the speakers. Then comes the speakers themselves, room conditions (such as whether it is a live room, or acoustically damped with carpet, furniture, and so on), and room size. All of these factors combine to determine how much power is needed. Once that's done, the output stage can be configured along with the type of output transformers to be used, followed by the needed front end to drive the output tubes. Depending on the sound desired, a decision is made about using or not using any negative feedback, and whether any high frequency filtering is desired for the "ear bleed" phenomenon which I've described in past threads. The power supply is designed last as it's capability depends on the rest of the circuit. So, it's obvious that my method is to design the amp backwards, starting with the listener's ears. Anyone who has followed the forum, and has tried to find a commercial amp which works well with their speakers, is aware that it can be a very difficult task indeed since the amp characteristics are fixed and can't be changed without warranty destroying modifications. So, let the discussion begin!Maynard 3 Quote Link to comment Share on other sites More sharing options...
minermark Posted March 30, 2014 Share Posted March 30, 2014 Nice looking work. Quote Link to comment Share on other sites More sharing options...
Deang Posted March 30, 2014 Share Posted March 30, 2014 You're the only person I'm aware of that takes the same approach as I do to this stuff. People are most comfortable with a "good, better, best" model, but it's really an awful way of going about it. Quote Link to comment Share on other sites More sharing options...
Tarheel Posted March 30, 2014 Share Posted March 30, 2014 (edited) Very nice work Maynard...would love to hear it! Edited March 31, 2014 by Tarheel Quote Link to comment Share on other sites More sharing options...
toolz Posted March 30, 2014 Share Posted March 30, 2014 Sweet. Quote Link to comment Share on other sites More sharing options...
beeker Posted March 31, 2014 Share Posted March 31, 2014 (edited) I love your amp Maynard. Everything about it. I enjoy seeing the home built amps around here. What i appreciate about yours i could finish it the way i prefer. The vrd(and i find it to be what others follow when finishing)has the wood base and if i want a wood base id rather make that choice once i get it home. Or plate or polish the case as i feel correct and throw some feet on it. I dont know how to make an amp just too much a curve i havnt taken on yet. I do know how to make my own base. And prefer not the cookie cutter look. good to see straight fresh tube builds ready to plug and play. i'd love to hear it as well. not acknowledging numbers and graphs other than the obvious i bet that 3.75@1khz watts sounds absolutely great! i am jealous edit to say:I imagined what you were doing and was correct as im sure others might of thought as well. very, very nice amplifier edit again to say:I absolutely love it the way it sits right now Edited March 31, 2014 by beeker Quote Link to comment Share on other sites More sharing options...
mustang guy Posted March 31, 2014 Share Posted March 31, 2014 (edited) I've added your amp to this thread: Tube art . . what lies beneath Very nice work Maynard! Edited March 31, 2014 by mustang guy 1 Quote Link to comment Share on other sites More sharing options...
tube fanatic Posted March 31, 2014 Author Share Posted March 31, 2014 Thanks for all of the comments. They are very much appreciated! And Beeker, I also appreciate your thoughts about the cabinetry. Some of the guys I build these for have woodworking skills (I have none!) and make cradles in which to drop the chassis. That said, most like the plain aluminum chassis. I hit them with some Mother's aluminum polish after they are drilled, and if the finish starts to dull a bit over time, it's an easy matter to spruce it up again. Mustang Guy, thanks for including the amp in the other thread!Mark, to address your questions: There's a limit, of course, to how small the amp could be made without going to much smaller output xfmrs. I tend to go somewhat larger than is absolutely necessary as I find it easier to get in there to do the work. Also, when using an onboard p/s, I like to keep xfmrs as isolated as is practical, particularly if any are open frame like the Hammond ESE series. The driver is a simple resistance coupled triode, the 6SQ7, which provides more than enough gain for this application with the approx. 250V B+ supply. It's contact biased (discussed in the thread about the parallel SEP) and has about -.7V on the grid. It's coupled to the output stage through a .22uf CDE DME series cap. I've been using these caps since they came out and have found them to not only sound great, but to be completely reliable with zero failures among the hundreds that I've used. As far as pentodes vs. triodes, I prefer the latter as output tubes as I find them, in terms of overall sound, to offer the best combination of all that I enjoy when listening. On the other hand, if someone wants a bit more bottom end, or needs/wants more power, I'll use a pentode. But, if given the choice, I try to get the midrange and top end of the latter to be as triode like as possible by playing with the nfb parameters, and/or including a fixed or variable low pass filter at the front end. For voltage amps, I try to use a triode whenever possible, although with some of the triode output stages it just isn't practical if a lot of voltage swing is needed. In those cases, I usually use a 6SJ7. I have the found the pentode voltage amps to be a bit more grainy, but with the triode output stage it isn't a problem. Now, I have a question for you! A recent discussion came up with a couple of my EE friends about the choice of diode piv ratings in a p/s using a full wave rectifier/capacitor input filter. One asserts that it's necessary, in the interest of preventing failure, to to go way above the peak voltage of the power xfmr secondary. The other believes that looking at a worst case, "normal", operating condition is sufficient. I'll use this amp as an example- the xfmr secondary delivers 445V with 120V from the line (268V at the input cap- a bit higher until the output tube starts drawing current). Peak is, of course, approx. 627V, so 1kv diodes should be fine. Even if the powerline was at peak of around 170V at the instant the amp is turned on (629V on the secondary resulting in peak voltage of aprox. 887), the 1kv diode should still be safe. I agree with this latter assessment. But, the first friend insists that I should run a couple of diodes in series on each leg to get the piv up to 2 kv. Obviously, if the amp takes a 5 kv powerline spike, even that won't make a difference. Your comments on this will be appreciated! You can't imagine the arguing that we do over here about issues like this; but it's fun regardless!!!Maynard Quote Link to comment Share on other sites More sharing options...
tube fanatic Posted March 31, 2014 Author Share Posted March 31, 2014 Mark, the point you raise about non mil-spec parts is a good one which I hadn't considered. So, I may pull the 1N5408s which are in the amp now, and replace them with 2 series connected 1 kv diodes which are somewhat smaller and will fit. As to your question about the power output, I'll add that it was measured into a non-inductive 8 ohm resistor with the feedback disconnected. The 3.75W was present just before any visible clipping was noted. As to distortion, get out the tar and feathers, I stopped measuring it in amps like this more years ago than I can remember (wish I still had a distortion analyzer, though)! Besides, measuring it into a resistor sure isn't representative of what's going to happen with a speaker load. Single pentodes/beam power tubes suck in terms of distortion, as you know. Between the varying speaker and output impedance, plus the frequency selective nfb which I use, the distortion is going to be wonderful or awful depending on the frequency (and the 2nd and 3rd harmonic outputs often increase or decrease inversely). In the days when I obsessed about measured distortion I always wound up having to change something to give the listener what he wanted even if the distortion was higher. So, now I simply voice the nfb parameters for the desired high frequency response, calculate and adjust the capacitor value to roll off the fb at the desired part of the audio band, and don't worry about what the distortion may be. It works every time. With triodes, I use a different approach, and graph operating conditions for the lowest possible 2nd harmonic level at the minimum speaker impedance (if known), or 4 ohms if it isn't. That way, the distortion goes down even further with the expected impedance rise across the audio band. And if the load line shifts upwards a bit from the 2nd harmonic rectification effects, the distortion drops along with it. As mentioned above, I like SETs! I'm not including the driver stage in this discussion as the distortion there is always far too low to be of concern. So, that's my heretical philosophy on distortion in single ended amp designs. Please feel free to tear me apart if you disagree (and I'm hoping Erik and Mike chime in on this topic as well!) Maynard Quote Link to comment Share on other sites More sharing options...
jason str Posted March 31, 2014 Share Posted March 31, 2014 Wow, what a meticulous job on the innards, i thought i was a perfectionist with my work but you Maynard would win first place. My hats off to you sir. Quote Link to comment Share on other sites More sharing options...
thebes Posted March 31, 2014 Share Posted March 31, 2014 Another terrific build! Quality craftsmanship to go with the in-depth knowledge. One simple question. I notice in your builds you always make a vent hole or two under some resistors. Do you think they generate enough heat to make affect anything, given that the really hot stuff, the transformers and tubes are on top? Or is this simply a belt-and-suspenders approach. Quote Link to comment Share on other sites More sharing options...
tube fanatic Posted March 31, 2014 Author Share Posted March 31, 2014 (edited) Jason, thank you for the compliment! I guess having a bit of ocd can be both a curse and a blessing depending on how it manifests!!! And, Marty, I try to minimize the under-chassis heat as much as possible; hence, the vents that you see in the amps. It would be less of a concern if I top mounted the electrolytic caps. It's just a matter of personal taste in the electromechanical layout. Mark, I didn't realize that's what you were looking for. But, George Anderson did some nice measurements on the 125CSE years ago which are representative of the whole 125xSE Hammond opts: http://tubelab.com/articles/component-testing/budget-output-transformers/ I also forgot to mention above that with pentodes I usually try to stay with published operating parameters if possible, or simply use the nomogram to calculate different ones if I need to move voltages around a bit. There was a time I would re-graph different operating conditions for pentodes, but don't have the patience to do it at this point! Maynard Edited March 31, 2014 by tube fanatic Quote Link to comment Share on other sites More sharing options...
mike stehr Posted April 2, 2014 Share Posted April 2, 2014 (edited) (and I'm hoping Erik and Mike chime in on this topic as well!) Maynard I was kind of wondering if you had a schematic of the parallel SEP, using the 6SF7 driver and 6AQ5 for output. I have a pair of James 6123HS universal output transformers, with 2.5K, 3.5K, and 5K primaries. The are 20 watts, but the primary current rating is 90 milliamps. I was reading that thread of the parallel SEP amplifier, and you mentioned you were running the output transformer right at the max rated primary current...but the Tubelab guy tested and determined the Hammond OPT gave a extra 10% above the rating. I'm not sure if I should run at the max rated primary current, or exceed it with these James OPTs. Of course, they don't list those sort of things in spec/measurement data. I'd like to breadboard the circuit for some reason. Mike Edited April 2, 2014 by mike stehr Quote Link to comment Share on other sites More sharing options...
tube fanatic Posted April 2, 2014 Author Share Posted April 2, 2014 Mike, if you can wait a week or so, I'll be able to get a schematic to you. After all of the years spent with this stuff, everything is part of my mind, so I don't bother drawing anything out as a rule. Regarding the James xfmrs, I've never had any experience with them. The whole issue is whether the dc for the output tube plates will cause the xfmr cores to saturate (and if they do, you'll know it by listening!). Running even 10% over spec certainly shouldn't cause any burnout concerns, especially at only 250V. But, in this case, the output stage plates should be drawing around 90 ma +/- a couple, so I wouldn't give it a thought. If you're concerned, you can always drop the plate/screen voltages a bit. The 2.5k impedance is what you will need. In any event, I'll pm you when the schematic is ready and can either post it for others to use also, or email it to you. Thanks for the interest! Maynard Quote Link to comment Share on other sites More sharing options...
mike stehr Posted April 3, 2014 Share Posted April 3, 2014 Mike, if you can wait a week or so, I'll be able to get a schematic to you. After all of the years spent with this stuff, everything is part of my mind, so I don't bother drawing anything out as a rule. Regarding the James xfmrs, I've never had any experience with them. The whole issue is whether the dc for the output tube plates will cause the xfmr cores to saturate (and if they do, you'll know it by listening!). Running even 10% over spec certainly shouldn't cause any burnout concerns, especially at only 250V. But, in this case, the output stage plates should be drawing around 90 ma +/- a couple, so I wouldn't give it a thought. If you're concerned, you can always drop the plate/screen voltages a bit. The 2.5k impedance is what you will need. In any event, I'll pm you when the schematic is ready and can either post it for others to use also, or email it to you. Thanks for the interest! Maynard I'm in no hurry Maynard. I've got 6V6 puss-pull mono amplifiers that I need to order parts for, and get that project out of the way first. Thanks! Quote Link to comment Share on other sites More sharing options...
JL Sargent Posted April 3, 2014 Share Posted April 3, 2014 Beautiful job Maynard. Fun to see some of your work after these years of tube amp discussion. Really like the amp. I guess we tend to like the big iron OPTs around here I was not aware of those guys you chose, the 125CSEs. According to the referenced article it sounds like they have clean reproduction on the low freqs. to about 5 watts? Wouldn't a larger Hammond have been more complimentary to your fine work though? I understand there are economics but I want bigger OPTs on that chassis LOL. We gotta have some bass! Quote Link to comment Share on other sites More sharing options...
tube fanatic Posted April 3, 2014 Author Share Posted April 3, 2014 Beautiful job Maynard. Fun to see some of your work after these years of tube amp discussion. Really like the amp. I guess we tend to like the big iron OPTs around here I was not aware of those guys you chose, the 125CSEs. According to the referenced article it sounds like they have clean reproduction on the low freqs. to about 5 watts? Wouldn't a larger Hammond have been more complimentary to your fine work though? I understand there are economics but I want bigger OPTs on that chassis LOL. We gotta have some bass! Thanks JL! Actually, the xfmrs in the amp are the 125ESE, a slightly larger unit designed for somewhat higher current demands. I had a pair of them on hand so they were used here. The CSE would have worked just as well in this particular amp. The frequency response of the two is identical. You would be amazed at just how low these budget xfmrs can go with some circuit tweaks, making it difficult at times to justify the greatly increased expense to go with the really big stuff. The bass of this unit, and the parallel SEP which I posted previously, is something to behold. But, since you like the larger "look" here's a picture of my reference SET against which I design all of my others. Enjoy!!! Maynard Quote Link to comment Share on other sites More sharing options...
JL Sargent Posted April 3, 2014 Share Posted April 3, 2014 (edited) Really like that one too. Here is one I rebuilt with some upgrades with Craigs help back in 2011. We used Hammond 1650Fs on this one. https://community.klipsch.com/index.php?/topic/127925-old-tube-amp-revived/?hl=%20craig%20%20and%20%20carlson%20%20and%20%20hammond Edited April 3, 2014 by JL Sargent Quote Link to comment Share on other sites More sharing options...
tube fanatic Posted April 3, 2014 Author Share Posted April 3, 2014 Love those NOS Stromberg-Carlson tubes!!! I only use old stock US/Canadian/Japanese (with the occasional Mullard, Telefunken, or Amperex thrown in) tubes and can truly say that I've never even tried any of the modern production types. Just got asked to design and build a 2-3 wpc SET around some Sylvania 6BX7s, as well as 2 other lower power SEPs. Those will have to wait until the spring/summer chores are under control and the radio restoration work is caught up; but, it's nice to have some challenging work waiting in the wings. Maynard Quote Link to comment Share on other sites More sharing options...
JL Sargent Posted April 3, 2014 Share Posted April 3, 2014 Maynard, what would really be cool if you would share your design on the SET and do a build thread!!! Quote Link to comment Share on other sites More sharing options...
Recommended Posts
Join the conversation
You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.