mike stehr Posted November 8, 2015 Share Posted November 8, 2015 Nuthin' yet...just got the parts. Figured I start my own thread. It won't be a complete build thread, just a bread-board circuit for now. 1 Quote Link to comment Share on other sites More sharing options...
cincymat Posted November 8, 2015 Share Posted November 8, 2015 Please post as many photos as you can. I'd like to do a tube amp build as a winter project. Need to learn more before I try so seeing someone else's project sure speeds the process. thanks, Mark Quote Link to comment Share on other sites More sharing options...
mike stehr Posted November 8, 2015 Author Share Posted November 8, 2015 Please post as many photos as you can. I'd like to do a tube amp build as a winter project. Need to learn more before I try so seeing someone else's project sure speeds the process. thanks, Mark The circuit will be set-up on a piece of wood, with some clip leads connecting things. The circuit won't have a power supply, as I'll be using a tube DC power supply for voltage and it's 6.3 volt heater winding for the 6FS5. Each parallel pair of 6AQ5 will have it's own 6.3 volt heater winding, which both will be powered from a variac. And if I decide to make a complete amplifier or a pair of mono amps, I won't be using the James output transformers. The are universal OPTs, and are nice for bread-boarding various circuits of interest. I'll have to source some OPTs, and that will have a difference in the overall sound of the amplifier circuit. So I don't know what you can really glom from the thread...I'll post what voltage numbers I can...I'll need to noodle with the RC feedback (with a scope) to adjust for the James OPTs, which would have to be accounted for and adjusted again if another OPT was sourced. Mike Quote Link to comment Share on other sites More sharing options...
mike stehr Posted November 23, 2015 Author Share Posted November 23, 2015 Got this far, with one side lashed up. I'm currently noodling with the RC feedback. Quote Link to comment Share on other sites More sharing options...
justinsweber Posted November 23, 2015 Share Posted November 23, 2015 looks great. The big James are is wonderful stuff. Quote Link to comment Share on other sites More sharing options...
tube fanatic Posted November 23, 2015 Share Posted November 23, 2015 Nice breadboarding Mike! Post some 'scope traces and other technical details as it evolves! Maynard Quote Link to comment Share on other sites More sharing options...
mike stehr Posted December 8, 2015 Author Share Posted December 8, 2015 Okay, Maynard and I have been conversing through emails, and he's been guiding myself along answering questions, and making sure I keep my output transformers within their DC current rating. He mentioned to start to the circuit out with 255 volts between the screens and cathodes of the 6AQ5s. I ramped up the voltage on the power supply, measuring until I had 255 volts. This put B+, at around 270 volts. This put the DC current for the OPTs over the 90ma rating, at around 100ma. So dropping the B+ on the power supply solved that. With the screen to cathode voltage of 6AQ5 at around 243 volts, puts the B+ at 255 volts. Within/at the current rating of the James 6123HS...90ma. (I think 90ma may be a conservative number, but I don't want to push it too much either.) Then for the RC feedback network, Maynard mentioned some cap and resistor values to try. Right now I'm using a 4700pF cap, a 500K pot, with 400K in series. With the 4700pF cap and a little over 800K, I show a good square-wave at 500Hz. I figure I can use the 4700pF cap and 750K, and still be fine. Maynard mentioned the square wave at 500Hz looked quite nice as well without the 4700pF cap, and the 800K or so resistance. I'm to the point of connecting everything to the main stereo, (Cornwalls) but I only have one side made. But I should just do it anyway, so I can listen to the difference with and without the 4700pF cap in the feedback network. (power it up, listen. power down, pull cap, power up, listen) I may not need that cap. Maynard has been great help pulling me along, here. Thanks again. I'm anxious to hear this circuit...but I also got a bathtub/shower surround project I have to get over with...it's gruesome... First pic is 500Hz with the 4700pF cap. Second is 500Hz without the cap. Third is 3kHz with the cap. Quote Link to comment Share on other sites More sharing options...
tube fanatic Posted December 8, 2015 Share Posted December 8, 2015 As I mentioned to Mike, those James opts are impressive. I've never had any experience with them but, looking at the 'scope pics he's sent me, they are very well made. This amp circuit, when used with CWs, is amazing. Will anxiously await Mike's listening impressions when the other channel is built so he can evaluate in stereo! Maynard Quote Link to comment Share on other sites More sharing options...
xxJPMxx Posted December 8, 2015 Share Posted December 8, 2015 Hello Mike, How much feedback are you using? 800k is a large value for a feedback resistor!! See the ringing on the 3kHz square wave? What you want to do is use a 10kHz square wave and change the cap value until the ringing is mitigated. Quote Link to comment Share on other sites More sharing options...
xxJPMxx Posted December 8, 2015 Share Posted December 8, 2015 My last push pull amps I used a good amount of feedback, 20db. I ended up adding a Lag compensation network at the plate of the input stage to improve performance. Here is a good article on calculating the compensation networks: http://www.angelfire.com/electronic/funwithtubes/Amp-Compensation.html Quote Link to comment Share on other sites More sharing options...
mike stehr Posted December 21, 2015 Author Share Posted December 21, 2015 (edited) Amazing bass with a Cornwall...for one side! Edited December 21, 2015 by mike stehr Quote Link to comment Share on other sites More sharing options...
mike stehr Posted December 22, 2015 Author Share Posted December 22, 2015 Thanks, echo. Hello Mike, How much feedback are you using? 800k is a large value for a feedback resistor!! See the ringing on the 3kHz square wave? What you want to do is use a 10kHz square wave and change the cap value until the ringing is mitigated. I'm pretty sure Maynard had emailed you a schematic of the circuit. The feedback is Schade, plate to plate, plate to grid, or whatever the hell they call it. High-pass RC feedback network is what Maynard calls it. The schematic of Maynard's shows 560K in the feedback network, I'm using 680K right now, and could probably drop the value down to 560K without much change of the square-wave response at 500Hz. I am aware of using feedback from a transformer secondary to the cathode of a driver tube, which is generally only a few volts involved. Establish the value of feedback R, and then beat down the ringing/overshoot with a compensation cap. I would use 5kHz and 10kHz to monitor square wave response. Maynard has given suggestions to possibly mitigate the ringing at the input, and I have tried them, but the open air nature of a breadboard with clip-leads to and fro, make for many noises. Where if all the stuff is in a shielded chassis, the noise issues would be less problematic for use of the pots. With a chassis, I can go ahead and try the given suggestions again. I skipped the pots, and used a 50K stepped attenuator between the source CDP and the amplifier circuit. Things sound best to my ears listening to this circuit, with the B+ at 250 volts. Nonetheless, I bookmarked the link regarding compensation networks, and have been plugging through it. Maynard has pretty much convinced me to build a complete amplifier. All I need to do to listen to the breadboard circuit in stereo is buy another 6.3 volt filament transformer. I can do that while I ponder...thinking of getting the chassis and PS xformers, build the amplifier, and then just lash in the James OPTs until I figure out what I'll do for output transformers. Quote Link to comment Share on other sites More sharing options...
xxJPMxx Posted December 22, 2015 Share Posted December 22, 2015 Hey Mike, Yes Maynard did email me the schematic. I do apologize for I thought you were using feedback with the transformer in the loop. The high value resistor and series cap makes sense now, thanks for clarifying. For tube amps 2kHz square wave testing is most important for it usually represents what the performance at the upper limit of our hearing. For real overkill I do like to run 10kHz square waves through for kicks but most people tell me that's overkill. 5kHz is a good compramise. And yes I would think flying leads from breadboarding would create a large parasitic capacitance at the plate of the driver tube. I have used an RC network around the plate load resistor in the past with good results, but I wouldn't worry about that until you get it in it's chassis with the permanent layout. Another thing you could try is shunting a 1nF 1kV cap from the plates of the 6BQ5 to ground and see if the square wave improves a bit. What you have is good so maybe no improvements will be had. Sounds like you only plan to use the James outputs temporarily? Those are nice, I would just build an amp around them. Quote Link to comment Share on other sites More sharing options...
mike stehr Posted December 22, 2015 Author Share Posted December 22, 2015 (edited) Hey Mike, Yes Maynard did email me the schematic. I do apologize for I thought you were using feedback with the transformer in the loop. The high value resistor and series cap makes sense now, thanks for clarifying. For tube amps 2kHz square wave testing is most important for it usually represents what the performance at the upper limit of our hearing. For real overkill I do like to run 10kHz square waves through for kicks but most people tell me that's overkill. 5kHz is a good compramise. And yes I would think flying leads from breadboarding would create a large parasitic capacitance at the plate of the driver tube. I have used an RC network around the plate load resistor in the past with good results, but I wouldn't worry about that until you get it in it's chassis with the permanent layout. Another thing you could try is shunting a 1nF 1kV cap from the plates of the 6BQ5 to ground and see if the square wave improves a bit. What you have is good so maybe no improvements will be had. Sounds like you only plan to use the James outputs temporarily? Those are nice, I would just build an amp around them. It's all good. The guy that mentored me back when he lived up here would recommend 5kHz-10kHz square waves for testing. But he also said don't be afraid to go up and down in bandwidth and at least see what's going on. I might be able do some things with the breadboard circuit to possibly makes things a bit more quiet. Maybe shorten the plate leads to the OPT, B+, move some things around. I haven't mentioned it to Maynard yet, but a lot of the noise is centered around the 6SF5. You get your hand near it, and it growls a bit...you touch it, and it growls. Dunno if being in a breadboard layout mounted on a testing type socket may be the reason. Can a 6SF5 be microphonic? Would it be wise to ground the shell? I could try the 1nF (nicofarad?) from plate to ground of the "6AQ5s"...6005s actually...never hurts to try, it's on the bench. The James are 20 watt universal OPTs with 2.5K,3.5K,and 5K primary impedances. They are fun to experiment with a variety of tubes. Today I decided to break the old Scott LK48B after a year and ram some NWA down it's throat, and finish the wood case for it. Edited December 22, 2015 by mike stehr Quote Link to comment Share on other sites More sharing options...
xxJPMxx Posted December 22, 2015 Share Posted December 22, 2015 Hello Mike, it looks like the input is the yellow wire to pin 3? If so I don't see a grid stop resistor. Maybe throw a big fat grid stop resistor as close to the pin as possible, 10k will do. Or maybe try grounding pin 1, the shell like you said. I would probably do both, easy enough to try. As far as the cap value it's 1 nanofarad or 1000 picofarads or .001 microfarads. I see now the James is versatile with all it's tap options. I have never used them but they look great and I bet they sound great too!! NWA through the Scott? You are cray cray!!!!! I haven't listened to them in years LOL. Actually you are kinda taking me by surprise with that musical selection. I don't judge because I listen to everything myself......I am all over the place. Anyway that Scott looks like it's in great shape I love it!!! Quote Link to comment Share on other sites More sharing options...
tube fanatic Posted December 23, 2015 Share Posted December 23, 2015 (edited) Hey Mike, Yes Maynard did email me the schematic. I do apologize for I thought you were using feedback with the transformer in the loop. The high value resistor and series cap makes sense now, thanks for clarifying. For tube amps 2kHz square wave testing is most important for it usually represents what the performance at the upper limit of our hearing. For real overkill I do like to run 10kHz square waves through for kicks but most people tell me that's overkill. 5kHz is a good compramise. And yes I would think flying leads from breadboarding would create a large parasitic capacitance at the plate of the driver tube. I have used an RC network around the plate load resistor in the past with good results, but I wouldn't worry about that until you get it in it's chassis with the permanent layout. Another thing you could try is shunting a 1nF 1kV cap from the plates of the 6BQ5 to ground and see if the square wave improves a bit. What you have is good so maybe no improvements will be had. Sounds like you only plan to use the James outputs temporarily? Those are nice, I would just build an amp around them. I haven't mentioned it to Maynard yet, but a lot of the noise is centered around the 6SF5. You get your hand near it, and it growls a bit...you touch it, and it growls. Dunno if being in a breadboard layout mounted on a testing type socket may be the reason. Can a 6SF5 be microphonic? Would it be wise to ground the shell? Mike, the 10 meg grid leak resistor body should be mounted as close to pin 3 as possible, with the other end grounded of course. Pin 1 should be grounded to allow the metal shell of the tube to act as an RF shield (a nearby magnetic field should simply flow around the shell). A convenient way to do it is to tie pins 1 and 2 together, run a lead to the ground buss, and simply put the 10 meg resistor between pins 1 and 3 or 2 and 3 as close to the socket as possible. And, of course, you must AC couple to the grid or the tube will not be biased properly. That should eliminate the problem, although the unshielded bread-board arrangement will still allow all kinds of "stuff" to get into the signal circuitry. Let us know if that helps. Regarding the upper freq. square wave testing, JP is correct. I generally use 3 kHz. In fact some opts, like the huge Hammond 16XXSEA series seem to drop off very quickly above 30 kHz, so looking at the 10 kHz square wave is kind of a moot point. I don't know about James, Transcendar, or Edcor in that regard. In actual practice, my own experience is that if the amp extends flat to even 12-13 kHz, the sonic presentation can be gorgeous. Regarding microphonics, the 6SF5 can be as prone to that as almost any other tube. Most are fine, but trying one or two others can never hurt. Maynard Edited December 23, 2015 by tube fanatic Quote Link to comment Share on other sites More sharing options...
mike stehr Posted December 24, 2015 Author Share Posted December 24, 2015 I may go ahead and swap out the test socket for a common octal socket on stand-offs, like I did with the 7 pin sockets. Then I can wire the socket as you explain, and hang the 10 meg G1 resistor right off the mentioned pins to the ground buss. I was listening to the circuit with the 6SF5 shell ungrounded, and think by merely grounding the shell alone would alleviate most of the noise. I wouldn't doubt this was the ripple I had seen when messing with the pots monitoring the scope. The Majority of the noise was heater growl from running the 6SF5 heaters floated, which I lowered quite a bit by running a 100 ohm R from one side of the winding to ground. So a pair of 100-300 ohm resistors from each side of the heater winding to ground, use a octal socket and ground the shell. It all must wait until Christmas has passed...and we've had little if any snow in these parts for the last two years. 18 inches all last year...and we just received that within the last two-three days. Bam! Now I gotta shovel my roof. yay Last Summer it was so dry you couldn't think about a wildfire without one breaking out, and now we're buried in snow. Climate change? Constantly! Last November we had a windstorm here with gusts over 70mph...any tall bull pine with a weak base went over, spruces as well...over 400 monster trees topple in the city, wiping all the power lines...then it got cold. Quote Link to comment Share on other sites More sharing options...
tube fanatic Posted December 26, 2015 Share Posted December 26, 2015 Mike, if you want an unbiased (I know, that's a terrible pun ) impression about the performance of the parallel SEP get in touch with Jim. He's now enjoying mine!!! Maynard Quote Link to comment Share on other sites More sharing options...
Jim Posted December 26, 2015 Share Posted December 26, 2015 I'm listening to the Beastie Boys right now and the bass makes me feel as if I'm in da hood back in Jersey City with my old car with the 15's in the trunk. The soundfield is spread about a good 2 feet from the outside of the speakers and I'm still playing with the set up. Volume is only at about 9 o clock as well.Bass digs deeper than the SET and imaging is almost 3D like a SET. Quote Link to comment Share on other sites More sharing options...
mike stehr Posted December 26, 2015 Author Share Posted December 26, 2015 (edited) I'm listening to the Beastie Boys right now and the bass makes me feel as if I'm in da hood back in Jersey City with my old car with the 15's in the trunk. The soundfield is spread about a good 2 feet from the outside of the speakers and I'm still playing with the set up. Volume is only at about 9 o clock as well.Bass digs deeper than the SET and imaging is almost 3D like a SET. From what I've heard just listening to one side of a breadboard setup, I can tell that that strong deep bass isn't no problem. At 6-7 watts per side, the circuit will have more bass as in louder compared to my directly heated 2A3 stereo SET amplifier. As far low frequency bass, I dunno...that'll be a close call. The 2A3 amplifier can get down in the lower 30Hz registers. But from hearing one side of the parallel 6AQ5 SEP amplifier circuit, it seemed to be getting down in the lower bass frequencies rather well. I only have one side lashed up, so I can't comment on imaging. Imaging should be close to the 2A3 amplifier, but that's one thing the 2A3 amplifier is really good at. I'll have to find out... I wanted a single-ended amplifier that can play some bass beats, and play some metal as well. It needs to delicate enough to play Jazz, and lighter genres of music without sounding thick/dark in the bass. This parallel SEP circuit can do it I'm thinking. I played 2live nasty as they wanna be first off, then Sublime. I knew lack of bass was no problem then. I played Brubeck "Take Five" last. the piano and horns stood out nicely, as they should...that's only one side. Just need to pony up for the PS supply x-formers, and a chassis, and a couple extra parts. It'll take some time. Edited December 26, 2015 by mike stehr 1 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.