I am looking at building the Kegger / Blueglow KT88 SET amp and I would like to see how well I can make it perform with upgraded parts. And, well, it is probably not fair to say I am willing to spend every dollar possible on it, I am open to a significant investment.
I plan to start with Hashimoto output transformers, power supply and choke. For the coupling capacitors, I purchased Duelund CAST PIO SN-CU. I am also considering adding Audio Note silver tantalum non-magnetic resistors and Cardas binding posts.
For those who built the amp or have thoughts regarding the circuit design, what else would you upgrade? Would you only use AN tantalum resistors in select positions? I have also heard carbon based resistors are better for the grid and wire wound resistors are better for the plate load positions, do you agree and have recommendations re brands?
I am newer to building tube amps, but I have a fair amount of experience in HiFi and have a very resolving home audio system. This is my reason for wanting to maximize the design.
Thanks for your thoughts. Here is a link to the BOM Mark shared. A big thanks to him for all his work on this version! https://blueglowelectronics.wordpress.com/diagrams/
I plan to start with Hashimoto output transformers, power supply and choke. For the coupling capacitors, I purchased Duelund CAST PIO SN-CU. I am also considering adding Audio Note silver tantalum non-magnetic resistors and Cardas binding posts.
For those who built the amp or have thoughts regarding the circuit design, what else would you upgrade? Would you only use AN tantalum resistors in select positions? I have also heard carbon based resistors are better for the grid and wire wound resistors are better for the plate load positions, do you agree and have recommendations re brands?
I am newer to building tube amps, but I have a fair amount of experience in HiFi and have a very resolving home audio system. This is my reason for wanting to maximize the design.
Thanks for your thoughts. Here is a link to the BOM Mark shared. A big thanks to him for all his work on this version! https://blueglowelectronics.wordpress.com/diagrams/
Attachments
''Upgraded'' parts will not improve this circuit. You need to start with one output tube and draw the optimum load line and build to that line with the best operating point you can plot for class A and find the parts that can reach that performance. This is a toy experimenter's amp that is just 1/2 of a PP amp, with a class AB operating point, not set up for optimal class A because the power tubes demand more idle current than the PS and OPT's can carry. Start from scratch if you want to put a lot of investment into an amp.
Thanks for your thoughts, 20to20. I ended up picking this circuit because it is listed as a SET configuration that is relatively simple to build. I have always found some magic in SET amps and in my humble experience have found that simple is often better (when possible).
If I were to run this with EL34 or 5881 tubes, would you have more confidence in the circuit to drive true class A power?
I recognize that the topic of upgrading parts can be a contentious one and sometimes seen as buying into snake oil. I am a musician and have pretty attuned ears and have found differences in my system when upgrading capacitors, cables and power supply components in other amps. But I know perspectives on the value of these changes can very.
If I were to run this with EL34 or 5881 tubes, would you have more confidence in the circuit to drive true class A power?
I recognize that the topic of upgrading parts can be a contentious one and sometimes seen as buying into snake oil. I am a musician and have pretty attuned ears and have found differences in my system when upgrading capacitors, cables and power supply components in other amps. But I know perspectives on the value of these changes can very.
If you choose the EL34 to build an SE Triode, you would pick a different PT to give a plate voltage under 375v. The OPT would be chosen for a primary Z @3000 ohms, no UL tap needed. Since the B+ would be lower you would need to redesign the driver supply, after you pick a particular driver tube with the gain to give you @ 26v peak drive to the EL34. That would mean ditching the LED's as designed now to get the driver biased as needed. And because you would be getting significantly less power out of the EL34 and less of that to the speaker, you could choose a 10W OPT instead of a 25W as specified. So, no. No confidence.
It's easy to throw big money at a new hobby with the expectation of fantastic results. I've been messing with SE amps for almost 5 years now and I've built some nice sounding amps, but also spent $$ I didn't need to and have accumulated a lot of parts that were not quite what a project required. So do your research before you spend that money.
I think it's nuts to match Hashimotos with a pentode. They are so nice they deserve a power triode. That's a 300b, 2A3, 6B4G or Soviet 6C4C. Those are all Directly-Heated Cathode tubes, so you'd need a more complex heater supply scheme than the KT88 amp with its Indirectly-Heated Cathodes. And of course power triodes are considerably more expensive than the lowly pentode/beam tetrode.
I would suggest you start with a project from CascadeAudio.com such as the Spalted Alder 6L6 Amp, and read through all those projects to get an idea of what's possible with a reasonable budget. Edcor CX series output transformers are nice but you can get great sound out of the GX 15w OPTs at 1/3 the price.
Make those first timer mistakes with a starter amp, then you can build your Baby Ongaku 🙂
w
I think it's nuts to match Hashimotos with a pentode. They are so nice they deserve a power triode. That's a 300b, 2A3, 6B4G or Soviet 6C4C. Those are all Directly-Heated Cathode tubes, so you'd need a more complex heater supply scheme than the KT88 amp with its Indirectly-Heated Cathodes. And of course power triodes are considerably more expensive than the lowly pentode/beam tetrode.
I would suggest you start with a project from CascadeAudio.com such as the Spalted Alder 6L6 Amp, and read through all those projects to get an idea of what's possible with a reasonable budget. Edcor CX series output transformers are nice but you can get great sound out of the GX 15w OPTs at 1/3 the price.
Make those first timer mistakes with a starter amp, then you can build your Baby Ongaku 🙂
w
I think the plate to plate feedback can be very effective but it's best between pentodes, or a least with a pentode driver stage. I've built an amp like this with very great subjective and measured performance.
https://hifihaven.org/index.php?threads/sakuma-schade-6v6-amp.8961/page-2#post-200479
Also have a breadboard of a similar but scaled up nature using GU-50 with a 6p15p driver and parallel feed output stage. It has very wide bandwidth and clear and detailed sound. My experience with the Edcor SE transformers is not so good. They have some serious HF resonance peaking. I've used MPS (Musical Power Supplies) output with great success and superior performance at a great price too.
Just 2 cents.
John
https://hifihaven.org/index.php?threads/sakuma-schade-6v6-amp.8961/page-2#post-200479
Also have a breadboard of a similar but scaled up nature using GU-50 with a 6p15p driver and parallel feed output stage. It has very wide bandwidth and clear and detailed sound. My experience with the Edcor SE transformers is not so good. They have some serious HF resonance peaking. I've used MPS (Musical Power Supplies) output with great success and superior performance at a great price too.
Just 2 cents.
John
Appreciate the different perspectives here. I am certainly open to taking more time to study and learn prior to the build. I have a background in quality and design, so I usually do pretty well with attention to detail/planning ahead.
If I can help it, I would rather start by making a good/great amp that I will use regularly than spend money on a project amp that will sit on the shelf. I had been looking at this build, as Mark mentioned in his videos that this was one of the better sounding amps he had built. And the distortion levels were consistently sub 1% with the right tube combination.
I am open to other SE design ideas (2a3, 300b etc), but am hoping to avoid something too complicated…Just looking for that purity of tone that makes the sound of the music like I am there.
If I can help it, I would rather start by making a good/great amp that I will use regularly than spend money on a project amp that will sit on the shelf. I had been looking at this build, as Mark mentioned in his videos that this was one of the better sounding amps he had built. And the distortion levels were consistently sub 1% with the right tube combination.
I am open to other SE design ideas (2a3, 300b etc), but am hoping to avoid something too complicated…Just looking for that purity of tone that makes the sound of the music like I am there.
Last edited:
In that case just build a 2a3 amp. 3.5 watts of clean Class A power with strong 2nd harmonics. Power triodes add a significant amount of 2nd and even harmonics, where the pentodes also add 3rd and the odds. Thus you end up trying to limit distortion, just like in a SS amp. DHTs just make an amp sound better.
JeLabs Simple45/2A3 is a good starting point, as is any Loftin-White-based circuit.
Here's the schematics to my 6B4G SET amp. Listen to it every day. NOS tubes for this amp are wayyyyy cheaper than 2a3s.
JeLabs Simple45/2A3 is a good starting point, as is any Loftin-White-based circuit.
Here's the schematics to my 6B4G SET amp. Listen to it every day. NOS tubes for this amp are wayyyyy cheaper than 2a3s.
Last edited:
The tubelab 300b (tseII) is a great amp. Relatively easy to build and sounds fantastic with 5k output transformers.
https://www.diyaudio.com/community/forums/tubelab.118/
https://www.diyaudio.com/community/forums/tubelab.118/
Thanks for the additional ideas @dubadub and @woodturner-fran. I will take a closer look at those.
I think I would like to stick with point-to-point wiring. I see JeLabs made a 300b with a 76 tube as the driver. That also seems like an interesting design too.
I think I would like to stick with point-to-point wiring. I see JeLabs made a 300b with a 76 tube as the driver. That also seems like an interesting design too.
Spend money on parts if you want. We build stuff for fun. Plenty of folks have extremely expensive tires on cars they never take to the track or Jeeps that they never run on real trails. Some folks spend as much on oil changes than what a pair of Tangos or Monoliths costs. Some of us spent large amounts on fancy parts for chip amps back in the day.
Do some research before committing to a circuit. Some of the older 300b designs mentioned above may not be what you want. Check out Skunkie Designs YouTube take on 300bs. You can point to point a Tubelab TSE, especially if you put SS bits on tiny perf board. But there are lots of options for good strong driver circuits. There are other options on this forum.
While I don't discourage fancy parts, do know that point to point layout and grounding can be tricky. So don't get too excited about getting it right the first time. Allow yourself plenty of space. Whether the first time or the tenth, they don't always go to plan.
If you can contemplate two, there are worse ideas than doing a Kegger KT88 it Tubelab SSE with decent parts first followed with a fancy 300b build.
Paul
Do some research before committing to a circuit. Some of the older 300b designs mentioned above may not be what you want. Check out Skunkie Designs YouTube take on 300bs. You can point to point a Tubelab TSE, especially if you put SS bits on tiny perf board. But there are lots of options for good strong driver circuits. There are other options on this forum.
While I don't discourage fancy parts, do know that point to point layout and grounding can be tricky. So don't get too excited about getting it right the first time. Allow yourself plenty of space. Whether the first time or the tenth, they don't always go to plan.
If you can contemplate two, there are worse ideas than doing a Kegger KT88 it Tubelab SSE with decent parts first followed with a fancy 300b build.
Paul
The only one I heard was a Paul build, and it had a triode connected 6EJ7 front tube. It worked quite well. I admit, I was trying to get Paul to run the circuit with a full perntode front end and use the E-Linear connection but he never got to it( his version did do better than I expected, so THERE! ).
Douglas
Douglas
I poked my head in here to say, if I were doing a cost-no-object build, I'd pick a 300B amp (I built a Tubelab TSE-II with 300B's that I absolutely love) and go nuts on the output transformers and tubes. Hell you can spend $1500 on a pair of new Western Electric 300Bs alone 😉
But I don't know what speakers you're running or what the rest of your system consists of, or your musical tastes, so my recommendation might be totally off the mark!
My advice is build whatever makes you happy. The Blue glow project is well documented which is a big plus, although I can't speak to how it sounds. There are many very well documented projects here that run the gamut from high power push-pull to fleawatt single-ended to projects designed around big glowing radio transmitter tubes or incorporating exotic (and potentially dangerous) parts like mercury vapor rectifiers. Before you commit to any project, look around at the threads with dozens of pages of replies and you'll find some of the better documented, honed, and iterated projects that have been well received by the forum. I doubt you'll be disappointed whichever you choose, I think the best sounding amp is always the one you worked hard to build yourself.
But I don't know what speakers you're running or what the rest of your system consists of, or your musical tastes, so my recommendation might be totally off the mark!
My advice is build whatever makes you happy. The Blue glow project is well documented which is a big plus, although I can't speak to how it sounds. There are many very well documented projects here that run the gamut from high power push-pull to fleawatt single-ended to projects designed around big glowing radio transmitter tubes or incorporating exotic (and potentially dangerous) parts like mercury vapor rectifiers. Before you commit to any project, look around at the threads with dozens of pages of replies and you'll find some of the better documented, honed, and iterated projects that have been well received by the forum. I doubt you'll be disappointed whichever you choose, I think the best sounding amp is always the one you worked hard to build yourself.
@pjanda1, @Bandersnatch and @metaphile that is all good advice. I have read a lot of good things about the Tubelab designs in other places too, so I will give that a good look. Skunkies 300B is also on the short list.
While I don’t have any fancy cars, vacation houses or high dollar liquor, I have invested more in my sound system than the Mrs. cares for. There is just something about continuing to refine and hone a hifi system and I enjoy trying different configurations/flavors. My primary amp right now is a Decware Torri MK5 and I am ultimately looking to build something that is on the same playing field. That said the Torri is an excellent amp.
While I don’t have any fancy cars, vacation houses or high dollar liquor, I have invested more in my sound system than the Mrs. cares for. There is just something about continuing to refine and hone a hifi system and I enjoy trying different configurations/flavors. My primary amp right now is a Decware Torri MK5 and I am ultimately looking to build something that is on the same playing field. That said the Torri is an excellent amp.
Lots of great advice! @metaphile brings up an excellent point, especially on the heals of your expectation of a build being “on the same playing field”. Need to define that field, the game and the rules. Speaker(s) being used, room(s) size, music choices, listening levels. You are on the heals of starting a glorious/frustrating/exciting/rewarding/learning/addictive journey assuming you decide to accept the challenge!! I love this hobby and the people here (ALL smarter than me) will steer you right! Keep initial build simple / frugal and well document! No doubt it will give you the experience to know whether class A SET will make you happy! Then you can scale up from there!
A link to an explanation why it is not advisable to apply plate to plate feedback between a triode as a driver and a power tube can be found on pages 396 to 399 of this book:
Fundamentals of Radio Valve Technique, Deketh, Philips, 1949
Crawl before you walk.........walk before you run..........built something that works (AND KNOW WHY, like or dislike!) before you get drug down "THE HOLE" of letting perfection (????) to get in the way of progress! OR else you will never sling solder, learn, experience and decide whether DIY is the right path!
Thanks for sharing, Robert. This looks like a good resource. I assume you are referring to the following section in particular?
“As a consequence the a.c. resistance in the anode circuit of the preamplifying valve is also much smaller than the internal anode resistance R^, to which is to be ascribed the reduction of amplification due to feedback. When a pentode is used as pre-amplifying valve only the amplification is reduced; the dynamic transfer characteristic remains essentially the same as the static characteristic. When a triode is used for pre-amplifying, the distortion, however, is greater due to the smaller external resistance.”
Last edited:
@emuffler that is exactly the kind of journey I am looking to take! I enjoy working with my hands, taking time to focus in a particular task and do that one thing well. That is part of the reason I like the idea of point-to-point wiring. It also feels like there is more freedom in it.
That is not to say one can not build a pentode. AKA the cascode. This assembly is generally a bit wasteful of B+. Using a MOSFET as the screening element reduces this use of available B+. For a SE amp, which has much more solid restrictions on B+( as it *HAS to operate in Class A ) the amount of B+ needed is half.
Here is a pic of the plate curves of a 6H6Pi and a 'FET. If another 6H6Pi section were used as the screen, the vertical line where the useful bit of the plate curves exist would be at 2x the voltage. As with a pentode, increasing the gate voltage from about 60 to ~75V raises the g1=0V current from 14 mA to ~20 mA in this example. This for the time being ignores the MOSFET gate-source voltage, and for a low voltage assembly like one could get with a lower triode element like a 7233 or 6C19pi, would matter a bit more.
Douglas
attribution; the 6H6Pi's plate curves came from the Klausmobile site... 🙂
Here is a pic of the plate curves of a 6H6Pi and a 'FET. If another 6H6Pi section were used as the screen, the vertical line where the useful bit of the plate curves exist would be at 2x the voltage. As with a pentode, increasing the gate voltage from about 60 to ~75V raises the g1=0V current from 14 mA to ~20 mA in this example. This for the time being ignores the MOSFET gate-source voltage, and for a low voltage assembly like one could get with a lower triode element like a 7233 or 6C19pi, would matter a bit more.
Douglas
attribution; the 6H6Pi's plate curves came from the Klausmobile site... 🙂
