Kyoto EL84 Mono Block build

[Jim Kiriakis]

Boy, you don't waste any time.

Thanks John. Next steps are to compare the Kyoto, I think first to the KG84. Will attempt this:

Audio source (phono and/or Magnavox CDB610 made in Belgium 80's CD player) into one Superphon Revelation preamp to Kyoto. Take tape out from it into another identical preamp driving the KG84. Each amp connected to two sets of econo bookshelf speakers, Polk RT-3s, pretty much identical though one set has new drivers. I will be able to toggle between the two amps while tracks are playing which may take some of the memory loss / perception delay out of the experiment.

 

[TriodeLuvr]

As we learned earlier the 100k resistor works great from grid to ground. Which then made me wonder about one of my recent SEP builds, the RH84 rev 2. It has a 1M driver grid resistor. Would you suggest changing that one? Schematic may be found here:

http://rh-amps.blogspot.com/2013/02/rh84-amplifier-revision-2_26.html

1 meg is too high. Every line stage I know of can drive 100K. The only possible reason I can think of to use a higher value is if you wanted to connect the old RCA phono preamp directly to the input of the amp. That design requires a 220K or higher load unless a buffer is added.

Jack

 

[Jim Kiriakis]

1 meg is too high. Every line stage I know of can drive 100K. The only possible reason I can think of to use a higher value is if you wanted to connect the old RCA phono preamp directly to the input of the amp. That design requires a 220K or higher load unless a buffer is added.

Jack

Jack -
That's completely helpful. In my RH84 build I will revise the driver grid resistor to 100k as I did with your Kyoto circuit. This exercise then prompts me to look back at my second most recent build, the KG84 (Kegger). Predictably, it looks like he got it right:

https://audiokarma.org/forums/index.php?threads/kegger-schematics-mega-thread.858662/

No need to change that one.

Thanks for educating us about input impedance. It's really nice having quiet amps.

Jim

 

[Jim Kiriakis]

Nice collection of single ended EL84 amps. How do they compare?

Currently toggling between the Kyoto (newest, least broken-in) and the Kegger. Auditioning Jeff Beck's "Truth" through two identical SS preamps (Superphon Revelation Basic, tape out from one into aux in the other). Identical sets of bookshelf speakers (Polk RT3). Both sound excellent.

Kyoto has the low cost Edcor OT's (XSE) vs. the Kegger with 8x more expensive Electra-Print OT's (partial silver). Kyoto has higher output Kegger slightly brighter. Both image extremely well. Please note the Kegger also has fancier tubes and the Kyoto runs the NECs harvested from an Akai M7. Also the Kyoto has two PTs is fully dual mono but a more minimalist power section 6x4 rectifiers vs. the Kegger w/ Amperex GZ34. Now listening to Bowie digital track through same inputs. Kyoto's low-mids a bit heavier.

This is a deliberately unfair comparison, as part of the goal was to see how good the Kyoto could perform with the minimalist OT's. The answer so far (and the amp is not broken in yet) is - very well.

If I walked out of the room left the controls to a third party and came back blindfolded it's entirely possible I couldn't differentiate them.

More to follow as I continue to fiddle. Need to try the RH84 in the mix next.

 

[TriodeLuvr]

Bear in mind that the Kyoto was "tuned" and tested with the Akai output transformers. So, in addition to differences in the transformers themselves, the amp may not be fully cooperative. Are you able to at least check the frequency response into a resistive load? That's a more essential test than listening when an un-vetted and newly built amp is first powered up.

Jack

 

[Jim Kiriakis]

Bear in mind that the Kyoto was "tuned" and tested with the Akai output transformers. So, in addition to differences in the transformers themselves, the amp may not be fully cooperative. Are you able to at least check the frequency response into a resistive load? That's a more essential test than listening when an un-vetted and newly built amp is first powered up.

Jack

Thanks Jack. Can you inform us how to check the frequency response into a resistive load? Happy and interested to try to.

And understand your point about tuning for the Akai OT's. Now I am thinking I may deconstruct my existing Akai's or try to find another pair.

Jim

 

[TriodeLuvr]

To test response, I first connect 8 ohm resistors to the outputs of both channels. Then I drive the input(s) with an audio generator and sweep manually from 20 Hz to 20 kHz while monitoring the output levels using a scope. There are other ways to do this, but that works for me as a quick test. You need to pay attention to the output level during a test like this. You'll want to first look at the small signal response, say with one or two volts pk-pk output. Then the test s repeated at full power. The latter will be a more severe test of low frequency performance. Just to be clear, frequency response is a very basic test that needs to be performed in order to get you closer to apples-to-apples comparisons. I do other testing on new amps as well, but it's more involved and requires additional equipment.

I think you'll want to try the Akai transformers sooner or later. They're not the equal of the Electra-Prints, but I'm certain they're part of the formula that makes these amps sound so good.

Jack

 

[Jim Kiriakis]

Take actual measurements from the PT and amp and then use the 3 charts on the data sheet, which serve as tests to see if a particular implementation is acceptable. If the amp passes all 3 tests you're good to go.

AC Plates measure 265V per plate. DC output @ filter is 295V, Per Rate Chart II I can ascertain rectification efficiency = .79

295/(1.41 x 265) = .79

This is within scale for Chart II on the horizontal axis

Not sure my multimeter can measure DC output per plate in milliamperes @ filter / choke input. That point to ground shows 146.5 which is way off the charts on the vertical axis so I am unsure it's an accurate measurement. Range on the chart is 0-50.

 

[Jim Kiriakis]

To test response, I first connect 8 ohm resistors to the outputs of both channels. Then I drive the input(s) with an audio generator and sweep manually from 20 Hz to 20 kHz while monitoring the output levels using a scope. There are other ways to do this, but that works for me as a quick test. You need to pay attention to the output level during a test like this. You'll want to first look at the small signal response, say with one or two volts pk-pk output. Then the test s repeated at full power. The latter will be a more severe test of low frequency performance. Just to be clear, frequency response is a very basic test that needs to be performed in order to get you closer to apples-to-apples comparisons. I do other testing on new amps as well, but it's more involved and requires additional equipment.

I think you'll want to try the Akai transformers sooner or later. They're not the equal of the Electra-Prints, but I'm certain they're part of the formula that makes these amps sound so good.

Jack

Thanks Jack. My son has an oscilloscope and mechanical engineering degree. Is it feasible to use an audio generator smartphone application, or do I need to source that piece of gear and if so, please suggest one.

I hear you re: trying the Akai transformers.

Jim

 

[FlaCharlie]

Not sure my multimeter can measure DC output per plate in milliamperes @ filter / choke input. That point to ground shows 146.5 which is way off the charts on the vertical axis so I am unsure it's an accurate measurement. Range on the chart is 0-50.

You're talking about Chart 1. Sounds like you're measuring voltage, not current.

According to the schematic, the current draw per channel is ~45mA. You can calculate the actual draw for both channels by taking the voltage drop between the two PS caps and dividing it by the DCR of your choke. Half of that will be the current per channel or per plate of the rectifier.

Or calculate it based on the voltage drop across the cathode resistors of the tubes in each channel.

 

[TriodeLuvr]

Thanks Jack. My son has an oscilloscope and mechanical engineering degree. Is it feasible to use an audio generator smartphone application, or do I need to source that piece of gear and if so, please suggest one.

I hear you re: trying the Akai transformers.

Jim

I've never used a smartphone for this. Just be sure the signal at the input is flat from 20 Hz to 20 kHz before measuring the output.

Jack

 

[TriodeLuvr]

Please note the Kegger also has fancier tubes and the Kyoto runs the NECs harvested from an Akai M7.

Don't sell them short. The NECs have proven to be the equal of NOS British Mullards in my subjective listening tests. I've also tried the Russian 6J32P, but didn't care for the sound. The other nice thing about using the original tubes is that '60s tape decks typically saw fewer hours of use than more universal equipment such as amplifiers and preamps. I've pulled many tubes from these decks that looked and tested exactly like new. The same is also often true of pulls from organs. People would buy an organ, piddle with it for six months or so, then lose interest. Bad for them, good for us.

Jack

 

[Jim Kiriakis]

To test response, I first connect 8 ohm resistors to the outputs of both channels. Then I drive the input(s) with an audio generator and sweep manually from 20 Hz to 20 kHz while monitoring the output levels using a scope. There are other ways to do this, but that works for me as a quick test. You need to pay attention to the output level during a test like this. You'll want to first look at the small signal response, say with one or two volts pk-pk output. Then the test s repeated at full power. The latter will be a more severe test of low frequency performance. Just to be clear, frequency response is a very basic test that needs to be performed in order to get you closer to apples-to-apples comparisons. I do other testing on new amps as well, but it's more involved and requires additional equipment.

I think you'll want to try the Akai transformers sooner or later. They're not the equal of the Electra-Prints, but I'm certain they're part of the formula that makes these amps sound so good.

Jack

Sooner as advised . Surgery complete on the Akais OTs, chokes and PTs are now uninstalled

 

[TriodeLuvr]

I've removed the iron from four of these amps so far for projects. Do I regret destroying the decks? Absolutely not, this is what I bought them for. In fact, I still have a few in storage that I picked up back when they were a little cheaper than now. I'll use those too eventually.

Jack

 

[Jim Kiriakis]

I've removed the iron from four of these amps so far for projects. Do I regret destroying the decks? Absolutely not, this is what I bought them for. In fact, I still have a few in storage that I picked up back when they were a little cheaper than now. I'll use those too eventually.

Jack

Jack -
Akai OTs are mounted on the Kyoto now I can re-wire.

Question: the Hammond 155h chokes measure 5H and 260ohm vs. Akais @ 6H 192ohm. Will it make a meaningful difference swapping the chokes as well?

Initially I will keep the Hammond 269jx PTs in place, unless advised otherwise.

Thanks,
Jim

EDIT - amp now running. I would agree that the Akai OT's sound superior to the Edcors.
EDIT 2 - listening to Bill Evans Trio Sunday at the Village Vanguard (1961). For a second I thought there were folks outside the house talking. Turns out you can clearly but faintly hear the audience in the room. Drums and upright bass sound quite good.

 

[Jim Kiriakis]

You're talking about Chart 1. Sounds like you're measuring voltage, not current.

According to the schematic, the current draw per channel is ~45mA. You can calculate the actual draw for both channels by taking the voltage drop between the two PS caps and dividing it by the DCR of your choke. Half of that will be the current per channel or per plate of the rectifier.

Or calculate it based on the voltage drop across the cathode resistors of the tubes in each channel.

Voltage drop (296.3 - 283.6) = 12.7 / DCR choke 262 = .0487328 / 2 = .02423664

Based on this looks like I am right in the center of the center of the Chart 1 graph

Thank you FlaCharlie

 

[TriodeLuvr]

The Hammond chokes should be fine. You can save the Akai chokes for another project. See what I mean about the sound quality? It's much better than it has a right to be. Hope you get many years of enjoyment from the amp!

Jack

 

[Jim Kiriakis]

The Hammond chokes should be fine. You can save the Akai chokes for another project. See what I mean about the sound quality? It's much better than it has a right to be. Hope you get many years of enjoyment from the amp!

Jack

Thank you Jack will leave the chokes be. I do hear your point about the sound quality and expect it will improve as it breaks in. Appreciate your schematic and follow up advice.

Jim

 

[TriodeLuvr]

Well, I might be making the first change to the original Kyoto later this year. The 7408 (6V6GT) outputs are getting tired, but I don't have more in my collection. OTOH, I bought a good quantity of 12V6GT about 10 years ago. They're identical to the 6V6 except for filament voltage, so in theory, no other modifications would be required. Also, the Akai power transformers are wired to use a separate 6.3V winding for the outputs. Any changes in that area won't affect the EF86s or rectifiers. Maybe this will be as simple as wiring the two 6.3V windings in series and revising the wiring to the two output tubes. If it works without introducing hum, I might even add a 6V-12V switch.

Jack

 

[Jim Kiriakis]

Well, I might be making the first change to the original Kyoto later this year. The 7408 (6V6GT) outputs are getting tired, but I don't have more in my collection. OTOH, I bought a good quantity of 12V6GT about 10 years ago. They're identical to the 6V6 except for filament voltage, so in theory, no other modifications would be required. Also, the Akai power transformers are wired to use a separate 6.3V winding for the outputs. Any changes in that area won't affect the EF86s or rectifiers. Maybe this will be as simple as wiring the two 6.3V windings in series and revising the wiring to the two output tubes. If it works without introducing hum, I might even add a 6V-12V switch.

Jack

Go for it Jack. Continuing to burn in the new Kyoto now. One note - OT lugs are fragile. Using a vacuum plunger solder sucker I managed to inadvertently dislodge one of the lugs. Fortunately there was enough of the tab remaining to render the OT usable.