We can make anything that you can dream of. Just shoot us an email at jnkaudiodesign@gmail.com or fill in the form below. Greetings from Penang, Malaysia!

Wednesday, February 25, 2015

Power supply capacitance myth

Hi guys,

Today, a customer asks us if we can increase the value of the class A tube rectified power supply capacitors for his custom made amplifier significantly so that he can get a better power supply with low ripple and deep excellent bass?

In his ordered design, we only chose values from 4UF to 100UF maximum for his 300B amplifier. He was questioning us why it is so low. He saw at least 220UF to 470UF in other designs, sometimes even 1000UF on some designs. If we use 100UF, will the bass be weaker compared to those with more capacitance?

What do you think? Do you use big capacitors and it equals to a great power supply?

We are sorry to tell you that it is a WRONG assumption. Huge capacitance does not equal to good power supply. Huge capacitance makes the bass weaker and looser due to the slow recovery time to charge the capacitors once the charge is used up.

What is important in class A tube amplifier power supply is to have a critically damped supply with excellent transient response with just sufficient capacitance to low the power supply ripple to an inaudible level.

Less is actually more! That's what we do with our power supplies. That's one of the things that separates us from others. You will find that our bass are taut, deep and fast! No woolieness!

J&K Audio Design

Tuesday, February 24, 2015

KR T1610 Triode

We were researching some high power tubes to be used as our next project to obtain single ended class A amplifier that is capable of driving most speakers and we looked into tubes such as GM70, 211, 845, 212, 300B XLS, 813, 833 and this tube pops up on our radar - KR T1610 Triode.

It is such a huge tube with a whopping of 150W of plate dissipation! It is 2x of a RCA 211 tube! Rules of thumb is that for class A1 amplifier, we would get ~25% of efficiency at reasonable distortion output. I like use 20% personally to arrive a real output power that normally people would operate at. 

Therefore, a single KR T1610 will get us about 37.5W (25% of 150W), and having 2 in parallel will get us 75W. We saw some people claimed 100W per channel using 2 KR T1610 in parallel. Is that possible? Well, let's try and see using theoretical calculation.

Rp (plate resistance) of KR T1610 is 370 Ohm. The primary impedance of the output transformer can range from 2.5x to 7x, depending on whether someone is looking for lesser distortion, more finese, or more power. When KR T1610 is used in parallel, the primary impedance of the output transformer should be halved to ~1.25x to 3.5x. This translates to 463 Ohm to 1.3K Ohm. Let's take the plate characteristics and draw out the loadline for maximum power.

Let's assume we are working with single tube first. The above loadline would be ~1.5K Ohm. That is about 4x of Rp. It is biased for maximum output power. Since the maximum plate swing voltage is not specified, I just ignore them for calculation sake to make things simpler. Vplate is 490V and biased at around 270mA, which is about 30mA above the absolute maximum allowable plate current (not recommended!). Yet, this operating point is still below the plate dissipation.

Let's just go with this first. In actual design, the Iplate should be lower, resulting a lower output power. Also, below is assuming full swing, meaning that to get such theoretical output power, the distortion figures will be very high since full plate swing is considered and operates at non-linear, saturation or cut-off regions. In real world, the output power should be reduced further to quote at a power rating with reasonable distortion, far away from the cut-off, non-linear and saturation region.

Using the a simple formula of output power = (Vp swing * Ip swing) / 8, that leads us to:

Power (single tube)  = (820 - 140) * (0.52 - 0.02) / 8 = 680 * 0.5 / 8 = 42.5W
Power (2 tubes) = 42.5W x 2 = 85W

This theoretical bloated output power is still 15% away from the quoted 100W figures. I have already exceeded the maximum Iplate to even get this result! This still has not factor in the discount for reasonable distortion figure. I wonder how the 100W is achieved. Bias super OVER the maximum specifications? Super bloated figures? Figure with super high distortion? Go figure yourself.

Let's try with 220mA per tube to be more reasonable, adhering to go below the maximum Ip, using a new operating point but with the same impedance.

Power (single tube) = (790 - 140) * (0.48 - 0) / 8 = 39W
Power (2 tubes) = 39W x 2 = 78W

Even lower output power! Still, we are operating at near cut-off or saturation at both ends. It should be much lower when we are at reasonable distortion figures.

If I'm the one designing it, I will choose a 2.2K Ohm primary impedance, and operate it at a more linear region, far away from cut-off or sturation, with everything within the manufacturer specifications. With the above operating point, that's what I will get, with reasonable distortion figures since I do not calculate the output power at maximum swing but stopping at reasonable region (green line).

P single = (690 - 150) * (0.38 - 0.1) / 8 = 18.9W.
P double = 37.8W.

This will be a more comfortable (or honest) figure I would specified for a Parallel Single Ended KR T1610! The tube operates in a more relax mode (more relax sound too!) as it is still under the maximum dissipation. Usually tubes that pushed to the maximum will sound tense.

What if I operate it at higher Vplate? That's not recommended since the positive swing will be compressed as it is close to the saturation region near the far right. Using a higher primary impedance OPT will help but it is still not optimal.

Well, that's it for today. This is just a rough calculation that I did without much thoughts just to see if what is quoted by some websites are reliable or trustworthy or not. I did not fully check the figures, so error might be present in the above calculation. I'm not diving further since I'm not going to design or commercialize this design due to the price of these tubes. I'd probably choose GM70 that is 125W that is much cheaper but just slightly less in power.

Conclusion is - do not always trust what you read online! It is better if you can prove it yourself. It is sad that nowadays we see such con jobs are being performed daily everywhere. There is just no free lunch!

J&K Audio Design

Monday, February 16, 2015

Audio Nirvana

Blog on Audio Nirvana Super 12 full range drivers from a friend of ours. We got a sample from Commonsense Audio quite some time ago and promised a write-up on their drivers. Apologies for taking so long. :)

Here goes the unaltered write-up from Thomas, a customer of our transformers, a seasoned audiophile friend, a long time friend and a distant brother-in-law of our founder!

My journey to audio nirvana

My journey to audio nirvana began a long time ago in varsity (1983 to be exact) starting with entry level gear (analog of course) – The AR turntable, Rotel 820 integrated amplifier, Mission700.2 speakers – stuff that would fit a student’s tight budget.

Upon graduation, my system and lifestyle evolved in keeping with my ever-increasing disposable income.  Vacuum-tube gear started taking over the backend in 1998 with some DIY efforts. My current system has since settled down to a DIY SET 6C33C-B power amplifier driven by a #26 tube preamplifier. Speakers are Musical Technology Kestrels, but I had always deemed the 86dB Kestrels to be too heavy a load for my 15W power amplifier. And so the quest to find an SET-friendly (read high efficiency) speaker was begun a few years ago. Before long I had on hand Coral and Fostex 8” drivers to try out.

As fate would have it, audio blogger Ken Loo (http://www.diy-audio-guide.com - obsoleted, now is this blog!) introduced me to Audio Nirvana’s Super 12 full-range drivers. OMG, a whopping 12-incher! So size really does matter and I could no longer resist the prospect of high-efficiency speakers doing justice to the many vocal oldies in my LP collection. Since I neither have the skill nor the equipment to cut perfectly circular holes for the drivers, Ken, a resourceful DIYer himself, located a cabinet maker to help me build the cabinets according to CommonSense Audio’s recommended specifications. I chose the dual-port version.

A few weeks later, they were safely in my house.  The cabs are constructed from 18mm MDF (middle density fiberboard). First I had to make sure that the all-important 279mm (11 in) holes for the drivers were accurately cut. Indeed they were. So far so good.

DIY Speaker Box
Audio Nirvana 12" Speaker Driver
That’s just the beginning. Next I had to decide what damping material was needed. After some trial and error, I settled on some readily available silicone floor mats and foam slabs. Two layers of silicone mats cover the insides other than directly behind the driver, where the foam slabs reside. At the bottom cavity, I placed polyester-filled pillow bags to further absorb reflected waves.
Padded speaker box with damping sheets and foam
Damping sheets

Initial impressions

Straight off, I loved the clarity of the mid-range. The music flowed effortlessly, with more than enough head room to spare. There was no in-the-face, fatigue-inducing highs that I had read about that characterized some high-efficiency full rangers. I was sold!

But with the much higher speaker efficiency, I could now hear a soft but audible 50Hz hum. It was traced to the dreaded affliction common to the #26 tube in my preamplifier – the hum induced by the heater supply. Due to my many other priorities (my audio hobby is never a priority), it took all of another year to sort that out. My #26 preamplifier re-vamp is another long tale in itself. For now, I’d just say I had to have it totally gutted – new (low noise) heater regulators (from Tentlabs), B+ shunt regulator (from KandK Audio), more bypass capacitors, B+ filtering re-design, grounding re-route, grid choke, etc.

With that solved, I settled down to more critical listening with the Audio Nirvana Super 12s.
Without the annoying hum, the enjoyment went up a big notch, but after a while I realized that the bass just didn’t sound right - there wasn’t enough of it in some of my favorite music when compared to my Kestrels. Even my non-audiophile wife preferred the old speakers, which, she commented, had more “enveloping bass”. I suspected it could have to do with the ports. Using an online port calculator, I obtained a length of 16.6cm for a 3-inch vent at 40Hz.

DIY Port

Raw appeal

It’s not easy to get proper speaker ports locally so I went to the hardware store to get some ideas. The closest the store had are pipe sections which, after some sawing and gluing, I had my ports. And what a difference the ports made!

The bass became more “directed” rather than dispersed, with more impact and it went deeper.  LP after LP, I found new enjoyment and purpose in pursuing this audio journey.

The music
Creedence Clearwater Revival

Willie Nelson - Stardust

Tracy Chapman
In my smallish 11ft x 12ft listening room, these speakers have no trouble pumping CCR’s “Have You Ever Seen The Rain” to ear-splitting level with just 15W from my 6C33C-B SET power amplifier. The bass really rocks! In Tracy Chapman’s original LP of the same name, “Talkin' bout a Revolution” has every breath and highs clearly fleshed out.  The rattling of the tambourine hangs in the air right in front of me, and drum beats are uncannily three-dimensional. Goose bumps! 
And how about vocal oldies?

Patsy Cline's
Patsy Cline is one of my all-time country crooner, with a melodious voice that transports you to another time. A system that’s not up the task will be excruciating to the ear. Analogue Productions’ latest reissue of her “Greatest Hits” (Decca DL74854) is sheer joy, with Patsy solidly planted in my listening room. For male country, Mobile Fidelity’s re-issue of Willie Nelson’s “Stardust” is just the ticket.  I have since been playing “All of Me” over and over again just to listen to the guitar and drum passage - the palpability just drives me crazy!

Nat King Cole

In Nat King Cole’s “Looking Back” (Nat King Cole Story Vol 3, Capitol W1928), in mono, Nat sounded really in-the-room, clear as day, and full-bodied; a relaxing and dreamy style that’s just perfect for a lazy Sunday afternoon. Yes, a lazy Sunday afternoon in audio nirvana. And all throughout my listening sessions, I did not for a moment feel any listening fatigue.
Thank you CommonSense Audio for bringing nirvana to my home.
HIFI system with Audio Nirvana Speakers

Audio Nirvana Speakers
Turntable: Vpi Mk 3 with AudioQuest PT-6 tonearm
Cartridge: Koetsu Rosewood Standard
Phonostage: Pentagon PS-3
SUT: Lundahl LL9206 at 10:1
Preamplifier: DIY #26  tube and Stevens & Billington Tx102 transformer attenuators
Power amplifier: DIY SET 6C33C-B
Speakers : Musical Technology Kestrel and DIY Audio Nirvana Super 12
Cables and Interconnects: Art Chamber custom-built with vintage Western Electric wires

Signing off!

J&K Audio Design

Friday, February 6, 2015

Spud Amplifier

Spud Amplifier - we love ya, we also hate ya!

We have always been mesmerized by the purest of all amplifier, a single stage single tube triode amplifier. 10 years ago, we tried building Spuds with 5842 or 6C45 PI. Sound was wonderful, but power is on the low side.

5842, or the WE variant, WE417, can give about 0.7W output. JAN5842 provides a bit more input headroom as it can be biased deeper. 6C45 is slightly more powerful than 5842 but it comes in only 1 variant, from Russia and tube rolling becomes impossible if 6C45 is used.

There are other tubes available in the market that can be used as Spud amplifier, such as the famous WE437A and EC8020, that are even better than 5842 or 6C45 but they are very, very hard to find, not to mention the stratospheric prices too.

Fortunately, there are still some other options, and one of them is the star of today's post - the E55L or 8233 Pentode. If Triode connected, it is very similar to the WE437 and EC8020. E55L has 10W anode dissipation, with biasing that can go as deep as -7W. Mutual conductance is whopping 45mA/V with gain of 30. That's good enough for driving a pair of high efficiency speakers even with a standard 2Vrms output from a CD player. For those lucky ones that has higher output for their CD player or DAC, it is even better.

Enough with the BS. We've built this E55L SPUD amplifier and we loved it. Here are all the techie details for some that love this:

  • Single stage
  • No coupling needed, interstage transformer or
  • All film capacitors in power supply
  • Choke in power supply
  • GZ32 tube rectification
  • Choice to select a 1:4 input transformer for more gain or direct in 

 It does not have the power of 300B but the transparency and clarity is just out of this world. Bass is not its strong point but we are happy with what we get out of using Zu Audio Soul Superfly speakers. You can't believe that it is merely 2W output. It was better than the J&K Nightlight that we built last year.

Enjoy! My daughter woke up. Time to go.

J&K Audio Design

Saturday, January 17, 2015

Amplifier Hum

Let us share a story related to amplifier hum with you guys. We think that it is worth reading as it may be applicable to some of the new DIY guys.

Recently, there's a guy that ordered a full set of Level 3 transformers from us. That includes - 1 Level 3 power transformer with Super Hi-B double C core, 2 Nanocrystalline double C choke, and a pair of Nanocrystalline output transformer.

According to him, his previous similar project ended up being too noisy and has pretty serious humming although he is using quite decent transformers from some company that we should not mention here. He is hoping that by getting high quality transformers from us, it will help to resolve all his problems.

We have heard this comment for many times and we always replied the same message to them, that we will write below. Since it affects a lot of the newbies, we decided to publish an article here so that everyone is clear on this topic.

A good set of transformers will no doubt bring you better signal to noise ratio, low hum, darker background, wider bandwidth, better THD and all those goodies, but, if your design look like these:

Amplifier layout that looks like Ampang Yong Tau Fu

Design that looks like Penang Char Koay Teow

Wiring that looks like Singapore Char Bee Hoon

A good set of transformers will only give you the edge and advantage to be at a higher ground than other low quality transformers, but if the design, ground, wiring, layout and all those important aspects of amplifier building are not followed, the end results are not going to delight you very much.

One thing that new DIY guys can do is to look at how others do the layout design on the WWW, and LEARN from them. There are thousands of nicely laid out design in the market nowadays. Their level of workmanship is simply stunning. We admit, we cannot come even close to some of the masters out there too but we'd always exchange photos internally when we come across those excellent designs out there and keep polishing our layout skills.

Layout is an art!

On schematic, or circuit design wise, beware of what you want to copy or clone. We've repaired many branded amplifiers and clones. There are a lot of "flawed" or purposely modified famous designs out there to mislead someone that by cloning these famous design, they get close to the sound of the OEM equipment. We found that a lot of those schematics floating online are NOT the actual design someone purchases on the shelf. They are mislabeled, values are changed, designs are changed purposely for you know what reasons. So, unless you know what you are copying, go ask someone that knows before you do something. If you are lucky, it works perfectly or substandard-ly. If not, kaboom!

Good luck in your DIY journey! Questions and discussions welcomed even if you are not purchasing from us. Audio is not just our business but our hobby too. Just use the contact us box in our blog to reach us.

P/S: No pun intended on using food pictures as references. It is just an analogy. These are no doubt great hawker food in South East Asia.

J&K Audio Design