Friday, 2 November 2007

PPA Amp Completed! (sort of...)

After a few years from starting the PPA v1 board, I finally had the time to finish off the board and give it a run through. Having forgotten what I was going to do with it, it was hard again, to figure out where I was at, and what I would need to complete it. The photos show the final setup, a real mess. But it works!

Initially I tested it with a TL072 amp, but it didn't work. There was no noise. My mistake is that the TL072 is a dual OPAMP and the board is for single OPAMP. Stupid me... Anyway, plugged a OPA227 in, and she was working like a charm.

Some initial thoughts. The setup is far from ideal, but the music was sourced from iTunes, using lossless AAC. All connectors were temporary, read poor quality connection and cabling. The headphones I used were my Sennheiser HD650.

With the bass boost on, as I put it in by default, the bass is too much for me. Maybe they would be well suited to smaller headphones, but that needs to be confirmed. With the bass boost off, the sound was well balanced, and has a certain 'fullness' and precision/clarity that was missing from my standard Maple Tree Audio EAR+HD tube amp. Mind you, having only used this amp for less then an hour, these comments are by no means definitive. So far, the sound has been very good with the 227, and I hope to test it with OPA627 and AD8610 and may even others. Which I think is the great beauty of the design. Find a single OPAMP and can give it a go.

On the side of noise, there was a low level of floor noise audible. But this could be due to a large number of reasons; dodgy wiring, use of a cheap regulated wall wart, lack of ground of the Alps control (as suggested by Tangentsoft) or maybe just bad wiring. I guess there will be more resolution on this, once everything is correctly wired.

I intend to use this also with my Shure SE210 IEM, and the pending Etymotic ER-4S. But right now, finishing off the case is the highest priority, as well as deciding on the power supply method. At present, using battery power seems to be the option (given I have the battery supply board), but determining a suitable voltage and charging method, that will be another decision in itself. Finally, PPA 2.0C arrived in the mail recently, and given I just soldered PPA1, it might be fine time to put together PPA2 with the diamond buffers.

Tuesday, 30 October 2007

PSU for the DRD45

The above is the schematic design for the DRD45 power supply. This was done using PSU Designer II. The above shows a residual 'noise' of around 6 mV, which with a 510 V gives a noise dB of around -98dB. This of course ignores the higher harmonics generated, but makes a good starting point. The design calls for a 50uF ASC oil cap, followed by Electra-Print's 20H choke, and then a dual combo of 100uF Solen aluminium can and a 47uF Solen PPE capacitor. There is scope to change the last 47uF to a 50uF ASC Oil. Wanted to avoid electrolytics not just for the rumoured acoustic difference, but also for reliability. The power supply transformer is 410-0-410V, which is an odd rating, but I am sure that Jack (Electra-Print) is able to wind one that is suitable.

dB Calculator

Here is another nifty webpage to assist with voltage noise and gain calculations.

Thursday, 11 October 2007

LC Filter Calculator

Here is another nifty calculator for determining the frequency response of a LC filter circuit. This could also be useful for PSU design calculations. I used this to aid in the understanding of the determination of the 'Ultrapath' capacitor. Using an example Plitron PAT-3035-SE, with a primary inductance of 28H and an Ultrapath cap of 51uF, the frequency cutoff (not sute if it is f-3dB) is 4.2 Hz. My understanding from this, is that any audio signals above 4.2 Hz will be returned back to the cathode to ground, rather then through the PSU capacitors to ground. Doubling the capacitance to 100uF, results in a frequency cut-off of 3 Hz.

Monday, 8 October 2007

From 300B to 45

After some weeks of research, dialogue and thinking, the 300B amp has been put on hold for the moment, to make way for a 45 DRD monoblocks. The main reasons behind this change is; lower operating B+ voltage and availability of quality new and NOS tubes (there are no NOS 300B easily available). To this effect, I have ordered the chokes and reactors from Electra-Print, and they arrived within a few weeks. In addition, I have started to place the other components on order. In the signal section, I am using Caddock MK132 resistors, Mills high wattage resistors, Ohmite adjustable rheostats, Blackgate NH driver cathode bypass, and Solen PPE Ultrapath. Tube sockets will be teflon type. In the PSU section, I will basically using the same design as Electra-print design. The diodes will be changed to hyper-fast diodes, the filter caps will change in value. At this moment, I am considering filter values in the range of 500uF to 1100uF. Apart from some minor changes in component specification, it is very similar to the schematic as provided by Electra-print. The irons will all be Electra-Print, except for the Power transformer, which may be sourced from another supplier. Though the change from a 300B to 45 reduces power significantly, from around 8W to 1.5W, I believe that tandem with a 95dB+ speaker, and given my listening room is a small room, I shouldn't be too concerned. Full schematic, photos and progress will be available shortly. About half the components have already arrived, so looking forward to completing all parts, ready for construction and audition!

Saturday, 6 October 2007

High Pass Filter Calculator

Here is a great link with a Java app that determines the frequency response of a RC high pass filter. Very useful for determining capacitor coupling values.

Friday, 21 September 2007

Link on Wire Gauge Info

Here is a great link that provides information on current limits, diameter and skinning effect of AWG wires.

Wednesday, 15 August 2007

Article on Loop Distortion

Here is a great article talking about different typologies, and loop distortion. Good read!

Saturday, 11 August 2007

Shure SE210

On a recent trip overseas, I purchased a pair of Shure SE210 in ear phones. I originally wanted a pair of Etymotic ER-6, but after going to a number of shops, and could not find it, gave up. The use of these phones didn't seem as dramatic and difficult as I have originally read. It does take a little bit of time to sort out the right size and shape seal though. The first impression is that they really do block out the outside noise. I found this especially so for the typical machine or repetitive noises. I was sitting under a very noisy air conditioning unit, and could not hear but the faintest noise. Sitting on a train and bus in Hong Kong was very quiet. With the music on, it was hard to hear the noise around you. However, I did find that PA announcements were audible. On the trip home, comparing with the noise canceling headphones supplied on the plane, these were not as quiet, but still, the aircraft noise was quite soft, and I could still hear the PA. In terms of sonics, I found these to have really goo bass and mids, but somewhat lacking in the trebles. So in terms of balance, these favored the mids and lows. It is not as 'crisp' or 'clear' as my Sennheiser HD650's, but this is just the first impression. I will now trial them over a range of amps and sources to see how they compare. The next step is to develop and build a small headphone amp running on two 9V D cells that has top sonics and decent battery life.

Tuesday, 31 July 2007

300B Dream

The DIY/Audiophile/Tube world seems to be filled with 300B designs. Many writers and builders place the 300B at, or near, the top of any 'most desireable' power output tube. Having never heard a 300B amp, good or bad, I have decided to follow the flock and build myself a 300B amp. There are a raft of designs around the Internet for the 300B, ranging from the original WE designs, RC designs (Angela, Audio Note Kit, Reinhardt, etc), exotic DC designs (Electra-Print), parafeed, transformer coupled, etc, etc. There are just so many!

Not being an audio designer, or electronics engineer, I have to rely largely on other people's designs, skill, knowledge and experience to put together something that will not; kill the builder, kill the end-user, burn or otherwise destroy property; and at the same time, sound decent and be reliable. It is therefore important that I not only use designs from people who really know what they are doing, but that I also understand the theory and design of the amp, so that I will not be simply soldering.

Not wanting to do the now conventional/standard RC approach, I decided that this amp should make use of technologies, designs, and topologies that are somewhat different. From reading the various information, it appears that the a 300B amp is best suited to its now 'classical' configuration, the Single-Ended (i.e., Class A) design. Further reading and research, suggests that the importance of the 'audio chain' begins backwards from the speaker. Taking this logical, the critical audio path, in regards to the power amp is as follows:

  1. Output Transformer
  2. Output Tube Stage
  3. Power Supply Stage
  4. Stage Coupling
  5. Driver Tube Stage
To thus 'rough-out' the basic design requirements. The output transformer should be as high quality as possible. Some of the commonly touted high-end transformer manufacturers include (in no particular order); Plitron, Tango, Lundahl, Sowter, Electra-Print and Tamura and Audio Note UK. Some of these transformers are in the hundreds of dollars, and some special versions, even more. More information will be provided as I research more deeply into this area.

In regards to the output tube topology, there seems to be four designs; Direct coupling between output tube-transformer-power supply), Ultrapath, Parallel Feed and a combo of Ultrapath/Parallel Feed. My main reference is an excellent article in a back issue of Vacuum Tube Valley. This is an important area that I will explore later.

The coupling between the two stages can take place in a number of ways; direct coupling; RC coupling, transformer coupled, Loftin-White, using active (semiconductor) devices, and others. Given that capacitors are the components that colour the sound the most, it would seem logical to use or not use only with full and proper understanding of its implications. An area that deserves far more attention.

Finally, the power supply and driver stage will be looked into much more detail later.

After all of that, a preliminary list of the designs that are being seriously considered is as follows; (this may grow or shrink)

  1. Electra-Print DRD 300B (A Loftin-White direct coupled 300B that also uses the Ultrapath technique for coupling the cathode to the output transformer).
  2. Andrea Ciuffoli's 300B PSE/SE (Transformer interstage between the 5842 driver and 300B).
  3. TubeLabSE (CSS 5842 Driver with a MOSFET stage between).

Monday, 30 July 2007

Headphone Amps

I originally started the DIY Audio scene with an interest in high quality headphones. My first project as Pete Millett's Low Voltage Headphone amp, followed by the Maple Tree Audio's EAR++ kit. Both these amps proved to be very nice to listen to through my Sennheiser HD-650's. And I continue to enjoy the intimacy of the listening experience that headphones offer. My upcoming trip has once again sparked an interest in headphones, but this time, in the mobile/portable hi-fi direction. So now, how can I mirror the quality/result of my current setup so that I can take it around with me? The search is now on for a small, low-power consumption (i.e., battery operated) and highest quality mobile/portable headphone and headphone amp combination. Some of the amp work has been extensively covered by sites such as Headwize and Tangentsoft to name a few of the popular sites. This would be a good starting place. The classic CMoy design is a very simple design, and having built a few of them, a great place to start. In regards to headphones, it appears that 'In Ear Headphones' or 'In Ear Monitors' are the best mobile solution. From reading various reviews, Etymotic ER-4S appears to be one of the best sets of in-ear headphones that are priced within reason (less than US$300). Other contenders include Shure and Ultimate Ears. The Headroom site offers a great selection of headphones, amps and information. Surprisingly, there are a number of Australian sites that stock various headphones and related equipment; Wicked Digital and Headphonic are some of note. More notes and progress to come!

Sunday, 22 July 2007

MathCAD Spreadsheet for FE-127E

Have now included a PDF print of the MathCAD models designed and made by Martin King that has the parameters for the FE127E speaker. It has been rn in for over 50 hours now, and some of the 'shine' or 'tinny' sound has been reduced. Still amazed at the level of bass that this has, given such a small Sd. But let me say that the bass is not 'loud' by any stretch of the imagination. It is audible, not soft, not recessed, but clean and balanced. The imaging is very nice indeed. Just watching a DVD movie with only 2 speakers, in a Dolby Stereo, the was a wide sound stage, and though not fully surround, the sound did appear to come from the entire front wall, with some sounds appearing to come from outside of the speakers.

The next project is to build another pair of the same, but with a higher front baffle so that it can be floor standing, and have the driver at 1000mm above floor, which is at ear level.

Friday, 20 July 2007

PCB Power Supply Boards

In my ventures, I stumbled across these PCB from DIY Club, an online shop in Hong Kong that sells some rather interesting audio DIY related items. Two of which I found great interest in, are the TP4 regulated HV and LT PCB, and A19, a LT regulated PCB. TP4 is described as: "Gold Plated double side pcb, MosFet output, delay function on the high voltage output, with four heater voltage supply output adjustable from 6.3V to 15V". Whilst A19 is described as: "output voltage adjust from 3V to 20V input voltage 5V to 25V AC use LT1083 as volatge regulator". The prices were very reasonable, US$10.50 for the TP4 and US$4.00 for the A19. I bought a few, and here they are.


The boards seem nicely made, with a thick board and nice tidy traces. It was noted however, that two of the pads to the A19 that shouldn't be touching, are in fact touching. This seems to occur (see photo above) with the variable resistor. This is not of great concern to me, as I plan to use a fixed resistor. The TP4 boards will mate nicely with my Aikido PCB, whilst the A19 would work nicely in any future tube amp projects.

Thursday, 19 July 2007

Irons arrived!

The Edcor output transformer and the Hammond choke has both arrived. See pics. The output transformer is definitely larger than I expected. The grey is not the standard finish, but I got it to match with the output transformer that is being 'custom' made by Edcor. The finish on the end bells is okay, not the best I've seen, but okay.

I have placed on order two custom power transformers from Edcor, with the following specs:

230V 50Hz

400V-0-400V @ 150 mA
5V-0V @ 3.0A (5.0V @ 3.0A)
6.3V-0V @ 3.2A (6.3V @ 3.2A)
12.6V-0V @ 0.6A (12.6V @ 0.6A)

That should complete the last of the big stuff for this amp. The chassis will be an all timber affair, approximately 450x300 in size, about 112 deep, with the tubes and transformers on the top side, and choke on the inside. I am thinking of constructing it with 12mm MDF top and bottom, painted black, and the sides of 90x19mm KDHW (hardwood). The connectors, IEC plugs, fuses, switches and posts will be the next to-do.

The tubes have already been purchased. I went with JJ KT88 after some review of stuff on the net and from Vacuum Tube Valley mags. The driver tubes are NOS 12SN7 (hence the 12.6 heater) and NOS JAN 5U4G for the rectifiers.

Friday, 6 July 2007

Fostex FE127E ML-TQWT Speaker in Progress

After a few tweaks using Martin King's MathCAD models, I have configured a ML-TQWT using a Fostex FE-127E. The design response is approximately -3dB at 40Hz. The size of the enclosure is externally 900x300x225mm with the driver 200mm from the top, and the port 150mm from the bottom. It is stuffed from the top to the top of the port. The port is 50mm diameter and 100mm long. Approximate total stuffing weight is around 500g. See the response graph below.
The cabinet is made entirely of 16mm MDF, which is readily available in Australia. The size of the cabinet was designed to minimise cuts required. Two sheets of 900x600 and 1200x450 was required. See the pictures below of the speaker in progress. Stuffing used was a product called 'Polyloft' from the local rubber(!) shop.

The speaker is now complete, and being 'run-in'. They are being driven by the Chipamp completed earlier, and a PCB build of Peter Millet's Low Voltage Headphone amp. At the moment, I can say that the sound is really nice. The imaging is very precise, and the separation and articulation of instruments and notes is very clear. I have yet to install the ports in, but the bass is nice and balanced. Generally, the sound is very well balanced and should improve with time!

Wednesday, 4 July 2007

Chipamp Complete

After two years of completing the PCB and getting a working amp, I finally managed to put together a case for chipamp! This was the kit purchased from and uses the LM3886TF. I have been running this in for the last few days and proved to be an excellent power amp. There is no hum and no noise. Quiet, and reproduces stereo very nicely. Even though it is not a 'mono-mono' PSU setup, the channel separation is very nice. The case is a fairly simple affair, with hardwood top and bootm panel, and the front and back made of MDF left over from renovations. The sides have been left open and allows the heatsinks to cool and air to enter into the case. So far, after running for 4 hours, the heatsinks are barely warm to touch, and the whole rig is very cool. The wiring may not be the best layout, primarily because the wiring was done without thought of chassis! However, as mentioned above, at normal listening levels, the floor noise is very low. As much advised by other people, the most difficult and expensive part of DIY power amp is building the chassis, and the transformers. It is very hard to get something that is suitable at a reasonable price.

Monday, 2 July 2007

Uniamp PCB almost complete!

This is a PCB build of Peter Millet's Uniamp, see his webpage for more info. I have decided to build this as my first true DIY/buy all the components and assemble yourself (except PCB) tube amp. These will be initially run with KT88. After a few months of delay, the PCB is finally assembled. All but 2 resistors have been soldered on and completed.

The parts used for this amp include:

  • NOS Allen Bradley Resistors from eBay.
  • Kiwame Power resistors (5W+) from DIY Hifi Supply.
  • Solen SM Tin Foil Coupling Capacitors from Parts Connexion.
  • Elna Cerafine cathode caps from Welborne Labs.
  • Sowter phase splitter from Sowter.
  • The power supply filter caps (Panasonic FC Snap-on) from Digikey.
Still working on an alternate design from Pete's webpage.