Monthly Archives: February 2011

Elekit TU-879S Stereo Valve Amplifier Kit Review

Posted on by

Update May 12, 2013

I was testing my Arduino powered voltage controlled oscillator, when I noticed that the Right channel was very soft. Almost, but not quite, inaudible. Immediately, I suspected that there was a problem with one of the valves but, no, it was something else. Read more at bottom of this post.

Choosing the Elekit TU-879S

Looking for a new project, I decided to build a valve (tube) amplifier kit. So, let’s follow the process and see where it takes us.

I was looking for something that around $100, as a toy to play with. But it seems that that price level is unachievable, so we have to pay more. But, as the purchase price goes up, this thing can no longer be just a toy. It must become a functional durable product, that can actually be used by anyone.

There are a couple of valve stereo amplifier kits on the market, such as this Stereo Tube Amplifier Kit for under US$200, or this Model 16LS Stereo Integrated Tube Amplifier for US$250. Both looked good, and are cost effective. But, both suffered from the problem that they are built on a breadboard, and leave 300V exposed on open wires. Ok for a toy, but unusable in the long run for family members, or even my own careless fingers.

Some further research on the Internetz turned up the Elekit TU-879S, which seems to fit the requirements exactly. Less than AU$1,000 so it can be imported to Australia with no GST issues. Kit looks very well made, and professionally presentable. No loose HV wires to electrocute anyone. And, most importantly a very strong user community who have built and use the Elekit TU-879S in their audio systems, and provide input on modifications and improvements.

Oh. Oh. Think I just hit the “BUY” button.

Victor Kung at VKMusic created the English instruction manual used by Tube Depot. I just learned that unfortunately, Victor is getting nothing for his translation and original market development work. I suggest that you buy the kit from him directly.

This kit uses three valves, one a single valve dual Triode voltage gain stage, and two power gain Pentodes driving the output transformers, to produce 8W per Channel. All of the valves are readily available using either New Old Stock (NOS), or new products equivalent manufactured in Russia, or China. My finished version is pictured below, wearing ’58 Telefunken 12AX7, modern production SED Winged “C” 6L6GC, and alternatively ’61 production GE 6L6GC.

Dsc04094Dsc04097Dsc04096
Dsc04098Dsc04101

Research while waiting (on back order)

There are a couple of excellent reviews of the Elekit TU-879S which I won’t reproduce here. Save to say that they provided some ideas for improving, or hacking, the Elekit TU-879S, and that even before I had built it.

The first noteworthy improvement is to substitute the decoupling capacitors between the input and the pre-amp stage, and then before the power-amp stage. The Positive Feedback review suggests using VCap capacitors in those locations, and raves about the results. Great, but I’m not about to spend 50% of the total purchase price on 4 components. Let’s look for some cost effective alternatives.

The Internetz seem to suggest that Mundorf has a good name and quality product, so an order was placed for 4x MCap Supreme capacitors at Madisound Speaker Components. These MCap capacitors have a musical heritage, with dual series counter-wound cores supposedly cancelling impedance effects, but don’t add too much to the project cost.

The second improvement is to replace the existing dual linear variable potentiometer with a higher grade unit. The provided unit is a $2 ALPS plastic film component. For a little more, with respect to the value of the end product, the TKD 2CP-601S 100K stepped dual log taper potentiometer provides a good alternative. One was ordered from HiFi Collective in the UK. At about $40 it bridges a logarithmic gap in price between the original $2 component, and the ultimate component being a $300 stepped resistor array.

So with these two modifications, we have effectively upgraded the “small signal” part of the amplifier, where noise and non-linearity have the most effect. There are a number of other modifications suggested, but they are more drastic than simple component selection and replacement, and can wait until after I’ve established what the base-line of performance is for the Elekit TU-879S.

Valve search

The 6moons review goes into some of the options available for the Elekit TU-879S. Really, there are so many options around that it is a lifetime of experimentation to find the best valve sound from this amplifier. In fact, half the fun of this kit is reading the Internetz opinions on which valve creates what sort of sound. Often, strong opinions on individual valves are expressed without reference to the rest of the signal chain (pre-amp, speakers, listening environment, etc) such that the argumentation takes on a religious air.

12AX7 (Voltage Gain)

There are many many reviews of this 12AX7 valve type. Within these reviews there is one type that stands alone at the pinnacle of sound reproduction, the 12AX7 / ECC83 Telefunken with smooth plates. Luckily TC Tubes has some (one only as I write this) in stock. These valves have a 40,000 hour life time, so there is no problem to buy a “premium new”  test device, with exactly matched triodes. The one that I received from TC Tubes was made in April 1958, and tests Gm 100%/100%, which is about perfect.

Also as an alternative, there are great reviews of the military version of the 12AX7 being the 5751 which has a lower gain (Mu of 70 vs. Mu100 for 12AX7), and is also extremely low noise. TC Tubes has these GE 5751 in stock too. TC Tubes shipped me a 1961 vintage GE 5751 with matched Gm 108%/108%.

There are many other alternatives of these valves, and as I mentioned, I think that they are all subject to a lore of selection.

6L6GC (Power Gain)

As the Elekit TU-879S accepts a variety of similar Pentode valves, it is difficult to determine which ones should be on the list to try. Cruising the forums, it seems that this amplifier works best with the 6L6GC, which is the same type that it ships with.

As the power valve has a limited (though long) life of around about 10,000 hours, I am not keen to spend a large amount of money on NOS for this position. So my ideal is to find the “best” modern valve and see what the service lifetime is in practice, before investing in a NOS alternative.

After looking at some reports at Watford Valves, I decided to get a matched set of SED Winged “C” 6L6GC valves, also from TC Tubes. These will complement the Electro Harmonix valves provided with the kit.

Once I’m sure that it is a good investment, in terms of sound quality and listening hours, I’ll certainly get some NOS power valves to use.

Update: I’ve decided to take the plunge on NOS Pentodes, and get some GE 6L6GC from TC Tubes. Reviews of these suggest that it has an excellent performance, exceeding any current production valve though possibly not the best available NOS.

The construction process

When you un-box the kit, with the components packed so neatly into 10 heat-sealed individual bags, and each piece of metal wrapped and separated by brown paper, you remember what the Japanese are famous for. OCD grade accuracy. Clearly, this kit is perfect.

The English translation of the instructions, and the inclusion of a 230V power transformer of the same type as the 100V original, are finesse added by Victor Kung at VKMusic. There is no need to add any further internal photographs of the process. Assembly, for someone who is used to soldering SMD and micro-electronic components, is a breeze. Absolutely everything is explained exactly, and everything is perfectly easily done.

The English translation has two small (tiny) typos in resistor numbering, which confused me for a minute (also OCD), but looking at the Japanese original instruction cleared up my confusion.

As the provided resistors are 5% accuracy specification, during the construction process I checked the actual resistance of all resistors before assembly using an accurate digital multimeter, and tried to match closest pairs of resistors to the same circuit position in both Left and Right channels. The thought behind that was that even if the circuit was very slightly off specification because of the tolerance in the resistor values, then at least Left and Right channels would be matched.

The modification process

With the Mundorf MCap Supreme capacitors being much larger than the standard polypropylene capacitors they have to be fitted onto the back of the PCB. No problem, there is plenty of space.

Dsc04081Dsc04080Dsc04079Dsc04078

The TKD 2CP-601S potentiometer is a little more tricky. Fortunately, the pin-outs are identical to the supplied device, so the volume control PCB can be used. I just needed to kink the pins towards each other slightly to get a fit. The volume circuit PCB shows some minor flux damage, as I had to remove the provided volume potentiometer fitted first for testing.

The axle housing on the TKD 2CP-601S is slightly greater diameter than that provided, but rather than drilling out the top cover I decided to use a pocket knife to cut “flats” on the two sides of the threads since it almost fit. As the housing metal is aluminium, and it is cut into a fine thread, it is easy to remove enough metal from both sides with a knife to get the potentiometer to fit cleanly into the slot. This could alternatively be done (better) with a small flat file.

Secondly, the tab on the front face needs to be removed so that it will fit flush with the front panel. Snip with side cutters, and it is gone.

Finally, the axle is about 5mm longer than that on the provided pot. I used large pliers holding the end of the axle to hold it still, whilst using a small hacksaw, resting against the pliers as a guide, to cut through the soft brass. Easily done. Don’t hold the body of the potentiometer when cutting, otherwise the axle will turn, and you will surely damage it. Use a file to tidy up the cut edges on the axle.

Testing & listening

The initial testing and listening was done using the provided Chinese 12AX7 and Electro Harmonix 6L6GC valves. Everything worked well. The amplifier is amazingly quiet on idle. No audible hum at all. If a line level source is connected into one input, with the other input selected, some crosstalk can be heard at high volume. But generally this is not noticeable, and can be removed simply by switching off the alternative source.

Both the SED and Electro Harmonix valves produce the lovely blue aurora surrounding the valve, when viewed in the darkness. The pictures below try to capture it, but don’t do it justice. The GE valves don’t produce the blue aurora, but that doesn’t affect their excellent sound quality.

Dsc04085Dsc04088

Once the proof testing was finished, I replaced the standard valves with the special NOS 1958 Telefunken 12AX7, and the matching 1961 GE 6L6GC valves. Amazing the difference in sound quality. After a few hours of burn-in, I started listening to the U2 War album, and I could swear that Bono started to sound incredibly like Elvis. Some of his spirit must be trapped inside the little glowing bottles.

Repair Update

Early May 2013, I was experimenting with a voltage controlled oscillator to see how sine waves would behave, and to test my speakers and my ears, when I noticed that the Right channel had become almost inaudible. I hadn’t previously noticed, or may be I had noticed but had put it down to my speakers being too old.

I did some standard trouble testing, and found all the valves to be good and also inputs and speakers to be good. It therefore had to be something inside the amplifier.

I gathered all my tools, slightly concerned that it would be hard to identify what was the problem. I needn’t have worried. The problem was obvious. Burnt black obvious.

Damage - Before

Electrically, R13 and R14 are 330 ohm 3W resistors, which provide the current load for the power valves. At idle they should present about 21V on Pin 8 of each of the power valves. In my measurements, the left channel was presenting at 28V and the Right channel at 58V. The significantly lowered current through the Right channel was causing the reduced amplification.

As soon as I turned the board over, the Right resistor sort of just fell off the board. Pretty well burnt out.

I have replaced the faulty pair R13 and R14 with 330 ohm 5W resistors, and have done a little bit of aerial bridging to try to support the weakened PCB tracks. The image below shows the result. It is not pretty, but it has returned the amplifier to working, with 22V showing on each of the valve’s Pin 8, and the volume level from each channel being similar.

My suggestion, use 5W resistors when you’re making this kit, and try to keep enough space around the resistors so that they don’t burn the nearby capacitors and sheathed wiring. In the photo the faulty resistors can be seen too, with the lower leg on the Right resistor almost turned to carbon powder.

Damage - After repair

Bagshot Level Crossing – April 2010 – Court Response

Posted on by

The former State Government of Victoria has been accused of using the sensitive issue of Road Safety to aggressively increase its taxation base by using Road Safety Cameras to raise general revenue. This accusation has previously been leveled by the current State Government and on 31 January 2011 the Deputy Premier said; “Victorians need to have confidence that the state’s traffic camera network is accurate and has proper oversight, which is why the Coalition has requested that the Auditor-General conduct an extensive investigation of speed camera operations and to report his findings to Parliament.”

For example I believe that the Road Safety Camera at Bagshot, which was installed at a cost of in excess of $900,000, is being used beyond its original remit to control level crossing related offenses to RAISE REVENUE from drivers traveling safely along a piece of open road, containing a level railway crossing.

Following the level crossing accident in Kerang where 11 lives were lost, there was a public outcry, and an investigation into safety on Victoria’s level crossings was initiated. Following the Parliamentary Inquiry into the accident and into the state of our level crossings, it was determined that a large number of changes were needed. This despite the fact that less than 2% of our state Road Toll occurs on level crossings, and that the general level of safety has improved by over 85% in the 30 years to 1999, taking the annualised toll to under four (4). “This is more than twice the reduction in the Road Toll over the same period”

One of the changes introduced in 2008 was to introduce a blanket speed limit of 80km/h on all sealed (but strangely not unsealed) major level crossings. This blanket is imposed irrespective of the lines of sight or Australian Level Crossing Assessment Model (ALCAM) rating applicable to the specific crossing.

Looking at the picture provided of the Bagshot level crossing, we can see that the environment is open grassland with level road and open sightlines, with no housing or other developments infringing on the crossing area. The crossing has been this way in my memory of over 30 years of travelling to and from Echuca. During all of those years, it has been safe to continue at up to the general speed limit whilst approaching the crossing, or slow down to whatever speed the driver deemed appropriate depending on the conditions at the time.

So, are we saying that the crossing has suddenly become unsafe?

To my mind, either a level crossing must be either more-or-less safe, or a level crossing is more-or-less unsafe.

Let us say that Bagshot crossing IS more-or-less UNSAFE.

IF the Bagshot crossing IS UNSAFE, it MUST be the duty of the Government and its agents to make it safe.

According to the Level Crossing Fact Sheet #2, the Government spent $1.8 million dollars on 2 Road Safety Cameras in 2007. One of these cameras was located at Bagshot.

Does a Road Safety Camera or a reduction in speed contribute to safety when crossing a level crossing? The Government’s own sources don’t think so. From “Towards Zero – A Strategy for Improved Level Crossing Safety in Victoria”, “ATSB research found unintended driver error was the most common cause of collisions and a factor in 46 per cent of crashes. Alcohol and drugs were less of an influence, as was excessive speed.”

Unintended driver error is the real contributor. The sort of thing that happens when a driver is forced to concentrate on several simultaneous activities at once, such as closely monitoring their speedometer, with attention inside their vehicle because they know there is a speed camera is imminent, whilst at the same time their eyes should be looking outside for warning lights, and casting up and down the train tracks looking for approaching trains.

So, IF Bagshot IS UNSAFE, what is the best way to make it safe? Well the Parliamentary Road Safety Road Safety Committee inquiry is clear that: “While detailed research is limited, the most effective treatments appear to be those that either eliminate the crossing through grade separation,.., or those which provide some physical restrictions such as boom barriers.”

But surely it must be too expensive to make Bagshot safe through the installation of boom barriers? Well no, as VicTrack announced in October 2010 that the $1.59M project to upgrade three Shepparton and Congupna level crossings (the crossings will now feature boom barriers, in addition to the existing flashing lights) is now complete.

Then with about $540,000 required for a set of boom gates the Victorian Government could have made Bagshot significantly safer. Yet, the Victorian Government chose not to make it safer, but rather to spend nearly double that amount (being $900,000) to install a Road Safety Camera at Bagshot. A cynical person could infer that boom gates, whilst being a primary level crossing safety mechanism, don’t produce general revenue.

But now, if Bagshot was safe, and is still in fact more-or-less SAFE

If the Bagshot levlel crossing is more-or-less safe, and scores too low on its ALCAM rating to justify any further immediate expenditures to significantly reduce the risk of catastrophic incidents occurring, then it must truly be asked, why the previous Victorian Government saw fit to spend $900,000 on this Road Safety Camera at a safe level crossing, when there are 2,267 level crossings in Victoria, and at the current rate of expenditures only 50 to 90 UNSAFE level crossings are able to be upgraded per year.

A cynical person could then further infer that, as the real risk of a catastrophic incident on a level crossing is already statistically extremely low, based on the Government’s own figures, that money spent on non revenue returning safety improvements at level crossings brings no general benefit, but that investing in a Road Safety Camera does produce general revenue.

In conclusion, I submit that I was photographed traveling at 99km/h when crossing the level crossing at Bagshot. I further admit that I was not concentrating on my speed at the time, as my attention was fully occupied scanning the railway tracks ensuring that I avoided the twice daily Echuca train.

The floods have badly affected the people of Victoria’s North. Rather than contributing a fine to general revenue, if required, I would be happy to contribute a similar donation to an aid agency of your choice. Thank you for listening.

Update

On 24th February 2011 this matter was heard at Bendigo Magistrate’s Court.

The Magistrate, whilst not willing to enter into a debate on the merits or otherwise of Road Safety Cameras, saw my point of view. She released me on my undertaking to donate a sum equivalent to the fine to the Flood Victims in Victoria.

Following the Hearing, the Prosecuting Police officer approached me. He apologised for “speed cameras” and said he “wished that they didn’t stamp ‘Police’ on the Infringement Notices, as is really nothing to do with them”.

Using this mechanism, I achieved the result of costing the public purse the salaries of the Magistrate, Prosecuting Police Officer, court officials, and court room time. Plus, the State did not get the General Revenue that would have been
sucked into the machine.

Generally, more people should object to Road Safety Camera Infringement Notices. The Magistrates won’t increase the fine over the Infringement Notice, nor will they reduce it. But, you have negated any revenue benefit to the State Government. If they lose the revenue, they will stop seeing speed cameras as easy revenue sources.

Oh, and by the way, a donation is Income Tax Deductible too, for what that is worth.

Footnote

If I’m ever feeling depressed about work, family, or life in general. I will go and sit in the Magistrate’s Court for a day. People attending that court generally have REAL issues, with alcohol, money, family, and a place to live. Which just makes me realise, I have nothing to complain about.

Links

http://www.premier.vic.gov.au/html/310111_001.html

http://www.transport.vic.gov.au/Doi/Internet/transport.nsf/AllDocs/8EE1EDA7067A3EE1CA2571AF0005EEFC?OpenDocument

https://www.victrack.com.au/statewide-projects/category/improving-safety/shepparton-level-crossing-upgrades

http://www.mainroads.wa.gov.au/UnderstandingRoads/Rail/Pages/Rail.aspx

Wigglesworth, Graham & Routley: Rail Related Fatal Accidents in Victoria Australia: 1990 – 2002.