Listening to Tweety

“Tweety”

Last modified: October 31, 2006

Tweety is here to stay...

I can recommend it to everybody: Get you own Tweety-bird. The first tests in the livingroom setup, comparing Tweety with the built-in Phono stage of the Marantz PM-14mkii amplifier made one thing very clear: Tweety is a winner. Tweety combines beautifully with this amplifier, both in sound, impedance and looks. It still keeps me from switching back to the internal phono stage of the Marantz (although the Marantz has better S/N ratio).

What helps is that I used two Lundahl transformers (about 100 euro/pair) as MC step-up, and this combination helps to get the best out of my Dynavector DV-17 cartridge.

At first I used a ECC81 for the input stage of Tweety, the sound was fine but the gain was lower. Also I did not have the 275-300 Volts power supply available from the first day. My patience was awarded though, Wil from AE-Europe delivered a very nice transformer and a 10H choke both were small in size but high in specs and they just fit in the box I used for the Power Supply (which was of-course identical to the amplifier housing).

I'm still in doubt about the details of the Power Supply. Shouldn't I be using a regulated B+ power supply? Or better stabilisation? When listening to Tweety I'm convinced that at the moment this would nog add value.

The power supply built already differs at some points from the original setup. I used a choke which I did not plan in trhe original design, and as a result I had to increase the values of several resistors in order to get the B+ back to 285 Volts.

At first I connected the mass of the amplifier circuit with a 100K resistor to the mass of the filament supply. As expected I continued to have some humm in the sound. After connecting both mass of amplifier and filaments together this humm disappeared immediately. But, but but ...

The two tube halves in the SRPP each work on halve of the B+ voltage; The upper halve has a plate voltage of 285 Volts and a cathode voltage of about 140 Volts. Since the filaments of all tubes are connected to 0 and 12 Volts, there is a cathode-filament difference of 130 Volts. According to the specifications of the ECC83 tubes I used (JJ-Tesla) this was well within specs. However, for those of you rebuilding Tweety I would recommend to change the setup of the filament supply such that it is lifted 70 Volts above ground. Use a voltage devider 1: 3 or 1:4 of the B+ and connect with the ground of the filaments.

• Here you will find again the schematics of Tweety and Granny (size 70KB)
• Here you will find the Parts list I used in Tweety (Attn500kB size)

I think sooner or later I might want to do some experimenting with a LM307 IC for power supply stabilization. This would not be too difficult and can be done without sacrificing Tweety by just building another power supply (after all the amp and the power supply are connected by means of a Hirschmann connector). This setup allows me to improve powersupply or amp or experiment with other amp setup without having to change the rest of Tweety.

Tweety sings

Meanwhile we're already listening to a singing Tweety bird for some time. Our own canary Tweety joins most of the time and apparently likes the music a lot. in general this is a good sign.

I have started getting my favourite collection of albums together for an in-depth musical recension of Tweety. We will listen to this music, as we did with the turntable and cartridge during their selection process, over a couple of weeks (next to other music of course) in order to come to a final conclusion about Tweety. Normally we listen to these albums amopngst others:

• Dire Straits: Brothers in Arms
• Eagles: Hotel California
• Kurt Rapf: Mendelssohn Orgelwerke
• etc...

I try to pay attention to at least the following aspects in the music: Does the music flow easily from the loadspeakers, and do the higher frequencies not contains too many "sch" or twinkel sounds. Also I want to make sure that Tweety is able to reproduce the lower freqencies.

On the last page I have explained how I tried to get the RIAA filtering correct with help of some simulation tools. Recently I started to do some lab exercises with Hounddog a scope and a tone generator. in the foreseeable future I will do the same exercise with Tweety, and the results will be posted to page 4 of Tweety. If you like to read about my RIAA filter setup and calculations please read "the background section" on this subject.

What more can I do .... ?

At some point in time one has to stop the current project knowing that there is always room for improvement. For me a difficult moment to realise that I have a long way to go in my goal to build the ultimate phono equaliser, after all it is just as much fun to keep these enhancements for future projects.

1. First of all, the powersupply could be built with no-compromise parts and use voltage stabilisation by means of IC or transistor. But I'm afraid that the room available in the used box is not enough to use a tube-rectifier, several chokes and mkp-in-oil capacitors.
2. As explained above, next time I will pay more attention to the filament supply. Especially in SRPP or Mu-Follower circuits you need to pay attention to the max allowable voltage difference beteen cathode and filaments. It will enhance the life of your tubes also.
3. Next time I would use Black Gate capacitors for decoupling of the cathode circuits. The audio capacitor between the input stage and the output stage is now an Audyn MKP capacitor. In future I would like to use Ultra cap or tin-foil caps.
4. The RIAA filter is built around MKP and MKT capacitor. Both are 5% tolerance and that may not be sufficient altough simulations and some initial measurements show that this is less critical as log as both are reasonably matched.
5. I decided that I would like to experiment with another type of tube in a next project. That does not mean ECC88 or other novals, but 12SL7, 12SN7 or similar.

© Maarten&Annemarie, March 2002, 2006