Tu-be or not Tu-be...?

 

 

 

HOME

PROFIEL

SPEAKERS

HOME CINEMA SET

BUIZEN VERSTERKERS

     * Annastasia PP

             * Annastasia overzicht

     * Oliver S E 

             * Oliver bedrading

A new High Performance KT88 PP amplifier, combining old techniques, with today's high performance electronics!

Old, meets new, in a state of the art amplifier....

What high-grade electronics are in:

-A state of the art "remote controlled attenuator" from Tent Labs.

-A "tube heater supply" from Tent Labs.

-A "NBS" (Negative Bias Supply) from Menno van der Veen, sold by (Tent Labs.)

-A self made Solid State ON/OFF module, controlled by the attenuator.

What more?

-Well, a heavy transformator set from AE-Europe (Schagen Holland)

-Self designed CNC produced, Aluminum chassis plates, for top-, front-, and back-panels, made by "Schaeffer" Germany.

-A rock solid, self-made cabinet, made of American Walnut.

     * Phono Dude

     * Oliver MK2

     * Bill

     * 300 B Exclusive

     * KT 88 PP

     * EML 300B

     * Cleo 6

     * CNO Project 2010

     * EML 300 B revised  

THORENS DRAAITAFELS

AUDIO MEUBELS

LINKS

 

As soon as this project reaches its end, I was very curious about the (electrical) performance of this amp. That's why I start this presentation, on some measuring results that can be done with my equipment. A distortion meter is something that's written on my wishing list....so not yet available:(
First a 1kHz sinus wave signal is connected to the input.

As can be seen 97mVtt at the input,                    ( 97/2=48,5Vtop ,48,5/1,41=34,3mVrms), equal to 34,3mVrms ,will produce a signal of 5,55Vtt at the output, which equals some 1,97Vrms.

On a output resistor of 3R9 this gives some 1Wrms at the output.

P=U^2 / R

(1,97^2 / 3.9 = 1Wrms)

Au=20log Uout/Uin=20log1,97/0,0343=35dB

 

To measure the bandwidht of this amplifier I have to determine the -3dB points over the frequency band. I will check frequency response at 1W output.

Au at 1kHz=35dB , so I have to look for the 32dB (-3dB) points at the lowest frequency, and at the highest frequency.

So I have to find the point, where Uout equals:

Au=20logUout/Uin >>>> 32=20logUout/0,097 >>>> logUout/0,097=1,6 >>>> Uout=3,86Vtop-top

It shows the -3dB point at 5Hz at the base of the frequency bandth, picture left.

 (Oscilloscoop is in digital mode)

I'll try to add some more measurements in the future, so stay tuned....

Now, the making of.....;

Again, it all starts on producing a decent housing, a walnut one, my favorit. There will be CNC lasered alu-anodised panels for front-, top-, and back of the amplifier.

A special bracket is made for mounting the 5H 500mA coils, this bracket is attached to a solid walnut block. The r
BYW96E Fast Soft-Recovery diodes are used as rectifiers in the power supply section. They are soldered on (NOS) ceramic solder tags. The green section on the left is used as a central soldering point for the 6,3Vac heating supply's. As there are very high voltages present on this soldering tag, I have made a plexiglas covering for it, see next picture.
On top of the plexiglas safety cover I mounted my self-made remote controlled power-up PCB. Build around the SHARP S202SE solid-state relay. The green LED gives the power-up indication when servicing the amp.

The bleu thick wires switch  the main transformer, the small black and red wires are 5Vdc driver signals comming from the Attenuator.

The large photo beneath illustrates the PCB placement of the "van der Veen" NBS PCB unit's. The grid supply-PCB placed as close as possible to the output (KT88) tubes. A complete application note ( In English) can be find here.

Some comments (taken from the Tent Labs web-page):

This negative bias supply is designed to bias (output) tubes,
such that their operating point(s) remain stable, independent of tube age,
mains voltage variations and music signals, without affecting the musical
and technical qualities of the amplifier.

Some major advantages, of which this unit contributes are:

The heavy coils take their places, on a self made alu chassis. Left a glimp of the solid state relay PCB.

• Higher resolution and correct staging due to lower hum and noise level
• Better bass due to near perfect (<1% error) current matching in pushpull
stages
• Better imaging due to optimized left / right channel matching

And I can only agree on this statements, this unit is just gorgious!

It has a onboard safety circuit, to protect you're tubes against failures. It has to detect high-voltage first before bias voltage becomes present. In this case it starts on -85V, and after some 40 seconds it regulates the bias to a pre-adjusted level of -50Vdc. Bias current will be some 60mA through the KT88 tubes.

 

 

Part 1

Notice the wiring to make it suitable for 220Vac use. The PCB is mounted on a self-made alu plate.

On the photo beneath, the JJ High-Voltage capacitors can be seen, attached with metal brackets on the wooden frame and on the alu top panel. Wiring is kept as short as possible, by mounting these two cap's close to each other. At the right of the cap's the High-Voltage power switch, with the two 100K resistors across. On the right, a piece of the NGS (Negative Grid Supply) PCB can be seen. This control unit is build of two separated PCB's, this one is the power supply part.

On the rear a glimp of the high-voltage supply, using Philips Long-Life capacitors for there excellent qualities.

eason for this ,is to maintain access to the nuts for mounting the output transformer housings.

Silver plated, teflon isolated wiring is used for all connections, exept signal and speaker wires.

The screen picture beneath shows a full power 1kHz output signal of 39,5Vtt over the 3R9 output resistor, supplying the resistor 50Watts RMS On the naked eye the signal stays very clean...To achieve this output rate, I put a 600mVtt signal on the input, equal to 212mVrms. By the way, the amplifier is capable of delivering 41,6Vtt before clipping occurs, some 55Watts!