The Kurt KR52BX story

The KR52BX saga continues :

Kurt tried a few other solutions , and he finished with a real Parallell outputstage using 2 output transformers along the way .It also took him a bit of time but he realised he doesn't need that much power , so he lowered the voltage on the tubes and he chose an operating point 550vdc 140mA but he notices that after long time listening the plates are red , but this was only noticed when all lights are turned off , my advice is to reduce current to 125mA and eventually also reduce voltage to 500volts .
He is very happy with the quality , its better then anything he heard but he needs power because he uses B&W 801 series and they are not so sensitive .(86dB/Watt/meter )

I am waiting for the pictures of the finished project .

The new schematic is given here and below is the story of the birth of this amplifier .

The Kurt KR52BX story :

Intro

A strange fellow named Kurt , speaking very dialectic Neanderthal dutch walked in the shop and asked for the best we have with lots of power , at that time we just received the krone tubes , so although very expensive I offered this as a solution 25watts rms from 1 tube in Single ended mode , what could you wish for more ? ( did I know what ?)
Well Kurt said I need more so I proposed a pse version that theoretically should be capable of putting out 50 watts together and this would run on a custom made transformer that could handle this amount of current .
Also in fashion at the time where (maybe still is ) the SRPP or mu follower circuits and interstage designs as to avoid capacitor coupling , so the scene was set .
The specs supplied by kr enterprises ( not to confound with mr Spocks enterprise ) mentioned
500volts anode and -95volt grid bias and a whopping 25watts Rms output power

Problems :

The main problem was that besides good quality sound very little power was coming out of the beast ,
about 9watts for this pse beast .
So a small check provided the following measured data :
U srpp = 244 volts Vin=2v peak Vout=25v peak
Uk srpp = 3,77v u anode = 121,2v Ianode = 3,77mA

U kt66 = 416v u anode kt66 = 403v uk = 149,2v ( into 3,75k)
Ianode = 39,7mA rdc LL1660 = 350

That seemed to work ok

With 1 outputtube :
The swing on anode with load = 200vpeak without load =350vpeak

2 outputtubes :
The swing on anode with load = 280vpeak without load =350vpeak

Saturating with best bias for max output at 14,5v peak on 8,25 ohms

The dc supply dropped to 480v on tube anodes at 240 mA
Due to choke DC resistance and damper resistor and DC resistance outputtransformers
Which is far too low to get maximum power
Also clipping occurred whenever grid approached ground .
This is a problem with bias obtained throughout a pot ( resistor ) , it cannot supply current
(high impedance )

Conclusion :

This is more suited to a traditional 300b because due to greater mu you'll get more power and the supply voltages are perfect for high power operation and will result in 25 watts of rms outputpower in pse mode .
However with the krones the main problem was 2 fold , since it was clipping whenever the grid was getting close to ground level , and a low impedance bias supply didn't help , the only way to get more power was through the increase of negative bias ( this is due to the low mu measured at 3,5 )
And following this idea a new powersupply was build .

How to solve this :

Negative bias supply had to be modified to allow grid current , = positive swing
In theory about 25% more power ( doesn't work , I still don't know why , anybody can explain please )
And plate voltage far too low , 600 volt on the plates seemed like a minimum
So from experiences of 845se , a powersupply was assembled on a wooden board
And a negative low impedance bias supply ( capable of delivering 10ma ) was also build on a wooden board .
Without load 880vdc high voltage ( this will drop to 600 volts under load)
And a negative bias up to -260 volts available ( -135 volts targeted )

The test :

At first for safety the voltage doubler was set to minimum and voltage on tube plate
Reached 560 volts , this already gave 35w rms
Then voltage doubler was modified to allow 720 volts giving a net of 680 volts on tube plates for 240ma idle current consumption at about -162 volt bias supply
Well achieved 50watts RMS or 28,5vpeak on load of 8,25ohm , but not entirely resolving clipping problems the grid still clips when it swings positively .must be the nature of the tube .

Notes on the revised powersupply :

Voltage doubler is necessary to obtain the required voltage using my std toroid multipurpose powertransformer . (cheap and excellent)

The caps that are charged by the diodes determine the outputvoltage :

Xc=omega C
The voltage drop across this impedance is in exact proportion to the current supplied
320vac x 1,41 = 450vdc
Xc = 14µF = 227 ohm
For 240ma a voltage drop of 55 volts

So loss is : 450-55 = 395 volts x 2 = 790 vdc
After going through 2 chokes and outputtransformer ( 435 ohm in total )
Net plate voltage over tubes is 685 volts

Bias is obtained through a rectified pre amp power toroid 230vac = 310 vdc
Lots of zenerdiodes in series to give different bias solutions and resistor setting current at 15mA

Filaments are of course rectified for DHT
But kept in ac for the SRPP and KT66

The output transformer is the LL1623/250mA
It is perfect for the series of krone 300B tubes as it can be wound to suit 680 ohm or 1440 or 2445
ohms to 8 ohms load .

All primaries in series
Secondaries connected as following :
In series : (11-19)- (20-12) - (13-21) - (14 -22)
In series : (26-18) - (25-17) - ( 23-15) - (16-24)
Then these 2 groups parallel

This is also shown as alternative D by the way

Know that the power output has reached the required value (kurt is smiling )
I can start on filtering the powersupply so that hum is within acceptable limits
This is done with heavy capacitors , and a few chokes .
Quality was best with 620volt on the plates and a maximum of 38 watts RMS
Of very clean outputpower , Kurt wants more , but the specs will not allow for it .
Also at 50watts RMS quality is lousy and the tube is rated for a maximum of 600vdc
The schematic is given , high quality parts used everywhere , the interstage tranny was a big worry but the Lundahl is really very good even at 50mA and the output transformer delivers very nice sound
Also the powersupply needs paper in oil caps at the voltage doubler because the load currents are very violent
And this needs to be stabilised because SE designs are very hum sensitive especially in the first stages .

So all in all this story ends on a good note however the KR factory is not very precise with its specs so beware of them , and measure before you build ( we had to start all over and the beautifull copper chassis are good for the garbage now )

To all who visit my website and to all customers a big thank you for without you're feedback these projects would not see the light of day .

Pictures of the schematic and of the prototype :

Sincerely
Benny Glass

Pictures of the schematic and of the prototype :

copyright webdesign 2002 aquablue Benny Glass