Facelift for the MU Stage in the KR300BXLS Golidav Amp
Normally you only give facelifts to older people , so why did I need to change something in the brand-new GOLIDAV? Let me explain: At the time I designed the amplifier, it was impossible to predict all working points and biases etc up front. So what you do, is to make the best possible effort on paper, built the project and than......... yeahhh, what than? Here the problems begins !! I know that a lot of people plug in the amp and without bad luck, MUSIC !! when it has been a design with better tubes etc, the sound will convince you, you did a great job. And the amp normally play well too. I did this in the far past as well, later on, I realized, it might be a good idea to have at least a signal on the amp tested with an oscilloscope. So, that is what I and some others thanks goodness do. You put a sinewave from 20Hz to 200kHz on the AMP, loaded with a resistor representing the lowest impedance of your speakersystem (in my case 2.7 Ohm) and checkout different levels. Finally a squarewave to check any ringing and oscillating and if every thing looks fine, which means: sufficient bandwidth and little rings outside the audioband on a squarewave, the amp get the stamp "Passed Quality Check" and moves into the living room. in 95% of the cases this is good enough, so what is the big deal than????
HOW DID I DISCOVER THE "PROBLEM"?
Ok, ok, I'll try to make the long story short: Since half a year or so, I have a digital scope with FFT analysis. Specially the FFT analysis, shows clearly how the distortion spectrum looks like. If you only have the scope, normally you say: well the sinewave looks fine, BUT in the frequency domain you see much more information and with more precision !! I did not checked the amp in the beginning with the analyzer. On purpose, I wanted to burn in the amp first. After a few weeks, I was very pleased with the sound, but I still had the feeling, that the Hulk was not coming totally out of his shirt........ 15 Watt was predicted. Well, at higher levels the amp start coloring the sound too much I found; this is pointing into the direction of too much uneven distortion. So I brought the amps down in the LAB for the FFT test........
This is an example from the FFT analyzer. The test signal is at 1kHz. The first (even) harmonic is called d2 and is clearly visible. the level is aprox at 1% (-40dB) the d3 is very low, than d4 comes up. the d5-7 are coming from my signwave generator, so we can discard them. This is a desirable spectrum. from d1 to d4 a nice line down and d3 below the rest. If the d3 pops up next to the d2, you know you have a grid current problem or you clip in the Supply Voltage. This is hard to see on a sine wave !!!! here you see immediately (turning the volume) at which point it kicks in. Literally kicks in !! and that is what you hear as coloration. Nice tool, ehh?
GOLIDAV AT THE DOCTOR's: "the diagnose"
I started to drive a 1kHz signal through the amp. When turning the "volume" of the sinewave generator, you could see the d2 slowly rise, till a certain point where suddenly the d2 jumped up and the d3 kicked in !!!! Quickly checked the RMS value of the signal: OOOPSSSS: only 37 Veff. This is too low. The KR300BXLS in my design needs for full A1 power (A1 power is the power driven by a signal on the grid without crossing the Ug=0V line. after this the amp goes in A2 power, using the rest of its current. This results in grid current. For the MU no problem, so distortion stays bearable, the AMP can sound much bigger so) approximately 50 Veff. Above the 35 Veff the MU stage Triode (the 227) started pulling grid current with the result of huge distortion. The triode is not linear anymore and is not on constant current. All the good reason to have the MU-stage for in the first place ! Actually my amp was limited to a lousy 4.5 Watt !!!!!! What do you mean Hulk??? This needed to be fixed quickly, Thank G*d I have a FFT analyzer ............................................................................. I do not sell them, so I am not biased !!
I needed more headroom at the output from the MU-stage, to be able to drive the endstage far into A2 without distortion! Therefore I reduced the resistor between the anode of the 227 and the cathode of the D3A from 15k to 4k7. To increase the Bias Voltage of the 227, I changed the Rk from 1k5 to 2k3. This resulted in Ua=190 Volt, Uk=12 Volt and Ia=5mA (rough values). FFT analyzer connected and there we go: in A1 power the MU stage swings with maximum d2 of 0.1% At almost full A2 power (my sinewave is limited to 7Veff) and 20kHz (!) maximum d2 is not more than 1% !!! The max output is now 85 Veff. Enough to deliver almost distortion free (well 5%-10% d2 in A2 power is considered as low in tube amps...) 25 Watt in A2 power. The A1 power is now 13 Watt with 3% d2. NOT BAD at all.
HOW DOES IT SOUND?
Good to hear all this science bulls**t, but any noticeable differences? Sure! No surprise that the amp can go much louder now, much more headroom for dynamic peaks. this aspect and also the fact that the MU stage never runs out of fuel, the amp sounds more open, bringing more rest in the whole soundstage, more control. The tonal balance stays the same, but it is really the authority that has gained seriously. No streetlength stuff etc. But now I finally encountered the Hulk in this amplifier.....
Ha ha! sound like that old FBI television series. At the end all the wise words and moral preaching came all over you. And that is exactly what I am going to do here. WITHOUT the FFT analyzer I would have had a less superb amplifier. The sine waves looked great and the sound was very good. So no real need for a change? Only with the FFT analyzer I was able to discover the distortion problems and tweak the amp, based on its internal physics. So my statement of the day is: If you seriously want to design and built Tube amplifiers, you cannot do without the right T&M equipment. Uhhhh, shall we set up a joint purchase for FFT analyzers? (just kidding...... or ????)
Happy building, listening AND testing.......
IMPORTANT: The information provided on this page is intended as guide for DIY activities and therefore free to copy and or publish. If any one wishes to use any of the information from my WEB site, please make sure to refer and footnote to my URL Link as source! Doede Douma