Tubelab SE Treble response

[acceler8]

After swapping my TSE (Sovetek 300b, Trancendar 20w OPT - TT-020-OT, several years on the clock) in and out of my system with another amp, I noticed that my trebble was severely lacking. I've always thought so, but the A/B comparison really highlighted the issue.
I checked the frequency response (QuantAsylum QA400) and it appeared mostly flat to 20khz, but in reality there is a large difference to other amps.

Looking for improvements, I found the following post by George:
"The grid stopper is normally used to prevent oscillation. It has a second function here. The grid stopper resistor determines the time constant under overload conditions since it is in series with the grid. It also forms a low pass filter pole with the tubes Miller capacitance so it affects the high frequency response of the amp. A big resistor (10K) will eliminate blocking distortion in most cases, but it will kill your treble response in triode mode. It works in pentode mode though."
at
http://www.diyaudio.com/forums/tubelab/177194-tubelab-se-coupling-capacitors.html

So, would playing with this resistor value (R33 and R34 on the schematic) be wise? If not, any other suggestions to increase the top end?

Thanks!

 

[wdecho]

I think not. I have just finished build the SE amp and it has a very good treble top end to my ears, not as good as say the Simple SE which I also have or some Class A SS amps. I think the treble is mostly a factor of the tube used than the circuit. A better 300B tube that is known to have better treble is probably what you need if this is what you are seeking. I built the SE because of the outstanding mids and vocals knowing that other amps are better at the lows and highs. To me music lives in the mids and I want outstanding mids over anything else. I do not think you will find any better mids with another amp over a 300B or other DHT. All of this is IMHO.

 

[wdecho]

I thought I would add that I am using these,

Genalex - Gold Lion PX 300B

You can find them on Ebay at a good price in pairs from a reputable seller. I read all the reviews I could find on 300B's before purchasing and these were the best bang for buck that I could find. So far after a short period of listening with this amp I am enjoying the sound immensely. I now see the allure of the 300B tube. The treble has more bloom than the simple SE that uses JJ tubes. The speakers are absolutely not there listening to music.

 

[acceler8]

Thanks, wdecho! I was hoping to stay away from new tubes (so very expensive), but I guess its inevitable.

 

[wdecho]

I now have both the simple SE and the SE and there is a very big difference in the treble between the two amps. Not that one is better than the other but very different. This is with my brand of tubes. I love the very smooth sound of the treble with the SE, very musical but then the Simple SE is more bright and quick sounding and I can see how others might prefer this sound over the SE. It probably comes down to personal preference and the sound you are after.

I do not know if buying new tubes are going to get the treble you are seeking but all the reviewers that I have found that have compared the Soveltek to the Genalex prefer the Genalex, not that there is anything wrong with the Soveltek.

I always lean to the better mids and live with the highs and lows of the one with the better mids.

 

[Evenharmonics]

Frequency response measurement would be helpful, if you've got em.

 

[acceler8]

Here is an older Freq. Resp. graph on ARTA (from before I had the QA400).

I thought I had the QA charts uploaded, but I guess I don't. I will try and find them tonight.

 

[w5jag]

Does that graph represent a single tone input at 1 KHz?

How did you like that software, btw?

Win W5JAG

 

[tomchr]

After swapping my TSE (Sovetek 300b, Trancendar 20w OPT - TT-020-OT, several years on the clock) in and out of my system with another amp, I noticed that my trebble was severely lacking. I've always thought so, but the A/B comparison really highlighted the issue.
I checked the frequency response (QuantAsylum QA400) and it appeared mostly flat to 20khz, but in reality there is a large difference to other amps.

There are a few things that can contribute to a "lacking treble". One is a bandwidth limitation, which I think is what you're trying to measure. The other could be that the "lacking treble" circuit actually has a cleaner transient response. Less ringing on the transient -> less "fizz", which could be perceived as a "lacking treble response". Finally, it could also be that your tube amp has a relatively high output impedance, hence, doesn't control the speaker driver as well. It may not excite the cone break-up modes the same way, which could smooth out high-Q resonances in the speaker+amp's combined frequency response. Again less fizz -> "lacking treble". With measurements it would likely be pretty straight forward to see what's going on.

Looking for improvements, I found the following post by George:
"The grid stopper is normally used to prevent oscillation. It has a second function here. The grid stopper resistor determines the time constant under overload conditions since it is in series with the grid. It also forms a low pass filter pole with the tubes Miller capacitance so it affects the high frequency response of the amp. A big resistor (10K) will eliminate blocking distortion in most cases, but it will kill your treble response in triode mode.

Playing with the grid stopper is generally a bad idea. What George is trying to say is that if you go extreme in one direction (say 0 Ω), you're likely to get oscillation. If you go extreme in the other direction (say 10 kΩ), you're likely to kill the treble. A good value would be between these extremes. Typically you see resistors in the range of 100 Ω*to a few kΩ used for grid stoppers. The actual value depends on the tubes involved. I wouldn't mess with it unless you have the equipment needed to detect HF oscillation (spectrum analyzer or wideband oscilloscope).

You can estimate the lowpass function of the grid stopper by looking at the input capacitance of the tube in the data sheet and using the following equation:

f = 1/(2*pi*R*C)

R is the grid stopper, C is the tube's input capacitance, f is the cutoff (-3 dB) frequency). All SI units.

Here is an older Freq. Resp. graph on ARTA (from before I had the QA400).

That's an FFT of the harmonic spectrum, not a frequency response plot. A frequency response plot shows the amplitude/gain as function of frequency. Here's an example.

An externally hosted image should be here but it was not working when we last tested it.

Tom

 

[acceler8]

W5JAG,
Yes, that was a 1Khz tone (with 60Hz hum & harmonics). I liked Arta well enough (esp. b/c it was Free!) But the QA400 is much easier to test with.


TomChr,
You are correct; sorry I posted the wrong image and not the F.R. Here is a screen cap from the QA400:


Unfortunatley, I cut the bottom off (with the X-Axis labels). If you look in the top left, you can see that the measurement range was 20-20Khz (with a Log Scale).

With measurements it would likely be pretty straight forward to see what's going on.

I do have REW and a EMC8000 but had never thought of comparing amps for some reason. Thanks for the idea!!!

 

[Evenharmonics]

Here is a screen cap from the QA400:

What's going on with those peaks at the lower half?

 

[acceler8]

Ugh, I never seem to pick the right link to images in the gallery. Here is the full-size:
TSE 300b Frequency Response - My Photo Gallery

I assumed the peaks were 60Hz hum (and harmonics). I've never really got a clean test without them.

 

[gcom]

That is also an fft.

Sent from my HTC One_M8 using Tapatalk

 

[acceler8]

I'll freely admit that I am unsure of the differences between an FFT and a Frequency Response graph. I assumed the the 'FR' mode of the QA400 referred to 'Frequency Response'. Can you (and/or TomChr) explain the difference? I (think) I know that FFT implies a mathematical conversion from a time domain to a frequency domain. But at the end of the day, doesn't that conversion imply a close relation to the frequency of a device under test?
Thanks for making me ask questions!

 

[gcom]

My understanding of it. Wide open to correction or improvement:
That analysis is of a single pure tone (in this case 1Khz) that is played through your amp.
The FFT then runs a spectrum analysis on what comes out to see it the amp has added other tones or harmonics. So you see the 60hz hum and some harmonics that have been added.
A freqeuncy response analysis plays a sweep of all the frequencies from low to high and graphs the level of what comes out to tell you if it is playing all frequencies equally or some frequencies louder or quieter than others.

 

[Evenharmonics]

FFT here. Frequency response here.

What amp FFT measurement typically looks like.

What amp frequency response typically looks like.

An externally hosted image should be here but it was not working when we last tested it.

 

[Francis Jansz]

Dear Audio specs.
I need to know what resistor values and the Capacitor values . Ñeeded to build a tube active 2 way ćrossover and change the filter to 600Hz low pass at 24 db i recently bought a pete millett 2 way tube crossover PCB and would like to get the filter working to low pass 600 hz cut out frequency at 24 db slope and would like my high pass filter to start from 600 hz to infinity please help me to achive this i need to know as to the values of the caps and the resusters to do this. So that i can order the right parts.
Kind Regards
Francis Jsnsz