Ian,
the room is 20' x 24', but this is misleading as the floor-plan is open with no doors and few segments of walls between the rooms on the 1st floor, so it is sort of 34x46. There is a lot of glass and not much furniture so it is live.
I assumed that I would need more wattage as I "listen" to music. It is not background. But my own tube amps from the 70's were gone by the time I could afford used horns so I have never listened to them in combination. Your experience with 1W into 104kb challenges my prejudice and is exciting!
I was intrigued by the curves which indicated that under-loading the output would shift distortion from odd to even harmonics, but did not know that this could impact the low frequency band. Can you direct me to links which illuminate my understanding of the compromises? My son sings opera, as a bass, so this frequency band is sacred!
the room is 20' x 24', but this is misleading as the floor-plan is open with no doors and few segments of walls between the rooms on the 1st floor, so it is sort of 34x46. There is a lot of glass and not much furniture so it is live.
I assumed that I would need more wattage as I "listen" to music. It is not background. But my own tube amps from the 70's were gone by the time I could afford used horns so I have never listened to them in combination. Your experience with 1W into 104kb challenges my prejudice and is exciting!
I was intrigued by the curves which indicated that under-loading the output would shift distortion from odd to even harmonics, but did not know that this could impact the low frequency band. Can you direct me to links which illuminate my understanding of the compromises? My son sings opera, as a bass, so this frequency band is sacred!
I have a similar room with Khorns
Don't settle for less than 5-7 watts for the live music experience
Don't settle for less than 5-7 watts for the live music experience
It's just a suggestion, to make it easier for the readers looking for certain tubes and/or designs, e.g., 6Y6 PSE Amplifier.
Good point. Egregious may be accurate, but is not a useful search term!
So "6Y6 or 26E6 PSE Amplifer"?
If I want to simulate the single-tube 2.5K load line on four tubes in parallel is there a better solution than a 600 Ohm primary (or a 1.25K primary with 8 Ohm secondary on my 4 Ohm load)?
Due to my 26V heaters I had already abandoned the traditional multi-tap transformer and planned to use a stack of DC power supplies for the heaters. I could add one more for the bias and then all tubes would receive the same bias voltage?
So "6Y6 or 26E6 PSE Amplifer"?
It seems that I can expect 1.75 W per tube in triode mode, so I will need 4 of them.
My only planned feedback was UL in pentode mode, with a switch to kill UL for triode.
If I want to simulate the single-tube 2.5K load line on four tubes in parallel is there a better solution than a 600 Ohm primary (or a 1.25K primary with 8 Ohm secondary on my 4 Ohm load)?
Due to my 26V heaters I had already abandoned the traditional multi-tap transformer and planned to use a stack of DC power supplies for the heaters. I could add one more for the bias and then all tubes would receive the same bias voltage?
I try to design for even, wide bandwidth, low distortion, and good damping. If you get good damping, then you will be surprised at what only a few Watts can do.
Of course I assume that your speaker is capable of doing its job too...
Rather the other way around: a typical 850:4R transformer might suffer when used as 1k7:8R. Primary induction of a typical 1k7:8R transformer should be plenty enough when used 850:4R.
Take care also though to keep primary DC currents within indicated limits.
Some load lines for your consideration, you might be able to get away with using your Edcor OPT after all. Using 1.7k:8, or 3.4k per tube, you get:
2x UL, Po = 8.6W
2x Triode, Po = 3.5W
I think that'll be plenty for the Cornerhorns.
2x UL, Po = 8.6W
2x Triode, Po = 3.5W
I think that'll be plenty for the Cornerhorns.
I like your load lines better than mine! (higher power and gain).
What software are you using? The only software I stumbled across did not seem to include the 6Y6 so I am using a pen and straightedge on printed curves.
But does the distortion curve in the following link indicate that a load of 3.4k per tube would increase distortion to about 6% even harmonic and 11% odd, for a total of 17%?
https://frank.pocnet.net/sheets/127/6/6Y6G.pdf
That curve is for 135V rather than 200, so I am trying to extrapolate. The pen, straightedge, and calculator may not be helping me.
The load lines are plotted with the pctube (code by Ayumi Nakabayashi), which runs on top of R, a statistical math program. Not sure about the distortion figures, in any case, the 3.4k isn't ideal but as the OP already has the Edcor OPT, it will have to do. Some gNFB will certainly help to knock the distortion down a bit.
Not so quickly, higher turns OPT have also higher parasitic L.
And higher this value is, the more malign it is - mainly on lower Z OPT. (70mH will do not harm on 10k; but this on 5k can be a small disaster)
Never heard of "parasitic L"; maybe you mean leakage inductance?
Could you explain why the difference between 10k and 5k?
I guess it has more to do with winding technique.
Very interesting!
http://www.mif.pg.gda.pl/homepages/frank/sheets/049/2/26A7GT.pdf
I had planned to run the plate of the 26E6 at 24V and was only expecting about 10mA. Your 26A7 offers 20.5 mA in each half of the tube
The dual tube was intended to be a single tube PP amp, but do you see any problems running them independently?
Is this it:
Ayumi's Lab.
R: The R Project for Statistical Computing
I have a few friends in Japan who are electrical engineers, so perhaps they can help me with translation! It is a shame that I did not learn more Japanese while attending the University of Tokyo.
Yes, that's it. All the codes are in English, only the instructions are in Japanese. Google translate does a good enough job, so you should be able to understand them.
First off I was confusing two different tubes. The 26A7 is an OCTAL and the 28D7 is the LOCTAL. Both were intended to operate as audio amps on aircraft 26 to 28 volt power systems. I had several of each and learned quickly that the plate voltage rating may be abused quite a bit, but the screen grid can not eat a lot of voltage. I was exploring maximum audio power output from higher voltages (100 to 150 volts). I can't remember all the details since I did all of this maybe 15 years ago. It doesn't matter on your chosen 24 volt supply especially when linearity is not important.
There is a common cathode, and common screen grid in both tubes. If this fits your design, then it should be OK.
> does the distortion curve in the .... 6% even harmonic and 11% odd .... That curve is for 135V rather than 200
If you do not run out of voltage-swing, a single (or parallel SE) amp makes no 3rd.
ANY amp will make 3rd out the wazoo when you hit the limits of voltage swing on both sides.
At 135V (and rather silly test condition to minimize the labors of the junior engineer), it runs out of voltage swing. At 200 in the same load, it may not. Jazbo8's curve sure does not smell of 3rd.
If you do not run out of voltage-swing, a single (or parallel SE) amp makes no 3rd.
ANY amp will make 3rd out the wazoo when you hit the limits of voltage swing on both sides.
At 135V (and rather silly test condition to minimize the labors of the junior engineer), it runs out of voltage swing. At 200 in the same load, it may not. Jazbo8's curve sure does not smell of 3rd.
Every curve I can find (any tube, any voltage) shows this pattern of heavy 2nd harmonic below the suggested load, and heavy 3rd above the suggested load. Some curves fail to resolve the distortion into its harmonic components, but the patterns look like the summation of declining 2nd and increasing 3rd. Two examples of more popular tubes which do include 2nd and 3rd content follow
https://frank.pocnet.net/sheets/127/6/6V6.pdf
http://www.tubeampdoctor.com/images/File/DATA SHEET 5881 Tung-Sol.pdf
Thus my fanatical pursuit of low OPT loads!
Are these charts reflecting an anomaly of test procedures and not indicative of real-world performance?
Am I misreading the charts?
https://frank.pocnet.net/sheets/127/6/6V6.pdf
http://www.tubeampdoctor.com/images/File/DATA SHEET 5881 Tung-Sol.pdf
Thus my fanatical pursuit of low OPT loads!
Are these charts reflecting an anomaly of test procedures and not indicative of real-world performance?
Am I misreading the charts?
- Status
- Not open for further replies.