OK, so I did some heavy fiddling, got it sounding a bit better. Load resistor is 22k, 1st cathode is 220, 2nd is 100. switched coupler to .1, got rid of some shrill mid-highs.
However, anode voltage is still in question. With 22k, I get 130ish. 2nd anode is still at 230. I placed a loading resistor in series with the OPT, 10k, which mostly equalized plate voltages, so I am assuming this OPT is simply way out of line. I have yet to measure voltages with a different OPT. How the heck do you figure the impedance of an OPT?
here's a question:
what about using a high impedance dummy load paralleled with the speaker, to increase the tubes' perceived imp.? Say, 20 ohms or so?
tyler
However, anode voltage is still in question. With 22k, I get 130ish. 2nd anode is still at 230. I placed a loading resistor in series with the OPT, 10k, which mostly equalized plate voltages, so I am assuming this OPT is simply way out of line. I have yet to measure voltages with a different OPT. How the heck do you figure the impedance of an OPT?
here's a question:
what about using a high impedance dummy load paralleled with the speaker, to increase the tubes' perceived imp.? Say, 20 ohms or so?
tyler
My circuit can be found in this thread http://www.diyaudio.com/forums/tubes-valves/156253-12au7-amp.html Used the 12ax7 as the first stage tube to get good gain for cathode feedback in the 12au7. PC sound card made it distort at about 2/3 of its output (less than 1 volt rms) so a 12at7 could have done the job but none on hand at the time.
What on earth are you trying to do? The anode voltage on your power valve will always be at or near the supply voltage- an OPT has almost no resistance! Putting a resistor in series with it will simply reduce the available power output. I'm not sure what garden path you've been led up, but anode voltage is the least important voltage in an amp, and is not something you should "design for".
I think you're still missing my point: according to the information provided the loadline was selected incorerctly so the operating point was near the bottom of the U-I graph where the curves bunch up. I suggested the loadline to be tilted up AND operating point be moved further up still. This is what my statement you quoted was referring to: operating closer towards the center of the SOA will produce more linear results. I wouldn't want my statement to be misinterpreted so I'm trying to clarify it.
Sorry, should've been clearer on this: I'm not USING a load resistor, rather it was temporarily placed to determine supply load on each plate.
Another point of interest, the "olson" signal caps I've been using in just about everything are leaky as hell. I've now had at least 5 of them fail, at much lower than rated voltage. Shame. Oh yeah, the amp sounds much better, got that leaky cap out, stabilized 2nd grid, and biased a bit cooler. Still can't figure EXACTLY how to set voltage correctly in first triode, but it's decent none the less.
tyler
tmaxwell,
You can't determine impedance ratio from DCR BUT you can from Turns Ratio which is voltage ratio.
Use a small transformer (6.3V filament winding will work "OK")
Put the low voltage thru the PRIMARY WINDING and MEASURE the voltage across it while powered up.
Then take a voltage reading on the secondary side. Be as accurate as you can with these two voltages.
Divide the PRIMARY VOLTAGE by the SECONDARY VOLTAGE (This is your turns ratio)
The "SQUARE" of this ratio is the IMPEDANCE RATIO multiply the square by the speaker load and you get the primary Impedance with that load.
Example. (assuming you have a 10K to 4ohm tranny) the numbers would look like this:
6.3 volts into the primary nets about 126 millivolts out.
So 6.3/126 = 50
50 x 50 = 2500
4 x 2500 = 10K
You're changing multiple parameters at the same time when you change one of resistors (anode or cathode) in first stage. Since you cannot listen to datasheet, do not bother yourself with trying to get the exact datasheet values, instead get your tube into sane area and enjoy the music !
18-20K anode resistor and 330-680R cathode resistor should be just fine.
I made an Excell / OpenOffice spreadsheet which will plot load lines for you. Just plug in the resistances / voltages and where the load lines cross tells you the bias point.
Excell
http://www.freewebs.com/valvewizard2/LoadLinePlotter.xls
Open Office:
http://www.freewebs.com/valvewizard2/LoadLinePlotter.ods
You can also play about with this calculator: http://www.pentodepress.com/tubes/12AU7/
although it won't tell you the bias point straight off.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.