Audio Asylum Thread Printer
Get a view of an entire thread on one page
|For Sale Ads|
In Reply to: RE: Driving Sound Labs with Atma's posted by throwback on June 13, 2012 at 16:34:27
Output Tubes 6AS7....
I have had the M60s for years...an i have a pr of Diy,ed MA1...these amps well run fine on the m60 4 tubes pr amp..the MA1 well run with 4 tubes pr amp!....the funny thing is the amps are the most stabel amps i have ever owened....The Sovtek 6as7...are great.. An all the 6as7s you get new MUST BE RAN ON THE HEATERS ONLY FOR i do 48 hrs...24hr well do.....An then thay well last for ever less there over bias.... but i have never lost but two tube in 8years an thay were the only match,et 6AS7 i ever got ....so i just dont get how you can lose tubes....well thay must be over bias,et...
Now as for this ESL mod ...i only have Acoustats but the interfaces are set up like the soundLabes some what ...i have had Acoustats from the 80s an by doing this mod i am geting the best highs i have ever got out of any ESL...even better than the Tube amps Made to drive the Acoustats.....Be for the removel of the res an caps in the primary of the high frc tranfourmers the MA1 coud not give good sound with the Acoustats... now the M60 give all the output i need...go figg..Thanks for your time an post on all the thing you done to you SLM1s...goodluck
Thanks, TYU. I thought the bad tubes might have been the answer, but if you were able to run the speakers with only 4 tubes, well, I dunno. I am still puzzled by the fact that I seem to be getting lower inpedance readings than everyone else.
It doesn't tell much to say that you are getting "lower readings than anyone else", since I am the only one I know who has actually posted data on line. (I believe you posted only SP levels vs frequency on SLOG.) If your readings are lower than mine, it is quite possibly due to the fact that Dr. West was lately experimenting with a new highly resistive coating on the membrane. My speakers were built probably last summer. But anyway, I don't know what readings you are actually getting, so I cannot judge. If you take a look at my data you will see that above 2kHz, even my speakers are 4 ohms and below. One thing that might help, if needed at all, is to reduce the bias voltage below the threshold level recommended by SL. I believe that Z will be to a degree inversely proportional to bias voltage. (I don't actually know this, but it stands to reason, since with no bias V the impedance would probably be infinite.) Anyway, of course first you need to replace those bad tubes, and like Tyu says, run the filaments for several days before applying B+.
I sent you a message with these readings and my whole Odyssey to date, but apparently you never received it. I'll try again.
Yesterday, I had a metrologist friend over to make impedance measurements. Configuration of back-plate: no resistors, no capacitor, no inductor, no brilliance control, midrange control +3. The results are below:
Impedance in Ohms
Freq (Hz) AU Black/Pos spkr term AU Black/Neg spkr term
40 4 9.8
50 5 12
80 8.8 5.9
100 12 4.3
200 23 2.5
400 24 2.2
500 24 2.3
800 20 2.4
1,000 17 2.6
2,000 7.4 4.5
4,000 3.4 4.8
5,000 2.7 3.5
8,000 1.9 1.6
10,000 1.6 1.2
16,000 1.3 .58
20,000 1.3 .46
In-phase hookup shows a more benign load (higher impedance) for the amplifier; but even then, the impedance at the higher frequencies is awfully low for any amp, but especially an OTL.
I found out about the tube problem after I did these measurements.
I cannot make out what you actually got with what hook-up at 40 and 50Hz, and at 100Hz I got a much higher number, assuming you saw 8.8 ohms (see SLOG), but above 100Hz, your readings with the "AU black pos terminal", meaning I hope that the doublet of screw terminals on one side of the AU primary were connected to plus, are very much consonant with mine as posted on SLOG. (I only took readings at 20, 50, 100, 200, 300, 400, 500, 1000, 2000, 5000, 10K.)
The right-most row of readings are no doubt the result of wiring the treble transformer 180 degrees out of phase with respect to the bass transformer. Mine were very similar; don't use that hook-up.
This is the point where I am wondering if the Bass transformer is rolled off by a choke. The readings look very much like the bass transformer is active above the crossover frequency.
Ralph, I believe the bass transformer has enough self-inductance to take itself out of the circuit pretty quickly above 2kHz. It cannot have much effect at all at 5kHz and above. My data suggest this, if I compare the AU90 transformer alone to the [bass + AU90] impedance readings at above 2kHz. With a 1.5mH inductor in front of it, the bass transformer alone gives 60 ohms impedance at 2kHz. Anyway, I am pretty sure Chuck (Throwback) is using an inductor, which is at least 1.5mH in value or higher.
Actually, no: no inductor. It came out clean when I removed the rest of the stuff--resistors, Brilliance Control, capacitors.
But I have to assume that it still represents a load at 20KHz. Have you done any measurements of just the bass transformer at that frequency?
The reason I ask is if there is self inductance, then its impedance should be going sky high if its not influencing the impedance at 20KHz. But all I have ever seen of ESLs is that impedance goes down as frequency increases. That really suggests to me that while the transformer may not be making any output, its still loading the amp and for no good purpose unless you like heat. Am I way off the mark??
You wrote, "The reason I ask is if there is self inductance, then its impedance should be going sky high if its not influencing the impedance at 20KHz. But all I have ever seen of ESLs is that impedance goes down as frequency increases. That really suggests to me that while the transformer may not be making any output, its still loading the amp and for no good purpose unless you like heat. Am I way off the mark??"
Here's the way I think of it, right or wrong: (1) Yes, I think the impedance of the speaker as seen via the bass transformer alone would by "sky high" at 20kHz, but no, I did not measure the bass tranformer alone (with no treble transformer in parallel) above 2kHz, because it was going straight up to the sky already at that frequency. I did it only from 20Hz to 2kHz in log intervals. Did someone else put out some data to the contrary? (2) I think the reason the speaker has a low impedance at very high frequencies is by far mostly to do with the speaker itself and its capacitative nature. So, when the speaker is seen by the amplifier, almost entirely reflected via the treble transformer at high frequencies, the impedance will be very low. The effect of the transformer should be a constant, related to the turns ratio, at all audio frequencies. I think R West himself recently wrote that the speaker is ~1600pF capacitor. (Can't remember, was some number between 1600 and 1900. These capacitances seem too low to me; I thought K Covi calculated the M1 as a 2uF capacitor.) Figure the impedance of a 1600pF capacitor at 20kHz and then divide that by the square of the turns ratio of either the SL toroid (75:1) or the AU90 (90:1). Gosh, I think the result would not correspond to reality, too high Z.
Post a Followup:
Post a Message!
This post is made possible by the generous support of people like you and our sponsors: