NFB is not a option . Maybe Hashimoto A-107 or apply zobel correction HF in Lundahl but I am not sure if affect to the timbre sound
Last edited:
snubbers tuned to that funny business can damp it......
fwiw, i have seen more tweeters fried using solid state amps than with tube amps...
you may be overthinking the problem...
fwiw, i have seen more tweeters fried using solid state amps than with tube amps...
you may be overthinking the problem...
yesterday I change the secondary load to 10k with good result:
40 khz +2.5db
30 khz + 1.9db
20 khz +0.8 db
rest 0 db
no more overthinking thanks
40 khz +2.5db
30 khz + 1.9db
20 khz +0.8 db
rest 0 db
no more overthinking thanks
Building a good interstage for SE to PP is very difficult. I used the 1660S to do that.
I believe that transformer is built and connected this way to do SE to PP: 2 sets of windings, one set over each core air gap (call them Left and Right gaps). Primary winding on Left in series with Primary winding on Right. Primary windings are coupled by the (series) current through them, and more loosely coupled by magnetics (so some leakage reactance). One primary is wound over Left secondary; one primary is wound over Right secondary. In relative terms, Secondary winding in Left is 'Push'; Secondary winding on Right is 'Pull'.
I also tried the 1660S with the 2 Primaries in Parallel, but the inductance is only 10 Henry. Harder to drive, but if I remember, that was a little better high frequency response.
I believe that transformer is built and connected this way to do SE to PP: 2 sets of windings, one set over each core air gap (call them Left and Right gaps). Primary winding on Left in series with Primary winding on Right. Primary windings are coupled by the (series) current through them, and more loosely coupled by magnetics (so some leakage reactance). One primary is wound over Left secondary; one primary is wound over Right secondary. In relative terms, Secondary winding in Left is 'Push'; Secondary winding on Right is 'Pull'.
I also tried the 1660S with the 2 Primaries in Parallel, but the inductance is only 10 Henry. Harder to drive, but if I remember, that was a little better high frequency response.
Last edited:
the LL1660S configuration ALT - B SE to PP with secondary load 8k to 15k has a pretty square wave with out HF peak. With load over 20k HF peak is too big.
Now need to see what happen with fixed bias to 300b the loop grid to cathode to bias are <15k to secondary
Now need to see what happen with fixed bias to 300b the loop grid to cathode to bias are <15k to secondary
Celsius,
Is that 8k to 15k Grid to Grid, or is it 8k to 15k from Grid to "center tap", and 8k to 15k from the other Grid to "center tap".
I say "center tap", because the windings are only connected by the user.
Is that 8k to 15k Grid to Grid, or is it 8k to 15k from Grid to "center tap", and 8k to 15k from the other Grid to "center tap".
I say "center tap", because the windings are only connected by the user.
I test with 10k secondary each coil, ALT B LL1660s, 5 to 10 10k and 9 to 4 10k, I need to check what happen when connect to grid
In Alt B, 5 and 10 are connected to each other and to ground, putting 10k across that does nothing;
In Alt B, 9 and 4 are push and pull grids, so you connected 10k across that?
In Alt B, 9 and 4 are push and pull grids, so you connected 10k across that?
If I am correct, a 10k loads 0n each '2' winding (5k) will reflect a 4.714k load on two '2.25' windings that are paralleled. Root (2/2.25) x 5k = 4.714k. 4.714k is what the plate of the driving triode sees at midband.
With 42H, the load is -3dB@17.9Hz
With 42H, the load is -3dB@17.9Hz
Last edited:
Next test will be on the amp to know exactly what roll of have. before only check simulating, the result will depend on the grid cathode impedance of the 300b, I am not sure if need extra load resistor.
But my 2A3 can do this- Waaah! Not what I expect from an expensive Lundahl! Nothing I can do to fix?
you are chasing flaws that you cant hear ....
oh btw... if u are obsessed with low thd, flawless sound - then go solid state... 😱 dont waste time with tubes....
I had problems before with ultrasound and a tweeter broken, you do not listen but tweeter can. Flawless sound, well I think that the tubes amp give the better sound possible far of SS, only that me, I am very perfectionist, if I use a 2a3 as driver, LL1660S splitter to 300B PP I am not looking the easy way, I am looking the best way or I try, if I use Duelund Cast CAP or amorphous core OPT I am not lazy to work in the best 🙂
note: Per Lundahl for this model does not recommends any Zobel compensation to show a beautiful square wave, he affirm that sound is is much more better without. " the sound will become uninteresting and dull."
note: Per Lundahl for this model does not recommends any Zobel compensation to show a beautiful square wave, he affirm that sound is is much more better without. " the sound will become uninteresting and dull."
No!
It should be root(2.25/2)*5 - a little more than 5K. As it it is a step down, it will be step up when you go the other way.
Thorsten 😉
I tried that device 15 years ago and saw similar waves as well as phase goofyness around 20khz.
I don't have any scientific data but think C Core is a poor choice for SE to PP. Having two sets of windings 180 degrees and physically separated didn't work out for this device.
I had my own ideas for a device with split bobben on EI where the primary was shared across the bobbens and secondaries had their own side of the bobben but wound in reverse. No one would build it so I'll never know.
Anyway, a centertapped choke praafeed worked well to about 30khz.
I don't have any scientific data but think C Core is a poor choice for SE to PP. Having two sets of windings 180 degrees and physically separated didn't work out for this device.
I had my own ideas for a device with split bobben on EI where the primary was shared across the bobbens and secondaries had their own side of the bobben but wound in reverse. No one would build it so I'll never know.
Anyway, a centertapped choke praafeed worked well to about 30khz.