The load on the voltage amplifying tube preceding the 300B is provided by the primary inductance (along with the DC resistance) of the interstage transformer. No reflected impedance is necessary nor desirable.
John
John
Hi John,
I understand what you're saying, but I don't completely agree.... not to say you're wrong, but... some preliminary testing of a 300B design has shown better performance (lower distortion, flatter, and more linear response) by providing a load on the interstage transformer. I think this is one of those areas where theory goes so far and real-world characteristics are different. Not just my findings but some others who aren't active on any forums.
Regards, KM
I understand what you're saying, but I don't completely agree.... not to say you're wrong, but... some preliminary testing of a 300B design has shown better performance (lower distortion, flatter, and more linear response) by providing a load on the interstage transformer. I think this is one of those areas where theory goes so far and real-world characteristics are different. Not just my findings but some others who aren't active on any forums.
Regards, KM
It often depends on the quality of the transformer. Many inexpensive ITs don't have enough primary inductance and may require a loading resistor on the secondary to measure well. A high quality IT with a decent size nickel core with >50 Hy on the primary usually doesn't benefit sonically from a loading resistor.
John
John
True, the quality of the transformer (and other components) always matters. I'll try and get some of this sorted while I'm back in the US early October for a couple weeks. I'm using all Hashimoto iron and I've found it to be excellent on all counts to date. I'll post results as I go along.
Regards, KM
Regards, KM
I have looked at their i/s transformers in the past, tho I found them a little too limited for what I wanted to use them for at the time. Too little current-handling ability and/or inductance, i.e. if I want to use over 15mA, necessitating parallel connection of the primary, nominal inductance is down to 15H (I think!).
I have been looking at i/s transformers for a while... I think I will be going with custom Intact, O-Netics, ElectraPrint, possibly SILK... Tango NC-20 if I have a moment of weakne$$.
Rambling over - back to regular programming.
kmaier said:
I too am testing Hashimoto A105's (5K, 15mA, 1:1) at the moment. My main concern here is driving an IT that is designed for a driver with an rp of ~5K with a tube with a far lower rp, like the 417A. From what I've seen so far, you're going to end up with a rising response and serious peaking, albeit with the Hashimotos above the audio band (~40K, +6 dB). But this will manifest itself as ringing within the audio band. Yes, you can load the secondary, but how far do you go and how will it affect the sound. Best results so far, WITH THIS IT, were obtained with a 6GK5 (nice tube! rp ~5300 ohms) at 200V and 9mA, Vg -2V, gain > 70V/V. Still to try: 8532/6J4, 6S4A, ECC99, 5687 and maybe a 6C45P just for the data.
jazzbo said:
Very interesting! If I may ask, which tubes have you already tested
and found to induce a HF peak in the response?
Thanks!
Michael
Hashimoto recommend a few different driver tubes for the A-105. Those being, 5687, 6S4A and 6AH4GT along with a 6BQ5 (aka EL84) in triode mode. The winding ratio is a 1x2 so with a 5K primary the secondary should be 20K IIRC. Note this is not to say that you load it at 20K.
I also think it would be an interesting test to to run the IT solo, i.e., no 300B tube and just use a pure resistive load and see what the performance looks like.
Jazzbo, be very interested to get feedback on your findings as you continue to test some different drivers. Thanks.
Regards, KM
I also think it would be an interesting test to to run the IT solo, i.e., no 300B tube and just use a pure resistive load and see what the performance looks like.
Jazzbo, be very interested to get feedback on your findings as you continue to test some different drivers. Thanks.
Regards, KM
kmaier said:
I think you are tackling two different situations here. The flatter response seen from loading comes from the design of the interstage and IMO if the IT requires loading to perform acceptably you should just get a better IT.
The distortion is a puzzling one since I have never seen a case where loading a transformer decreased distortion in a single stage. However if the measurements were taken at the output of the amp, then the small increases in distortion of the driver stage caused by loading could offset distortions of the 300B netting overall lower distortion.
dave
Michael Koster said:
kmaier said:
I've looked at three candidates so far: 45, 46 (triode strapped) and 6GK5. The secondary is unloaded. The 45 and 46 were run at the 15mA max. for the IT and both were tried with and without a compensation network across the secondary of 100K in series with 150pF. Both exhibited significant peaking at about 43kHz (~6 dB). For the 6GK5 the secondary is unloaded. Don't have the 6GK5 data to hand but I recall the results were pretty flat and -1 dB at 40Hz which is pretty close to what the formula given in RDH4 predicts. I'll can't attach an Excel file of the plotted results for the 45 and 46 but can email them if you contact me. I can continue to report my results if anyone is interested. Perhaps I should start a new thread. Am also open to suggestions for setup.
Sorry guys, but have I missed something? I think the A-105, with the secondaries wired in parallel, has a winding ratio of 1:1; not sure how the primary arrangement (series or parallel) would affect the final ratio.
Also, if it is a 5K transformer, would it not be more suitable to load a tube with a plate resistance ~ 1/3 of that value (vague rule of thumb)?
Transformers, espeacially i/s transformers, are not really my bag - so please, if I am totally off-base please set me straight.
Cheers
Also, if it is a 5K transformer, would it not be more suitable to load a tube with a plate resistance ~ 1/3 of that value (vague rule of thumb)?
Transformers, espeacially i/s transformers, are not really my bag - so please, if I am totally off-base please set me straight.
Cheers
The A-105 with the secondaries in parallel gives you a ratio of 1:0.5, so steps down. The primary inductance is dependent on how much current you're running through it. The specs are 60H at 7 mA and 50H at 15 mA. Take the formula given in RDH4 to calculate the low end -3 dB point: 2*R = 2*pi*f(-3 dB)*L where
R = rp + Rprimary. So a tube with a 5K rp running at 7 mA will give you an f(-3 db) of 28 Hz, which is pretty much what I see with the 6GK5. If you want to extend the lower end, yes, you need to use a tube with a lower rp, but you will pay for it by raising the Q of the resonant circuit formed by the leakage inductance and shunt capacitance of the transformer. The rp of the tube effectively more or less damps this resonance. So testing a 46 (rp of ~2400 ohms) at 15 mA gave a bit more low end response at the price of a 5 - 6 dB response peak at 43 kHz. So it's a trial and error thing and the exact response depends on the particular IT. Anyway, it's all about how it sounds in the end.
R = rp + Rprimary. So a tube with a 5K rp running at 7 mA will give you an f(-3 db) of 28 Hz, which is pretty much what I see with the 6GK5. If you want to extend the lower end, yes, you need to use a tube with a lower rp, but you will pay for it by raising the Q of the resonant circuit formed by the leakage inductance and shunt capacitance of the transformer. The rp of the tube effectively more or less damps this resonance. So testing a 46 (rp of ~2400 ohms) at 15 mA gave a bit more low end response at the price of a 5 - 6 dB response peak at 43 kHz. So it's a trial and error thing and the exact response depends on the particular IT. Anyway, it's all about how it sounds in the end.
Thanks for that - it adds a few extra constraints I had not been considering. Wen this year is over I may have time to do some more reading.
Cheers
Cheers
- Status
- Not open for further replies.