|
Vintage Audio (record players, hi-fi etc) Amplifiers, speakers, gramophones and other audio equipment. |
|
Thread Tools |
28th Jul 2021, 9:37 am | #21 | |
Moderator
Join Date: Mar 2012
Location: Fife, Scotland, UK.
Posts: 22,800
|
Re: Valve amplifier design 1965
Quote:
What's more transistors have some added problems with miller-related effects. Ccb is a diode junction and the capacitance is that of a varactor diode, so the HF roll-off pumps with strong signals... hello intermods! then there's the Early effect. David
__________________
Can't afford the volcanic island yet, but the plans for my monorail and the goons' uniforms are done |
|
28th Jul 2021, 10:16 am | #22 | |
Dekatron
Join Date: Sep 2007
Location: Oxfordshire, UK.
Posts: 4,310
|
Re: Valve amplifier design 1965
Quote:
Then again, perhaps the trade-offs that have had to be made to get decent bass capability, which we wouldn't need, make these transformers sub-optimal for us. I note that the NTM1 has a secondary winding resistance of more than 100ohms for example. Cheers, GJ
__________________
http://www.ampregen.com |
|
28th Jul 2021, 11:25 am | #23 |
Moderator
Join Date: Mar 2012
Location: Fife, Scotland, UK.
Posts: 22,800
|
Re: Valve amplifier design 1965
Most dynamic microphone transformers are voltage step-up jobs to suit more typical input impedances of amplifying devices.
The lack of need for bass response is a great freedom in the design of this transformer. Remember that you want its behaviour to be good far above the audio range, to and beyond the unity loop gain point which has to be as far above the audio range as the amount of feedback you want to be still there at the top of the audio range. So the feedback balun has to go a lot higher in frequency than the output transformer. But it doesn't have to go so low. So the number of octaves it needs to work well over might not be quite so arduous. BUT the reduced impedance is a gift made in heaven. Transformers at low-Z are an awful lot easier to design. A long time ago, I had to develop a transformer-based impedance measuring bridge. The wanted frequency range was 60kHz to 18.6MHz. It had to be to full accuracy spec over this range, so the transformers had to be good well over an octave beyond each end. There was one design freedom, it was used between 75 Ohm source and detector, but the detector was ludicrously sensitive, so I could afford a lot of insertion loss. So I used it. A resistive pad reduced that 75 Ohm Zo as far as I needed, and in the new low Z regime, I didn't need quite so much primary inductance. Transmission line techniques handled the top end (Ruthroff: "Some transmission line transformers") David
__________________
Can't afford the volcanic island yet, but the plans for my monorail and the goons' uniforms are done |
28th Jul 2021, 2:58 pm | #24 |
Nonode
Join Date: Jun 2016
Location: Bognor Regis, West Sussex, UK.
Posts: 2,288
|
Re: Valve amplifier design 1965
This thread is fascinating and I am sure it has a long way to go yet.
Mods, perhaps it should be moved from "wanted" to vintage audio. Peter |
28th Jul 2021, 4:07 pm | #25 |
Dekatron
Join Date: Feb 2007
Location: Lynton, N. Devon, UK.
Posts: 7,060
|
Re: Valve amplifier design 1965
Well, the feedback from the output transformer primary starts rolling off at 6.77kHz (the series 47kΩ and 500pF).
As David points out, the little balun transformer is loaded by 68Ω - the dominant load. So allowing for the transformer itself introducing negligible LF phase-shift, we'd like the reactance of the half-winding to be at least 680Ω at this 6.77kHz. That then puts the inductance at 15.9mH. I tend to think that saturation will not be an issue - the voltage across a half-winding will be well under a volt (47kΩ and 68Ω forming a 690:1 voltage divider from the voltage on the output transformer primary); the frequency is high, so flux swing will be small. There's the DC cathode current flowing through it which will tend to polarise the transformer core, but a possible refinement might be to put a 1μF capacitor in the line from the transformer to the junction of the 68Ω and 680Ω resistors, which will eliminate that. 16mH is easily achievable, with a high-permeability ferrite core. Ed's suggestion of nickel-iron would probably be better still - inductance achieved with fewer turns, which would mean less problems with parasitic capacitance and leakage inductance at the HF end. |
28th Jul 2021, 4:25 pm | #26 | |
Moderator
Join Date: Mar 2012
Location: Fife, Scotland, UK.
Posts: 22,800
|
Re: Valve amplifier design 1965
Quote:
Presto! Changeo! Done! David
__________________
Can't afford the volcanic island yet, but the plans for my monorail and the goons' uniforms are done |
|
28th Jul 2021, 5:28 pm | #27 | |
Nonode
Join Date: Jun 2016
Location: Bognor Regis, West Sussex, UK.
Posts: 2,288
|
Re: Valve amplifier design 1965
Quote:
The Bailey EE design crosses at 6.4kHz and -3.26dB The TT100 crosses at 1.446kHz and -2.7dB Peter |
|
30th Jul 2021, 10:55 am | #28 | |
Nonode
Join Date: Jun 2016
Location: Bognor Regis, West Sussex, UK.
Posts: 2,288
|
Re: Valve amplifier design 1965
With reference to the inverting transformer in the feedback, Kalee has observed
Quote:
I have been comparing the Bailey / Radford circuits that used the ECC88 Cascode front end, the Bailey design we are discussing here, the STA100 and the STA25 The Anode load resistors are different in each case and the Anode currents are consequently different. Circuit Anode resistor Anode current (approx) STA25 22k 3.5mA STA100 47k 2.5mA Bailey 100k 1.5mA Perhaps Bailey chose a lower value of Anode current (and hence Cathode current) to minimise this polarisation. Peter |
|
30th Jul 2021, 11:22 am | #29 |
Moderator
Join Date: Mar 2012
Location: Fife, Scotland, UK.
Posts: 22,800
|
Re: Valve amplifier design 1965
The need for only low winding inductance plays nicely into tolerating some direct current.
It needs watching and checking, but shouldn't be too difficult. I'm wondering about mumetal dust or sen dust toroid cores? David
__________________
Can't afford the volcanic island yet, but the plans for my monorail and the goons' uniforms are done |
30th Jul 2021, 12:05 pm | #30 |
Nonode
Join Date: Jun 2016
Location: Bognor Regis, West Sussex, UK.
Posts: 2,288
|
Re: Valve amplifier design 1965
Bobdger and I are experimenting with ferrite transformers, I can try checking the bandwidth at 1V ac both with some dc and without.
Peter |
30th Jul 2021, 1:33 pm | #31 | |
Dekatron
Join Date: Feb 2007
Location: Lynton, N. Devon, UK.
Posts: 7,060
|
Re: Valve amplifier design 1965
Quote:
A toroid, of mumetal or amorphous alloy, permeability approaching 100,000 would allow the inductance to be achieved with the fewest turns - and few turns gives the lowest self-capacitance (remembering that this is the transformer which handles the HF stuff). Polarising the core then becomes more of an issue. It could well be that the numbers indicate it's a non-problem, and even a larger, lower-permeability core which is saturation-proof with 1.5mA, still has capacitances in the low tens of pF. All the same, if it's a choice of DC or no DC, I know which I'd prefer! |
|
31st Jul 2021, 6:55 pm | #32 |
Moderator
Join Date: Feb 2006
Location: Wembley, Middlesex
Posts: 7,219
|
Re: Valve amplifier design 1965
I have never heard of a KT90, Are these new developments or do they go back to the early days, ie late 40’s and 50’s?
|