I'm now retired, but over the years I dealt with unemployment and it SUCKS. I hope the situation improves, RAPIDLY.
How about getting what's available out of the puny "iron", for now? When economic conditions improve, buy decent O/P transformers. Eventually, the small trafos go into guitar amp projects. 😉
SS rectify the B+. UF4007 diodes are what I advise, but (given the $ situation) 1N4007s, especially if already on hand, will make do. Use a modestly sized 1st filter cap., to avoid stressing the power trafo and access to much of the RMS current capability. Smaller 1st filter caps. induce larger conduction angles and less I2R heating in the rectifier winding. Employ sufficient capacitance to keep the rail voltage up and call it a day. At approx. $16, a Triad C-14X choke should be feasible. Plenty of capacitance, to suppress ripple and store energy follows the choke.
The 6DJ8 preamp board comes reasonably close, but no cigar, to what's needed in the small signal circuitry dept. The provided 6Y6 amp schematic shows the way. As this project is not as volts shy, a 6DJ8 section can be used for the "concertina" phase splitter.
Triode wire the O/P tubes and forego GNFB. More power and better linearity can wait for more favorable times.
How about getting what's available out of the puny "iron", for now? When economic conditions improve, buy decent O/P transformers. Eventually, the small trafos go into guitar amp projects. 😉
SS rectify the B+. UF4007 diodes are what I advise, but (given the $ situation) 1N4007s, especially if already on hand, will make do. Use a modestly sized 1st filter cap., to avoid stressing the power trafo and access to much of the RMS current capability. Smaller 1st filter caps. induce larger conduction angles and less I2R heating in the rectifier winding. Employ sufficient capacitance to keep the rail voltage up and call it a day. At approx. $16, a Triad C-14X choke should be feasible. Plenty of capacitance, to suppress ripple and store energy follows the choke.
The 6DJ8 preamp board comes reasonably close, but no cigar, to what's needed in the small signal circuitry dept. The provided 6Y6 amp schematic shows the way. As this project is not as volts shy, a 6DJ8 section can be used for the "concertina" phase splitter.
Triode wire the O/P tubes and forego GNFB. More power and better linearity can wait for more favorable times.
Attachments
Thanks, Eli. I'm near retirement age, but I can't afford to not work, not yet anyway! Would swapping over to better OTs at a later date require much/any modification?
No UL taps is not really an issue. Regulate g2 B+ at some fraction of anode B+ and tend to the NFB. I can see the 20 W./8K model in combination with "12" W. power types, like the 6AQ5 or 6Π15Π (6p15p). Most definitely, filter infrasonic noise out at the amp's I/Ps. The power trafo the OP salvaged just happens to have nice bias supply capability. 😀
So, the PT is up to the job and worth building around? It's a whopper, but then it did power a whole bunch of tubes.
We've been disparaging the small "nutone" O/P transformers. While taking my 1st serious look at the OEM schematic, I see something that is an indication of some quality in the "iron".
Mass is needed for bass extension. There's no getting around that. However, take a close look at how the secondary is constructed. A segment that feeds a (sic) GNFB loop is present. Some money was spent putting the feature in. So, the "iron" is not total garbage, after all. Corporate "bean counters" have been plaguing the consuming public for a long, long, time and they would have balked at a "frill".
For the initial (0 $ budget) build, a 120 Kohm grid to ground resistor and 0.018 μF. cap., at the very I/P, places the 3 dB. down point at approx. 68 Hz. Not great, but not ghastly too. A "standard" double bass' lowest note is 41 Hz. Less than an octave of that instrument's capability is being suppressed. Plenty of music should be enjoyable, until better economic times allow for the purchase of better O/P "iron".
Mass is needed for bass extension. There's no getting around that. However, take a close look at how the secondary is constructed. A segment that feeds a (sic) GNFB loop is present. Some money was spent putting the feature in. So, the "iron" is not total garbage, after all. Corporate "bean counters" have been plaguing the consuming public for a long, long, time and they would have balked at a "frill".
For the initial (0 $ budget) build, a 120 Kohm grid to ground resistor and 0.018 μF. cap., at the very I/P, places the 3 dB. down point at approx. 68 Hz. Not great, but not ghastly too. A "standard" double bass' lowest note is 41 Hz. Less than an octave of that instrument's capability is being suppressed. Plenty of music should be enjoyable, until better economic times allow for the purchase of better O/P "iron".
I agree with Eli's conclusion. Build something now.
His circuit in post # 21 will work.
Build now, listen now, enjoy now.
You can modify later.
I listen to a lot of music. Most of recordings do not go below about 80Hz.
Yes rock bass guitars do, and definitely pipe organs do.
But many composers do not use low frequency bass, they do not want the bass to be more than an octave below the other instruments and the other singers.
And most music does not include Thunder Drums, etc. either.
One thing about the MOSFET concertina.
1. The Zener is a real good idea
2. The Zener and Gate do not draw current in normal music operation (when not clipping).
If they are clipping, it sounds bad, so turn the volume control down slightly.
3. When there is no Zener current, and no Gate current, that means:
Source current = Drain current.
Source signal voltage = Drain signal voltage.
So there is no need for the 1.5 k potentiometer in the Source's series resistors to ground.
The only other cause of Gate / Grid current is a bad MOSFET, or a bad tube (filament to cathode leakage or a gassy tube).
The balance of a Concertina is intrinsic, as long as the Source / Cathode load resistance, and Drain / Plate load resistance are equal.
Do not forget, that also includes the need to have the following stages grid resistors equal.
Precision match loads and next stage grid resistors will Warrant balance (not Guarantee, there is a difference, outer space beings may put a large magnetic field over your house).
His circuit in post # 21 will work.
Build now, listen now, enjoy now.
You can modify later.
I listen to a lot of music. Most of recordings do not go below about 80Hz.
Yes rock bass guitars do, and definitely pipe organs do.
But many composers do not use low frequency bass, they do not want the bass to be more than an octave below the other instruments and the other singers.
And most music does not include Thunder Drums, etc. either.
One thing about the MOSFET concertina.
1. The Zener is a real good idea
2. The Zener and Gate do not draw current in normal music operation (when not clipping).
If they are clipping, it sounds bad, so turn the volume control down slightly.
3. When there is no Zener current, and no Gate current, that means:
Source current = Drain current.
Source signal voltage = Drain signal voltage.
So there is no need for the 1.5 k potentiometer in the Source's series resistors to ground.
The only other cause of Gate / Grid current is a bad MOSFET, or a bad tube (filament to cathode leakage or a gassy tube).
The balance of a Concertina is intrinsic, as long as the Source / Cathode load resistance, and Drain / Plate load resistance are equal.
Do not forget, that also includes the need to have the following stages grid resistors equal.
Precision match loads and next stage grid resistors will Warrant balance (not Guarantee, there is a difference, outer space beings may put a large magnetic field over your house).
Last edited:
Thanks guys, good to hear. So, what would be a better choice (when I have a job) for the OTs? Nothing too crazy please, maybe a Hammond? I've been looking at them but I can't decipher which ones would A. work, and B. work better.
Why would the solid state phase inverter be an advantage? I'd certainly be more comfortable using a tube given my experience.
Why would the solid state phase inverter be an advantage? I'd certainly be more comfortable using a tube given my experience.
Last edited:
Many are successfully using a MOSFET in tube amps, as a Source Follower, and as a Concertina Phase Splitter.
That might be worth investigating, proper application should work.
It has different problems, such as transient voltages at turn on and large signal operation.
It may have capacitance problems (capacitance varies versus the voltage drops across the part).
A tube concertina has voltage swing limits.
It can also have noise injected from the filament to cathode leakage.
That might be worth investigating, proper application should work.
It has different problems, such as transient voltages at turn on and large signal operation.
It may have capacitance problems (capacitance varies versus the voltage drops across the part).
A tube concertina has voltage swing limits.
It can also have noise injected from the filament to cathode leakage.
Because of the 6Y6 O/P tube in that amp, the B+ rail is comparatively short. The volts needed to operate the DC coupled configuration were in very short supply. The FET is (more or less) a heaterless pentode and, like a pentode, it can swing closer the the B+ rail than a triode can. Even with the FET splitter, things were quite "dicey". You should have the volts available to use 6DJ8 sections in both the voltage amplifier and "concertina" phase splitter roles. Do keep in mind that the hybrid setup would keep a good sounding OS tube on the "pile" for the day the 1 in use wears out.
IMO, it's very important to distinguish between current controlled BJTs and voltage controlled FETs. Getting tubes and FETs to "play nice" with each other is easy enough and frequently leads to a better end product than would be obtained with either device type by itself.
Edcor sell for less than Hammond and is, at a minimum, just as good. Edcor's CXPP25-7.6K works with "12" W. multi-grid power types and has ultra-linear taps. Sufficient magnetic headroom is present to allow, with confidence, the use of global NFB.
Select FETs whose reverse transfer capacitance (Crss) is low and reasonably constant. A low Crss is important for avoiding HF roll off.
A seminal document about incorporating FETs into tube circuits is MOSFET Follies. If you haven't already read it, do so.
A protective Zener diode between gate and source has proved essential for reliability. The Zener diode protects against those nasty turn on transients. It is noteworthy that some useful FET types have a built in gate to source protective Zener diode.
A couple more things:
A zener might have more capacitance than the MOSFET has.
Parts selection.
What ever a schematic shows for B+, when the amplifier first turns on, it may be much larger, the output tubes are not loading it down.
A zener might have more capacitance than the MOSFET has.
Parts selection.
What ever a schematic shows for B+, when the amplifier first turns on, it may be much larger, the output tubes are not loading it down.
You might check the usability of the STF13N80K5 MOSFET. It’s well protected, insolated (easy mounting) and with low CRSS. I’m testing them now in a concertina, as well as source follower. Measurements are quite good and distortion is quite low (< 0,2% THD) for max output.
In the source follower I’m trying them with a 10M90S CCS with 5mA (instead of a source resistor to ground). So far so good, but further testing needs to prove if this can be the final configuration.
Regards, Gerrit
In the source follower I’m trying them with a 10M90S CCS with 5mA (instead of a source resistor to ground). So far so good, but further testing needs to prove if this can be the final configuration.
Regards, Gerrit
Are the strings of resisters feeding the mosfet dropping resistors? Are there multiples to dissipate more heat? Also, what is the component marked "CCS 8mA" above the 6922 tube diagram? The symbol is two intersecting circles.
The resistance strings on both the source and drain electrodes of the FET are the "concertina" (split) loads. Both net value and heat dissipation capability cause the technique to be employed.
The interlocking circles schematic symbol indicates "constant" current. CCS = constant current source/sink. CCS circuitry can be simple or complex. A simple 10M45S is plenty good, in this situation. Constant current loading of active devices yields a highly desirable large AC impedance. A consequence of constant current is a horizontal load line. Triodes "love" horizontal load lines.
Last edited: