The Triad N-77U is not available in my country: Due to government regulations, Mouser is unable to sell this product in your country. I get your point though.
@Eli: I appreciate your list of amps. Now I can search more focussed. However... Could you point me to the best info w.r.t. el Cheapo? I read you were involved. It is not that easy to find concise info about it on the internet.
@Eli: I appreciate your list of amps. Now I can search more focussed. However... Could you point me to the best info w.r.t. el Cheapo? I read you were involved. It is not that easy to find concise info about it on the internet.
That amazes me about hifi tube amps: the output transformers are mostly very close to the power transformer. I used the headphone trick on my guitar amp and I had to position the OT over 30cm (12") away from the PT to hear no hum. Why is that not the case for hifi?!
Generally, headphones on your ears are many times more sensitive, versus loudspeakers.
I often ask why HI Fi Stereo manufacturers do what they do.
Lack of knowledge.
Average customer base that is fairly easily satisfied.
Customer's speaker efficiency, distance to listener, room size and absorbent items, etc.
Price to a Point (save money, more profit)
Think of your own reasons that they act that way.
I try very hard to reduce the hum and noise on my amplifiers.
Almost all of them have less than 100uV at the output;
I use them with medium efficiency loudspeakers that are only 2 or 3 feet away from me.
But . . . I do not consider my amplifiers to be good enough for most headphones . . .
Unless the output is sent to a resistor load, plus a pair of resistors to attenuate what goes to the headphones. That way, you turn the volume up a little, and the signal level, versus the hum and noise level is improved (good enough for almost all systems).
In the 50s and early 60s, the accepted way to get a good Hi Fi Stereo, was to tie together a number of individual specialized products into what was called a 'component system'.
Many thought that all of the individual components were compatible with each other; not always true.
Now, in 2020, so many think they can tie a number of products together, and get what thay want.
Again, that is not always true.
A system is complex, involves signal voltage levels, magnetic pickup of nearby fields (like a magnetic cartridge and a turntable motor, or placing it on top of a power amplifier), gain, input and output impedances, speaker efficiency, room size, music selection, what volume is desired, etc.
One of the most important ways to get a system that satisfies, is to completely and accurately describe both the wants and needs of such a system.
Without that, it is like shooting an arrow, blindfolded, in the dark, at an unknown target (the target is 'safe' it hardly ever gets hit).
Many a dollar has been $pent on all kinds of gear, that is then replaced with other gear, and never satisfies.
That is because of two things:
The target was never hit.
Some are never satisfied.
Listen to the Music. Do not listen to the sound of the system.
Enjoy the music.
Then you might like the system.
I often ask why HI Fi Stereo manufacturers do what they do.
Lack of knowledge.
Average customer base that is fairly easily satisfied.
Customer's speaker efficiency, distance to listener, room size and absorbent items, etc.
Price to a Point (save money, more profit)
Think of your own reasons that they act that way.
I try very hard to reduce the hum and noise on my amplifiers.
Almost all of them have less than 100uV at the output;
I use them with medium efficiency loudspeakers that are only 2 or 3 feet away from me.
But . . . I do not consider my amplifiers to be good enough for most headphones . . .
Unless the output is sent to a resistor load, plus a pair of resistors to attenuate what goes to the headphones. That way, you turn the volume up a little, and the signal level, versus the hum and noise level is improved (good enough for almost all systems).
In the 50s and early 60s, the accepted way to get a good Hi Fi Stereo, was to tie together a number of individual specialized products into what was called a 'component system'.
Many thought that all of the individual components were compatible with each other; not always true.
Now, in 2020, so many think they can tie a number of products together, and get what thay want.
Again, that is not always true.
A system is complex, involves signal voltage levels, magnetic pickup of nearby fields (like a magnetic cartridge and a turntable motor, or placing it on top of a power amplifier), gain, input and output impedances, speaker efficiency, room size, music selection, what volume is desired, etc.
One of the most important ways to get a system that satisfies, is to completely and accurately describe both the wants and needs of such a system.
Without that, it is like shooting an arrow, blindfolded, in the dark, at an unknown target (the target is 'safe' it hardly ever gets hit).
Many a dollar has been $pent on all kinds of gear, that is then replaced with other gear, and never satisfies.
That is because of two things:
The target was never hit.
Some are never satisfied.
Listen to the Music. Do not listen to the sound of the system.
Enjoy the music.
Then you might like the system.
Last edited:
The method employed by Hegeman and others was used in "El Cheapo" (schematic provided). The LC reservoir section is critical to high performance. A refinement is inserting a LC "hash" filter made from a high current RF choke and a 1000 pF. mica or high quality ceramic (NP0/C0G) cap. between the doubler stack and the reservoir section.
Large valued caps. at the I/P of a PSU filter "crush" the ripple fundamental, but they generate ripple overtones up into the RF range in doing so. (TANSTAAFL strikes again) I've discussed the mechanism many times both here and AA. A "hash" filter keeps the trash out of the ultimate PSU product.
Large valued caps. at the I/P of a PSU filter "crush" the ripple fundamental, but they generate ripple overtones up into the RF range in doing so. (TANSTAAFL strikes again) I've discussed the mechanism many times both here and AA. A "hash" filter keeps the trash out of the ultimate PSU product.
Attachments
Eli,
Of course that schematic of El Cheapo is a good power supply . . .
Using 3 separate power transformers is sometimes considered a luxury.
More real estate, weight . . . and expensive.
1 B+, and 2 separate filament transformers.
That helps to isolate the noise from one power supply getting into another supply
(has to cross from the 1st secondary, to the 1st primary;
and then onto a 2nd primary, back to a 2nd secondary.
Of course that schematic of El Cheapo is a good power supply . . .
Using 3 separate power transformers is sometimes considered a luxury.
More real estate, weight . . . and expensive.
1 B+, and 2 separate filament transformers.
That helps to isolate the noise from one power supply getting into another supply
(has to cross from the 1st secondary, to the 1st primary;
and then onto a 2nd primary, back to a 2nd secondary.
Last edited:
Believe it or not, 3 transformers were (sic) the inexpensive route. 
Triad's N-77U and N-68X are bargains for getting 300+ V. of B+.
Things change over time. Anybody in North America considering an "El Cheapo" build can get a complete set of power "iron" from Allied Electronics.
Stock # 70218190 = main B+ $32.59
Stock # 70218344 = 2 "12" VAC windings for B+ boost and heater power. $16.73
Stock # 70009000 = B- feed and 12AT7 heater power. $16.83
Stock # 70218145 = B+ filter choke $8.62
Approx. $75 for a reasonably sophisticated set of power "iron" is (IMO) a pretty good deal. FWIW, Triode Electronics charges approx. $100 for a ST-35 power trafo and that chunk of "iron" will not completely satisfy the needs of an "El Cheapo".
Something additional is provided to show the versatility of the Triad isolation transformers.
Here's a vacuum rectified "300" V./60 mA. B+ supply, for preamp projects.
Triad's N-77U and N-68X are bargains for getting 300+ V. of B+.Things change over time. Anybody in North America considering an "El Cheapo" build can get a complete set of power "iron" from Allied Electronics.
Stock # 70218190 = main B+ $32.59
Stock # 70218344 = 2 "12" VAC windings for B+ boost and heater power. $16.73
Stock # 70009000 = B- feed and 12AT7 heater power. $16.83
Stock # 70218145 = B+ filter choke $8.62
Approx. $75 for a reasonably sophisticated set of power "iron" is (IMO) a pretty good deal. FWIW, Triode Electronics charges approx. $100 for a ST-35 power trafo and that chunk of "iron" will not completely satisfy the needs of an "El Cheapo".
Something additional is provided to show the versatility of the Triad isolation transformers.
Here's a vacuum rectified "300" V./60 mA. B+ supply, for preamp projects.Attachments
So the only expen$ive thing about the EL Cheapo having 3 transformers . . .
is the extra large chassis to be able to fit those on.
1 transformer = more hum and noise cross coupling; but a smaller chassis.
3 transformers, and 2 output transformers that need to be spaced away from them = a larger chassis.
Tradeoffs.
is the extra large chassis to be able to fit those on.
1 transformer = more hum and noise cross coupling; but a smaller chassis.
3 transformers, and 2 output transformers that need to be spaced away from them = a larger chassis.
Tradeoffs.
The N-77U is pretty good sized, but the B- trafo and twin "12" VAC items are not especially large.
Consider mounting some power "iron" item on the side wall.
Consider mounting some power "iron" item on the side wall.
Artosolo lade this suggestion for a power transformer on another thread ...
tsl180_001/transformers-with-fastening/indel/tsl-180-001/
For €57 it has lots of versatility.
tsl180_001/transformers-with-fastening/indel/tsl-180-001/
For €57 it has lots of versatility.
Keep 'm coming guys. I'm busy researching El Cheapo, Baby Huey and the ST35. All have different pros and cons.
El Cheapo:
Active helpful forum member involved in design
Multiple PT's. Suggested ones are not always available. Don't like the naked look of 'm.
I (think I) don 't need the amp to be extremely cheap. I think I prefer a more straightforward amp, that costs a bit more.
Baby Huey:
Nice looking and pleasingly symmetrical layout. Does nothing to the sound, but somehow I like it.
Quite some info present on the internet.
ST35:
Kit available, pcb's available. I thought I would do it all myself, but currently I'm a bit overwhelmed and thinking about playing it safe.
Lots to learn and decide...
El Cheapo:
Active helpful forum member involved in design
Multiple PT's. Suggested ones are not always available. Don't like the naked look of 'm.
I (think I) don 't need the amp to be extremely cheap. I think I prefer a more straightforward amp, that costs a bit more.
Baby Huey:
Nice looking and pleasingly symmetrical layout. Does nothing to the sound, but somehow I like it.
Quite some info present on the internet.
ST35:
Kit available, pcb's available. I thought I would do it all myself, but currently I'm a bit overwhelmed and thinking about playing it safe.
Lots to learn and decide...
Last edited:
Start with a proven circuit. I'd recommend a variation of
the Williamson circuit, or the Loftin-White SE 2A3 circuit. You can find schematics for the Williamson with 12AU7 or 6SN7 driver/phase inverter tubes, and/or 6L6, 807, or KT66 power tubes. One of my first amps was the Healthkit W5-M mono blocks. Some good info here:
The Williamson Amplifier
The Loftin-White SE 2A3 is a simple, straight forward design. If 4 watts will driver your speakers, this would be a great way to start. I've ran this circuit for 20 years and haven't found a reason to move away from it.
loftin-white - American Radio Historyworldradiohistory.com › Radio-News-1930-03
the Williamson circuit, or the Loftin-White SE 2A3 circuit. You can find schematics for the Williamson with 12AU7 or 6SN7 driver/phase inverter tubes, and/or 6L6, 807, or KT66 power tubes. One of my first amps was the Healthkit W5-M mono blocks. Some good info here:
The Williamson Amplifier
The Loftin-White SE 2A3 is a simple, straight forward design. If 4 watts will driver your speakers, this would be a great way to start. I've ran this circuit for 20 years and haven't found a reason to move away from it.
loftin-white - American Radio Historyworldradiohistory.com › Radio-News-1930-03
I've been running the 2A3s on a pair of Altec VOT w/ sensitivity ~ 103db. It pushes those very nicely. These amps can be made inexpensively since the part count is so low. I wouldn't recommend this amp unless you have efficient speakers.
A SE KT66, 6550, or EL34 amp will give you at least 2X power of the 2A3. I believe you can use the same SE transformer for the 2A3, 6550, and KT66 - at least in some circuits.
Building a Williamson amp can also be inexpensive, depending on which version of the circuit you build. My preference has been one that uses 6SN7 up front and 807 in the back. Good 807s can be hard to find. The military version of the 807 is 5933. Very solid tube. The heathkit version is nice because it uses 12AU7s up front and you can uses 6L6 instead of KT66.
I believe you can also use the same PP transformer for 6L6, KT66, 6550, and 2A3. You can build different circuits and use the same transformer.
You can keep your costs down on your build if you opt for a stereo rather than mono blocks. Keep parts costs down by using ok components for everything except signal path. Get the best parts you can for signal path within your budget. The most important part is going to be the iron, the output trans. Starting in-expensively will get your foot in the door - so to say - and you can upgrade parts as you go.
The "Achilles Heel" of Williamson style topology is low frequency instability. Costly O/P "iron" relatively free of phase shifts is part of the Williamson "package".
Mullard style topology is more tolerant of O/P "iron" aberrations and (IMO) is a much better choice, for a 1st time builder. Mullard style is "idiot resistant".
The OP's budgetary limits (unfortunately) tend to exclude tubes more capable than the "12" W. multi-grid types. In the interests of completeness, I'm uploading examples of the Williamson and Mullard styles.
Mullard style topology is more tolerant of O/P "iron" aberrations and (IMO) is a much better choice, for a 1st time builder. Mullard style is "idiot resistant".
The OP's budgetary limits (unfortunately) tend to exclude tubes more capable than the "12" W. multi-grid types. In the interests of completeness, I'm uploading examples of the Williamson and Mullard styles.
Attachments
I've got a few questions:
El Cheapo
- Why are different resistor types mentioned (CC, CF, non-inductive, Dale)? What I've taken from my guitar amp journey is that I would use metal film for most locations for low noise and metal oxide in/around the power supply for safety.
- Somewhere 'thermal isolation resistor'(10KOhm carbon film) is mentioned. What does that mean. A google search returns 3 (!) hits, but no explanation.
- How is the LTP biased? I would expect the tail resistor to be replaced by a CCS, but it looks slightly different.
All:
- Why are guitar amps always in a closed metal chassis (Faraday cage) and most hifi amps use a wooden box. I would expect higher demands from the hifi community. Or is it the larger amount of gain in a guitar amp that needs that?
- I thought 6.3V heater supply would be quietest (Blencowe). Do you agree?
- I used a humdinger in my guitar amps and it seems to work quite well. What is your opinion on them in hifi amps?
- There are only a few power supply filter stages (less than what I'm used to). Capacitors are mostly bigger, that helps. But I don 't get the 100R + 50uF filter stage in the left and right supply (El Cheapo). I would think that would result in a low pass filter with the -3dB point at 32Hz. Am I missing something? Is it the CMRR by virtue of the CCS's?
- The coupling capacitors are quite large resulting in a -3dB point at 1.5-3Hz. Are they that low to limit phase shifts?
That's it for now.
El Cheapo
- Why are different resistor types mentioned (CC, CF, non-inductive, Dale)? What I've taken from my guitar amp journey is that I would use metal film for most locations for low noise and metal oxide in/around the power supply for safety.
- Somewhere 'thermal isolation resistor'(10KOhm carbon film) is mentioned. What does that mean. A google search returns 3 (!) hits, but no explanation.
- How is the LTP biased? I would expect the tail resistor to be replaced by a CCS, but it looks slightly different.
All:
- Why are guitar amps always in a closed metal chassis (Faraday cage) and most hifi amps use a wooden box. I would expect higher demands from the hifi community. Or is it the larger amount of gain in a guitar amp that needs that?
- I thought 6.3V heater supply would be quietest (Blencowe). Do you agree?
- I used a humdinger in my guitar amps and it seems to work quite well. What is your opinion on them in hifi amps?
- There are only a few power supply filter stages (less than what I'm used to). Capacitors are mostly bigger, that helps. But I don 't get the 100R + 50uF filter stage in the left and right supply (El Cheapo). I would think that would result in a low pass filter with the -3dB point at 32Hz. Am I missing something? Is it the CMRR by virtue of the CCS's?
- The coupling capacitors are quite large resulting in a -3dB point at 1.5-3Hz. Are they that low to limit phase shifts?
That's it for now.
CC resistors are both non-inductive and non-metallic. That's great for grid stopper duty.
Vishay/Dale 1% tolerance RNnn milspec metal film resistors are quiet and non-magnetic. They are excellent where low noise is important and the tight tolerance is (obviously) good.
During the development of "El Cheapo" DIY Audio "guru" SY pointed out that the NFB was working into a high impedance non-inverting control grid. An inductive NFB resistor could cause trouble. Both MF and CF resistors are spiral trimmed and that introduces inductance. The low cost answer was employing meter matched (between channels) CC resistors. Use Caddock parts, if your bank balance can tolerate the "hit".
The 10 Kohm CF resistor between the LTP's tail and the CCS drops some volts and also blocks the conduction of heat from the tube (via the wiring) to the CCS. Unwanted heat is the mortal enemy of all things electronic, especially SS. I took no prisoners. BTW, that resistor also slightly increases the net load the LTP tail works into.
IMO/IME, amps (including "El Cheapo") which employ GNFB loops that the encompass the O/P "iron" should filter infrasonic noise out, at the unit's I/P. Suppressing said noise is prophylaxis against core saturation. When the "iron" is of decent quality, a 1 pole RC filter whose F3 is in the 16-18 Hz. range is adequate. The "El Cheapo" schematic shows a no longer recommended guitar amp O/P trafo that can't provide full bass extension at full power. F3 for the 0.056 μF./100 Kohm combination shown is 28.4 Hz. For any rational signal topology, better O/P "iron" = better amp.
Please notice the the "El Cheapo" schematic shows triode wired O/P tubes. Approx. 6 WPC are available with low cost guitar O/P "iron". The "Gold Standard" O/P trafo for "El Cheapo" builders is the "Dynaclone" Z565, which allows for both UL power and full bass extension. The Z565 is the price/performance champion "iron" for building PP amps that employ "12" W. multi-grid O/P tube types. TANSTAAFL!
Many DIY builders find a metal plate (usually aluminum) and a wood frame convenient. A commercial kitchen baking pan can be convenient too. Adhesive metal shielding foil is available, should wooden construction prove troublesome.
Decent B+ rail filtering and stage to stage decoupling has proven quite sufficient.
Good sized interstage coupling caps. avoid deleterious phase shifts, which can be particularly troublesome in units containing feedback loops. Heaven forfend that most definitely unwanted fulfillment of the Barkhausen Criterion for oscillation occur. I adhere to the 2 octaves away from the passband notion. Some folks follow a slightly more rigorous rule of a decade away.
Filament voltage definitely is an issue, when AC heating is employed with directly heated types. An AC heated SE 2A3 will usually be fine. OTOH, time and again, AC heating of SE 300Bs has been problematic.
Vishay/Dale 1% tolerance RNnn milspec metal film resistors are quiet and non-magnetic. They are excellent where low noise is important and the tight tolerance is (obviously) good.
During the development of "El Cheapo" DIY Audio "guru" SY pointed out that the NFB was working into a high impedance non-inverting control grid. An inductive NFB resistor could cause trouble. Both MF and CF resistors are spiral trimmed and that introduces inductance. The low cost answer was employing meter matched (between channels) CC resistors. Use Caddock parts, if your bank balance can tolerate the "hit".
The 10 Kohm CF resistor between the LTP's tail and the CCS drops some volts and also blocks the conduction of heat from the tube (via the wiring) to the CCS. Unwanted heat is the mortal enemy of all things electronic, especially SS. I took no prisoners.
BTW, that resistor also slightly increases the net load the LTP tail works into.IMO/IME, amps (including "El Cheapo") which employ GNFB loops that the encompass the O/P "iron" should filter infrasonic noise out, at the unit's I/P. Suppressing said noise is prophylaxis against core saturation. When the "iron" is of decent quality, a 1 pole RC filter whose F3 is in the 16-18 Hz. range is adequate. The "El Cheapo" schematic shows a no longer recommended guitar amp O/P trafo that can't provide full bass extension at full power. F3 for the 0.056 μF./100 Kohm combination shown is 28.4 Hz. For any rational signal topology, better O/P "iron" = better amp.
Please notice the the "El Cheapo" schematic shows triode wired O/P tubes. Approx. 6 WPC are available with low cost guitar O/P "iron". The "Gold Standard" O/P trafo for "El Cheapo" builders is the "Dynaclone" Z565, which allows for both UL power and full bass extension. The Z565 is the price/performance champion "iron" for building PP amps that employ "12" W. multi-grid O/P tube types. TANSTAAFL!
Many DIY builders find a metal plate (usually aluminum) and a wood frame convenient. A commercial kitchen baking pan can be convenient too. Adhesive metal shielding foil is available, should wooden construction prove troublesome.
Decent B+ rail filtering and stage to stage decoupling has proven quite sufficient.
Good sized interstage coupling caps. avoid deleterious phase shifts, which can be particularly troublesome in units containing feedback loops. Heaven forfend that most definitely unwanted fulfillment of the Barkhausen Criterion for oscillation occur. I adhere to the 2 octaves away from the passband notion. Some folks follow a slightly more rigorous rule of a decade away.
Filament voltage definitely is an issue, when AC heating is employed with directly heated types. An AC heated SE 2A3 will usually be fine. OTOH, time and again, AC heating of SE 300Bs has been problematic.
You state that CC's are great for grid stoppers as they are non-inductive and non-metallic. I don't argue that. By why does that make them perfect for grid stoppers?
Blencowe states that inductances only become important at radio frequencies. He also states the inductances of metal film resistors and leads, where leads are easily more inductive because of their length. I've looked at resistor specs and found a formula for the inductance of a wire and I come to the same conclusion. Furthermore, as I understand it, the grid stopper forms a low pass filter with the Miller capacitance. The resistance of the inductive resistor rises with frequency thus making the low pass filter even more effective at high frequencies.
Why is it benificial for a grid stopper to be non-metallic?
I look forward to learn a few new English words like prophylaxis...
Blencowe states that inductances only become important at radio frequencies. He also states the inductances of metal film resistors and leads, where leads are easily more inductive because of their length. I've looked at resistor specs and found a formula for the inductance of a wire and I come to the same conclusion. Furthermore, as I understand it, the grid stopper forms a low pass filter with the Miller capacitance. The resistance of the inductive resistor rises with frequency thus making the low pass filter even more effective at high frequencies.
Why is it benificial for a grid stopper to be non-metallic?
I look forward to learn a few new English words like prophylaxis...
You mount the body of a grid stopper as close to the tube socket lug, as circumstances permit. Definitely keep that lead short! Metal mass is an "enemy", for more than 1 reason.
Grid stoppers are used to suppress parasitic oscillation. Because carbon composition (CC) is non-metallic, it will not act as an antenna for RF. Because CC is non-inductive, adverse resonant interactions with all sorts of capacitances do not occur. The downside to CC resistors is noise and drift. Fortunately, exact value is not critical in a stopper and CC noise is current related. The current present in a control grid circuit is vanishingly small.
Various forms of resistor construction excel at differing circuitry roles. Selection is a matter of "horses for courses". The notion that "1 size" can fit all is highly defective in numerous (perhaps all) situations.
Grid stoppers are used to suppress parasitic oscillation. Because carbon composition (CC) is non-metallic, it will not act as an antenna for RF. Because CC is non-inductive, adverse resonant interactions with all sorts of capacitances do not occur. The downside to CC resistors is noise and drift. Fortunately, exact value is not critical in a stopper and CC noise is current related. The current present in a control grid circuit is vanishingly small.
Various forms of resistor construction excel at differing circuitry roles. Selection is a matter of "horses for courses". The notion that "1 size" can fit all is highly defective in numerous (perhaps all) situations.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.