Hi Gentlemen..
Progressing... just got back my Aliexpress-chassis after "some" work on it.. it's not perfect, I'm not a big fan of typo-s but hey, it's made for myself so I try to digest this little glitch in the project.
Lesson learned: never ever try to modify on the layout, change text, then revert back to the previous text at around 2:00 AM after a long day.
Anyway, here it is at the moment:
Top plate has been screwed up by the company doing the CNC and engraving work, a new one will come soon from another team.
I might rotate the sockets a bit to be in-line with those ventillation gaps between the 6S33S tubes.
Such a shame.. one eye is happy, the other is crying of anger.. I can't believe I have to tell a CNC machinery handling person how to do such tasks... unbelievable.
(
That's it so far. Hoping the best from the new team. The company who wrecked my original top plate, made a 2nd one for me, this time a good one. Almost. Colour, surface is fully different, no sand-blasting and 3 screw holes are a full 2mm shifted from their supposed-to-be position. I'm pissed off. (Sorry).
Ah... I just wanted to give a quick status about my project.
At least I can begin cutting the stainless steel plate with the special plate cutting scissors, then sand the sharp edges, fold them and create some stable inner structure, the inner "box" for the input tubes. They'll be built into the chassis in a horizontal position right above the RCA/XLR sockets.
Btw this thing is heavy like a beast. I put it onto my computer desk and almost pinched a finger, huh. I think I'll continue building + soldering without transformers.
They will be the last thing I'll install inside.
Depending on arrival of the new top plate, I expect finishing with the project within 2-3 weeks from now.
Hope the best, fingers crossed.
Progressing... just got back my Aliexpress-chassis after "some" work on it.. it's not perfect, I'm not a big fan of typo-s but hey, it's made for myself so I try to digest this little glitch in the project.
Lesson learned: never ever try to modify on the layout, change text, then revert back to the previous text at around 2:00 AM after a long day.
Anyway, here it is at the moment:
Top plate has been screwed up by the company doing the CNC and engraving work, a new one will come soon from another team.
I might rotate the sockets a bit to be in-line with those ventillation gaps between the 6S33S tubes.
Such a shame.. one eye is happy, the other is crying of anger.. I can't believe I have to tell a CNC machinery handling person how to do such tasks... unbelievable.
( That's it so far. Hoping the best from the new team. The company who wrecked my original top plate, made a 2nd one for me, this time a good one. Almost. Colour, surface is fully different, no sand-blasting and 3 screw holes are a full 2mm shifted from their supposed-to-be position. I'm pissed off. (Sorry).
Ah... I just wanted to give a quick status about my project.
At least I can begin cutting the stainless steel plate with the special plate cutting scissors, then sand the sharp edges, fold them and create some stable inner structure, the inner "box" for the input tubes. They'll be built into the chassis in a horizontal position right above the RCA/XLR sockets.
Btw this thing is heavy like a beast. I put it onto my computer desk and almost pinched a finger, huh. I think I'll continue building + soldering without transformers.
They will be the last thing I'll install inside.
Depending on arrival of the new top plate, I expect finishing with the project within 2-3 weeks from now.
Hope the best, fingers crossed.
I have thought occasionally about adding a dc blocking capacitor in the various OTL amplifiers that I have built. I have nothing against capacitors "in the signal path"; after all, the power supply capacitors are in the signal path too. But one issue that arises in the case of a totem-pole type OTL is that there would seem to be no clear choice for the polarity needed for the blocking capacitor. Depending on the nature of the fault condition it is protecting against, it might involve the "live" speaker output terminal either going to a large positive potential relative to speaker ground, or else the live speaker terminal going to a large negative potential relative to speaker ground. Since the required capacitance will need to be of order a couple of thousand microfarads or so, it clearly has to be done with an electrolytic. Are you using a bipolar electrolytic? (Or, equivalently, a back-to-back series pair?)
polarity never bothered me, and sound quality is the same with or without the cap...
I can well believe that the sound quality would be the same. I have never subscribed to the belief that capacitors are villains.
The polarity of the capacitor would only matter if a catastrophic fault of the sort you are trying to guard against were actually to develop. I suppose if the fault were such as to drive the live speaker terminal to the negative HT supply voltage rather than the positive HT supply voltage, a single polarised capacitor might not be of much help. (Assuming your capacitor has its positive terminal connected to the amplifier.)
Having had a fault in my Tim Mellow OTL recently that did send the live speaker output high, I think I probably will install a blocking capacitor before re-commissioning the amplifier. But I'll use back-to-back series capacitors, to ensure protection for both polarities of failure mode. (In fact, when mine failed, a couple of fuses blew and the speaker survived.)
Interesting to read your conversation about protections. Actually, the author says the amp is current-limiting by nature (this was one of the design goals) and theoretically, no additional speaker protection is needed, the fuses are enough to protect the speakers from an amp failure. Anybody I asked so far here and there, it seemed to work or I just haven't met a single case yet where it was killing the speakers.
- - - - - - - - - - - - - - - - - - - -
Has anybody who built this amp substituted N1-N2 with diodes ?
1- according to this short explanation, I might be better off with Schottky-s instead of Zeners. Or am I wrong ? (For N1, N2).
2- according to this short explanation, isn't the Avalanche diode type the seemingly best suited here ? (Again, for N1, N2).
What do you think ?
- - - - - - - - - - - - - - - - - - - -
Has anybody who built this amp substituted N1-N2 with diodes ?
- N1 serves to limit the heater-cathode voltage on both halves of V1 to about 65V maximum during warmup. It is unlit during normal operation.
- I intend to use 2 zener diodes facing eachother, 62V/5W types. Next would be 68V so I rather go for the 62V spec'd.
- N2: If the output exceeds 90V, it strikes and then clamps the output to a low safe voltage until the output voltage has dropped below this
- I would use again 2 zener diodes facing eachother, 82V and 100V types available to me, I'd go for the 82V type.
- N3 - just an indicator lamp for the amp's ON state, I'll leave it out completely (using LED).
1- according to this short explanation, I might be better off with Schottky-s instead of Zeners. Or am I wrong ? (For N1, N2).
2- according to this short explanation, isn't the Avalanche diode type the seemingly best suited here ? (Again, for N1, N2).
What do you think ?
N1 replacement: you don't need 5W because current is limited by the 470K resistor to 1mA max:Has anybody who built this amp substituted N1-N2 with diodes ?
What do you think ?
and you don't need back-to-back diodes because voltage can only be negative, never positive during start-up;
I use simple small neon indicator (those without resistor of course) ...
N2 replacement: Tim used a gas discharge limiter for a purpose, when fired they drop less than 10V, in other words they crow-bar; I think the reason why he fitted them is that he did not want to publish something which could develop unsafe potentials on touchable parts like speaker connectors; high voltages develop not only in the event of a tube flash over but also when driven hard without speaker connected and terminals exposed; zeners don't crow-bar, they would just limit the voltage, in your case to 100V instead of 150V ... shottkys are just diodes with lower forward drop, not suitable; avalanche diodes are just like high-voltage zeners, they limit but don't crow-bar ... thats why gas discharge devices (or gas plasma arresters) are still being produced today, they don't have a SS equivalent; I don't use any ...
Last edited:
Hi Vortex! You can use these on output as well https://www.aliexpress.com/item/4000881714983.html?spm=a2g0o.9042311.0.0.27424c4dbOSPEa
Thanks for remembering. I use 3 of them, 2 for the 2 sides' (L/R) monoblocks for the amp circuit, 1 for the heater transformer (heating all 4 6S33S tubes).
Used in the primary of each transformer, triggered and kept "ON" by the microcontroller.
I think these SSR-s are oversized anyway, even with the 16A triac inside, or do I miss something ? At 235V in the primary I don't think I will ever have more than about 1.1A need, per relay.
Sorry, I have to correct myself here:
you do need back-to-back zeners for N1 because after start-up there will be a small positive voltage at the wiper of the pot where N1 connects and also some signal; I forgot that there is life after turn-on ... sorry again.
If you can easily buy from china, go for it;
where I live they changed the rules recently and now everything from china has to go through customs with all the hassle and cost ...
Littlefuse CG90L: a little more expensive but there is a european source from TME (Hungary !) : https://www.tme.eu/hu/katalog/?search=cg90L&s_field=1000011&s_order=desc
I bought transformers from TME susidiaries in Poland and Germany and they had very reasonable shipping cost.
Last edited:
Thank you all, I'll try the lamp and the GDT then, as in the original article.
Sorento, thanks for the very exact link, what a hit Nice. I'll go for this then, it makes sense.
(Do they get warm or absolutely not at all ?)
PS: China - same here. It's an EU regulation I think. Everything is inspected now and taxed, even the smallest things which previously slipped through customs unoticed or without remark.
Sorento, thanks for the very exact link, what a hit
Nice. I'll go for this then, it makes sense.(Do they get warm or absolutely not at all ?)
PS: China - same here. It's an EU regulation I think. Everything is inspected now and taxed, even the smallest things which previously slipped through customs unoticed or without remark.
Last edited:
In normal operation they just sit there doing nothing, they don't conduct at all, so why should they get hot ?
Unless there is a catastrophic tube failure of course, in which case a fuse will blow quickly.
Good morning !
Tim mentioned in the original article the quality of the filter capacitors (4+4 pcs in series for the +150V and -150V rails, per side). For stereo altogether 16pcs, this is a lot..
A 4-cap block in series (for 1x +/- 150V rails) gives an equivalent capacitor of 1700uF, used at 150V (so better spec'd higher than 200V when substituting with 1).
Due to space constraints, I'd like to substitute them with these two possible models:
Both fit in the chassis nicely, with respect to ventillation too. (4pcs needed for stereo then).
Not sure if Panasonic is that much better than the JCCON. The latter is made in China, and to be honest, everything's what's not made in China is a fake (we used to say).. so of course the Panasonic is also just a normal basic electrolytic I think, no big magic here to expect.
In our local markets (online) I only saw the cheapest bargain parts, which isn't a problem at all if they manage it via immense mass-production without quality suffering, but this is not always the case.
Do you think I can go the JCCON route, supported with +1 quality-made 1-10uF film capacitor on each electrolytic, or is there maybe a U.S. brand capacitor (or anything actually) in this same league I can use here for replacing the original 16 caps ? Anode voltage is separate, imagine my build as 2 monoblocks within 1 chassis. Heating is shared, nothing else.
Tim mentioned in the original article the quality of the filter capacitors (4+4 pcs in series for the +150V and -150V rails, per side). For stereo altogether 16pcs, this is a lot..
A 4-cap block in series (for 1x +/- 150V rails) gives an equivalent capacitor of 1700uF, used at 150V (so better spec'd higher than 200V when substituting with 1).
Due to space constraints, I'd like to substitute them with these two possible models:
- JCCON 2200uF/250V (105˚C long life) - Ø35mm,L60mm and I have stainless steel mounting rings for Ø35mm
- Panasonic 2700uF/250V (85˚C) - Ø40mm,L63mm
Both fit in the chassis nicely, with respect to ventillation too. (4pcs needed for stereo then).
Not sure if Panasonic is that much better than the JCCON. The latter is made in China, and to be honest, everything's what's not made in China is a fake
(we used to say).. so of course the Panasonic is also just a normal basic electrolytic I think, no big magic here to expect.In our local markets (online) I only saw the cheapest bargain parts, which isn't a problem at all if they manage it via immense mass-production without quality suffering, but this is not always the case.
Do you think I can go the JCCON route, supported with +1 quality-made 1-10uF film capacitor on each electrolytic, or is there maybe a U.S. brand capacitor (or anything actually) in this same league I can use here for replacing the original 16 caps ? Anode voltage is separate, imagine my build as 2 monoblocks within 1 chassis. Heating is shared, nothing else.
JCCON capacitor is an audio grade type, so low ESR/ESL is good to prevent undesirable effects, good temperature margin. But not in list of high tiers but good reason why you buy them.
https://www.diyaudio.com/community/...ly-of-tube-preamp-what-is-your-choice.330074/
I am contemplating to buy them as well (250v/1000uf) as new stock. I use 4pcs in each rail, a total of 4000uf. There is still a 40V (mine is +-170v rails) drop under full power output condition, also depends on how well the transformer regulation is.
https://www.diyaudio.com/community/...ly-of-tube-preamp-what-is-your-choice.330074/
I am contemplating to buy them as well (250v/1000uf) as new stock. I use 4pcs in each rail, a total of 4000uf. There is still a 40V (mine is +-170v rails) drop under full power output condition, also depends on how well the transformer regulation is.
having the filter caps in series also have their esr's in series, i suspect the parts that were available to Tim Mellows those times was the reason for it, but 250volt rated caps are available and easy to acquire. i used two 10kufd/100vdc caps in series because i happened to have them in quantity...i have a quad of 15kufd/200vdc caps on hand for future builds...