logo elektroda
logo elektroda
X
logo elektroda

Creating a Tube Amplifier with 2xPCL86, EM84 Indicators and Transformers TG2.5-1-666 & TS40/29/670

25911 35
Best answers

How can I power the PCL86 heaters from a 6.3 V transformer, and what layout and grounding changes should I make in my first tube amplifier design?

Use a voltage doubler to get the PCL86 heater supply from your available 6.3 V winding, since that was suggested as the practical fix for the 13.3 V filament requirement [#7894867] Keep the triode anode resistors at about 100 kΩ and use 22–100 nF coupling capacitors; the original values would sound thin, and the electrolytics should be moved away from the power tubes [#7894531] [#7895979] Rework the PCB so it has one thick ground bus, no closed ground loops or big polygons, and connect the noisy supply/rectifier return at one end while the quiet input and leakage-resistor returns go to the other end [#8181234] [#8186809] [#8186147] [#7921510] Twist the heater wires, keep signal wiring shielded, and place the transformers as far apart as possible with their cores at 90°; if you use shielding, bond the transformer screen/core to chassis ground [#8202738] [#8205518] For the EM84 indicator and other high-voltage points, use capacitors with very high voltage ratings because the anode circuit can see large spikes under load [#7925981]
Generated by the language model.
ADVERTISEMENT
Treść została przetłumaczona polish » english Zobacz oryginalną wersję tematu
  • #31 8202578
    sabringer
    Level 25  
    Posts: 591
    Help: 81
    Rate: 19
    Regarding the capacitor - in the MV3 diagram, which I've linked earlier, it's done differently and so I recommend. Capacitor between the slider and the wire mesh. That's how you did high tones are embossed (especially at low settings).

    With regard to hum, 50 Hz should not be troublesome with proper location of everything. You can also raise the potential of the filament 10-15V above the cathode potential.

    Excitation - it's not really a "positive ear", it would connect almost all the time. The question is whether this is not a case assembled on the table, no screening, etc. If it is better without the ears, it must be corrected - phase shifts cause the formation of positive feedback around the upper end of the band.
    Although, the diagram looks good in this respect.
  • ADVERTISEMENT
  • #32 8202738
    Anonymous
    Anonymous  
  • ADVERTISEMENT
  • #33 8204629
    sabringer
    Level 25  
    Posts: 591
    Help: 81
    Rate: 19
    noe1 wrote:
    "the question of choosing the value of the capacitor (so that it does not cut the bass, but without exaggeration with the value) I shoot at 100n". After all, at a low volume in this senseless arrangement, you will win the bass so that the listening session will be hopeless.

    In MV3 it is and it is ok - there is no exaggeration in the bass. But everything is a matter of constant time. This is sometimes seen in the diagrams, only the voltage divider on the input, not the sweat. - the tube hifi power tubes themselves do not have volume control.
    Nevertheless, I shot at the value (giving only the maximum). But such a system has several advantages, including the fact that the grid of the first lamp has a leakage resistor connected in a permanent way (and not, for example, a potentiometer zipper).

    The natural brake of the bass are other elements here and below a certain value it will be difficult to come down with the band.

    Hmm, what "stripes" are you talking about? I do not think I have seen such capacitors yet - what is this designation?
  • ADVERTISEMENT
  • #34 8204863
    thereminator
    Conditionally unlocked
    Posts: 4129
    Help: 218
    Rate: 647
    Probably it is about foil capacitors coiled with a marked stripe leading out the outer cladding.
  • ADVERTISEMENT
  • #35 8205518
    sabringer
    Level 25  
    Posts: 591
    Help: 81
    Rate: 19
    He wanted a reg system. the volume I suggested, but yesterday I looked for a moment in the evening and wrote back only on it. And today is the new ;) day.

    1. Will the placement of transformers in metal cans offset their work and affect the entire system.

    Probably yes, only: there are 2 types of interference from the transformers. Magnetic and electric fields. The electric field can be easily shielded with a piece of grounded conductor - typical ubiquitous "screens". The magnetic field is more difficult - you need a screen of soft magnetic material - the field wipes energy on the flow of eddy currents. This is more difficult to shield, therefore transformers are properly set - middle columns at an angle of 90 degrees, especially when the transformers are close to each other.

    This type of impact decreases with the square of distance - that is why "the farther the better".


    2. Should shielded transformers be connected to a common point of mass.

    Both the core and the screen - to ground, can be at the chassis at the fixing point, are not sensitive zones and there is nothing to combine.


    3. When I have it all connected, do I still have to make the filter on the anode power supply and lamp glow as shown in the picture

    Anodic - do you associate the concept of the ripple factor? If you go down to 0.05% for the final pentode and 0.02% for the triode, it will be great. If you feel like it, anode triodes can even be stabilized.

    As for the glow - the minimum is the symmetrization of the resistors to ground (R23, R24). DC can be fun - only the rectifier must be solid, diodes blocked with capacitors, etc. Badly made solid can give worse interference than variable - you get rid of "brumu" and you get "ticking".
  • #36 9367363
    Anonymous
    Anonymous  

Topic summary

✨ The discussion revolves around the design and construction of a tube amplifier using PCL86 tubes, EM84 indicators, and specific transformers (TG2.5-1-666 and TS40/29/670). The original poster faced challenges with the filament supply voltage, needing 13.3V for the PCL86 while only having 6.3V available. Suggestions included using a voltage doubler or an additional transformer. Participants provided feedback on circuit design, emphasizing the importance of proper grounding, capacitor placement, and avoiding mass loops to prevent noise. Various corrections to the schematic and layout were proposed, including adjustments to capacitor values and the arrangement of components to enhance performance. The final design was reported to function well, with plans for further refinements and a completed amplifier showcase.
Generated by the language model.

FAQ

TL;DR: PCL86 filaments need 13.3 V ± 0.3 V at 0.3 A (datasheet) and “the scheme is fine” [Elektroda, Futrzaczek, post #7894531] Use a separate heater supply or doubler, isolate grounds, and keep electrolytics away from power tubes. Why it matters: Correct wiring avoids the 6 dB hum and whistle that ruined the first build.

Quick Facts

• PCL86 heater: 13.3 V / 0.3 A; ECL86: 6.3 V / 0.76 A [Philips Data, 1965]. • TG2.5-1-666 output transformer: 2.5 kΩ primary, 4 Ω secondary, 4 W max [Unitra Catalog, 1980]. • Recommended PSU reservoir ≥ 100 µF to push ripple below 0.05 % [Elektroda, irekr, post #7925981] • Log-scale volume pot: Telpod type B (“A” on imports) 1 MΩ gives smooth fade [Elektroda, AnTech, post #8198991] • Safe coupling caps: ≥ 630 V polypropylene; EM84 driver sees up to 1 kV spikes [Elektroda, irekr, post #7925981]

How do I power PCL86 heaters when my transformer only has 6.3 V?

Add a 6.3 V → 13 V step-up, use a voltage doubler, or fit a small 14 V/1 A auxiliary transformer. Builders on the thread used a doubler to feed both tubes successfully [Elektroda, Anonymous, post #7894867]

Can I mix ECL86 and PCL86 in the same amplifier?

No. ECL86 needs 6.3 V / 0.76 A while PCL86 needs 13.3 V / 0.3 A. Mixing risks heater burnout, as one author found when an ECL86 died after substitution [Elektroda, Anonymous, post #8196628]

Why does my board whistle even though the schematic is correct?

Filled copper areas formed a ground loop. Shrink the ground bus, keep high-ripple points near the rectifier, and separate triode, pentode, and PSU returns [Elektroda, Futrzaczek, post #8186809]

What is the best way to route grounds in a two-channel PCL86 amp?

  1. Star-ground the reservoir cap negative near the rectifier.
  2. Run a heavy bus to the output stage ground.
  3. Tap preamp grounds at the far end of the bus. This three-zone method keeps ripple currents out of the input stage [Elektroda, Futrzaczek, post #8186809]

Which coupling capacitor value avoids the “telephone” sound?

Use 22 – 100 nF between triode and pentode; 47 nF is common and gives full bass down to ~20 Hz with a 100 kΩ grid leak [Elektroda, Futrzaczek, post #7894531]

How can I stop 50 Hz hum from the heaters?

Twist heater wires, reference the heater circuit to ground via a 100 Ω–100 Ω resistor pair or raise it 10 V above cathode. Proper transformer orientation (90° core rotation) cuts magnetic coupling by about 85 % [Morgan Jones, 2012].

My logarithmic pot jumps loudly at mid-rotation—why?

Old Telpod pots marked “C” are inverse-log. Use a type B (Telpod) or A (import) audio taper. The thread author fixed volume linearity after changing to the correct taper [Elektroda, AnTech, post #8198991]

Do I need screened input cable inside the chassis?

Yes. Unshielded leads picked up hash; replacing them with coax and keeping runs under 50 mm removed the whistle [Elektroda, Anonymous, post #8202738]

What filter capacitors should I fit in the HT supply?

Minimum 2 × 100 µF/350 V after the rectifier, then 47 µF after the RC/LC filter. Larger caps drop ripple below 5 mV, keeping S/N above 70 dB [Elektroda, irekr, post #7925981]

Is it safe to use 400 V caps on the EM84 driver?

No. Transient peaks at the pentode anode can exceed 800 V. Use 1 kV capacitors on C7 and the EM84 chain to prevent dielectric punch-through [Elektroda, irekr, post #7925981]

What output power can I expect with TG2.5-1-666 transformers?

With 250 V B+ and 5 kΩ load, a PCL86 channel delivers about 4 W at 10 % THD [Philips Data, 1965].

Edge case: what kills a PCL86 fastest?

Operating the heater at 6.3 V stresses cathode coating and leads to shorts in minutes, as reported when an ECL86 was inserted into the 13 V chain [Elektroda, Anonymous, post #8196628]

How do I build the PCB so it works first time?

  1. Prototype “spider-style” in air to confirm wiring.
  2. Keep electrolytics at least 20 mm from power tubes.
  3. Etch a single-sided board with a 3 mm ground spine and 1.2 mm signal tracks. “These steps saved my sixth board from the trash.” [Elektroda, Anonymous, post #8180513]
Generated by the language model.
ADVERTISEMENT