Single Ended Parallel Lead - review my planned overhaul

[GlideOn]

I got a reply back from AnTek basically stating "If you do not use more than 200VA of power, it should not be heat up. It will get warm only."

Since I'm only going to be using half of this PTs potential, that implies it may only barely get warm to the touch. I may be okay with the covers.

However, my other concern is the mounting, or the way I opted for it.

I'm not clear on how to mount it without tangling up the wires in the process, which happens easily when attempting to thread the cover onto the mount.

(don't worry, I fixed this so the bolt would center)

PXL_20240810_182849758.jpg

PXL_20240810_190155965.jpg

So I used spacers to elevate the threaded mount, then bolted them underside with 6-32 screws and nylon locking nuts through the threaded washer.

PXL_20240812_124532406.jpg

The main M8 bolt that would ordinarily go through the hollow toroidal core and bolt other side against rubber pad - now threads into to the cover itself from the underside. I had to get a shorter 1" M8 1.25 30mm bolt as the included 40mm was too long.

To my recollection, is this a dangerous way to mount for Torodials? Supposedly the magnetic flux is very powerful down the center and my old physics class tells me that conductive materials on both ends forms a loop which could basically spot-weld this thing in pace of voltage/amperage was high enough.

It has to couple firmly against something, my logic told me that the bolt was not sufficient enough to couple the entire thing stably, so both the mount needed to be bolted to the chassis so as to not wiggle around (especially for a combo) and bolt threaded thru underside. Given that the mount has two pilot holes on side and had steel screws and bolts included, implied this was okay to do?

PXL_20240812_131651340.jpg

The included rubber pads I put on either side of the toroidal both insulating the inside top end of the cover and the bottom against the threaded washer, which is what the protective covers screws into. I was thinking these were meant as cushions, not so much insulators. If they were, then one of the rubber pads would have pilot holes going through them, no? But they do not, so I didn't use it under the mount but ontop/under the toroidal and ontop/under the cover.

Was this the correct way to mount this?

[maxkracht]

If you're not using the cover, you would usually mount it flat to the chassis, rubber disc on either side of the transformer, then the cone thing clamping it down. Wires would go out to the side then through the chassis, preferably through some rubber grommets to avoid damaging the wires. Never mount directly to the epoxy core if your transformer has one, they can separate fairly easily and cause damage.

The only thing that looks potentially problematic with your setup is the two screws that are a bit proud of the mounting plate. I would countersink those holes so the screws don't damage a winding.

[GlideOn]

If you're not using the cover, you would usually mount it flat to the chassis, rubber disc on either side of the transformer, then the cone thing clamping it down. Wires would go out to the side then through the chassis, preferably through some rubber grommets to avoid damaging the wires. Never mount directly to the epoxy core if your transformer has one, they can separate fairly easily and cause damage.

The only thing that looks potentially problematic with your setup is the two screws that are a bit proud of the mounting plate. I would countersink those holes so the screws don't damage a winding.

Believe me, it was a bear getting it to stay on with just the mount as it is quite heavy with the cover on. I could not get the wires to stay in their holes as with rubber discs in place, they make it grip the toroidal too as you're tightening the cover. Without some gaps under via the mounted bracket, you're just grinding the wires' away when tightening.

I had zero issues with the Output Transformer and Choke as the mounting screws go through the cover and are held in place by plastic/nylon top screws and rubber pads both side of course.

Right, we don't want the windings exposed to the bare chassis, which Torodials have arguably weakly wrapped covers to begin with. That's makes much more sense. That's what the rubber pads are primarily for I gather?

I can easily use nylon screws in place of the steel screws and pure nylon locking nuts - or countersunk/conical screws are a good idea too!

[maxkracht]

As far as I understand, the rubber discs are to prevent mechanical damage to the windings, they don't have much protection from the outside world. I would avoid using nylon screws, they might not be strong enough long term. Either countersink, or add an additional pad so the heads don't press against the transformer. Anything sticking up could put stress points on the windings and/or rub away some insulation. It's going to live in a vibrating box. Probably fine with the rubber pad for protection, but better safe.

[GlideOn]

IMG_20240619_231714.jpg

Before

PXL_20240819_032359693.jpg

After

[GlideOn]

PXL_20240819_210050606.MP.jpg

Hmmm, getting a blown fuse (3A slow blow) after standby is flipped up after approx 13-15 seconds. The fuse doesn't look too terribly damaged either.

This would indicate more current is being drawn than it can supply and/or shorting, no?

See (after) picture above. The Output Transformer has a 3 primaries:

Red
Purple
Blue

Red goes to B+1

Blue goes to Anode of V4 (parallel jumper wires to V5).

Purple is going to Earth/IEC ground (under and between power tubes).

I was thinking Purple was the shield wire for the metal cover I opted for as it is the case with the Power Transformer which has purple shield wire. My instinct was to ground it.

Now that I had a quick look at the specs, this may be the "Ultralinear" tap as it is in the middle of the Red and Blue and labeled "Screen." I don't plan on utilizing it.

Multimeter measuring continuity at the Anode/Plate to chassis confirm this is shorting to ground, some resistance of 2ohms though.

Should this purple wire be snipped and capped off instead?

I see no mention of any guide anywhere of what to do for unused Ultralinear/Screen tap.



(P.S.- These AnTeks are driving me nuts!)

[maxkracht]

Should this purple wire be snipped and capped off instead?

Yes. If you don't want to use a UL tap, don't use it.

[GlideOn]

Should this purple wire be snipped and capped off instead?

Yes. If you don't want to use a UL tap, don't use it.

Okay! Snipped, shrink wrapped and ziptied!

It's also been brought to my intention that using the stock first power tube cathode as reference for heater ground/center tap could create voltage instability, so instead I'll go with the more proven, idiot proof 100ohm resistor to ground method off the Pilot Lamp. That way they can act like fuses too if anything goes wrong.

[maxkracht]

It's also been brought to my intention that using the stock first power tube cathode as reference for heater ground/center tap could create voltage instability, so instead I'll go with the more proven, idiot proof 100ohm resistor to ground method off the Pilot Lamp. That way they can act like fuses too if anything goes wrong.

What instability are you referring to? I haven't heard of this. You should still use a pair of resistors or the centertap with elevated heaters, just like you would connecting to ground.

[GlideOn]

It's also been brought to my intention that using the stock first power tube cathode as reference for heater ground/center tap could create voltage instability, so instead I'll go with the more proven, idiot proof 100ohm resistor to ground method off the Pilot Lamp. That way they can act like fuses too if anything goes wrong.

What instability are you referring to? I haven't heard of this. You should still use a pair of resistors or the centertap with elevated heaters, just like you would connecting to ground.

I'm also DM'ing someone over at TDPRI, someone experienced with single ended amps and whom recommended the Torodials.

Apparently center tapping from pin 7 (heater) to pin 8 (cathode) on just one can invite a voltage swing of up to 9v + on the opposing 6.3v lead wire. I have no idea how this might be possible as I technically have the heater wires reverse phasing to each power tube, so that may be a moot point as all of the heater wires forms continuity, no matter where measured.

But I also know that resistors can act as safety fuses, so it seems to me that Center Tapping using 100ohm of the Pilot Light to PT ground is smarter bit added bit of extra security vs depending upon a single power tube socket, wired to the middle of the circuit board.

[maxkracht]

I didn’t see your schematic before. Connecting one side of the heater string to the cathode is odd, I’m not sure why they chose to do it that way or if it was an error.

The heaters need a DC connection to ground to limit hum. Usually, you would connect the center tap or artificial center tap (100r resistors) to: ground, power tube cathode, or another positive voltage source. Here, center tap refers to midpoint of your 6.3v winding. There is no center tap used if only one side is connected, but there is a ground reference. You can still connect your 1/2w 100r resistors to the cathode for potential hum reduction. The heaters are still 6.3vac, but that 6.3vac is riding on +15vdc or whatever your cathode voltage is.

[GlideOn]

Yes, it was originally wired to the positive side of cathode as the original SEL schematic calls for that.

[GlideOn]

Small thread update - had some trouble with signal shorting in and out, plus oscillation in the power section when turned up louder than a whisper, so I took to tech as it was beyond me to figure out.


Turns out is was just two simple issues:


- the Input Cliff Jack was shorting to the chassis

- the Fake PI Stage: Presence Control doesn't feedback into other half of the tube not used, so it was generating positive feedback and oscillation around 50hz.


The Fake PI Stage schematic I followed was bupkis. It doesn't work! Normal PI is actually inverting phase, so the insertion of presence and NFB is negative to positive and so forth. Not so when you only use one half!

However it did work when the wire was moved to the pin or start of the Cathode Resistor. As a consequence the presence control isn't dramatically effective this way, but the Negative Feedback is! NFB in amp this greatly helps keep the hair down a bit and more importantly tighter and more detailed. Very worthwhile to have done after all.

Now I am just trying to extract more power out of it all.

I know it's Single Ended, but I was hoping to get more than 15w RMS, more like 20-22w would be great.

I have each EL34 only biased with a 330r each for 42ma, I suppose I could lower these for more bias?

Also, revisiting the B+ line, I have no idea why I still have the additional 5k dropper for the Screens. They already have dedicated 1k Screen droppers anchored from pin 6 to pin 4.

1000067444.jpg

AX84_SEL_100616 (1)_240226_185735_2.jpg

I'm not sure why the original schematic has it. Is it some necessity to regulate screen voltage in Class A? Nothing I can read tells me this.

Would removing the 5k dropper raise the voltage...without side effects?

See pics.

[lonote]

Looks to me like the 5K (R6) & C5 (10uF) are creating a B+ side node, lowering the screen voltage without affecting the downstream main B+ rail voltage by making it all pass through the 5K. I would have to guess that was an intentional design choice.

[GlideOn]

Looks to me like the 5K (R6) & C5 (10uF) are creating a B+ side node, lowering the screen voltage without affecting the downstream main B+ rail voltage by making it all pass through the 5K. I would have to guess that was an intentional design choice.

Yes, the Screens must get filtered but many amps I see do not have a dropper until before the PI. The tubes get their own individual 5w resistors, either 1k/5w for EL34/KT88 or 470r/5w for 6L6/KT66.

The original schematic is merely a jumping off point. Its otherwise a new design now. There's new PT putting out 400v @ 300ma + and OT rated 2.5k @ 30w. Whatever voltage I can throw into the circuit, it can handle it.

I actually removed the 1k/5W that was coming off the B+2 filter, now that has need has been facilited by a new 4.7k prior to V3/preamp. 10k V1 dropper still there.

I'm really scratching my head as to why it was needed. I know of no information as to any restrictions of screen/plate voltage disparities in Class A. Also, it's highly likely under wattage with 400v+ coming at it and tubes drawing 85ma+