Converting Filaments AC/DC

[MarkusBass]

Now it's time to build the circuit and report results.

Mark

[lowell]

Will do!

[tedmich]

Software

PSU Designer II works well for me!

[Gregg]

You can also use a voltage doubler followed by a 12V regulator (LDO if necessary) to get 12V regulated for the filaments but then you should omit the artificial CT point resistors or if it's a CT winding to disconnect the CT from ground. This is the schematic posted in a similar thread:

[MarkusBass]

You can also use a voltage doubler followed by a 12V regulator (LDO if necessary) to get 12V regulated for the filaments but then you should omit the artificial CT point resistors or if it's a CT winding to disconnect the CT from ground. This is the schematic posted in a similar thread:

[ATTACH=CONFIG]35849[/ATTACH]

Can you provide a link to this discussion? Why the resistors should be removed but they were there initially? I'm working on Mesa Mark IV that uses this solution but for me it's one of the worst solutions. There are many problems with this circuit. The voltage provided to the input of 7812 (this is not LDO) is very close to 12V and it is not sufficient for the regulator to work correctly (it should be at least 14.7V). Mark IV has also double power switch (Full Power/Tweed Power) that switches primary winding of the power transformer. But this also switches heaters voltage and in this case the voltage is below 6V (which is again not sufficient for 7812).

Mark

[J M Fahey]

Can you provide a link to this discussion? Why the resistors should be removed but they were there initially?

Because the initial filament wiring was CT 6.3VAC, while the newer one has a voltage doubler.
Start analyzing the doubler fully floating, later we think about grounding.
* You have 2 caps in series, C442/443 .
* one end of filament winding is connected to capacitors center tap and nowhere else (we need to lift artificial center tap resistors).
* the other winding end charges alternatively C442 ~8V positive and C443 ~8V negative relative to the center point, so ~16V end to end.
They claim 15VDC , fine with me, probably the real world loaded value.
Now you ground necative of C443 and have very useful raw +15V available from positive of C442 .

Artificial ground resistors offer nothing useful, load the center ~8V to ground and unbalance the center point ... who needs them?

I'm working on Mesa Mark IV that uses this solution but for me it's one of the worst solutions. There are many problems with this circuit. The voltage provided to the input of 7812 (this is not LDO) is very close to 12V and it is not sufficient for the regulator to work correctly (it should be at least 14.7V).

Well, that's a problem.
Maybe an LDO could help.

Mark IV has also double power switch (Full Power/Tweed Power) that switches primary winding of the power transformer. But this also switches heaters voltage and in this case the voltage is below 6V (which is again not sufficient for 7812).

Harebrained design.

[MarkusBass]

I don't want to hijack Lowell's thread - that's why I was asking for the link.

Because the initial filament wiring was CT 6.3VAC, while the newer one has a voltage doubler.

This still does not explain why on the official Mesa schematic the resistors are present (along with voltage doubler and 12V regulator).

They claim 15VDC , fine with me, probably the real world loaded value.

15V is the optimistic version. In reality I get close to 12V. This is maybe due to artificial ground - I have to check it.

Maybe an LDO could help.

LDO requires at least 12.5V to work correctly. Here, depending on the power switch sometimes I get less than 12V on the input of the regulator.

There are several other problems with the amp design. The power transformer has 220 and 240V windings (and no 230V). But there are some countries (including Poland) that have 230V mains power voltage. With 240V winding I get 7V for the heaters. With 220V I get 6.6V for the heaters with Full Power settings and less than 6V with Tweed Power settings. No matter what I do, most of the voltages (or at least many) are incorrect.

Mark

[raiken]

Use Schottky diodes in your DC filament supply to get less voltage drop. Typically around 0.35-0.55V instead of 0.7V per diode.

[Gregg]

Can you provide a link to this discussion? Why the resistors should be removed but they were there initially? I'm working on Mesa Mark IV that uses this solution but for me it's one of the worst solutions. There are many problems with this circuit.

There are many threads about DC filaments in the forum and I don't remember which one exactly that was. Bottom line was those resistors shouldn't be there. I can tel you from experience that if you leave them they get very hot and don't help reducing hum in any way sometimes it's the opposite. I used this circuit many times and had no problems with it. Use Schottky diodes, an LDO and 10000uf caps and you're not going to have any problems with the circuit provided you have 6.3VAC on your filaments' winding.
It definitely works for two tubes (~300mA) but I'm not sure it can handle more than that.

There are several other problems with the amp design. The power transformer has 220 and 240V windings (and no 230V). But there are some countries (including Poland) that have 230V mains power voltage. With 240V winding I get 7V for the heaters. With 220V I get 6.6V for the heaters with Full Power settings and less than 6V with Tweed Power settings. No matter what I do, most of the voltages (or at least many) are incorrect.

It should be the opposite: 230 mains -> 220V winding - 7V, 240 winding - 6.6V. Also you can always lower higher voltages but the opposite could be impossible.
I don't understand why they are switching the filament voltage. May be it can be undone.
Getting only 12V after the doubler means there is a problem. I would start with the electrolytic caps.

[MarkusBass]

It should be the opposite: 230 mains -> 220V winding - 7V, 240 winding - 6.6V. Also you can always lower higher voltages but the opposite could be impossible.

Yes, maybe I specified the voltages not in the order I intended, but you get what I mean. With 240V windings the voltages are lower than expected, with 220V winding they are higher. In either case heater voltage is incorrect because they haven't foreseen that he amp may be used in a country with 230V mains voltage. And 230V is used in whole Europe, almost whole Asia, Australia, half of Africa and half of South America. And they just forgot about it .

I don't understand why they are switching the filament voltage. May be it can be undone.

They are switching primary windings. The filament winding is not switched but when you switch primary windings it affects secondary winding too (including filament winding). Of course, you can use the FULL POWER position only and in this case the problem does not appear.

Getting only 12V after the doubler means there is a problem. I would start with the electrolytic caps.

Thanks, I will check it - starting from virtual ground resistors and then capacitors.

Mark

[g1]

With 220V I get 6.6V for the heaters with Full Power settings and less than 6V with Tweed Power settings. No matter what I do, most of the voltages (or at least many) are incorrect.

That sounds reasonable for the design. Using a PT primary tap for the full power/tweed switching like that means both positions must be a compromise. The Dual Rectifier bold/spongy switch works the same way. If your heaters were 6.3V on full power, then they would fall even further less that 6V on the tweed position.
In north america with the Dual Rect. set for 120V, I've found the same issue, "bold" setting gives heaters more than 6.3V, "spongy" setting puts them below 6.3V. I don't think it can be any other way with this design.

[Gregg]

MarkusBass, if I were you I would use the 220V winding and lower the filament voltage to specs using fat ass 0R1-0R22 Ohm resistor attached to chassis. Lowering anode voltage to preamp is not a problem and I don't think the plate voltage will go up too much. This way you'll get the filament voltage to specs being able to feed the voltage doubler circuit but check its components first.

[J M Fahey]

I don't understand why they are switching the filament voltage.

Oh, but that's the point !!!!

Don't look for much Logic into that.
They are just blindly following what (Guitar) GOD said, period, in this case EVH .

I still have somewhere the dog eared yellowing (by now) old copies of Guitar Player where at interviews he spewed Tech nonsense by the truckload, with the interviewer copying it without understanding himself.
Not too different in concept to Howard Dumble's famous "fragile harmonics not surviving in the crystal lattice of transistors" and similar BS .

Yet once printed and attributed to him it became part of the Gospel.

This "spongy/bold" thing comes from his mention of using a Variac to get his trademark sound, which he called "the Brown Sound" ... **clear** reference to the reduced voltage you get during a "brownout".

Which is what spongy/bold literally does ... and of course "reduced primary voltage" means "all voltages inside the amp" .are reduced.

FWIW I also remember him saying that he "grabbed one of those large filter thingies inside the amp, those which look like a stick of dynamite" and supposedly put them in series with the primary to permanently achieve such "brown sound" ... which *might* be technically possible *if* those were not electrolytics which of course can't stand AC.

Also saying the "during his solos there was a roadie by the Variac wheel, and "when he played a blistering solo the guy rised voltage to 130 , 140 and even 160V, that tubes started to shine like Christmas tree lamps and that they didn't just explode because he was sucking all that energy into his Solo, if he stopped shredding even for a second they would melt down" .

Not quoting literally but what I remember, but the general idea was that.

Notice that the ideas are not fully nonsense, that there is a grain of truth behind them, he's saying that with reduced voltages sound becomes mushy ("spongy" anybody? ) while with improved voltage and maybe improved cathode emission tubes would work better (even if short lived) ("bold" anybody ? )I always had the idea that either he or some assistant had at least some idea of Electronics but he said that tongue in cheek, probably an internal joke directed at those who think they can play like him just by copying his setup.

Thanks God he didn't say that his best solos were played using soiled undies or something .

FWIW I am old enough to remember (even found a dusty one in my transformer stash) "(CRT) tube revivers/reactivators" (loose translation from Spanish "reactivador de tubos (de imagen)"
when old trusty BW Tube TV in the Living Room started to give up the ghost , and brightness/contrast controls lost effect to improve the greyish washed out image, then Mr Tech came to the rescue and applied one of them, which magically restored everything ...for a short time.

They are simply filament voltage autoformers, with 6.3/7/8/9/10VAC taps.
No mains primary, which made them very cheap.

CRT filament was boosted , as much as needed, to compensate for worn cathodes reduced emission.
It worked ... sort of , the first time for a couple months, the second one, and so on.

It bought some time to raise cash to buy a new TV.

Maybe Eddie was aware of that and used it both ways, either to boost or to pad the full amp.

And probably Mr Randy Smith added it to the 2756 amp options.

"Qué le hace una mancha más al tigre?" (does an extra spot/stripe worry the tiger?)

[MarkusBass]

This is what is called "thread hijacking" . Lowell are you still with us?

Juan, what you write about the "Brown Sound" makes sense for me. The HT voltage is lowered probably by about 50V (I may measure it tomorrow), which changes the sound but there is one (big) problem. The first tube is supplied with voltage doubler and 12V voltage regulator. And this voltage regulator (7812) does not work correctly if it gets below 14.7V on the input. Thanks to Gregg (more on this later) the voltage regulator work correctly now but only in the FULL POWER mode. In TWEED mode I get 11.47V DC on the first tube. It means that the regulator is not working.
Regarding the virtual ground resistors: Gregg was right - they are not needed. I did a simulation of the circuit and I measured that with the resistors in the circuit the first resistor dissipates 0.54W and the other 0.93W. This is a lot since the resistors are 1W only. It's no wonder that the PC board below them is carbonized. The ripple is 0.7V (with 4700uF capacitors). Most probably they should be increased to 10000uF but there is no space to do it. I removed the resistors and now on the input of the regulator I have 15V (in FULL POWER mode).
This is very interesting because there were two versions of Mark IV and the one that I'm fixing is version "A" (the same as the schematic posted above) and in version "B" they added another winding (12VAC) for the first tube and footswitch power supply. The other tubes are supplied with CT winding (without resistors). So they noticed that there is something wrong with the first version. I don't know who removed the two resistors on the schematic posted above, but it looks like version "A" was assembled at the factory with both resistors on the board. And this was a serious mistake.
But the life is not so easy. Without those resistors heaters of other tubes are not balanced. One side of the heater winding is grounded with a diode. Juan said that the resistors are not needed but you find them in almost every amp. So someone considered them to be important. It seems that the version "B" is correct and version "A" is wrong no matter whether the resistors are removed or not. I wonder what people from Mesa would say about the problem.

So big "thank you" to Gregg . I solved another problem with the amp. Actually, I solved three major problems but the amp still has one or two more. I thought that one of them is related to the voltage regulator but it's not. I think that I'll create a new thread about the amp.

Mark

[Leo_Gnardo]

I solved another problem with the amp. Actually, I solved three major problems but the amp still has one or two more. I thought that one of them is related to the voltage regulator but it's not.

Good to see you're making progress on the Mk IV. I've been having a go at a IV someone spilled their soup on & it dripped thru the handle mounts, apparently a pungent phô loaded with plenty of nguoc-mam. Every time I solder something on the circus board it releases a cloud of funky fish odor. pee-yew! At this point most everything works, got a good clean & crunch channel but the overdose channel sounds like a geiger counter reading a piece of plutonium. Must have used a hundred cotton swabs and alcohol cleaning greasy crud off the board. The things you have to do to fix amps...