Best way to get different voltages from a tapped transformer

[tizman]

I had a look inside the organ amp I repurposed, and the B+ does come from the centre tap of the main transformer and not from the 5V transformer centre tap. Sorry for my confusion. With respect to the power transformer being able to supply the required 70 mA using the 0-395 winding, there is no way of checking wire diameter, or anything else inside, as the transformer is fully potted and encased in a welded enclosure with terminals on the exterior. It is a very stout Freed transformer, and appears well made. As I am drawing 70 mA from the winding rated at 35 mA, and I will not be using the other 520-0-520 Volt @ 250 mA winding at the same time, I think it’s worth trying out.

 

[jhstewart9]

jhstewart9: What sort of switch would you recommend? As an alternate to the switch, I could install and wire two seperate rectifier sockets, one for each circuit, and insert the corresponding rectifier tube when I use the circuit. Is this a sensible alternative to a switch?

The type of switch required to operate on the several leads & combine to result in the voltages you need may be difficult to source. First of all it would need to be break before make. Because of the voltages are large it would need to be 3-position, keeping the voltages well spaced from each other. Might be something out there but hard to find.

With regard to the B+ output from the rectifier & filter, allowance needs to be made for IR drop in the choke(s). The forward drop in vacuum rectifiers could be 10s of volts, depends on the loading conditions. The 5R4 family has a larger drop than most of the other 5V rectifiers.

So if we take 520vrms as the starting voltage, we would get the 520*0.9 volts of the choke input filter but less the rectifier & choke losses. So in a practical cct, 440V is a good guess to start the analysis of the rest of the circuit.

Here is an alternative. Altho it may appear somewhat complicated, removing all the meters it becomes much simplified. It amounts to running the 300B amp straight from your 5R4 rectifier & filter system. And the 45 Amp derives the correct voltage from a simple fixed regulator, in this case set to about 260 Volts. A 6L6GC can dissipate 30W in the triode mode, so OK here.

Other voltages are possible simply by substituting in other gas regulators. There are several to choose from. Last time I looked they were still low cost & plentiful. And they are more stable than Zeners. I’ve found HV Zeners have a very large coefficient of temperature vs voltage.

Building on this a variable voltage output is also possible. If you are interested, ask. I’ve designed & built everything up to a 4.5KV, One Amp regulated monster with fail safe for the operator. The fail safe cct saved my *** around 1963!

 

[jhstewart9]

I had a look inside the organ amp I repurposed, and the B+ does come from the centre tap of the main transformer and not from the 5V transformer centre tap. Sorry for my confusion.

Should be other way round, B+ from the rectifier cathode. Check again.

With respect to the power transformer being able to supply the required 70 mA using the 0-395 winding, there is no way of checking wire diameter, or anything else inside, as the transformer is fully potted and encased in a welded enclosure with terminals on the exterior. It is a very stout Freed transformer, and appears well made. As I am drawing 70 mA from the winding rated at 35 mA, and I will not be using the other 520-0-520 Volt @ 250 mA winding at the same time, I think it’s worth trying out.

Don't need to check the wire diameter at all. An Ohmmeter will reveal all.
If the transformer HV is would serially it will be the same wire, end to end.
Most are for these kind of applications. Just measure resistance of the HV wdg from tap to tap. The ratio of R from the CT to the 395V tap & then the end of that wdg is same a voltage spec.

 

[jhstewart9]

Your "Does This Work?" clearly does NOT work. It gives 2.5V AC where you seem to be expecting large DC.



That IS the "ideal" case for a naked-filament dual rectifier. Taking B+ from one end or the other adds/subtracts 5V on alternate waves, adding some 50/60Hz tone to the big 100/120Hz ripple.

But I once prodded a friend to 'scope it, and neither of us could see the difference (for a 400V supply, 5V rect).

And so many rectifiers are indirect-heat cathode, in which case the "ideal" connection is to that end of the heater.

Here is the paper that explains that in detail. And the remedy.

 

[PRR]

Here is the paper that explains that in detail. And the remedy.

Thanks-- but that's not the effect I was talking about.

Taking one side or the other of a 5V filament introduces a 60Hz component. Using a CT on the 5V filament or its winding nulls that.

You have solid-state diodes, and an unbalanced HV winding. You found a 60Hz component, and figured why, and a way to null that.

Yes, it is possible that my no-result was really a combination of two unbalances, which only happened to complement each other. Maybe if we had swopped the plate -or- the filament leads we would have seen 60Hz pop up.

 

[tizman]

transformer resistance vs voltage

jhstewart9: I checked the resistance of the taps on three of the five identical transformers I have, and found the following:

taps 3-6 1040 Volts 122.1-118.5-117.5 Ohms (average is 119.37 Ohms)

taps 4-5 395 Volts 43.5 - 42.5 - 42 Ohms (average is 42.67 Ohms)

There is continuity between all the secondary taps. The voltage ratio is 1040/395 or 2.63, and the resistance ratio is 119.37/42.67 or 2.79. These ratios are fairly close to each other. If the mA ratio is proportional, this would indicate that I can expect the 395 volt winding to be capable of 250 mA plus 35 mA for 285 mA (the combined ratings of the two windings operating simultaneously). This works out to 285/2.79 or 102.2 mA. I am using only 70mA for the two 45 tubes, and there is a large portion of the winding not being used at all in this situation. It would appear that I am good to go, assuming I did the math right.

The two halves of the 1040 Volt winding were consistently off by 10 Ohms, with 3 to 5(the CT) averaging 54 Ohms, and 6 to 5 averaging 65 Ohms. I haven't measured the offload voltages on all the transformers, but the one I did measure was very close to equal with respect to the voltages from each of the two ends to the centre tap.

 

[jhstewart9]

We may not be aware of how a particular instrument measures resistance.
That can be a factor when measuring resistance of inductors such as chokes & transformers

Some of the newer ohmmeters using active electronics as opposed to the ordinary analogue VOM use an AC signal, usually one KHz while in operation.That is OK for an ordinary resistor but may lead to errors on most iron cored devices. And some crossover inductors as well.

I can't give any examples of makes or models that operate this way. But it is worth knowing should the measured answer looks unrealistic. Double check with another meter, preferably one of the older analogue VOMs, such as Simpson, Triplett or AVO.

The attachment shows a very useful piece of test gear I built many years ago to make some of the more difficult resistance measurements. It is low cost, easy to build. I've found it very useful.

 

[PRR]

I too have had trouble measuring DC Ohms on big iron with a DMM.

The first trick is to short any other winding, kill the bulk inductance that is confusing the DMM's auto-range.

Yes, a needle-meter is a good friend.

More elaborate methods may be needed in many cases, especially heater and speaker windings. But 1000V 200mA power winding should not be a problem. I agree your measurements already suggest this is all one winding all the same gauge. (It is not worth switching from #37 to #36 for a minor load, just slows production.)

 

[tizman]

As an afterthought, if this transformer was designed to operate the full winding and the taps simultaneously, that implies that the 0-395 winding is rated to carry its own 35 mA as well as the portion of the entire winding’s mA that it shares with the 520-0-520 Volt @ 250 mA winding, when both are used simultaneously. It is the same piece of wire. Wouldn’t this imply that the 0-395 volt @ 35mA winding spec would reflect a derating to take this sharing of the full winding into account?

43/119 Ohms = .36

.36 X 250mA = 90mA

The 1040 Volt winding is rated at 90mA within the shared 0-395 volt portion of the entire winding. The 0-395 volt portion of the winding is rated at 35 mA while in simultaneous use with the entire winding. 90mA+35mA=125mA. If the same gauge of wire is used throughout the winding, this would imply that entire 1040 Volt winding’s mA rating reflects the derating required by the sharing and simultaneous use of the 0-395 Volt portion of the entire winding. This also implies that the 250 mA rating would be higher if the 0-395 Volt portion is not used.

With respect to thermal loading, there is a large portion of the entire 1040 volt winding not being used when just the 0-395 portion of the winding is in use.

 

[bluerooster]

jhstewart: Like this?

Also, I need to purchase a 5 volt transformer for the rectifier filaments. If I buy a centre tapped 5 volt transformer, can I use its centre tap for the B+ and avoid the two extra diodes?

Yes, But... you will also need a center tap on the hv transformer. Otherwise it will be no different than any other two diodes, whether silicon, or tube. AKA half wave rectifier.
I may be slightly off here but; A transformer with no center tap requires a bridge rectifier, and a center tap transformer can use a full wave rectifier, unless you don't use the center tap, then it's treated as a transformer without ct, and needs a rectifier bridge.
You can also use one leg, and center tap, and disregard the other leg, but it will require a bridge rectifier. so in theory you can take a 520-0-520 transformer, and using a fullwave rectifier, have 520x1.414= ~735DC. or you can take that same transformer, with a bridge rectifier, disregarding the CT, and have 520+520x1.414= ~1470DC, or you can use half of the secondary taps (520 and 0) with a bridge rectifier to get ~735dc.
Attached is a data sheet for an 872 diode. On page 2 is given the various circuits for full wave, and Bridge rectifier circuits. The 4 diode, full wave circuit, is commonly called a "rectifier bridge", the 2 diode circuit is refered to as a "full wave rectifier".
it is not necessary to take the B- from center tap of the filament supply, it can be taken from one side of the filament.
I've found that some folks can get confused about silicon rectifiers, and tube rectifiers.
But just think of them as Diodes, period. They do the same thing: only allow electrons to flow in one direction.
Now, with the double diode rectifier tube that you want to use, it is a full wave rectifier, you can use two of them, and have a rectifier bridge, for the transformer you have shown. Instead of a 2.5-0-2.5 transformer for the filament supply, you can use a 5-0-5 transformer, to feed the filaments on the two tubes, by using one 5 for each tube, and the 0 as common for both tubes, and draw B+ from one leg of the filament.