Different rectifiers and proportion of voltage dropped

[tubeswell]

Hi there

I am after a simple chart or formula(e) or rule of thumb to help me work out what to expect when I hook up various center tapped HT secondaries with various rectifiers. Something that sorta looks like the following:

Sec. winding of xyzVAC with Full wave diode rect = ratio in VDC
Sec. winding of xyzVAC with 5AR4 = ratio in VDC
Sec. winding of xyzVAC with 5U4G = ratio in VDC
Sec. winding of xyzVAC with 5Y3Gt = ratio in VDC

and then one for HT windings without a CT that says:

Sec.winding of xyzVAC with Full wave bridged diode = ratio in VDC
Sec.winding of xyzVAC with Half wave diode = ratio in VDC

Or is it not that simple?, (because do I have to take into account the particular loads)?

Cheers

[Enzo]

Theoretically when you rectify AC, the resulting DC once filtered will be about the peak voltage of the AC waveform. In other words, 120VAC would make about 170VDC. That is with ideal lossless rectifiers, of course.

SOlid state diodes lose about a half a volt, so that drop can be ignored.

If you are not using a center tapped power transformer, I tend to doubt you are using rectifier tubes.

To figure the loss through a rectifier tube, just assume the AC rectifies to the ideal DC, then subtract the tube loss. SO a given tube will yield peak minus 30 or 50 or whatever. When you start to draw current through the tube, then things get more dynamic.

I am too tired to get into the tubes at the moment, I am sure someone will chime in with more specifics.

[Rob Mercure]

Tubeswell,

The peak voltage Enzo refers to is 1.41 times the RMS. But you'd be better off getting a copy of one of the later RCA Receiving Tube Manuals or the ARRL Handbook which pretty well break out power supply design. With the tube manual you can refer to the characterics of particular rectifiers. And these are not free on the internet - 20-30 years ago we used to have to buy them at the equivalent (constant dollars) of about $30 US.

Rob

[daz]

I am too tired to get into the tubes at the moment, I am sure someone will chime in with more specifics.

Uh, wait a minute...Enzo TOO TIRED TO TALK TUBES?????!!!!! I am apparently having a bad dream here. tell me this isn't true !

[hasserl]

DC voltage from various rectifier's:

Rectifier * AC V per plate
5Y3 1.13
5R4 1.17
5U4 1.27
5AR4/GZ34 1.3
SS Full Wave 1.414

I've got an Excel spreadsheet with the formulas plugged in, all you need to do is insert the AC voltage and it will list the voltages from the various rectifiers. Send me an email addy if you want a copy.

[MWJB]

Following on from Rob's post, some guys use the following proportions for tube rectifiers:

5Y3GT 1.1x unrectified AC
5U4G 1.2x unrectified AC
5V4 1.26x unrectified AC
GZ34/5AR4 1.3x unrectified AC

As Enzo touches on, things won't necessarily work out "to the volt" as guitar amp PTs are unregulated and current draw inversely affects B+ voltage - high power tube plate idle currents & high 6.3VAC heater current draws may pull down final voltages, conversely the differences between B+ voltage (if note in envelope of attack & decay) in something like a Champ will be minimal (even if not in envelope of attack & decay) because the power tube draws relatively little current and is a small load on the heater winding (from memory, voltage difference between 5Y3 & SS, with the same power tube, in my Champ is only 10v or so) assuming a reasonably robust PT. Subbing the 6V6 for a 6L6 may have a much greater effect, like 450v at 30mA fixed bias for 6V6 vs 390v at 50mA with 6L6 (no adjustment to bias pot - remember, as well as increased plate current, 6.3VAC current is also being greatly increased in this example).

Biasing cold with low idle currents may cause the B+ voltage to rise beyond that 1.41x "specified" AC voltage. E.g. I've seen an amp with the same rectifier show +/- 60vdc on the power tube plates at extremes (but still usable) of bias. Differing brands of the same type of tube will also affect B+ differently, for these reasons.

So those multipliers are probably best used as a guide for push-pull, fixed bias, with moderate idle current draw and reasonable headroom on the 6.3VAC heater winding (33-50%)...then again, it's the same scenario for SS rectifiers too. As a rule of thumb I use the rectifier multipliers, then assume I'm going to lose 20v-ish with an SS rectifier...

E.G. 345-0-345VAC x 1.4 (SS rectifier) = 483vdc, drop 20v for a reasonable current load and you'll probably see 465v after the rectifier...sounds spot on to me! That 20v drop may progressively decrease as you sub down to lower voltage rectos, less demand on the B+ winding again.

The reality is don't throw your toys out the pram if your 20v short due to the number of variables. Also be aware that PT's spec'd to a given B+ secondary voltage rating may vary somewhat due to manufacturing tolerances/wall AC in your region/"in between" PT primary taps, etc.

For cathode bias you also have to take into account the typically higher idle current draw, that might mean losing another 10-20v, before you even consider subtracting cathode voltage from plate to ground voltage. E.g. my 2x6V6 amp runs 430vdc in fixed bias, 410vdc in cathode bias, but idle current is pretty well doubled. (JJ 6V6s in all cases, I wouldn't run RCA/valuable NOS this high voltage-wise).

Overdesigned PTs with larger B+ current ratings and excess heater current supply (100% overrated?) tend to exhibit less voltage swing, downside is "off the shelf" types with good margins for 2x6L6 tend to lack 5V rectifier windings...no problem if you're winding your own/custom ordering. Likewise, proprietry 6L6 types may have too much voltage on the B+ secondary to allow you to use the rectifier of your choice with 6V6?

[tubeswell]

Excellent information on the full-wave rectifier types thanks everyone

I am still wondering about half-wave rectfier voltage drop.

I have heard thru a grapevine that you should use .707 when working out a Half Wave voltage drop, but am not sure if this includes the filtered DC level or not.

I am interested in trying to solve this problem for a contemplated 6G15 Stand-alone reverb build. (scheme attached)

I have a 2nd-hand PT with an Ht winding of 376VAC that is not centre-tapped, so I could only use it with a Bridge rectifier or a half-wave rectifier. A bridge rectifier would produce much more B+ than I'd need.

If I use this tranny with a half-wave rectifier in the kind of choke filtering system envisaged on the 6G15 Scheme, would the B+ get down to a ball-park where I could use a lower-powered tetrode or pentode (or dual triode?) to drive the reverb xformer? And how do I figure out the voltage drop with half-wave rectifiers? and why are there three diodes in series in the 6G15 recifier scheme? (Is that merely to ensure the diodes can handle more voltage without self-destructing?)

(If I simply can't use the PT for this purpose I'll stow it for a meaty bass amp build later.)

[Enzo]

Think about how a rectifier circuit works. The AC waveform comes to the diode, and from there only one polarity comes through - assume positive for discussion. The AC waveform peaks at 1.414 times the RMS voltage of the AC. That doesn't change based upon rectification. The AC has no idea how it got sliced into polarities. You now have pulsing DC going up to 1.414 x RMS. That being the same as peak. Ignore for now losses in the rectifier or tube sag.

Whether it is full wave - in which we have 120 positive pulses per second - or half wave - where we have only 60 pulses per second - the output from the rectifier(s) goes to the first filter cap. That cap charges up to the peak voltage and stays there until some load draws it off.

And there is the difference. When you draw current off the power supply, it comes from the cap. The cap is continuously being recharged 60/120 times a second. The cap discharges into the load and its voltage drops until the next pulse recharges it. The more current you draw, the lower that cap voltage drops between erctified AC pulses. That is ripple.

With recharging pulses only 60 times a second, the cap discharges a lot further under load before the recharge occurs. That means more ripple. Ripple in an audio curcuit means hum. SO with half wave, the voltage will be the same, though it will sag more, but there will be increased hum under load. The sag in voltage will be due to the rippled DC averaging slightly lower than DC that stays right at the top. Nothing liek tube rectifier sag though. Transformer winding sag is a separate issue.

SO for a similar level of cleanness on the DC, a half wave rectifier will require more filtration.

In an amplifier, the output stage draws a lot of current, and it varies with demand. SO we almost universally use full wave as a LOT cheaper and simpler than tons more filtration. In the little reverb unit though, nothing draws much power adn the demand is steady since all the stages are class A. SO they decided to use half wave and an extra filter stage.

SO half wave is not going to be the way to drop the voltage. The basic voltage will be the same either way.

That filter in the 6G15 is a cap input filter with a choke crossing the pi. Eliminate the first cap and you have a choke input filter. That puts out a lower voltage than a cap input filter. Look up choke input filters, I think that is convered over at Aiken's. Now going half wave, losing that cap might require you to add it back later in the B+. Lower voltage is separate from filtration itself. You still need enough filter. Even full wave might like something added if that cap is removed. How does 330-340VDC sound? Better than 530?

[tubeswell]

Thanks Enzo

You know that's the kind on answer I was looking for, but I should've probably kept it simpler and asked a more direct question at the start, however I now have so much more information.

Yes I have been studying up on different rectifiers and filters and I understood that a choke filter didn't have that cap in 'front' of the choke. So you're saying that if I take that 40uF and put it in parallel with the other 40uf on the othe side of the choke , it'll help get rid of the ripple more, or should I best be advised to have a Pii RC filter after the choke filter (if I want to keep the 1/2 have recto with this beefy PT)?

[tubeswell]

Enzo -Please check schemes

Hi Enzo

Further to my last post is either of these gong to lower 376VAC to about 340?(in your last post) and will the bottom one be quieter in theory?

[Enzo]

You have two conversations melting together here.

You can make excellent power filters with just caps and resistors, or with a choke added. It can be cap input or choke input. All these methods will reemove the ripple, and thus hum, from the circuit.

The difference between cap input filter and choke input filter is the resulting voltage from the same AC input to the rectifier. That is why I suggested it and why I suggested looking up choke input. We don't see that very often. The result will be lower voltage from the transformer you have.

It was never a matter of one filtered better than the other.

When I suggested removing the first cap and that it might have to go back in later, I was not thinking so much of doubling it up parallel, I was thinking more like your lower example where you added an extra stage of pi filter. The whole entire point of considering the choke input filter was so you could use the transformer you already have that would otherwise result in too high a voltage.

Once you select the filter TYPE to determine the voltage it makes, THEN you can decide how much filtration you will need for clean sound.