So everyone with normal brains knows a 7805 won’t work correctly with a 6.3V transformer, normal value filter cap (i.e. the meanwhile famous 2200 uF cap of the rule of thumb) and normal diodes. Then, to kill time, people mix in Schottky diodes, LDO regs and large value caps to prove a 5V regulator works OK on 6.3V AC!?!?
Such people are probably perfect politicians 😀
Such people are probably perfect politicians 😀
Last edited:
You can even do it with normal diodes and caps. A doubler with a couple of 3300uf works perfectly too.
Fact: a 7805 can work with 6.3VAC.
Apologies for my abnormal brain.
Not with a standard single winding 6.3V transformer and not without tricks. Fact. Apologies accepted 😉
Last edited:
Standard single winding 6.3VAC secondary, two standard rectifier diodes, two standard 3300uF caps, one standard 7805. Where is the trick? The voltage doubler (1908 tech)?
Schottky diodes aren't a (expletive) trick dude. LMAO There's NOTHING exotic about this at all. Just a stubborn person who can't move past 1N4001 diodes.
@jcalvarez To make jean-paul happy, please model a 6V3 primary, and use 1N4007 as bridge and a 2200µF cap so we can see he's right 🙄
I am not your dude. Dudes just do something. All this talk came from the rule of thumb of 2000 uF/A that supposedly was not a good guideline. I then explained where it came from and why it then worked perfectly fine then (although rules of thumb are not advisable). That was of course with a standard diode bridge rectifier circuit and standard parts and choices of the time that the rule of thumb was invented.
Of course I know Schottky diodes, LDO etc. as I design PSUs and only so with currently available parts. I have made quite some LDO PSUs for tube amplifiers filaments as tube guys still try with 78xx and find out it won’t work 😉
Yes make that model. I calculate by head that it won’t work reliably/all the time/under load. Already did that earlier on just like Marcel did.
Of course I know Schottky diodes, LDO etc. as I design PSUs and only so with currently available parts. I have made quite some LDO PSUs for tube amplifiers filaments as tube guys still try with 78xx and find out it won’t work 😉
Yes make that model. I calculate by head that it won’t work reliably/all the time/under load. Already did that earlier on just like Marcel did.
Last edited by a moderator:
I didn't say "MY dude" I said "dude"... As in bro, pal, buddy, guy, kurwa etc... I can also not find any definition of "dude" that reflects just doing something - please clarify.
If you use archaic rectifiers and small caps "That was a good guideline" in the 1970's, then yes, you should use at least an 8V primary. If you can use modern stuff, and maybe go full wave instead of bridge to save one diode drop, you can use 6V3 for bridge or 12VCT FW. I really don't understand what all the fuss is about.
If you use archaic rectifiers and small caps "That was a good guideline" in the 1970's, then yes, you should use at least an 8V primary. If you can use modern stuff, and maybe go full wave instead of bridge to save one diode drop, you can use 6V3 for bridge or 12VCT FW. I really don't understand what all the fuss is about.
What a surprise. The standard circuit of the 2000 uF/A rule of thumb indeed needed the then standard 9V transformer for 5V output. The then expensive and physically very large filter cap could be 2200 uF for 5V 1A. Ripple voltage was low enough, 2 x 1V diode dropout was not felt and things worked reliable. The rule of thumb simply was OK then in that context.
Back to 2022. Forget the 2000 uF/A, forget 78xx with its now 2V dropout voltage. Choose transformer voltages wisely, use low dropout voltage diodes, large filter caps in CLC or CRC for lower ripple voltage and uLDOs. Calculate and design for low loss/low heat/low noise. Design stuff that does not waste energy.
Last edited:
It was designed to fail, by using the wrong components. I got the expected result. No surprises here.
Sorry, that is not true. And repeating the same thing over and over doesn't make it any truer.
Everyone on this thread knows how to multiply 6.3 by 1.414 and subtract 2 volts, as you have been doing. There is nothing very clever in that. However, it doesn't match reality with any mains transformer I have ever used.
(It doesn't match reality with any silicon rectifier diodes I've used either: I've rarely seen a voltage drop bigger than 0.8 volts across any of them. I don't think I've ever seen a full 1 volt; that must be a conservative datasheet upper bound.)
In reality, with every transformer I have tried, a 6.3V AC RMS transformer, and ordinary silicon rectifier diodes, work just fine to power a 7805. The reasons for this have been explained several times in this thread, but you refuse to acknowledge them. Many people have successfully built and used power supplies using 6.3V AC transformers, ordinary 1N400x rectifier diodes, and 7805 regulators, but you refuse to acknowledge that, too. You're literally refusing to acknowledge reality.
Hammond transformers were mentioned in a recent post; they are a particularly good example, as many of them produce far more than their rated voltage. And they have low DCR as well. For example, the last "48V RMS" Hammond transformer I used produced over 75 volts DC after rectification and filtering. According to "by the book, ignoring reality" simulation / calculation, it should have produced less than 65.7 volts DC peak. In fact, it produced 10 volts DC more than the theoretical value.
I also used a smaller Hammond transformer designed for tube amplifiers, with a 6.3V AC heater winding, on one of my last tube projects. The "6.3V" AC winding actually put out more than 7.5V AC RMS - under full load, with the heaters in my tube amp connected and drawing current. I had to insert a series power resistor to drop the tube heater voltage down to 6.3V AC RMS.
That particular "6.3V" winding would have put out over 8.5 volts DC after filtering and diode drops, if I had used it that way.
And that is not at all unusual. As I said, that has been my experience with every 6.3V AC transformer I have ever used.
Meantime, you keep suggesting the use of a 9V AC RMS transformer. In practice these typically put out 14 volt peaks, which translates to over 12 volts DC at full load current after rectification and filtering. If you feed such a high voltage to a 7805, the poor voltage regulator has to dissipate over 7 watts of heat at 1 amp load current. This will overheat and shut down the 7805, and repeated overheating will shorten its life.
Reality is that using a nominally 9V AC RMS transformer with a 7805 regulator abuses the regulator, requires a huge heatsink, and limits maximum DC output current to considerably less than 1 amp.
Reality is that a 6.3V AC transformer, which looks too low a voltage on paper, works perfectly well, particularly if you're not planning to abuse the regulator by trying to draw the maximum datasheet DC current of 1 amp from it.
LTSpice, calculators, and paper-and-pencil only produce the right answers if you put the right numbers into them. At the moment, you are not putting the right numbers into them.
-Gnobuddy
Either there is something wrong with the mains voltage at your place or you need to find a better brand of transformers. It is perfectly normal for the voltage to be a bit too high under light or no load, but more than 7.5 V out of a 6.3 V winding under full load is crazy.
Post #92 says it all. No more words needed. All I wrote was with reference to the time the "rule of thumb" was invented (as explained numerous times) and 7805 had a 3V dropout voltage. Apparently context is difficult.
Even if it works in random cases it still is stupid to use a recent 7805 regulator with 2V dropout voltage on a 6.3V transformer with a reasonable load as a low mains voltage will then cause effects. Since I was educated in this I will notify my former teachers that they did a bad job. I wish you luck with using 7805 😀
Even if it works in random cases it still is stupid to use a recent 7805 regulator with 2V dropout voltage on a 6.3V transformer with a reasonable load as a low mains voltage will then cause effects. Since I was educated in this I will notify my former teachers that they did a bad job. I wish you luck with using 7805 😀
Last edited:
Nope, I have measured the same results in three different countries, over a period of at least 30 years.
In North America, AC line voltages have increased quite a bit in the last few decades. Many transformers whose design was standardized a long time ago put out much more voltage with today's higher AC line voltages. Most 6.3v transformers were designed to power tube heaters, so they are older designs.
This is the same reason why old Fender tube guitar amplifiers now have stunningly high DC voltages. My Princeton Reverb has 445 volts B+, and that is on 6V6 tubes that were originally rated for 350 volts DC max plate voltage.
By the way, Hammond is one of the best brands of transformers you can buy. They are made in Canada, and have been made here for decades.
-Gnobuddy
Last edited:
Ever hear the saying "A little knowledge is a dangerous thing"? Look a little deeper, and the simple textbook calculations intended for beginners to electricity aren't always accurate.
As a very smart fellow on the Aussie Guitar Gearheads Forum used to say, "If theory and practice do not agree, you need more theory."
I trust actual data from my DMM and my 'scope more than simplified textbook formulae that assume ideal behaviour from every component. The simple textbook formula is useful to get a starting point, but you have to check it against reality. Reality is that in real life, a 9V AC RMS transformer puts out too much voltage to be a good match to a 7805 regulator.
And yes, I know where the magic number 1.4142... comes from. Yes, I can do the trig integrals to prove it, and yes, I still know by heart the trigonometric identity used to deal with the sin^2 term. We learned those formulae, and that integral, in high school. You don't need a great deal of education to understand the rather basic mathematics behind a simple bridge rectifier circuit.
Of course the formula works perfectly, as it should - if you measure the actual AC RMS voltage of the transformer and plug that into the formula.
But if you assume that a transformer always puts out exactly its nominal voltage, then the formula gives you the wrong answer, because you haven't used enough theory. You haven't allowed for the fact that people who wind quality transformers try to err on the high side of the nominal ratings.
Since we were discussing rules of thumb earlier, one of my rules of thumb is to expect the DC voltage from a full-wave rectified transformer to be 1.5 times the nominal AC RMS rating. Not (1.414 x Vrms - 2 x Vdiode ). In my experience, this is a lot closer to reality, so I use it in the initial design phase.
Then, of course, you build the thing and measure it, to see if reality matches the rule of thumb or not.
AC line voltage in the USA seem to have been typically around 110V a few decades ago. Then it was standardized at 115V, then later the standard was increased to 120V. I just measured the voltage at the nearest AC outlet, and got 120.7 V RMS, almost exactly what it's supposed to be.
But 120.7 volts is 10.6% higher than the typical 110V that many older transformer designs were intended to operate on. And it is 5% higher than the 115V standard.
On top of that, a good quality transformer is designed to put out its rated voltage at full load. Many 6.3V AC transformers were rated for much more than 1A RMS, in order to feed multiple thirsty tube heaters. When lightly loaded by a 7805 regulator (which tops out at 1 amp) these transformers are not working hard, current output is well below their maximum rating, and as a result, they will put out more than 6.3V AC, even if the incoming AC voltage was exactly 115 volts.
Now, add in the fact that the mains voltage is likely to be 5% - 10% higher than that. And you get the real-world results that I (and several others) have been reporting in this thread.
If theory and practice do not agree, you haven't used enough theory.
-Gnobuddy