Current capability full wave bridge rectifier

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Administrator
Joined 2007
Paid Member
Those should be fine providing you are not using a massive low impedance transformer and massive oversized reservoir caps.

In other words a reservoir cap of say 3300uF and a smallish transformer suitably rated for your chosen voltage/current. If you are using say 20,000uF and a 500va transformer then you need something with a higher rating because the charging pulses would be higher than we might normally expect.

1N400x series diodes were often used in lab power supplies of 1amp rating and with no reliability issues at all.
 
Thanks Mooly!

The reservoir cap is 10,000uF though, but the transformer is suitably rated. This whole contraption is for dc heating two 12ax7's, which should draw 0.6A dc.

Isn't the "peak forward surge current 8.3 ms single half sine-wave superimposed on rated load" of 30A enough for the charging pulses?
 
Administrator
Joined 2007
Paid Member
I tried doing a quick simulation to see what it looked like. This gives approx 6.3 volts and 0.6amp load current and the conduction angle of the diodes is pretty low.

First image is the transformer current (which equals each diodes current) and output voltage.

Second image is current in one of the diodes. No series resistance was assumed for the voltage source (and that does make quite a difference to reducing peak currents). The diode is conducting for around 3ms and the peak current is well within the limits for a 1N400x type device which is around 10A repetitive peak current.
 

Attachments

  • Annotation 2019-04-10 145828.jpg
    Annotation 2019-04-10 145828.jpg
    368.7 KB · Views: 218
  • Annotation 2019-04-10 150049.jpg
    Annotation 2019-04-10 150049.jpg
    342.2 KB · Views: 167

PRR

Member
Joined 2003
Paid Member
My opinion: OVER-sizing the rectifier is less hassle than maybe going back to fix it *if* it ever smokes.

While 0.6A DC (0.3A per diode) appears fine for 1N4007, cold tubes are hard loads for much of a Second.

A 2A or 5A bridge is not real expensive, much less than tech-time to do any future repair.

(I should point out that I once earned my bread BECAUSE many designers under-sized rectifiers, 100 or 1,000 starts and fwoom, the job fell to me. As long as I was on salary, that was fine; but my own work always oversized generously. And never failed.)
 
If you have bought a batch of UF4007, then you have the option to parallel 3-4 parts for each 'diode' in the bridge.
If possible, seperate the diode bodies a bit. They will share reasonably well if they have the same temp, and from same batch.

You would need to PSUD2 sim the supply to appreciate the peak current per diode and see how that compares with datasheet limit. Some discussion on how to interpret the datasheet in:
https://www.dalmura.com.au/static/Power%20supply%20issues%20for%20tube%20amps.pdf
 
Last edited:
...10,000uF...UF4007s...suitably rated transformer is...
<snip>
This whole contraption is for dc heating two 12ax7's, which should draw 0.6A dc.
In my opinion, a modern DC switching power supply is a far superior solution for your 12AX7 heaters. It's much smaller, much lighter, much cheaper, and produces much better quality DC, without needing enormous filter caps.

Twelve-volt switching power supplies are everywhere, and will happily power 12AX7s with the heaters wired for 12V.

Or you can find a power supply rated for a little over 6.3V, and drop the extra DC voltage by using a series resistor. A thrift-store 8.2 V Sony power supply, with a series resistor, has been powering all the heaters in my little 2-watt 6AK6 amp for a few years now.


-Gnobuddy
 
Administrator
Joined 2007
Paid Member
Yes it is 8.5 volts but remember that in LTspice that is the peak voltage we set. So the RMS value is just 6 volts.

The cap gives a non sinusoidal ripple and so the big question is what is the real equivalent DC heating power in that resistor. Thats the sort of question that could be endlessly debated ;)
 

Attachments

  • Annotation 2019-04-12 191839.jpg
    Annotation 2019-04-12 191839.jpg
    363.1 KB · Views: 97
...The cap gives a non sinusoidal ripple...
And a lot of ripple at that, about half a volt peak-to-peak from the graph. Even with a huge 10mF (10,000 uF) filter cap.

This is one of several reasons why a small switching power supply is a much better solution for this particular task. Ripple voltage tends to be much smaller. Because today's switchers operate at tens of kilohertz, it takes relatively little additional filtering to remove even that little bit of ripple. For heater power, I've never had to add any additional filtering.

With a switching power supply for power, the thermal inertia of a valve heater completely removes any ripple in the actual heater wire temperature, because of the high ripple frequency.


-Gnobuddy
 

PRR

Member
Joined 2003
Paid Member
...The cap gives a non sinusoidal ripple and so the big question is what is the real equivalent DC heating power in that resistor.....

In Probe: RMS(Vheater). Let run several cycles.

On fingers:

The sawtooth is symmetric between 6.07V and 6.54V. Average is 6.305V.

RMS is something near the root of half of 6.07V^2+6.54V^2 which has numbers 36.845 and 42.77, giving near 6.309V.

The sub-0.1% discrepancy is clearly beneath practical notice.

The other question is whether the half-volt of ripple-spike may be more annoying than 6V of pure hum.
 
(I can see the attraction of an SMPS for this kind of application)
Here is some very interesting data on the subject. Post #32 by ruffrecords on this page ( REW Update ):
ruffrecords said:
...I found the biggest culprit was the 12V dc heater supply...Big toroid (transformer) followed by big bridge and reservoir cap followed by LDO (Low Drop Out voltage regulator) on big heatsink...The toroid is asked for some serious current pulses by the rectifier...
<snip>
The electronics was in a screened box and I even invested in toroids with mu-metal screens which helped but hum was still affecting the noise floor
<snip>
Almost in desperation I replaced the linear heater supply with a Meanwell 12V SMPSU (Switch Mode Power Supply Unit). Bingo, no more hum. REW (Room EQ Wizard) measured the best noise floor I have ever achieved with a tube mic pre.

There it is, a documented case where a cheap, light, small, efficient switching power supply did a better job of reducing hum than a hefty (and expensive) toroidal main transformer, a huge filter cap, and a linear voltage regulator!

This was the missing piece in the golden era of vacuum tubes. Back then, DC heaters meant battery power. Even after silicon rectifier diodes arrived, there was no practical way to produce clean DC power for heaters - even with the biggest and most expensive electrolytic filter caps available at the time. But now, decades later, switch mode supplies are ubiquitous, and they are the perfect solution for clean, hum-free, heater power.


-Gnobuddy
 
Thanks everybody. The DC heaters I wanted to implement are on this site The Valve Wizard

Unfortunately they start to look like a bad plan now, too much ripple spike. I elevated the heaters with 55V and this helped but there is still some hum.

Smps's sounds like the way to go .. does anybody knows a smps that is easy to screw in my amplifier chassis? Can I take just any led power supply or the Meanwell ones?
 
...there is still some hum...SMPS sounds like the way to go...
I suggest first doing a quick test with whatever suitable SMPS you have lying around - use a series resistor to drop the voltage to 6.3V if necessary. The test is to make sure the hum really does go away when you use an SMPS.

For instance, if you want to power a 12AX7 heater, you can wire it for 12V operation and connect to a 12V DC SMPS. If you only have a 7.5V SMPS lying around, wire the 12AX7 for 6.3V operation, then use Ohm's law to estimate the series resistor to use:

R = voltage drop / current = (7.5 - 6.3)/ 0.3A = 4 ohms

Power dissipated in resistor = voltage dropped x current = (7.5 - 6.3) x 0.3 A = 1.2V x 0.3 A = 0.36 watt.

So you could put a 3.9 ohm, 1 watt or 2 watt resistor in series with the 6.3 V heater, and power it with the 7.5 V SMPS.

Remember to ground (or elevate to known DC voltage) one end of the heater voltage - you don't want the DC output of the SMPS to float to some high voltage and damage the heater-cathode insulation in the 12AX7, so do something to prevent that.

The second question you asked - the cheap and ugly solution is to wire the AC prongs of the wall-wart to your incoming switched, fused, AC mains voltage, inside your amp enclosure. Use heat-shrink tubing to make sure there is no exposed metal at mains voltage!

The slightly more expensive solution is to buy the guts of an SMPS, but not packaged in a wall-wart with a built-in AC plug. For example, Digikey has this very affordable 7.5V, 2A SMPS on a bare PCB you can mount inside your chassis: EPS-15-7.5 MEAN WELL USA Inc. | Power Supplies - External/Internal (Off-Board) | DigiKey

You will need connectors to fit that SMPS, and it should be mounted inside a plastic housing - there are deadly voltages on parts of the SMPS board.

I have also seen PCB-mount SMPS modules at Digikey and Mouser. Connections are made with pins sized to fit through typical PCB holes - you can use a generic proto-board PCB. And then there are the chassis-mount SMPS modules, which are already housed in some sort of metal enclosure, like this one: https://www.digikey.ca/product-detail/en/mean-well-usa-inc/EPS-45-7.5-C/1866-1708-ND/7703186

While the more expensive solutions are elegant, I usually just hunt through the junk-box or visit the nearest thift store to find an SMPS with suitable ratings, and just use that.

For me, good engineering is all about performance, not appearance. (It should look good when the case is closed, but I don't care if it's "pretty" inside the box.) But I understand that not everyone feels the same way, particularly in DIY!


-Gnobuddy
 
Thanks everybody. The DC heaters I wanted to implement are on this site The Valve Wizard
The tiny, cheap, reliable, SMPS that are everywhere these days are a pretty new phenomenon. Twenty years ago (very recent, in historical terms!), they were pretty rare, except inside you big desktop PC, or outside your expensive laptop.

It's only in the last ten years that these SMPS were suddenly found in all sorts of new consumer electronics products, and only in the last five years that they have become ubiquitous. Nowadays, most people have a small collection of wall-warts just lying around the house, left over from dead electronics products.

I don't know exactly when Merlin Blencowe (aka the Valve Wizard) originally put that page together. It might have been a few years ago, when cheap ubiquitous SMPS still didn't exist.

I also note that I've found on this very forum that many grizzled DIY valve veterans become quite angry if you suggest that a cheap modern SMPS is in fact the best heater power solution. They react as though you had just insulted their religion. They are emotionally invested in ancient technology, and do not want to even think about the possibility that something new might actually be a perfect complement to the old technology they love. And they want to show off their beautifully twisted and laid out heater wiring. :)

The fact is that, from the engineering (rather than emotional) point of view, AC heater power has never made any sense. Running several amps of audio-frequency current through the guts of a sensitive audio amplifier? Absolute insanity! It's the equivalent of putting 50 kilos of rotting fish in the back seat of your car.

Of course there was a reason why engineers from the golden era of valves were forced to use AC heater power - there was no affordable, practical alternative at the time. If there had been, I am quite sure they would have rejoiced and immediately switched to it.

But lacking anything but noisy AC heater power, they came up with all these clever tricks to manage the stinky fish, sorry, noisy AC: twisted heater wiring, heater wires carefully routed around valves and pushed into corners of the chassis, et cetera. It's like putting three trash bags, one over the other, around the 50 kilos of rotting fish in the back seat. If you bag and seal everything perfectly, it will only stink a little.:eek:

Times have changed. Today, we can use clean cheap DC from an SMPS - and it's not only better, it's even cheaper than paying for the extra transformer windings needed for AC heater power. We have no reason to keep rotting fish in the back seat any longer at all! :)


-Gnobuddy
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.