Using multiple rectifying bridges from same transformer secondary

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

johnheath

Well-known member
Joined
Jul 31, 2014
Messages
890
Location
Sweden
Hi all...

I have some ideas to build a 8 channel mixer. Each channel strip is built with two 12XX7 tube preamp with an EQ with another 12XX7 tube... each channel has a current draw of 11mA so no major worries there.

However thinking about the construction of PSU some concerns are arising. First of all I'd prefer DC for the heaters and I have a a total of 3,6VDC from 24 12XX7 tubes.

Also I'd prefer to use as few transformers as possible for the heater supply.

Using a secondary of 12VAC rectified would make the filtering tricky to create so I was thinking if the "problem" could be solved by using multiple rectifying bridges tapped from a secondary of 12VAC?

Maybe a complete novice question? But what are the facts behind such a solution?

// John
 

Attachments

  • PSU Test.png
    PSU Test.png
    7.5 KB · Views: 47
Thank you sir

Yes that is one of my questions what really happens? I was wondering if the VAC would remain more or less constant and therefore the voltages after the bridges would be "high" enough to filter with resistors... I guess I could use chokes but that would kind of become a new or at least a similar problem by having a lot of iron all over the place,

Best regards

// John
 
johnheath said:
Hi all...

I have some ideas to build a 8 channel mixer. Each channel strip is built with two 12XX7 tube preamp with an EQ with another 12XX7 tube... each channel has a current draw of 11mA so no major worries there.

However thinking about the construction of PSU some concerns are arising. First of all I'd prefer DC for the heaters and I have a a total of 3,6VDC from 24 12XX7 tubes.

Also I'd prefer to use as few transformers as possible for the heater supply.

Using a secondary of 12VAC rectified would make the filtering tricky to create so I was thinking if the "problem" could be solved by using multiple rectifying bridges tapped from a secondary of 12VAC?

Maybe a complete novice question? But what are the facts behind such a solution?

// John

First some clarification:
Each channel draws 11mA *of high voltage B+*. 
You have a total of 3.6*AMPS* DC *for the heaters of* 24 12AX7 tubes

Rectifying the 12VAC will give you around 16VDC so you can lose a couple volts in filtering.  This works better if your winding can provide more current than needed, say 5A?

Another option is to run the heaters at 6.3V instead of 12.6V.  Use a 12VAC winding then rectify, filter and even regulate if you are going nuts with it.  Again, you may need a higher current 12VAC winding - not sure of the math on that.

In my experience, rectifying and filtering a 6.3VAC winding does not give you enough headroom to properly filter before the voltage gets too low.
 
johnheath said:
Using a secondary of 12VAC rectified would make the filtering tricky to create so I was thinking if the "problem" could be solved by using multiple rectifying bridges tapped from a secondary of 12VAC?
Maybe a complete novice question? But what are the facts behind such a solution?
Rectifying 12Vac produces about 14Vdc after rectification, but you need about 20 000 uF for 3.6A. Then you can drop 1.4V in a 0.33 ohm resistor, with another 20 000uF cap to reduce ripple.
Now, you could split rails like you showed, by scaling the components, e.g. if you create 4 rails, each would have two 4700uF caps and a 1.2ohm resistor. Beware dropping 1.4 V at 3.6A generate some heat
 
mjrippe said:
First some clarification:
Each channel draws 11mA *of high voltage B+*. 
You have a total of 3.6*AMPS* DC *for the heaters of* 24 12AX7 tubes

Rectifying the 12VAC will give you around 16VDC so you can lose a couple volts in filtering.  This works better if your winding can provide more current than needed, say 5A?

Another option is to run the heaters at 6.3V instead of 12.6V.  Use a 12VAC winding then rectify, filter and even regulate if you are going nuts with it.  Again, you may need a higher current 12VAC winding - not sure of the math on that.

In my experience, rectifying and filtering a 6.3VAC winding does not give you enough headroom to properly filter before the voltage gets too low.

Thank you sir

Yes 12VAC to rectify for 12,6VDC... "going nuts"? Naeee... but I do so in single units and it works fine... I have tested with AC but surely you often get hum so DC it is.

Best regards

// John
 
abbey road d enfer said:
Rectifying 12Vac produces about 14Vdc after rectification, but you need about 20 000 uF for 3.6A. Then you can drop 1.4V in a 0.33 ohm resistor, with another 20 000uF cap to reduce ripple.
Now, you could split rails like you showed, by scaling the components, e.g. if you create 4 rails, each would have two 4700uF caps and a 1.2ohm resistor. Beware dropping 1.4 V at 3.6A generate some heat

Thank you sir

Yes this is the math ... and that's why I look for some solutions not to use one rectifying circuit for all the heaters.

So I guees that the overall answer is that having multiple circuits tapped from the secondary would work... as long as the transformer have enough current to supply all the circuits?

Best regards

// John
 
johnheath said:
So I guees that the overall answer is that having multiple circuits tapped from the secondary would work.
But isn't optimal. I would use a single rectifier consisting of 4 cheap Schottkies for 10A/45v followed with four networks Abby already suggested.
 
Heater voltages are relatively low. Diode and wiring drops matter. It is generally more efficient to series the heaters. Same current, higher voltage. Three 12V heaters per audio channel suggests 38VDC. Two or four might be more convenient because those are standard power supply voltages.

Have you ever sat with twenty four tubes??? My Langevin had almost that many and my knees still sweat just thinking about the heat under the desk.
 
moamps said:
But isn't optimal. I would use a single rectifier consisting of 4 cheap Schottkies for 10A/45v followed with four networks Abby already suggested.

Thank you sir

I reread the post from Abbey and realized that I could even work on my english :)

I'll have a look at it  for sure... nothing is yet finally decided ... still on the drawing board.

Best regards

// John
 
PRR said:
Heater voltages are relatively low. Diode and wiring drops matter. It is generally more efficient to series the heaters. Same current, higher voltage. Three 12V heaters per audio channel suggests 38VDC. Two or four might be more convenient because those are standard power supply voltages.

Have you ever sat with twenty four tubes??? My Langevin had almost that many and my knees still sweat just thinking about the heat under the desk.

Thank you Sir

No, nothing is decided yet... actually I am stuck in Africa since early January and thinking about DIY has kept me mentally sane.

I will consider to series the heaters... never done that before though but I suppose it would require a transformer with higher voltage secondary to deal with the number of heaters?

This is the reason for my initial question since I find it not too easy to create the heater solution.  :eek:

Best regards

// John
 
johnheath said:
I will consider to series the heaters... never done that before though but I suppose it would require a transformer with higher voltage secondary to deal with the number of heaters?
There are two caveats regarding using heater series connection. 1st one is obviously if one of the tubes has an open heater, or if the tube is missing for any reason (troubleshooting, awaiting replacement...) a large number of stages won't work. In particular, in a multi-channel unit, having series heater connections that cover several channels is a bad idea.
2nd is that submitting heaters to an elevated DC voltage can induce drift of operating point, noise and distortion. This is due to the non-linearities of the material that insulates the heater from the cathode. These effects may or may not be significant, but you'd better be ready to prepare for a plan B.
I'm not saying it's a bad proposition, just that it will take some experimentation, if only to check the performance is optimized.
 
Wouldn't it be easier to just use a SMPS? Ian (ruffrecords) was using them for heaters, and was reporting good things about it.
For example a meanwell RS-75-12 is for sale for about €20, and it should deliver enough current for this application.
 
wlinart said:
Wouldn't it be easier to just use a SMPS? Ian (ruffrecords) was using them for heaters, and was reporting good things about it.
For example a meanwell RS-75-12 is for sale for about €20, and it should deliver enough current for this application.
Indeed smps make more and more sense. However there's always a concern they introduce HF crap in an otherwise clean environment. In most cases they need to be properly shielded and completed with a passive LC filter.
 
I have used SMPS's in some work for a couple of reverbs before, but to me in this case they also present the samt type of problem with extra need of space... I would rather stick to a bunch of separate transformer in that case to create all wiring for the heaters.

abbey road d enfer said:
Indeed smps make more and more sense. However there's always a concern they introduce HF crap in an otherwise clean environment. In most cases they need to be properly shielded and completed with a passive LC filter.

If that is the case then they are out of my equation...

Best regards

// John
 
johnheath said:
I have used SMPS's in some work for a couple of reverbs before, but to me in this case they also present the samt type of problem with extra need of space... I would rather stick to a bunch of separate transformer in that case to create all wiring for the heaters.
I don't get it, smps are typically smallerthan the equivalent xfmr, and the LC filters are quite small too. A 10mm coil and an electrolytic cap paralleled with a ceramic isn't that big.
 
abbey road d enfer said:
I don't get it, smps are typically smallerthan the equivalent xfmr, and the LC filters are quite small too. A 10mm coil and an electrolytic cap paralleled with a ceramic isn't that big.

No, I know that but they need mounting and fixing the boards for these filtering... I did not mention that I am building most of my stuff without PCB's... so while at it with creating the boards for caps and a resistor I could stick to a couple of toroid transformers with multiple 12VAC secondaries.

I know it might sound stupid to not to acknowledge more modern building techniques but it has worked for so far :)

Best regards

// John
 
abbey road d enfer said:
Indeed smps make more and more sense. However there's always a concern they introduce HF crap in an otherwise clean environment. In most cases they need to be properly shielded and completed with a passive LC filter.

Second problem when someone like to use SMPS for heaters supplying is a cold resistance of heaters at power on, which can be about 4 times lower than nominal. Some SMPS do not wont to start with this temporary (over)load. 
 
PRR said:
Have you ever sat with twenty four tubes??? My Langevin had almost that many and my knees still sweat just thinking about the heat under the desk.

I have never really understood statements of this kind. Mixers are basically class A  line level devices and typically dissipate two or three watts per tube. So 24 tubes is barely 150W which in a box the size of a Langevin is nothing - certainly not enough to make anyone sweat.

Cheers

Ian
 

Latest posts

Back
Top