Externally programmed and adjustable regulated outboard amp power supply - diyAudio
Go Back   Home > Forums > Amplifiers > Power Supplies

Please consider donating to help us continue to serve you.

Ads on/off / Custom Title / More PMs / More album space / Advanced printing & mass image saving
Reply
 
Thread Tools Search this Thread
Old 5th June 2010, 10:35 PM   #1
exeric is offline exeric  United States
diyAudio Member
 
Join Date: Mar 2008
Default Externally programmed and adjustable regulated outboard amp power supply

I'm trying to design a regulated tube PS. I'm new to this but have some basic, but very rusty, background in it. It seems to me that an outboard PS is the way to go to exclude most magnetic interference to any amp. Since it's going to be an outboard PS I'm going to use 2 mains transformers for HV and another transformer for bias supply and, if I can do it, for heaters.

I plan on using industrial control xfms that can be converted beween 220 and 440 AC RMS output. They have 500 VA each and I already have two, so that part is defined and won't change. The reason I got 2 such big honkin xfms is I got this idea in my head. Tell me if this is crazy, because I can't believe that if it isn't crazy that someone hasn't done it before. It just makes so much sense.

The idea is this: since I'm building an outboard PS anyway I should maximize the ability to make it flexible for all my future amp projects. Then I only have to build an umbilical that will connect to any future amp. I know what you are all saying to yourselves at this point. That's crazy! You're going to need to change voltages for different amp designs! This is where I can't believe someone here hasn't done this, and that there must be some reason I don't know about why it won't work. If one uses a typical (say Maida) high voltage adjustable regulator there is always an adjustable resistor in the output voltage sensing. Is it possible to just run a separate wire through the umbilical to a resistor in ones amp? You would most likely have to return it back through the umbilical through a separate wire unless one picks up a ground supplied by the amp.

When you are building an amp one generally knows what HF voltage you will need. To get that voltage first one would have to find the resistor needed in the output sensing divider that is needed to create that voltage. In this case you would just make a test resistor that you plug directly into the correct pins of the umbilical on the amp side of it. If the regulated output is correct then you install the resistor into the amp and connect it to the corresponding pins on the amp connector that connects to the umbilical.

This just seems like a no brainer idea to me. Since that will be two regulated HV power supplies, one for each mains xfmr, one can have two regulated high voltage supplies for any amp you build. You wouldn't have to adjust anything on the PS when you plug a new amp into it because both regulators will automatically use the sensing resistors you (correctly) installed.

Tell me this is crazy. Or not.

Eric
  Reply With Quote
Old 5th June 2010, 11:43 PM   #2
exeric is offline exeric  United States
diyAudio Member
 
Join Date: Mar 2008
Make "hf" "ht"
  Reply With Quote
Old 7th June 2010, 12:22 AM   #3
exeric is offline exeric  United States
diyAudio Member
 
Join Date: Mar 2008
Hi all,
I've been thinking more about this idea of mine and it seems to be a good idea. I'm by nature a non-conformist so it leads me to investigate ideas others don't think about. But that doesn't mean I'm as technically proficient in those areas as I should be. So I hope that having a good idea doesn't mean people will also assume that I know the best way to implement it. Many people here will have more proficiency there than I do. I'm sort of depending on all of you to help with this idea. I'm not greedy about credit for an idea coming to fruition that will benefit all of us.

There seems to be an overriding theme to creating an outboard PS that can be programmed for 95% of all amplifier requirements. And that is this: you have to find the sweet spot between "enough" complexity to handle 95% of requirements while realizing that that other 5% cannot be handled without creating a complex monster. Here's my analysis, for now, of where the priorities should be.

1. Two high power externally programmed regulated power supplies. This will provide suffient flexibility for final amplification. They can be used as one per channel in stereo or one can be used for a separate function, such as screen regulation. Remember, even if you can't think of a good use for that second high power regulated voltage right now this PS is designed to be the last one you design so you should make it powerful and somewhat redundant for future needs you can't think of now. However, that second HT is a personal choice and can be reasonable done away with by reasonable people if they elect.

2. A programmable voltage for the heaters that is also developed through a sense resistor in the external amp returning to the regulator in the PS.
Ideally you would like to have the same ability as with HT to have at least two different voltages available but I think that is going overboard here. So if more than one heater voltage is required in a given amp design then one simply developes it inside the amp from the heater voltage you already have. For instance if you have 6.3 volts and need a separate 5v heater supply then you drop 1.5 volts. You could just use a resistor in that case to drop it and still have a reasonable wattage being dropped through it. If the wattage that would have to be dropped is excessive then you drop it though a regulator in the amp. You want an outboard PS that is reasonable, not excessive and overly complex. There's some wiggle room here but common sense comes into play. Part of the common sense is using a dedicated high VA transformer in the PS. Perhaps putting out 12 volts P to P.

3. You should have a negative supply for bias. My understanding is that typical negative bias for fixed bias output stage vary between around -11V (el84) to around -150V . This is a real problem to program from one xfmr and maintain reasonable efficiency with a regulator. I haven't figured out how to do that yet and stay within reasonable bounds of complexity. If you are using a follower circuit to drive the final stage it will draw some current, perhaps 5 to 10 ma per channel up to 10 to 20 ma for push pull. That's a total of 10 to 20 ma draw. That isn't much for a regulator to regulate. If any of you have any ideas of how one would design this for flexibility let me know. Maybe there's a way to do this by combining it with the heater current for the lower bias voltages, say -11v to -50v. This part of the design I haven't figured out. One direction combines an overly complex solution while the other direction does not provide enough flexibility for different amp designs. Any help or good ideas here would be gratefully appreciated. It should include a way to utilize more current draw to provide better regulation, if possible.

4. Finally one should have a fixed 12 volts that is non programmable that is provided by the PS. One doesn't neccessarily have to use a regulator for this, just a good filter in the PS. This will be used for sandy devices. If a different voltage is required, say 5 volts, then it should be developed in the amp. I think most amp designs will be able to avoid that.

5 Optional. It would be nice to be able to control automatically how the two output windings on each HT transformer are wired together. This would include an automatic way to orient ground to those connections. Here are 4 examples:

1.Secondaries connected in parallel -340V peak out, assuming 240 rms. Ground connected to one end of the paralleled secondaries.

2.Secondaries connected in series -680V peak. Ground connected between the two windings.

3. Secondaries connected in series -680V peak. Ground connected on one end of the two windings connected in series.

4. Finally, both HT transformers connected together all winding in series.
Ground on the bottom. This will provide 1360 volts unregulated for transmitter tubes.

One would have to use high voltage relays, solid state or otherwise to implement this option. You wouldn't want that connection to go through the umbilical. You could use a variety of schemes to enable the correct relays to select the connection you required. Again, you would hardwire the selection of the connection method through the amp you are building. I don't have the expertise to implement this idea but I'm pretty sure there might be individuals here who do. Again, any help would be greatly appreciated.

The number of wires in the umbilical that would be required would be divided into 3 categories. I used the worst case scenario where individual grounds are paired with each voltage carrying wire.

High Tension: 6 each, including grounds for each. 4 for high power and 2 for low power (biasing)

Low Tension: 4 each, heaters and 12volts fixed.

Control: 2 for each selectable voltage - 8 total (3 HT, 1 LT)
Optional :2 wires for binary control of relays connecting HT secondaries to each other.

Thats a total of 20 wires and I know I'm missing some. Also I need to study the grounding article here some more to find out the best way to use grounding. It would be pretty complex and one can't just assume it will all work. In fact you can assume the first attempt WON'T work! That's why one has to persevere in these things. It would be nice to be able to eliminate some of those individual ground wires.

Last edited by exeric; 7th June 2010 at 12:35 AM.
  Reply With Quote
Old 7th June 2010, 02:22 AM   #4
diyAudio Member
 
Sch3mat1c's Avatar
 
Join Date: Jan 2003
Location: Milwaukee, WI
Send a message via ICQ to Sch3mat1c Send a message via AIM to Sch3mat1c
Bah, linear? Go with switching.

My senior project was a programmable power supply. Though low voltage, the design can be adapted to any reasonable output. I don't have a webpage for it but the schematics are available here.
http://myweb.msoe.edu/williamstm/Ima...409_Schem1.png
http://myweb.msoe.edu/williamstm/Ima...409_Schem2.png
http://myweb.msoe.edu/williamstm/Ima...409_Schem3.png
http://myweb.msoe.edu/williamstm/Ima...409_Schem4.png
http://myweb.msoe.edu/williamstm/Ima...409_Schem5.png
The microcontroller provides PWM outputs, which are filtered for the control voltage setpoints the TL598 reads. For more outputs, you'd want to use chained TL598s (so the oscillators are synchronized), one for each set of linked voltages, and use a microcontroller with lots of DAC outputs instead.

On the other hand, if you want plug-and-go portability, you can put the voltage feedback resistors inside the amp, so the feedback ratio is set by the amp. This can be used with switching or linear supplies, with no need for a microcontroller.

You could also use the same approach with a microcontroller, replacing the feedback networks with a single serial EEPROM. This way, you could store the actual voltages and currents digitally and read them with only two wires.

Tim
__________________
Seven Transistor Labs
Projects and Resources
  Reply With Quote
Old 7th June 2010, 03:49 AM   #5
exeric is offline exeric  United States
diyAudio Member
 
Join Date: Mar 2008
Well, I've been avoiding switching power supplies because I have a learning curve I need to overcome. However, I'm open to that direction because it is much more efficient. Also, the problem of bias supplies which can have so much variability really exacts a cost in either efficiency or complexity. So it seems complexity is involved in any solution and it would appear that switching supplies would improve on efficiency in the other supplies also. So I would be open to it but I would just have to get up to speed on it. Any good books on switching power supplies that you can suggest that are not extreme on the math side of it?

I think I would be much more interested in a plug and play solution that uses resistors within the amp for feedback, instead of a microcontroller and EEPROM . There is a good reason for that besides reducing complexity. One can then experiment with rheostats installed in the amp to find the optimum operating point for any circuit. That's a really big advantage. Besides, with 5 or 10 wires (depending on how you do grounding) whats 5 or 10 additional wires added to the mix. No big deal it seems to me for the added flexibility and reduction in complexity.

Eric
  Reply With Quote
Old 7th June 2010, 06:00 AM   #6
exeric is offline exeric  United States
diyAudio Member
 
Join Date: Mar 2008
Quote:
Originally Posted by exeric View Post
I think I would be much more interested in a plug and play solution that uses resistors within the amp for feedback, instead of a microcontroller and EEPROM . There is a good reason for that besides reducing complexity. One can then experiment with rheostats installed in the amp to find the optimum operating point for any circuit. That's a really big advantage. Besides, with 5 or 10 wires (depending on how you do grounding) whats 5 or 10 additional wires added to the mix. No big deal it seems to me for the added flexibility and reduction in complexity.

Eric
I should add that one could make a separate panel, say with 5 rheostats that are labeled for 5 different voltages. One then just wires it with a connector that mates to the umbilical in place of the amp. One could then use it to set up the voltages for any amp you're building. Disconnect the panel and read the resistances for each rheostat. Then get resistors identical to those resistances and wire them to the corresponding pins on the amplifiers umbilical connector. All voltages are then programmed.

Probably I confused people about what the good idea actually was. It wasn't in my implementation using linear supplies. (Though that is what inspired me). The good idea part of it is the plug and play aspect of it where you can use the same power supply over and over again for any amp you build. Actually using switching supplies to implement it as Tim suggested makes a little more sense.
  Reply With Quote
Old 7th June 2010, 07:14 PM   #7
exeric is offline exeric  United States
diyAudio Member
 
Join Date: Mar 2008
Here's my conclusions so far about implementing a plug and play PS solution.

It's obvious its a good idea but I don't have the expertise presently to implement it. Some person or persons here will have to take it to the next level. However there are "broad strokes" that should be in any solution for whoever takes it to the next level.

1. Enough different externally programmed voltages for any foreseeable amp construction without going overboard. My present estimate is at least 5 different voltages with one of them being a fixed voltage for sandy devices. There's wiggle room here.

2. There should be minimal complexity that needs to be added onto the amp one is constructing to implement this idea. If you have to go and add a bunch of extra stuff to any new amp you build just to use this external PS it doesn't make any sense. This probably eliminates an EEPROM installed in every amp just to reduce the wires in the umbilical. Resistors are the easiest to install in a linear supply. For now I'm taking it on faith from Tim that a simple resistor in a feedback network can also be used for SMPS power. I hope he's right because if he isn't it just won't be a practical idea using a SMPS topology. Who wants to add a bunch of extra junk to every new amp you build just to be able to use your hot, sexy new PS. Not me. Better to build it into the PS itself. One can justify it because you will be using constantly if it simplifies building any given amp. The ultimate goal is simplifying our lives, not making it ever more complex.

3. I thought about that last goal. It means that the PS should have a way of reading the voltages being put out. I originally avoided that because it made it seem more like a stand alone standard PS. However it makes a lot of sense to include both readouts and a way to adjust the voltages on the power supply AT THE POWER SUPPLY. Its much more convenient that way. That way you can adjust the voltages optimally at one convenient location that doesn't change for each amp. This means instead of having an external panel you temporarily connect to the amp end of the umbilical the voltages are instead adjusted and the resistances read directly at the PS. This requires a relay that can select between the external amp resistors and the internal PS selectable resistors. When the PS is electrically powered down it should fail safe to external amp resistors, or if there are none installed then to a safe output condition. The reason for this is because if you just want to play an amp that is already built you don't ever, ever, ever want to power it up with randomly selected voltages from the last amp build you did. There should be the equivalent of a held relay that you actually have to switch on after power up to use the adjustable internal resistors. Again, a little extra complexity on the PS but fantastic convenience and simplicity and safety for just building or playing amps.

4. I'm really thinking a switch mode PS is the best way to implement this whole idea for the wide variances that are required of the different voltages. It just to make sense.

I don't have the expertise or knowledge to go all the way with this but still think its a fantastic idea. (I did order some books though and I stayed at a Holiday Inn so maybe that helps. ) I hope the people who do have the expertise will chose to go after this idea. Many many people let their egos get involved too much and can't stand implementing others ideas. I'm man enough to say I don't have the expertise to build this presently. I hope others here are man (or woman) enough to say, "no, it wasn't my idea but its a damn good idea and I'll help with it for the good of the DIY community. Enough preaching already.
  Reply With Quote

Reply


Hide this!Advertise here!
Thread Tools Search this Thread
Search this Thread:

Advanced Search

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are Off
Pingbacks are Off
Refbacks are Off


Similar Threads
Thread Thread Starter Forum Replies Last Post
Adjustable Regulated B+ power supply question Captn Dave Power Supplies 9 21st October 2009 06:13 AM
Outboard power supply grounding ? woody Power Supplies 5 3rd May 2008 10:24 PM
Looking for Outboard Power Supply for B & K Sonata MC-101 Preamp dtm1962 Swap Meet 1 23rd January 2008 05:54 AM
outboard power supply elaar Parts 2 3rd April 2006 05:48 PM
Adjustable regulated power supply Ford_V6 Solid State 2 1st April 2003 06:03 PM


New To Site? Need Help?

All times are GMT. The time now is 08:14 PM.


vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2014 DragonByte Technologies Ltd.
Copyright 1999-2014 diyAudio

Content Relevant URLs by vBSEO 3.3.2