I'm getting my stuff together for a pair of new KT88 SE monoblocs and have come to the PS.
The most basic question is SS or tube rectification. Tubes have the advantage of slow turn on of B+. SS is smaller and less expensive and I already have some 35A 600V bridges.
The next question is regarding chokes. If I go with tube rectification how do I go about sizing the choke?
Some basics about the circuit-
* KT88 SE monoblocs
* 6N1P driver
* 400V B+
* 75mA Bias
I was thinking of using Hammond 272BX PT and Hammond 1628SE OT. My first thought regarding chokes was Hammond 193D (8H, 150mA, 75 ohms, 600VDC).
Can someone with more experience (I have zero with designing/building tube rectification) help?
1- Tube or SS
2- If tube, choke size?
3- Choke even if SS?
4- Smoothing cap type and size?
Thanks!
The most basic question is SS or tube rectification. Tubes have the advantage of slow turn on of B+. SS is smaller and less expensive and I already have some 35A 600V bridges.
The next question is regarding chokes. If I go with tube rectification how do I go about sizing the choke?
Some basics about the circuit-
* KT88 SE monoblocs
* 6N1P driver
* 400V B+
* 75mA Bias
I was thinking of using Hammond 272BX PT and Hammond 1628SE OT. My first thought regarding chokes was Hammond 193D (8H, 150mA, 75 ohms, 600VDC).
Can someone with more experience (I have zero with designing/building tube rectification) help?
1- Tube or SS
2- If tube, choke size?
3- Choke even if SS?
4- Smoothing cap type and size?
Thanks!
What he said.
The 8H choke will be fine as long as it handles the current. Typically, anything from 1 to 20H will be used. IMHO, the higher values (>5H) are only useful for choke input; any filtering advantage with the larger values can be handled better (i.e., a stiffer rail as a result...[hey stop giggling!]) with a larger second cap.
If you don't want to wind one yourself, you might only find larger (10H range) values in stock. That's fine. If you do want to wind, likely you won't want to put on 5000T, so you'll do a 1-2H size. Fine for CLC but no good for choke input unless you have a high current demand.
SS vs. tube: slow turnon is a bunch of bull (cathode stripping, etc.), and only works on the first shot anyway. If you turn it off then back on within a few seconds, it comes back in the same time as with SS. Tubes burn a very large amount of heat, 5V at 3A is 15W just for heater, plus the 10-40W (SWAG) plate dissipation caused by the voltage drop. Regulation is also horrible in comparison.
Tim
The 8H choke will be fine as long as it handles the current. Typically, anything from 1 to 20H will be used. IMHO, the higher values (>5H) are only useful for choke input; any filtering advantage with the larger values can be handled better (i.e., a stiffer rail as a result...[hey stop giggling!]) with a larger second cap.
If you don't want to wind one yourself, you might only find larger (10H range) values in stock. That's fine. If you do want to wind, likely you won't want to put on 5000T, so you'll do a 1-2H size. Fine for CLC but no good for choke input unless you have a high current demand.
SS vs. tube: slow turnon is a bunch of bull (cathode stripping, etc.), and only works on the first shot anyway. If you turn it off then back on within a few seconds, it comes back in the same time as with SS. Tubes burn a very large amount of heat, 5V at 3A is 15W just for heater, plus the 10-40W (SWAG) plate dissipation caused by the voltage drop. Regulation is also horrible in comparison.
Tim
what Sch3mat1c forgot to sait is that tube rectification sound better.
the best power supplie you can do is a LC (big choke, small caps)
then, for SE, you need better regulation, so LCLC (big choke small caps)
I try the opposite (big caps, small choke) but it`s not as goot to my ear...
Martin
the best power supplie you can do is a LC (big choke, small caps)
then, for SE, you need better regulation, so LCLC (big choke small caps)
I try the opposite (big caps, small choke) but it`s not as goot to my ear...
Martin
You may not care at this stage but SS rectification generally sucks sonically. Once you've built it you may want to investigate how vacuum rectifiers, damper diodes and mercury compare. Vacuum devices generally drop a lot of volts so comparisons won't be fair but you'd get the idea.
Chokes usually help the sound (apart from doing filtering); in particular the bass. Using SS rects you may be able to get low ripple with large caps but often the sound is better with small caps and large chokes. As you may notice i generally avoid making strong statements and comparing chokes to sliced bread. Many PS chokes don't sound good at all, probably due to high parasitic capacitances and high dc resistance. Sonically i was about to give up on them until i heard some decent examples.
Finally, you may want to investigate a choke input supply. For some strange reason these are not very popular although usually sound better.
Chokes usually help the sound (apart from doing filtering); in particular the bass. Using SS rects you may be able to get low ripple with large caps but often the sound is better with small caps and large chokes. As you may notice i generally avoid making strong statements and comparing chokes to sliced bread. Many PS chokes don't sound good at all, probably due to high parasitic capacitances and high dc resistance. Sonically i was about to give up on them until i heard some decent examples.
Finally, you may want to investigate a choke input supply. For some strange reason these are not very popular although usually sound better.
With SS diodes, the B+ might rise to 500V before the output tubes warm up. So make sure your filter caps are rated to at least 600V or so.
Valve rectifiers are much more fun to work with. How often do you see SS diodes glow.
If you are going to use a valve, use a GZ34 for slow warm up and low voltage drop. As it is class A you don't need to worry too much about regulation, the current draw is reasonably constant. If you use a tube, the maximum first cap will be 47uF or less.
I would use a choke whatever topology you go for, just make sure it is rated for the current.
Valve rectifiers are much more fun to work with. How often do you see SS diodes glow.
If you are going to use a valve, use a GZ34 for slow warm up and low voltage drop. As it is class A you don't need to worry too much about regulation, the current draw is reasonably constant. If you use a tube, the maximum first cap will be 47uF or less. I would use a choke whatever topology you go for, just make sure it is rated for the current.
OK, then; tubes all the way
It sounds like the consensus is that tubes would be the way to go. Even though I haven't done tube rectification before I find myself happy with this result and I'm happy to give it a go.
There are other chokes available with similar specs (current capacity, inductance etc.) as slightly lower cost (about 1/2) but I think I want the iron up top with the tubes (gotta show off that expensive hardware!) and those open frame chokes are ugly !
So it seems I need to-
* use a choke with plenty of current capacity for both the driver and output valve
* use as high an inductance as I can afford (10 or 12 H)
* and balance the above with the lowest DC resistance I can find
correct?
Now for the PS caps-
* for 400V B+ it seems the caps should have 600V minimum capability
* what about capacitance?
Thanks to everyone for their input! I'm looking forward to this project.
It sounds like the consensus is that tubes would be the way to go. Even though I haven't done tube rectification before I find myself happy with this result and I'm happy to give it a go.
There are other chokes available with similar specs (current capacity, inductance etc.) as slightly lower cost (about 1/2) but I think I want the iron up top with the tubes (gotta show off that expensive hardware!) and those open frame chokes are ugly !
So it seems I need to-
* use a choke with plenty of current capacity for both the driver and output valve
* use as high an inductance as I can afford (10 or 12 H)
* and balance the above with the lowest DC resistance I can find
correct?
Now for the PS caps-
* for 400V B+ it seems the caps should have 600V minimum capability
* what about capacitance?
Thanks to everyone for their input! I'm looking forward to this project.
Hi,
Correct.
Choke input is also my prefered way to go about it but it just wastes an awful lot of volts.
Use chokes that are designed for this app. "Swinging chokes" is what we call them over here.
Enough for the task at hand, it's basically a constant current circuit. You won't be able to cancel ripple so that rail should be clean for the word go.
More importantly use good quality caps, think MKP, ASC oilers, BGs if you have deep enough pockets.
Cheers,
No one mentioned the RFI pollution those SS diodes spit out?
Correct.
Choke input is also my prefered way to go about it but it just wastes an awful lot of volts.
Use chokes that are designed for this app. "Swinging chokes" is what we call them over here.
Enough for the task at hand, it's basically a constant current circuit. You won't be able to cancel ripple so that rail should be clean for the word go.
More importantly use good quality caps, think MKP, ASC oilers, BGs if you have deep enough pockets.
Cheers,
No one mentioned the RFI pollution those SS diodes spit out?
A LA CARTE........
Hi,
Oh...as for the circuit proper:
I'd replace the 6N1P and use a ECC88 of decent pedigree, ommit the cathode bypass on that one too.
Use a Mundorf silver in oil coupling cap for the 0.220µF (600VDC or thereabouts), use a BG 220µF/160VDC not 63VDC.
If the 100K volctrl is a pot put a gridleak of about 470K to 1M behind it too and pray the wiper of that pot won't fry on you.
Try a carbon film, they tend to sound better than those plasticky thingies.
Put a 10K (carbon comp.) gridstopper on the 6N1P or ECC88's grid unless you like to listen to birdies...
A 1K carbon comp. as a gridstopper on the KT88's grid would help overload recovery on that KT too....
Voila, my two pence worth of free advice.....
Cheers,
Hi,
Oh...as for the circuit proper:
I'd replace the 6N1P and use a ECC88 of decent pedigree, ommit the cathode bypass on that one too.
Use a Mundorf silver in oil coupling cap for the 0.220µF (600VDC or thereabouts), use a BG 220µF/160VDC not 63VDC.
If the 100K volctrl is a pot put a gridleak of about 470K to 1M behind it too and pray the wiper of that pot won't fry on you.
Try a carbon film, they tend to sound better than those plasticky thingies.
Put a 10K (carbon comp.) gridstopper on the 6N1P or ECC88's grid unless you like to listen to birdies...
A 1K carbon comp. as a gridstopper on the KT88's grid would help overload recovery on that KT too....
Voila, my two pence worth of free advice.....
Cheers,
Re: A LA CARTE........
Wow! Thanks for the input. I'll look at those changes. I was thinking of an Auricap for the .22uF signal cap but I'm tempted also by Jupiter beeswax. I'll have to start counting pennies!
And I was thinking of Angela polyprop in oil caps for the PS (40uF 515V). They look like a pretty good deal.
I'm planning on doing these as monoblocs and I really would like to eliminate the pot altogether and use a valve pre-amp (currently under construction) for the input.
My first thought was putting a largish resistor to earth in place of the pot but I haven't worked out the details and value yet. Not being versed in these things I have to pick a schem that is similar and then do the math.
I'm getting away from the topic I started this thread with, but I truly appreciate the advice!
fdegrove said:Hi,
Oh...as for the circuit proper:
...Voila, my two pence worth of free advice.....
Cheers,
Wow! Thanks for the input. I'll look at those changes. I was thinking of an Auricap for the .22uF signal cap but I'm tempted also by Jupiter beeswax. I'll have to start counting pennies!
And I was thinking of Angela polyprop in oil caps for the PS (40uF 515V). They look like a pretty good deal.
I'm planning on doing these as monoblocs and I really would like to eliminate the pot altogether and use a valve pre-amp (currently under construction) for the input.
My first thought was putting a largish resistor to earth in place of the pot but I haven't worked out the details and value yet. Not being versed in these things I have to pick a schem that is similar and then do the math.
I'm getting away from the topic I started this thread with, but I truly appreciate the advice!
Hi,
Since you appreciate it....
Yeah...I lost track of the fact that you plan on building monoblocks, my bad.
You could just replace that pot (or whatever it was meant to be) by a resistor of the same value. In casu 100K, a Dale RN55 or similar would do nicely + a 10K Dale gridstopper.
I assume you know the ropes on how to use a gridstopper properly: resistor body at one end as close to the grid connection solder lug as possible without breaking the resistor's endcap loose, gridleakresistor at the other.
Having monoblocks with a volctrl. in them kind of defeats the purpose anyway...........
Just in case: should the preamp overdrive the input tube of the amp (the unbypassed cathode is the culprit), try a little larger cathode resistor. That should cure it.
The Zout of the unbypassed ECC88/6DJ8 is similar to the bypassed 6N1P but the current feedback should reduce distortion of the entire amp a little.
The ECC88/6DJ8 has better much transconductance and that makes a world of difference with such a design.
I think those are ASC oil and polyprops, not sure.
The Jupiters should go well with the UL KT88 according to my little cookbook especially if you can find a meter or so of 24AWG solid core silver wire (enamelled is fine) and put it in as hook-up wire
After all that it's all down to the quality of the output iron but I reckon these should be great little amps.
Happy building,
Since you appreciate it....
Yeah...I lost track of the fact that you plan on building monoblocks, my bad.
You could just replace that pot (or whatever it was meant to be) by a resistor of the same value. In casu 100K, a Dale RN55 or similar would do nicely + a 10K Dale gridstopper.
I assume you know the ropes on how to use a gridstopper properly: resistor body at one end as close to the grid connection solder lug as possible without breaking the resistor's endcap loose, gridleakresistor at the other.
Having monoblocks with a volctrl. in them kind of defeats the purpose anyway...........
Just in case: should the preamp overdrive the input tube of the amp (the unbypassed cathode is the culprit), try a little larger cathode resistor. That should cure it.
The Zout of the unbypassed ECC88/6DJ8 is similar to the bypassed 6N1P but the current feedback should reduce distortion of the entire amp a little.
The ECC88/6DJ8 has better much transconductance and that makes a world of difference with such a design.
I think those are ASC oil and polyprops, not sure.
The Jupiters should go well with the UL KT88 according to my little cookbook especially if you can find a meter or so of 24AWG solid core silver wire (enamelled is fine) and put it in as hook-up wire
After all that it's all down to the quality of the output iron but I reckon these should be great little amps.
Happy building,
SHiFTY said:...
Valve rectifiers are much more fun to work with. How often do you see SS diodes glow.
I would use a choke whatever topology you go for, just make sure it is rated for the current.
OK, now that I've decided on valve rectification how about using a 5U4G? No particular reason other than they look cool
with the shape echoing the KT88!
Re: Re: A LA CARTE........
A search on ebay using the keywords "run capacitor" will get you essentially the same things for much less money...Angela's are just rebadged ASC motor run capacitors.
Sherman said:
A search on ebay using the keywords "run capacitor" will get you essentially the same things for much less money...Angela's are just rebadged ASC motor run capacitors.
Re: Re: Re: A LA CARTE........
Are motor run caps a bad thing? Should I not use them in this PS? I can always go with BlackGate. (Or maybe those lovely robin's egg blue Xicons! )
Seriously though, this project is going to cost so much that an extra 10 or 15 bucks for caps will be a small percentage. I want to have good quality components and am willing to pay a bit more if the ones I'm thinking about won't work.
My budget is not unlimited though! (Note that I'm planning on Hammond iron, not Tamura!)
pedroskova said:
Are motor run caps a bad thing? Should I not use them in this PS? I can always go with BlackGate. (Or maybe those lovely robin's egg blue Xicons!
)Seriously though, this project is going to cost so much that an extra 10 or 15 bucks for caps will be a small percentage. I want to have good quality components and am willing to pay a bit more if the ones I'm thinking about won't work.
My budget is not unlimited though! (Note that I'm planning on Hammond iron, not Tamura!)
Re: Re: Re: Re: A LA CARTE........
There is nothing wrong with motor run caps, some are really good.
Black Gate caps for a tube supply are going to be over $100 each not $10 or $15.Sherman said:
Are motor run caps a bad thing? Should I not use them in this PS? I can always go with BlackGate. (Or maybe those lovely robin's egg blue Xicons!
Seriously though, this project is going to cost so much that an extra 10 or 15 bucks for caps will be a small percentage. I want to have good quality components and am willing to pay a bit more if the ones I'm thinking about won't work.
My budget is not unlimited though! (Note that I'm planning on Hammond iron, not Tamura!)
There is nothing wrong with motor run caps, some are really good.
Re: Re: Re: Re: A LA CARTE........
Motor runs are very good as PSU caps. I would choose them over BG's...not only for sonic preferences, but because of reliabilty issues as well. I only pointed out ebay because the ASC's, GE's, and Aerovox's are the equivalents of Angela's, but cheaper.
On the bright side, Angela is cheaper than my local HVAC supply house
PS - Angela says that the oil may act as a vibrational damper. Not so - a number of people have reported that they can "sing" if you pump enough signal thru them. I think Jack Elano(sp) was the first to note this.
EDIT: Stay the hell away from motor START capacitors. In spite of sharing the term "motor", they are entirely different beasts. The motor start is not meant for continual DC duty.
Sherman said:
Are motor run caps a bad thing? Should I not use them in this PS? I can always go with BlackGate. (Or maybe those lovely robin's egg blue Xicons!
Motor runs are very good as PSU caps. I would choose them over BG's...not only for sonic preferences, but because of reliabilty issues as well. I only pointed out ebay because the ASC's, GE's, and Aerovox's are the equivalents of Angela's, but cheaper.
On the bright side, Angela is cheaper than my local HVAC supply house
PS - Angela says that the oil may act as a vibrational damper. Not so - a number of people have reported that they can "sing" if you pump enough signal thru them. I think Jack Elano(sp) was the first to note this.
EDIT: Stay the hell away from motor START capacitors. In spite of sharing the term "motor", they are entirely different beasts. The motor start is not meant for continual DC duty.
Re: Re: Re: Re: Re: A LA CARTE........
Thanks for that. Although I once had a motor with a centrifugal switch and starting capacitor, that thought hadn't occurred to me.
pedroskova said:EDIT: Stay the hell away from motor START capacitors. In spite of sharing the term "motor", they are entirely different beasts. The motor start is not meant for continual DC duty.
Thanks for that. Although I once had a motor with a centrifugal switch and starting capacitor, that thought hadn't occurred to me.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.