|
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
63.155.48.243
I'm still an amature at this and I can't figure out PSUD-2, so I'm wondering if this is an optimal PS for my amp. I do understand that you are basically listening to your power supply. I was taught this around eight years ago on this forum, so I see it is absolutely necessary to have a good foundation to build upon.
I've been messing around with rebuilding vintage amps for a few years and I would like to finally try something new. This will be my first amp build and I don't like wasting money if it is not the most efficient, fastest, and best sounding PS I can build (on a budget) without any PS buzz or hum. I have very sensitive ears and love to hear the details, but can't stand any harshness. The recommended PT is a Triode Electronics PA060. The B+ & B+1 are both 330 VDC. I won't be using any lytics. They'll be either polys and/or oilers.
I've included the PS schematic. I don't doubt that the design sounds good, but I'd like for my first to be exceptional. This may be as good as it gets.....I don't know. Help a first-timer be a success at his first build. Thanks!
Larry D.
Follow Ups:
Applying a 15% current step to the Finals at 4.1 seconds. we can simulate your supply. It is wasteful as a design. Unfocussed. It has a dynamic Z of 296 Ohms, and it settles in ( yawn ) 385 mS.John Swenson who did measurements of recovery times and posted it previously on this Forum sez, "50 mS. or less is nice".
My suggestion, the second simulation, with the same 15% current step to the Finals, settles in about 50 mS., over seven times faster, and has a dynamic Z of 229 Ohms, over 20% lower, using tube rectifiers that are very common and inexpensive compared to two 5AR4s.
Specific off-the-shelf industrial parts of my suggested supply is as follows : One Hammond 372KX power trannie, 600 VCT at 575 mA., 23 Ohm HV secondary, Shared L1, one Signal Transformer Co. CL-1-2, 72 mHY at 1.5 A at 1.4 Ohms, 20 uF L1, .
From the shared L1/C1, you go to a "Y" to each channel, with a L2 and C2 to each P-P output section. L2 becomes the new Edcor 2 Hy 200 mA. 21 Ohm device, and C2 is 50 uF to feed the finals.
For the front end, one can use either a third choke or a R-Dropper to get adequate LOW ripple. Speaking of ripple, my above-suggested "fast dynamic recovery" supply has about 1 VAC of ripple, and in a P-P amp, you will NOT detect that feeding the finals. My supply comes in at 323 VDC assuming a 120 VAC line, if your home's line 122 VAC, about 2% more, you will have your 330 VDC target to the finals.
Your supply uses two 40 Ohm DCR chokes in series, (80 Ohms total) and that DCR (40 Ohms in a choke to the Finals) KILLS the amp dynamically. Both of my suggested chokes in series total only 23 Ohms, and that will do the job nicely. Your 220 uF caps to the Finals will deliver energy out-of-time, and hog current from the Finals when the music instantaneously needs it, as per JLH's concise posts on this in about 2008 to 2010.
My suggestion is off the top of my head. It should boggie woogie with the BIG boys very very well Lektrik.Jeff Medwin
Edits: 07/10/12 07/10/12 07/10/12
I will try this first and if I like it (and I probably will) I'll leave it at that. My whole point in beginning this thread was that after initially looking at the PS schematic, it simply looked too slow with the 220uf caps (and I didn't want to use electrolytics anyway).
If I'm going to bother building an amp it needs to be foundationally 'right' at the power supply. I can mess around with the amp section (breadboarded) after I get the PS built right.
Thanks for the schematic to build by. I appreciate the effort you put into it.
Larry D.
I use hybrid bridge with two 6AX4GT damper diodes. Choke input design using same chokes as you spec'ed. 4HY-30uF-4HY-to 2x Cree diode to 2x 100uF (one cree diode & one 100uF for each channel. I would then feed each channel front end with same 1.5HY choke and a 22uf capacitor. Unless you have a very high current demand front end 22uf is plenty of filtering.
Capacitor sizes simulate well for fast recovery without voltage overshoot, yet low noise (assuming 150-200ma for total average current draw). The Cree diode & 100 uf for each channel isolates the last capacitor from the other channel. Use a low DCR power transformer of about 10-20 ohms on B+ winding.
Neff,
I had a PP 40KG6 amp that used the 6AX4 damper diode. I liked its sound really well, but sold it because I wasn't thrilled by the build quality.
What size (amp/volt) of Schottkys would you use in this application? Two 100uf caps using polys or oilers? Seems like a lot of capacitance. Is all that necessary?
Man, I've got to get working on figuring out the PSUD II simulator. Maybe I can work on it this weekend. I'll shoot my friend an email and see if he's available at all.
L.D.
Larry hi,
Here's my take on the PS.
As a starting point:
C2, 50uf - 100uf
C3, 50uf
C4, 30uf - 50uf
L1,L2 are fine but L3 needs to be bigger say 40H( DCR may be an issue with this size choke ) or so, as the FE is drawing only around 30mA. 1.5H will act as a resistor in this position.
More current - lower Henry and less current more Henry for the chokes.
You'll need a couple of different cap values (uf) for fine tuning.
I too prefer more H and less uf.
I would take L3,C4 directly off C3 before the OPT CT feed.
Enjoy,
Wheeze
I see that I'm going to have to play around with this PS a bit to get it where I like it and where it works well. I will pick up a couple extra (various value) 630V Solen polys and various chokes to work with and fine tune it.
Looks like this will be more of an ongoing experiment than a straight-forward build. That's okay. I want to learn in the process and also be very happy with the end product. I don't mind trying several ideas and hearing the differences.
After I get this amplifier completed I would like to build some single-ended amps using the 6AV5, 40KG6A, and 6AQ5. Hopefully by that time I will have figured out how to use the PSUD program and just post for any recommendations on the amp section.
Larry D.
Do you know current draw from the OPTubes and from the FE tubes at idle?
Wheezer,
I think these are the maximum current ratings (which I got from the tube data sheets). The current draw for the 6AV5s is 1.2A each. I'm not going to use the 6FJ7, but use a 6BL7 or 6BX7 instead @ 1.5A. The 6BQ7 draws 0.4A each, so it looks like the total is 7.1A. I do have some seperate filament transformers if I need to use them.This is another person's finished build which they sent me the schematic for and not one that I put together. I'd had an Ideal Innovations 6AV5 amp and really liked the sound of it and since I have several dozen of the 6AV5 tubes I would like to build one of my own. I'll probably breadboard it first, then play around with it until I get it where I like its sound, then put it on a chassis.
Larry D.
Edits: 06/20/12
Larry
A better choice for the 6FJ7 is the 6SL7.
Just change the plate resistor R11 to 56K.
Also the 6SN7 would be better than the 6BQ7.
Change R27 to 1.5K and R17 to 5.6K.
DanL
Dan,
Thanks for the recommendations. I will try those tubes out when I breadboard the amp section and see (hear) what happens. I'm sure there are lots of things that can be done there and it will be fun to play around with it for a while before I build it onto a chassis.
Larry
Wheezer
According to the schematic it draws 125mA per channel.
Output tubes draw 52mA each.
First stage draws 13mA.
LTP draws 8.65mA.
DanL
Thanks Dan.
W
.
Hi Amateur,Quick analysis and comment. The supply you posted is neither fast or optimal. High HYs and high Cs are NOT your friends. Use as little as possible to get the job done.
In Push Pull, you can have 2 VAC of ripple to the Finals and it will not be a problem.
I have simulated your supply, guessing at the DCR of the 300 mA. rated power transformer. It is pretty poor at a 15% current step test. It settles in about 550 mS., and has a dynamic Z of (12 volts drop divided by 31.2 mA. current change, or,) 384 Ohms.
My suggestion has under 1 VAC of ripple, inaudible to the Finals in P-P, and it settles in 30 mS (18 times faster) and drops only 3.8 VDC on a similar 15% current step. The Dynamic Z is (3.8 VDC drop divided by 15.75 mA. current change, or,) about 241 Ohms.
Your supply uses two expensive 5AR4 cathode type rectifiers, my suggestion is to use two less expensive readily available 5U4GBs. Such a tube rectifier has NO cathode, and it sounds good. Buy a good Hammond 300 series (highest possible current rating) transformer, a shared L1, which can do 2 A. of current, and a "Y" arrangement of two inexpensive EDCOR 2 HY 21 Ohms open frame chokes, one for each channel.
I have drastically reduced the DCR, HYs, and uFs in the supply.
What I suggested, what I would term a "modern " power supply will ROCK on playing music and it will be worthy. The amp will now kick butt.
I have only quickly considered the supply to the finals, I would use a R/C to the front end, and those values can be easily be determined in PSUD, shoot for 1 or 2 mVAC of ripple.
This is just my initial idea, I know it will play well, and I avoid L-Critical supplies, unlike many others here.
Jeff Medwin
Edits: 06/20/12 06/20/12 06/20/12 06/20/12 06/20/12 06/20/12
The amp schematic I posted was one built by Kegger on AK. Its one of his personal amps he uses. He sent me the schematic of his finished amp which he likes the sound of very well, so this is not something I designed.
I don't know that I'll make any changes in the amp section, but the power supply looked .....let's say, not right. Maybe a bit too big and slow (compared to the vintage stuff I'd been rebuilding). I know there is a wealth of collective experiential knowledge here and since I don't know what I'm doing, but know what I want to hear, I thought 'why not post?'. Heck, I don't really prefer tube rectification. Some Schottky diodes might do the job better.
I'm not a builder or designer. When it comes to that I'm a total newbie, so that's why I came to this forum. I don't mind trying ideas out and taking time to get it right. My ears will tell me if things are right or not.
I bought a set of 1978 Klipschorns and everyone was telling me how great they were. I did not like them. Too harsh, even after replacing the crossover caps. Sold 'em. Got a set of Von Schweikerts now. Seller told me how great they were. I didn't like them at first, but after rebuilding the crossover I love them. So I'm not an opinion follower, except for my own and that is derived from my own ears.
This 6AV5 amp should put out around 12-15 WPC according to Kegger which is a bit more than the 8.5 WPC single-ended amp I'm using with the VS speakers now. I want a little more power than 8.5 watts for these speakers. You'd recommended to use the Hammond 372 series PT. Does it put out enough B+ voltage as is needed for this amp? I'd like to overbuild a bit and not have things run too hot.
Larry D.
Jeff
The hole in your argument is that at 20Hz
the current will increase for only 12.5ms.
Then the current will decrease because
it will be on the downside of the sinewave.
So he will see a drop of only 2.6V during use.
I am not saying anything about the sonics here.
DanL
Dan,
Did you notice the DCR specifications on the Hammond PT I posted ?? Could you MEASURE the BFT-1b on both sides, and please post the DCRs so we will know what that is. The BFT-1B data sheet totally ignores the specification, Hammond certainly does not, and I am guessing the designer was unaware of low DCR.
You have the power trannie, right there on your premises to easily measure. Thank you !!
As to "holes" in analysis (at 20 HZ.!!), that is fine with me, the analysis is only there to get DIY types interested enough to LISTEN to such a supply. Then it usually becomes " I'm ALL IN " as per deathtube667, rage, Drummerwill, lakerfan, etc, etc. :-)
DCRs si'l vous plait monsieur.
Jeff Medwin
LarryOptimal is quite arbitrary.
There are 2 extremes and most fall in between.
You have the LSES which is very small chokes and caps.
The idea here is a quick recovery from power surges.
And you have massive capacitance for flat B+.
The idea is for the PS to be stable no matter what.
Most fall in between those extremes.
Typical is the LCLC with critical inductance
and capacitance of 60uFd and below.
LCLC tends to be the most musical.
CLC tends to have stronger bass.
Your design is somewhere in between.
What you will like is up to you.The PA060 is rated at 300mA and you are pulling
250mA with a capacitor input filter.
The transformer won't like that.
It will get quite warm.
The One Electron BFT-1B would be better.
here is the spec -
http://www.one-electron.com/Trans/BFT1B_11.pdf
Link below is for the dealer.
BTW I use this on my amp.DanL
PS You have way more filtering than you need.
60uFd would be plenty instead of the 220uFd.
Edits: 06/20/12
Wow! That One Electron PT looks like a real beast! I would prefer to not run the PT too hot. I like things overbuilt so that they run cooler, if possible. Is it available to check the DCR of your One Electron PT? I'm curious of the different PT's DCRs. There are a lot of things to consider in building a good power supply. I'm just trying to find out the specifics and find a good balance of what will work well for me.
Heat. That was one thing I did not like about the 6AV5 amp I got from Ideal Innovations. It ran very hot and got there quickly. I'm wondering if the Hammond 372KX PT Jeff recommended would be a cool runner or if its running close to its limits. Like I said, lots of things to consider....
Larry D.
Look at the Hammond spec URL. The Hammond 372KX will loaf and sound great as it is rated 575 mA. continuous, about double the current capability of what your stereo amp uses.
We will hopefully get a measurement answer from Dan on the DCRs of the BFT-1B. My guess is the Hammond will "trounce it" on being low in DCR.
Jeff Medwin
Jeff
I am not ignoring you.
I had to get another meter because
mine is so far out of calibration
that the reading would be meaningless.
Shorted out I read over 20 ohms.
I got another one today.
It is over 95 in here right now
and my work room is much hotter.
I have to dis-assemble a few things
to get at the BFT so am going to
wait until it's cooler.
DanL
That sounds reasonable to me, thank you.
I have a old Fluke 8060A, belonged to Mr. Bob Fulton, which is really accurate. Just need the primary and 1/2 of the high voltage secondary measured reasonably well. You can also measure the entire high voltage secondary, at the rectifier tube socket, usually Red-Red wires, but you know that !!
Thanks again Dan.
Jeff
JeffNow that it is below 80 in here ...
Primary is just over 1 ohm.
HV Secondary is just under 32 ohms.
That is an end to end measurement.DanL
Edits: 06/21/12
Hi Dan,
That is reasonably respectable and I thank you very much for taking the time to supply us with the data.
Half of 32 Ohms would equate to 16 Ohms, (outer winding to the center tap.) Under 20 Ohms, in my experiences, IS "high fidelity". Under 10 Ohms becomes "ultra high fidelity", as suggested several years ago up here.
To head-off Mikey at M.Q., all these DCRs assume a balanced transformer design, core wise, winding wise, etc., just overbuilt in general.
Cheers.
Jeff Medwin
Jeff,
Some days the brain just does not want to function properly. This is one of those days. I looked at the Hammond PT specs and doubled the ma specs (thinking in stereo). In the back of my mind I knew something in my thinking was out of kilter. Must be this between-the-eyes sinus headache I've been dealing with all day today. Just took some ibuprofen believing that may clear things up a bit.
Yes, I've been eyeing that Hammond 372 PT for a while now. It would work well for some of the other single-ended projects I plan to build. Would like to hear what Dan finds out also, if its available to check out the One Electron PT he has.
L.D.
Yes, the 372 KX is a honey, and there is one 600 VCT one even more stiff. The KX has a DCR on the primary of about .6 Ohm when windings are paralleled, and half of the secondary, to the center tap is only 12.5 Ohms in DCR, and advantageously, it is an E-I core so it will have no high frequency generation problems of a toroid. The multiple primary taps, and secondary low voltage windings, are useful to a DIYer also.
I am guessing, where the 372 KX is 12.5 Ohms, (half the HV winding to center tap) the other PT MIGHT be 40 Ohms plus !! Dan will tell us.
Jeff Medwin
Dan,
Hammond now makes a new set of power transformers, in their 300 series, that would be very attractive, I am thinking more-so than what you suggest above and use. They have continuous ratings of around 350 to 500 mA.
As I recall, the DCR on your power transformer is around 45 ohms across (I forgot which, likely it is HALF,.... may be entire) secondary. Since 2010, Hammond makes these new E-I core PTs, in 50 VAC increments (on center tapped PTs) at 40 Ohms and 26 Ohms across their ENTIRE center tapped secondary !! Furthermore, their primary winding DCRs are very low, usually about one ohm, which is excellent.
(For improved SE audio use, one only needs to replace the Hammond's lead outs with double TCSS Kimber wire, to get a decent result in the single ended world.)
Dan, would you please measure the DCRs of your BFT-1 primary and secondary HV windings, and post it up here for us to see and know??
That would be exceptionally helpful. This data was not provided to end users. Thanks Dan.
Jeff Medwin
Lektrik, I understand that the PS is for a 20 Watt per channel amplifier.
You may want to model it on PSUD with only one 5AR4, I believe that there is no need to double-up the rectifiers.
Also, a 5-10 Henry 50 mA choke for the input stage will certainly sound much cleaner (less audible hum).
Your schemo does not include R13, R15, C13 and C15, which are included in the amplifier schematic but are actually part of the power supply. Make sure to insert those in the PSUD model (actually a total of four RC filters for a stereo amp).
I have read that a PS should not have identical filters in sequence, in order to avoid LC ringing. The easiest way out is to change the value of either C2 or C3.
I like a smaller C2 to reduce heat and possible buzz on the first choke. Say 100 uF ?
I assume you have designed a quasi choke input power supply. While playing with PSUD, be aware that C1 will see huge AC spikes. Click on its voltage graph and use a cap rated at least double the peak-to-peak voltage shown on the graph, otherwise the cap will heat up and may explode. Make sure that it is a high-ripple or AC-rated cap: paper in oil or polypropilene and oil.
Have fun!
This is someone else's design. I have designed nothing on PSUD yet. I tried, but I'm missing something somewhere and it didn't seem to work for me. I have a friend who's supposed to help me with it (over the phone), but we haven't connected yet.
I plan on using at least 630V caps just to be on the safe side. Got a box of 30uf 1000V oilers and sixty or so 5uf 700V polys which I'll try and use some of too. Looks like there will be a lot of experimenting done before I get a final product.
L.D.
Thanks, Letrik.
A quasi choke loaded power supply generates huge high voltage spikes on the first cap.
I have seen 1000 V peak-to-peak spikes in the first (small) cap of a 400V power supply.
To be on the safe side, C1 should be a 1000V or 1200V AC-rated oil cap.
I suggest that you ask your friend to use PSUD and click on the Voltage at C1 box.
The RMS voltage is not too bad, however the peaks on the screen graph are very spiky.
Have fun!
Some more thought after re-reading my previous post.....
I loathe the infighting and bickering on this DIY forum, so if you have any helpful information that 'might' start such wasteful time and space usage on this thread, then send me a PM. I like to keep things simple/uncomplicated.
Additionally, I'm thinking that since I don't plan on using any lytics in the PS then I should be able to use less capacitance. A friend of mine from Michigan has been doing this successfully in several amps he's reworked with much improved results. Does PSUD-2 accomodate the differences in using poly or oil caps instead of lytics?
Larry D.
Hi Larry,
there is no single optimum way to good power supply design. There are different schools of thoughts and each claims good sound results. You will very likely get contradicting suggestions.
PSU design is a topic which often triggers heated debates. Way too much focus on PSU design IMHO
Looking at the power supply in isolation without any information about the actual amp circuit attached, does not make a lot of sense. It is true that the power supply is an integral part of the amp and will contribute a lot to the sound. But it is not the only part. The amp itself is just as important.
Here are my preferences:
- get the best possible power transformer (within your budget) for your design. An electrostatic screen between primary and secondary is very important.
- I prefer full wave bridge rectifiers with 4 diodes over the center tapped configuration
- resort to TV dampers, they are cheaper and better rectifiers IMHO
- I prefer higher inductances and smaller caps 220uF oilers will be huge. I stick with values around 50uF, often smaller.
- I would use a higher inductance to decouple the front end. If on a budget, rather save this inductance and do RC decoupling
Best regards
Thomas
Thanks Thomas! I will eventually figure out the PSUD program and maybe put together a PS for the amp that I'd be happy with, if I don't get a PSUD derived answer here. Its a 6AV5 PP wired in triode amp and I've included the amp section in this post. I don't think I want to change anything on the amp section, but I thought the PS was not exactly what I would want to be feeding it.
Also, when I said I was on a budget all I meant was that I didn't want to spend ultra bucks on handwound Japanese transformers and the likes. I don't mind spending a little more on something that I can readily hear the difference in.
I did have an amp that used the 6AX4 damper diode. They seem to be just slightly cheaper than a pair of NOS Mullard 5AR4s. Hah!
Larry D.
Mate even I get a dog fu$K understanding of Duncan's psu Sim. I think you will find that the second cap is too big when you Sim it. Just muck about with it. Don't change the cap resistance values get the dcr's right. Set the trans voltage value and type. Tic the the second box at the bottom left basically you want to look at the ripple at the final load R in the PS. Set that R to get the right current draw. Another tic box shows current at that R. That gives you voltage all so. Make sure you monkey with the caps until you get no ripple or big peaks at the start. If this advice is wrong I am sure somebody will relish telling why then I might learn something too .:)
BTW In stating that you loath arguing you are being confrontational IE inviting an argument. For eg. If I loath loathing. I would be a loather. I would loath being a loather therefore I would be all uptight and full of self loathing probably stating at the start of every post that I loath loathing. So loathing would be unhealthy for me. So I don't really care about arguing. But I do I care about you and your loathing. So don't loath please. It will spoil your arguments.
When I posted this question I knew it had a great possibility of becoming another bickering session, but this is 'the' place (more than any other place) to get the answer to my question. My assumptions on finding good useable information proved to be true.
I'm of the thinking that when I have a disagreement with someone I state my point and then drop the subject. I don't argue because I don't have to prove anything to anyone. I have dealt with too many people who will not consider anything you have to say because they have already made up their minds on what is their version of 'truth'.
I try to listen to everyone and glean from their experiential knowledge of what works and what doesn't and then try out their ideas (if they're sensible/logical) and see/hear what results I get. I don't like to 'follow' anyone, but I can learn something from almost everyone here that has a little or a lot of background in this DIY field.
I haven't messed with the PSUD for a while since I had an unproductive time with it a few months ago. I've got someone who is supposed to walk me through it (over the phone) as we design a power supply together for an single-ended 6AV5 amp that we both plan to build. After I get all the info from this posting compiled together I may take another try at it on the PSUD program and see how it fares.
L.D.
Thankfully, I don't think this post has seen much disrespect or bickering.
Mr. Mayer and Neff described a traditional supply, and me a non traditional supply, and you are in the cat bird seat and can do as you like !! Pretty cool, me thinks !!
Jeff Medwin
Hi!
As I siggested, I would increase the inductances and lower the caps. Like 10Hy/50uF. You can tweak the output voltage with the first cap after the rectifier.
Best regards
Thomas
Post a Followup:
FAQ |
Post a Message! |
Forgot Password? |
|
||||||||||||||
|
This post is made possible by the generous support of people like you and our sponsors: