 
|
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
|
|
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
40.136.241.227
I used Gordon Rankin's schematic for his "Bugle" 45 amp and modified a few things to try it with a 71a output tube instead of a 45.
I used my test mule and set the operating points for the 71a per what is shown on the schematic.
There is music coming out the other end, but it does not sound as smooth as it should once you get above about half volume, it starts distorting or breaking up. Since I am pretty close to the design operating points for the 71a (see schem ), I am inclined to think that the way I have done the 6072a srrp does not work well with the 71a (too much gain?).
I tried following through Gordon's explanation of how he designed the amp that is in the article in sound practices issue 13, but my level of knowledge is not quite up to the task yet. If anyone has any ideas on what might be done differently with this design to improve things, I would appreciate the input.
Read changes to schem in following posts.
Edits: 08/30/16 08/30/16Follow Ups:
Decided to wait another day and see what Hermine is going to do before heading to Shelby. Right now she seems likely to dump a lot of rain on the hamfest. Don't know if the dealers will show up if the forecast is for heavy rain on Friday. Should have a better idea of the weather tonight. It's more than 3 hours from here so I want it to be worth the trouble.
Anyway.... while I am waiting I decided to try some of the other suggestions. I tried taking the signal off the plate of the lower triode, but I did not think it sounded as good as the original way. Then I tried using just the lower triode as a driver instead of an srrp. That actually sounded better. What I have been doing is A/B tests of the different combos against the 26/26/45 amp I just finished a short while ago. I feed a mono recording into my preamp and have each amp hooked to a separate channel. Then I can go back and forth fairly quickly between them with the volume controls. Up on one down on the other, then reverse it. Doing this lets me hear subtle differences that it would be tough to pick out just using memory, even if you played them a few minutes apart.
I found out that the srrp sounds a teeny bit muffled, as if the highs are rolled off, when compared to using one half of a 6072 only. I also could hear that there was more detail and clarity from the 26/26/45 amp than the one with 1/2 of a 6072. So, long story short, I have decided to give up on the 6072 for now and try using two 26's for drivers as I did with the 45 amp. If they do what they did for the 45, it should be interesting.
As Jeff had pointed out, my B+ for the 6072A's was too low. I decided to work on that first before trying anything else. I went back and checked Gordon's article and the design operating points are listed on the attached schem.
I had to put a dropping resistor in front of the 71a to hold the B+ down for it while I raised it for the 6072a. Turns out that the 275 vdc b+ is just right for the 6072 with no resistor on the plate of the second triode to give me the 140 vdc design plate volts. Since I had not changed the cathode resistor of the 6072 the grid volts and grid current were both ok.
What a difference that made! It sounds really good now. No distortion all the way up to just short of full volume. The bass on the 71a is really amazing too. I had forgotten that from the last time I built one. It is on a par with any of my other amps and better than a couple of them. Sweet little amp if anyone wants a fairly easy SET amp project. I used ac on the heaters and a hum pot on the 71a. Quiet as a mouse.
I think I will try a couple of the suggestions folks made just to hear what difference they might make, but there sure isn't anything wrong with it as is. I will report back after I have played with it some more, but this weekend is the Shelby, NC hamfest and it is about the best one near to me for parts, so I will be busy with that for a few days.
Thanks for all the help folks, I really appreciate it.
Edits: 08/31/16 08/31/16
Vinnie,
You want to avoid adding resistance to the B+ feeding the Finals, at all costs and whenever possible. It hurts the dynamic response of the Finals tube, as it powers the speaker. Recall all my emphasis on LOW DCR power supplies, to the Finals ?? Adding resistances to the Finals is the opposite of good.
So, Feed the 71A directly with the higher 275 VDC, just raise the Rk resistor, on the 71A, so that it is operating as you want it to ( plate to cathode voltage differential, and desired current )
You may wanna have some mild decoupling of the Driver and Finals stage, so they don't interact. Hopefully, a higher B+ than 275 VDC will allow for a better design. AS I pointed out, Rankin used 330 VDC on the Bugle, across two tubes, and one might like to see even higher VDC across both, to get optimum fidelity.
Can you raise the 275 VDC some ? IF your supply is choke input, maybe use a NON-critical L1 choke, like a Hammond 1/2 HY at 11 Ohms unit, will get ya maybe 20% higher B+ without even PSUDING it. Use a L1/C1/L2/C2 to the 71A, no " R" in series to the Finals' filter.
Have fun.
Jeff Medwin
That was just a quick lash up to get some noise out the other end before I started tweaking. I agree that it would not be a good way to build the final amp.
As for the B+, Gordon got the 330vdc for the bugle by adding the 275vdc max plate volts and the 55vdc for the bias. He just used what was needed.
I tried all the suggestions, and the only one that made an improvement I could hear was removing the cathode resistor from the srrp. That made quite a bit of difference. I can now go most of the way to full volume without distortion. Still some in the upper reaches of volume though.
Have updated and corrected the schematic per attached. I see I cut off some of the operating point data. Grid voltage is -33.5 vdc, current is 18.6 mA. Also the grid resistor on the bottom triode is 1k.
I think I will play with the operating points a bit and see what happens. Anyone have any other suggestions?
Edits: 08/31/16 08/31/16 08/31/16 08/31/16
that the thing that made a difference was removing the cathode bypass capacitor, not the "cathode resistor"? I ask because that was my suggestion, and I am interested in the outcome.
I also second what Fred76 wrote. Running that same configuration as a plate-loaded triode (by taking the output off the plate of the bottom tube) may sound better than SRPP. What you give up is low output impedance, but probably/possibly you could drive the downstream tube anyway with no problem.
What's up with the 10uF capacitor dangling off the plate of the top tube?
Yes, it was the cap I removed per your suggestion. I could definitely hear a difference. Less distortion at higher volumes.
I am going to try all the new suggestions one at a time and see what happens. I will report back on the results.
Vinnie,
Goodness !! You are using a B+ of only 208 VDC , and splitting it across TWO tube sections in series as the driver. Gordon used 330 VDC in the Bugle as a B+ !!! Even THAT VDC is far below optimum, IMHO.
Has any old experienced DIYers ever told you what the letters SRPP stands for? It means " Sounds Really Pretty Poor", even when executed ideally.
But the theorists ( usually non - listeners ) may " dig " it, conceptually . KISS rules in SE design.
Jeff Medwin
The b+ for the 71a is right, but I may have to adjust the b+ for the 6072 to match what Gordon had for the bugle.
Can not agree with srrp's sound sounding poor though, Gordon's Bugle is a great sounding amp
I just want to point out that the totem pole topology reduces the B+ across each tube by 1/2. It also reduces the available voltage swing at the output. I suggest you try it with one tube and see if it doesn't seem a little more effortless.
--------------------------
Buy Chinese. Bury freedom.
Do you mean just one half of the 6072A?
nada aqui
--------------------------
Buy Chinese. Bury freedom.
ok, I will give it a try.
Are you measuring maximum voltage swing or distortion anywhere in the amp when you make changes like this? It's easy to misjudge a particular combination of parts if those things aren't verified first. About the totem pole, my own 211 SETs use this topology in the pre-driver. The two stacked tubes run at more than +500V B+. That provides the voltage swing needed to drive the 211 fully into A2. All the tubes are barely wiggling at the critical levels below 1W. :)
--------------------------
Buy Chinese. Bury freedom.
Not sure I know how or have the gear to do either? Is there some place on the net that describes the procedures?
do you have a scope and a signal generator?
then read pages 195-204 after learning basics
http://www.tubebooks.org/Books/Atwood/Crowhurst%20Cooper%201956%20High%20Fidelity%20Circuit%20Design.pdf
That's an excellent book, but it might be a bit much for someone who's new to all this. I think a couple books on test equipment and/or basic troubleshooting of a more general nature might be beneficial.
Vinnie, here are a few pubs I found in a quick search on eBay. These are basic books on using test equipment to measure and troubleshoot common circuitry. I read tons of books like these when I was getting started in electronics.
Radio Shack "realistic guide to oscilloscopes" 1972
eBay # 361514967147
$9.99
T A B Books "Electronic test equipment--and how to use it"
eBay # 252514011665 1974
$3.00
Prentice Hall "Complete Guide to Electronic Test Equipment and Troubleshooting Techniques" Ex-Library 1975
eBay # 222202722645
$3.97
Prentice Hall "Handbook Of Electronic Test Equipment" 1971
eBay # 231978021650
$14.99
--------------------------
Buy Chinese. Bury freedom.
Thanks TK. Actually though I have been using my scope and signal generator for quite a while for trouble shooting, signal tracing mostly. It's the other parts of checking performance, etc. that I have not done much of, like the post Mike just put up above for measuring power output.
Edits: 09/01/16 09/01/16
you have used a scope for troubleshooting, but always come here for help troubleshooting things that would have taken less time to fix ,with said scope, than it took to type your post.
Did it occur to you that I might not know how to use the scope in the way you are talking about? Ownership of a scope does not automatically mean the owner knows how to use it for everything it can do. Like I said, I have used the scope mostly for signal tracing, i.e. to find out where the signal was stopping before it got to the speaker. That's about it at this point. Most of the books I have found on trouble shooting with a scope don't seem to cover the problems I run into. I am going to see if i can find some of the books TK mentioned in his post, maybe they will be better.
Edits: 09/01/16 09/01/16
Thanks, I have both.
Hi,
I remember reading Paul Joppa's comments before, that in an srpp circuit you could try to take the signal fom the bottom triode's anode of the totem pole instead of the top triode's cathode, making the top triode a form of CCS (albeit not the most efficient one iirc)..
Regards...
Might be worth a try.
185v at the plate, - 40v at the cathode gives you 140v across the tube. Your output is likely around .4W Low volume is all that you will get. It's not clear how you lost 23v across the output xfmr.
Wrong number by a little. The grid voltage actual is -35 vdc. -40 was the design goal.
Edits: 08/31/16
I'm not saying its not a good idea, but I am curious about what precisely are the advantages of an SRPP? If it is to live up to its name, as a push-pull stage, it only operates as a balanced PP if it is driving into precisely the right load impedance. I'd strongly suspect that with the schematic you have shown (once the grid of the upper tube is drawn and connected!), the impedance it is driving into is far too high for it to be operating in PP mode; essentially, the currents through the two triodes will be more or less tracking in phase, rather than in anti-phase.
The circuit, operating in such a mode, may be perfectly fine but it is not going to be push-pull. I'm interested to know what the advantages of the SRPP are, in such cases. Again, I'm not saying there aren't any; I'd just be interested to know what they are.
Chris
To put it simply, I do not know. Maybe others here can explain it. I used it because I built Gordon's original design (bugle) quite a few years ago and it was my main amp for 3 or 4 years. I liked the sound of it very much. Thinking that the 45 and 71A might be somewhat similar I wanted to try this with a 71A. My problem is I do not have the skill to determine what other changes may need to be made to the circuit to have it operate at it's best, or to know if that can not be achieved using the 71A.
Edits: 08/30/16
You need to connect the grid of the top triode to the plate of the bottom triode. Don't forget the grid stop resistor.
Tre'
Have Fun and Enjoy the Music
"Still Working the Problem"
Edits: 08/30/16
Opps, drafting error. That connection was made, forgot to draw it in. Where would you expect the grid stop resistor to be? There is a 1k resistor between the grid of the first section of the 6072a and the input, and there is another 1k resistor between the plate of the first section and the cathode of the second that I forgot to label. There is nothing else on Gordon's original schem that I see.
Dang found another error! There is a 10uf cap going from the plate of the second section of the 6072a to ground.
Edits: 08/30/16 08/30/16 08/30/16 08/30/16
Don't you need to reference 6072 heaters to something other than ground?
Not sure what you mean.
The cathode to heater potential in the upper tube of an SRPP circuit should be within spec.
Did you check the tube manual?
I use a typical resistor divider off the power supply to set the AC heaters on a DC bias above ground for SRPP.
I measured the voltage on both heaters and it is 6.3vac on each. Is that what you mean?
No, in the tube manual there is a voltage spec. for heater to cathode potential, this is for IDH tubes where the heater is a seperate element.
For 6072, the spec is 100 volts.
In SRPP, the upper tube cathode is at the same DC voltage or close to it as the lower tubes plate.
It is a good idea to raise up the AC heaters on a DC bias above ground to stay within the heater to cathode voltage spec.
I am a little confused on how to measure this. The 6072 has a common pin and then one other pin for each filament.
I will have to check on that. Thanks.
Connie, your coupling cap from SRPP to 71A is not right. .01 is too small. Find a calculator online and recalc.
Who's Connie? Actually that value of cap is what I generally use for couplers these days. I have found by trial and error that the smaller the cap I can use the better the sound.
Sorry, spellchecker decided to rename and regender you. Check that top triode cathode voltage.
No, the cathode pin of the upper triode measures what DC voltage.? If this is different by anywhere near 100volts, which is max rating of 6072, then you need to reference the center tap of the heaters transformer to a higher DC potential. You are just changing "sea level".
http://www.tubebooks.org/tubedata/HB-3/Receiving-Type_Industrial_Tubes/6072.PDF
See where it says peak heater-cathode voltage. The original Bugle must have done this.
Grid-stop resistor goes right ON the grid pin, with nothing else connected to the grid pin.
There is a good theoretical argument for dumping the Rk bypass capacitor in SRPP. You might want to try it. It equalizes current flow around the circuit. This is assuming that Rk1 = Rk2, where Rk1 is the cathode resistor on the lower tube section, and Rk2 is the resistor between the plate of the lower tube and the cathode of the upper.
They are the same, I just did not mark Rk2 like I should have. I will give that a try and see what happens.
What value should the grid stop resistor be? Would it tie to the grid on the grid side of the 1k resistor on lower section?
If 1k is good enough for the bottom triode it's good enough for the top triode.The grid of the top triode gets a grid stop resistor that connects to the plate of the bottom triode.
Tre'
Have Fun and Enjoy the Music
"Still Working the Problem"
Edits: 08/30/16 08/30/16
So the 1k resistor to the grid of the bottom triode is a grid stop resistor? That explains what it is doing there. I thought the value was too low for a grid stop. I will give it a try, but Gordon's original schem does not show a grid stopper on the grid of the second triode.
Edits: 08/31/16
"So the 1k resistor to the grid of the bottom triode is a grid stop resistor?"
Yes.
"Gordon's original schem does not show a grid stopper on the grid of the second triode."
I see that but if the bottom triode "needs" a grid stop resistor then the top triode does also.
Tre'
Have Fun and Enjoy the Music
"Still Working the Problem"
I can not tell you why, but it sounded worse with the grid stopper on the grid of the top triode.
Well, that's interesting.
It shouldn't have made any difference except stopping oscillations (if there were any).
Tre'
Have Fun and Enjoy the Music
"Still Working the Problem"
Maybe I will try it again later when I get the rest of the wrinkles out.
Tre' did not comment on this, which makes me question myself in saying it, but 1K actually seems like a high-ish value for a grid stop resistor on a 6072, assuming the dual triode shown IS a 6072. If you can hear it, then you might consider going to a lower value or even starting with 100R, which would probably be sufficient. Carbon.
I thought 100k was the usual starting point for grid stops?
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:
Top of Page ]
[ Contact Us ]
[ Support/Wish List ]
[ Copyright © 2024, Audio Asylum, all rights reserved ]
