Ok then. I think you all agree on not touching the coupling cap, so that's what I'ill do. Thanks everyone for taking the time to help others!!
I just wish I knew at least something about this to make things better for my amp:
I just wish I knew at least something about this to make things better for my amp:
From the science you have modern caps. The .22F is a metal film so replacing this will have little effect. Again the cathode resistors are many times bigger than the ESR of the electrolytic caps so again very little effect. I would personally remove the 820K FB cap which may limit the drive of the first stage add some global negative feed back instead.
You could drop the UL resistor to say 470R 2W. May help. Could do an ltspice simulation although wound need to guess the OPT parameters. Can I just check that is a SE OPT you have - the UL is rather high at 50% which made me wonder if it is a CT PP. I see the 2K resistor has done the heat shrinking for you Ha-ha.
Last edited:
Get a schematic for the amplifier and some type of multimeter. If you take readings at various points we can see if anything is wrong and go from there.
ALL answers were Technical!!!
None compared Brand A silkyness to Brand B micro granularity to Brand C authority on Female voice and so on 😉
Which Castle Conways do you own? Are they the large floor-standing models? It's hard to imagine that they would sound good with an amp that probably makes 3-4 watts at the absolute most. From the tests I've seen these start rolling off at around 6kHz! There's no feedback around the output transformer so the damping factor is probably terrible. Really, TRULY, boutique caps are not going to fix the problem. If you MUST try a different cap get some Cornell-Dubilier 942C series. These are very good caps at a normal price. If you don't hear an improvement then just stop there. Don't waste hundreds of dollars on V-Caps or Duelunds or whatever.
@grovergardner they are Conway III. I'll probably follow your advise and try some not costly caps first. Thanks.
Sharing what Stephe from Skunkie Designs already did for all of us Reisong A10/12 owners. There are some measurements there. Not sure if that's what you meant. That's exactly how I have my amp right now.
Attachments
@grovergardner I would appreciate if you can share how to fix this. I've just posted the schematics for the amp. Remember I understand almost nothing about this, so the more specific the better so I can just go and do it.
How certain are the output plate and screen voltages? If you look at the current flow from B+ to plate, from highest to lowest my expectation would be decreasing voltage from B+ to screen to plate.
Re: caps, as someone who does cap swap I've never run into a one that made an amp terrible. Save it for last if you really need to scratch that itch and try an inexpensive Solen.
The plate to plate feedback can also work fine and does for one of my amps. Even open loop a properly operating UL shouldn't be terrible. I tested this just yesterday. Definitively has a character but it's towards brightness tracking the rising high frequency impedance of most dynamic loudspeakers.
Re: caps, as someone who does cap swap I've never run into a one that made an amp terrible. Save it for last if you really need to scratch that itch and try an inexpensive Solen.
The plate to plate feedback can also work fine and does for one of my amps. Even open loop a properly operating UL shouldn't be terrible. I tested this just yesterday. Definitively has a character but it's towards brightness tracking the rising high frequency impedance of most dynamic loudspeakers.
Ouch!
Never checked Reisong but expected it would be "an amp".
Meaning , say, PP EL34, a reasonable OT, at least 20W RMS, overall NFB, etc.
But this SE 5-8 W amp (tops) requires a very efficient speaker (a Klipschorn would be perfect), even with that funky local NFB very important OT is outside the loop, damping will be very poor, it needs a KILLER OT just for reasonable performance.
No surprise it sounds poorly.
Triodes can get away fine with similar designs, but ... A Pentode? ... EL34? ... and no adequate NFB to help it?
Its problems are WAY beyond coupling cap replacement.
Lots of iffy choices in that design.
Sorry..
Never checked Reisong but expected it would be "an amp".
Meaning , say, PP EL34, a reasonable OT, at least 20W RMS, overall NFB, etc.
But this SE 5-8 W amp (tops) requires a very efficient speaker (a Klipschorn would be perfect), even with that funky local NFB very important OT is outside the loop, damping will be very poor, it needs a KILLER OT just for reasonable performance.
No surprise it sounds poorly.
Triodes can get away fine with similar designs, but ... A Pentode? ... EL34? ... and no adequate NFB to help it?
Its problems are WAY beyond coupling cap replacement.
Lots of iffy choices in that design.
Sorry..
@JMFahey 🥵. You are most likely right (I don't say 100% because I understood half of your comment, hehe). I just remembered one thing though. No idea if this is important either. When I did the mods, it was specified to use a 5AR4 rectifier tube and I have a 5V4GA in there. Could that be another issue that could fix some of it's problems? Oh, and I'm using Gold Lions KT77 instead of El34s as well.
I wasn't kidding when I said I could rebuild the whole amp with another schematics from here. If at least the transformers could be saved, I "could" try something like that. I have some good tubes, so moving around/replacing some resistors and caps should be that hard. This Reisong is PTP so...
I wasn't kidding when I said I could rebuild the whole amp with another schematics from here. If at least the transformers could be saved, I "could" try something like that. I have some good tubes, so moving around/replacing some resistors and caps should be that hard. This Reisong is PTP so...
Last edited:
It seams as if you are a builder rather than a design engineer. Nothing wrong with that. Given that, build a proven design with the parts list specified. No substitutions. Then listen and measure, at least with a multimeter. If you can predict the measured values ( not guess ) then you are really beginning to understand the design. Do two mono blocks and employ blind a/b listening. If then you still want to play the component upgrade game, only do one thing at a time to only one amp, keep the other as is for a reference.
You will learn things that will protect your pocketbook and about listening bias.
Good luck on your journey.
You will learn things that will protect your pocketbook and about listening bias.
Good luck on your journey.
The amp has local feedback between the output tube and the driver, and that presumably lowers the output impedance to some extent. You could add global feedback from the output tap to the driver cathode, but you'd need to have a scope and some other tools to make that work. It's not something you can "ballpark" without possibly making things worse. ;-)
Well, now, a 5V4 will drop a good bit more voltage than a 5AR4. Anywhere from 10-30 volts off the B+ in this amp. And every volt counts here. That's why someone asked if you've checked your own voltages against the schematic. A 5AR4 might well bring the performance up to a tolerable level. I would certainly try that.
🙌 @grovergardner, will try that out (when I buy one). For now I have factory 5Z4PJ. Will it make things worse or slightly better till I get the 5AR4? Can't A/B test right now as I'm on vacations far away from my amp.
A better pot is more enjoyable to use, and maybe even audible, so why not? Nothing wrong with actually improving things.
If I were to attempt a simple mod of the design, I'd remove the driver stage's cathode bypass cap, and probably increase the anode load resistor to 150K or so, but a more effective change would need for feedback application to be moved from the driving valve's anode to its cathode. Not trivial because DC conditions change (current from the feedback resistor significantly effects driver bias) and also not as well focused (existing feedback path puts it at exactly the best place, around the output valve).
The issue with the existing feedback path is that it heavily loads the driving stage, rotating its loadline very far clockwise (effectively adding another parallel load equal to the feedback resistor's value divided by gain of the output stage plus one), reducing its gain and increasing its distortion. On balance, I'd guess that degenerating the cathode and moving idling point up and left would be net positive, but not a panacea.
Maybe interesting, O. O. Schade did not propose this type of feedback - his original paper of 1930 used series applied feedback over the output stage, possible with transformers, and didn't have this liability. Someone somewhere in the recent (?) past misunderstood the distinction and his name got stuck to the current scheme. But, not his fault.
All good fortune,
Chris
Hi Erlandes,
If that capacitor is an Orange Drop capacitor, your description appears so, it is one of the most inaccurate, poor
sounding capacitors on the market. If it is a red chinese capacitor, I would not bet on it.
I would not bypass. Instead, might try a 0,22 ufd Mundorf EVO in oil aluminum. Don't go smaller ufd, go better quality.
Concerning the 220uf 25 volt cathode bypass capacitor; if you can parallel, say, 4) 47ufd + a 33ufd capacitor for 221ufd.
The sound/bass will be less muddy. Could also try 3) 68ufd + 15ufd or 22ufd, the sound will still be less muddy.
cheers and all the best.
pos
Last edited: