Hi FriedMule,
So sad the way your side went out (of World Cup) -- way sorry . . . .
To give a direct answer to your P.M. question, "I'm honored to be asked, and would be glad to help."
But I think we should stick to presenting our ideas on 'the Forum'. That way the other members can keep me honest, when I mess up, or overlook something, or am just too damn slow.
Also sorry that you had to P.M. me to get my attention -- speaks to one reason I hope you'll be O.K. with continuing your project 'on the forum', rather than by P.M.'s -- I'm awfully, awfully slow -- in EVERYthing . . .
Also, PRR, DF96, jameshillj, Eva, Mark Whitney, abraxilito, SemperFi, scottjoplin, and others may very well have different ideas -- they've already contributed some excellent advice, and I welcome their opinions.
If you could let the Moderators know which thread you want to go with your build, and we can all climb aboard that little dude and make it special.
Oh, and, if my arithmetic isn't in the ditch, the gain of this beast is pretty close to (1+(50K/2K)) -- 'cause that's where the gain-clamping feedback ends, at the front the triple EF output stage - the Vbe bias multiplier.
Let's build this animal and start enjoying the sound . .
Best regards,
-Rick
So sad the way your side went out (of World Cup) -- way sorry . . . .
To give a direct answer to your P.M. question, "I'm honored to be asked, and would be glad to help."
But I think we should stick to presenting our ideas on 'the Forum'. That way the other members can keep me honest, when I mess up, or overlook something, or am just too damn slow.
Also sorry that you had to P.M. me to get my attention -- speaks to one reason I hope you'll be O.K. with continuing your project 'on the forum', rather than by P.M.'s -- I'm awfully, awfully slow -- in EVERYthing . . .
Also, PRR, DF96, jameshillj, Eva, Mark Whitney, abraxilito, SemperFi, scottjoplin, and others may very well have different ideas -- they've already contributed some excellent advice, and I welcome their opinions.
If you could let the Moderators know which thread you want to go with your build, and we can all climb aboard that little dude and make it special.
Oh, and, if my arithmetic isn't in the ditch, the gain of this beast is pretty close to (1+(50K/2K)) -- 'cause that's where the gain-clamping feedback ends, at the front the triple EF output stage - the Vbe bias multiplier.
Let's build this animal and start enjoying the sound . .
Best regards,
-Rick
Yes that was not what we had hoped for at all!
Thank for your answer! It is properly just my mentality but I after a while feel bad to ask for help, when the only thing I do is to leach on other without giving anything back.
It is totally ok to continue her on the forum!
I just feel that I keep asking the same questions, without knowing it, and that I have to be tiresome for everyone.
YES!!! They have made a huge and fantastic afford to help!!
That's also why I feel that I just beg for help and help and help but without giving anything back and I can look as if there answers were all to no avail since I do ask the same questions, 10 minuts later because I did not completely understand the English or the details in the answer..
Great idea and thanks but how do I do that best?
Here is a great example on what I mean by not understanding and asking over and over again!
I asked about the gain her on the great forum and after what I could understand, they both did tell me that the gain was about 25 times.
Then I say it to you, as if there answer was not fine enough.
And you use your time to calculate it all over to help me!
And just to put salt in the wound, I am not sure that I do understand this part of your answer.
Thanks a lot, hope that anyone will tell me how to best ask so it is more fun to help and please tell me if I can help in any way!!
Hi FriedMule,
Try not to worry about lots of asking -- technical jargon is confusing, even when it is in one's first language. Also helping is fun, especially when someone has 'smart curiousity'. Besides, you'll surely help someone else along their path.
One thing that would help both questions and answers would be Component Designations -- the C4, R12, Q6, etc that uniquely identify a particular component. Any chance you have, or can find, a version of the schematic that has 'em? It would probably be worth the time and effort to edit the schem you have if not.
Sorry, but I don't know how to ask a Moderator to choose or merge threads. I'm kinda new to this excellent place myself.
Regards,
-Rick
Try not to worry about lots of asking -- technical jargon is confusing, even when it is in one's first language. Also helping is fun, especially when someone has 'smart curiousity'. Besides, you'll surely help someone else along their path.
One thing that would help both questions and answers would be Component Designations -- the C4, R12, Q6, etc that uniquely identify a particular component. Any chance you have, or can find, a version of the schematic that has 'em? It would probably be worth the time and effort to edit the schem you have if not.
Sorry, but I don't know how to ask a Moderator to choose or merge threads. I'm kinda new to this excellent place myself.
Regards,
-Rick
That's the gain 'calculation' -- so really not a big deal. At the highest and lowest frequencies it gets a little more complicated, but for pretty much the whole audio range it is that straightforward.
The right one of the two BC556's (the input differential pair) sees a scaled-down version of the output signal. It comes via the two 100K resistors straddling the BC547, in effect providing a 50K signal path since the transistor maintains a temperature-compensated DC voltage span. Then it gets attenuated approximately 26:1 by the 2K resistor to an AC ground provided by the 100 uF Silmic II. After the BC547 there is no additional voltage gain, only current gain.
Sorry I messed up the line:
"at the front the triple EF output stage - the Vbe bias multiplier."
it would make a lot better sense with a properly placed 'of':
"at the front of the triple EF output stage - the Vbe bias multiplier."
'EF' is shorthand for Emitter Follower -- all 6 transistors after the BC547 (which is the Vbe multiplier) have their Collectors connected to power rails. The current at its Base increases by it's beta (approx - varies with Vce), with the result available at the Emitter. The voltage at the Emitter closely tracks that at the Base. These stages are also often called Common Collector.
The more you work with these types of circuits, the easier it'll become to make sense of them.
Happy building,
Rick
The right one of the two BC556's (the input differential pair) sees a scaled-down version of the output signal. It comes via the two 100K resistors straddling the BC547, in effect providing a 50K signal path since the transistor maintains a temperature-compensated DC voltage span. Then it gets attenuated approximately 26:1 by the 2K resistor to an AC ground provided by the 100 uF Silmic II. After the BC547 there is no additional voltage gain, only current gain.
Sorry I messed up the line:
"at the front the triple EF output stage - the Vbe bias multiplier."
it would make a lot better sense with a properly placed 'of':
"at the front of the triple EF output stage - the Vbe bias multiplier."
'EF' is shorthand for Emitter Follower -- all 6 transistors after the BC547 (which is the Vbe multiplier) have their Collectors connected to power rails. The current at its Base increases by it's beta (approx - varies with Vce), with the result available at the Emitter. The voltage at the Emitter closely tracks that at the Base. These stages are also often called Common Collector.
The more you work with these types of circuits, the easier it'll become to make sense of them.
Happy building,
Rick
Thanks a lot!!
So about 26 x gain (except in the extremes)
This amplifier is originally specified as a 60W 8Om amplifier
26-0-0-26 secondary dual rail transformer.
Many have told me that something is wrong in these number, but I think it maby have something to do with early clipping?
Do you know the input impedance?
So about 26 x gain (except in the extremes)
This amplifier is originally specified as a 60W 8Om amplifier
26-0-0-26 secondary dual rail transformer.
Many have told me that something is wrong in these number, but I think it maby have something to do with early clipping?
Do you know the input impedance?
It's about 40K. But again only over most of the audio range, since the 220pF starts to factor in at higher frequencies. A properly designed differential input pair will have both transistors tracking together, resulting in comparatively low Base currents, which gives a high reflected impedance. The 220pF seems a little heavy to me -- it has a capacitive reactance (offset 90 degrees from a purely resistive load) equal to the 39K resistor to ground at 18,550 Hz -- I'd put it higher. Maybe 180, 150, or even 120pF (34KHz -3dB frequency) would be better, and still provide plenty of attenuation to out-of-band signals that have leaked in.
Not sure what you mean by 'early clipping', but whoever it was that insulted the power supply design (earlier in the thread) had a valid point IMHO -- the design shows a conspicuous misunderstanding of the basics of bridge rectifiers and transformer secondary effects. It's the kind of thing that 'Marketing guys' are known for, over the objections of engineers who know better. My best recommendation would align with another member's: Standard Center-Tapped secondary, ONE bridge -- and forget all that foolish 3-bridge drama. All it does (besides waste power and add cost) is reduce peak to peak output -- using the same transformer.
A 26, 0, 26 transformer for a 60W @8 Ohm amp does seem a little light, at a glance, especially considering the triple EF outputs and the extra diode drops in the original PSU design. So let's do a little arithmetic:
60W at 8 Ohms is 21.9Vrms, which for a sine is +/- 31.0V peaks. A lossless rectifier could charge the main rail caps to 36.8V from a 26VAC secondary -- if there was no load. But an 8 ohm load will draw 3.88 amps from a 31V source (kinda heavy load). A (more typical, standard) single bridge would provide about +/-36V lightly loaded; the original design would be a little over +/-35V.
Now for the output losses: A triple EF reduces the peak-to-peak swing of the output by at least 2.2V or so, in each direction -- so about 4.5Vp-p minimum. The VAS and its load, the CCS (about which I cried 'mis-designed' in an earlier post) have an Ic of ~2.1mA to work with, allowing them to get decently close to the rail as they saturate. But remember, the pre-driver, driver, and final are all loosing beta as their Vce drops to one or two volts, too -- and this is when the highest current is asked of them (again, assuming a resistive load).
So it's really just an educated guess without a simulation, or build, but it looks to me like a continuous 60W @8 ohm output would be at least slightly modulated by 100/120 Hz ripple from the PSU. Those big 22,000 uF main rail caps will help, as would a very stiff transformer. Others here can advise better than me, as to what VA rating might be sufficient.
Your last post sounded as though you still have questions before you're ready to build.
Hope this covers a couple/3 at least.
Regards,
-Rick
Not sure what you mean by 'early clipping', but whoever it was that insulted the power supply design (earlier in the thread) had a valid point IMHO -- the design shows a conspicuous misunderstanding of the basics of bridge rectifiers and transformer secondary effects. It's the kind of thing that 'Marketing guys' are known for, over the objections of engineers who know better. My best recommendation would align with another member's: Standard Center-Tapped secondary, ONE bridge -- and forget all that foolish 3-bridge drama. All it does (besides waste power and add cost) is reduce peak to peak output -- using the same transformer.
A 26, 0, 26 transformer for a 60W @8 Ohm amp does seem a little light, at a glance, especially considering the triple EF outputs and the extra diode drops in the original PSU design. So let's do a little arithmetic:
60W at 8 Ohms is 21.9Vrms, which for a sine is +/- 31.0V peaks. A lossless rectifier could charge the main rail caps to 36.8V from a 26VAC secondary -- if there was no load. But an 8 ohm load will draw 3.88 amps from a 31V source (kinda heavy load). A (more typical, standard) single bridge would provide about +/-36V lightly loaded; the original design would be a little over +/-35V.
Now for the output losses: A triple EF reduces the peak-to-peak swing of the output by at least 2.2V or so, in each direction -- so about 4.5Vp-p minimum. The VAS and its load, the CCS (about which I cried 'mis-designed' in an earlier post) have an Ic of ~2.1mA to work with, allowing them to get decently close to the rail as they saturate. But remember, the pre-driver, driver, and final are all loosing beta as their Vce drops to one or two volts, too -- and this is when the highest current is asked of them (again, assuming a resistive load).
So it's really just an educated guess without a simulation, or build, but it looks to me like a continuous 60W @8 ohm output would be at least slightly modulated by 100/120 Hz ripple from the PSU. Those big 22,000 uF main rail caps will help, as would a very stiff transformer. Others here can advise better than me, as to what VA rating might be sufficient.
Your last post sounded as though you still have questions before you're ready to build.
Hope this covers a couple/3 at least.
Regards,
-Rick
Last edited:
Rick PA Stadel you are a fantastic help as always!!
You have answered most of my questions but I am thinking a lot about all the cretic of the design of this amplifier!
As I did wrote earlier I choose this amplifier because it's musicality.
It did only lacked in the bass and maybe was a tad much in the high's but properly the most musical amplifier I have hard.
But I am now thinking if your suggested changes would change its musicality, give more bass control or what It will do?
By the way I am fare from good enough to draw a new schematic from your descriptions by now but even if I was, I do not know how to simulate the circuit and what to use. I am using KiCad but it looks like it can't simulate.
You have answered most of my questions but I am thinking a lot about all the cretic of the design of this amplifier!
As I did wrote earlier I choose this amplifier because it's musicality.
It did only lacked in the bass and maybe was a tad much in the high's but properly the most musical amplifier I have hard.
But I am now thinking if your suggested changes would change its musicality, give more bass control or what It will do?
By the way I am fare from good enough to draw a new schematic from your descriptions by now but even if I was, I do not know how to simulate the circuit and what to use. I am using KiCad but it looks like it can't simulate.
There's LTSpice, built and maintained (I think) by Linear Technology, and it is free. Trouble is, all that it will help with is to adjust a resistor value here and there, in pursuit of higher output before clipping and/or lower distortion. It won't help even a tiny bit with the attribute that (it sounds like) you value most highly -- musicality. There may be no alternative to building it.
Feel free to build it just like it's drawn -- I won't be offended, and I doubt anyone else on here will be, either. If there are still issues you want help with, I'll still be here.
Bet it'll sound great!
Best regards,
Rick
Feel free to build it just like it's drawn -- I won't be offended, and I doubt anyone else on here will be, either. If there are still issues you want help with, I'll still be here.
Bet it'll sound great!
Best regards,
Rick
I am thinking of building an "original and a Rick-version
I do not know if I am asking to much, but if I make a new schematic with designations on every component, would you then write down, the changes that you feel would make it better?
An other thing is that many recommend a impedance of minimum of 47kOhm to about 100kOhm, to make it easier for the source to work and because I have read that higher impedance give a better bass.
I do not know if I am asking to much, but if I make a new schematic with designations on every component, would you then write down, the changes that you feel would make it better?
An other thing is that many recommend a impedance of minimum of 47kOhm to about 100kOhm, to make it easier for the source to work and because I have read that higher impedance give a better bass.
Glad to hear you are ramping up your ambition - sounds like a winning approach.
That doesn't sound like 'asking too much' to me. I will -- but should warn that what I've suggested isn't about 'what I feel would make it better'. It isn't about taste or personal preference, just following where the arithmetic leads.
The change from 39K to 47K is harmless -- but you'll have to be sure to reduce the 220pF cap to ground. Also unlikely to improve bass, unless what you're driving it with has too small or low quality of capacitor on its output.
All the best,
Rick
That doesn't sound like 'asking too much' to me. I will -- but should warn that what I've suggested isn't about 'what I feel would make it better'. It isn't about taste or personal preference, just following where the arithmetic leads.
The change from 39K to 47K is harmless -- but you'll have to be sure to reduce the 220pF cap to ground. Also unlikely to improve bass, unless what you're driving it with has too small or low quality of capacitor on its output.
All the best,
Rick
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.