I do agree the simpler circuits are better. With work it is possible to attain distortion numbers comparable to the McIntosh units with them. The main crux is the iron. You need to have it custom wound for best results, and specify what exactly you want. With general run of the mill, the upper frequencies will distort pretty badly above 10k. In reality that does not matter as harmonic power is inaudible above 10k anyways.
For some, numbers do not matter, just the sound and tone.
For some, numbers do not matter, just the sound and tone.
One thing about many local NFB. Every time you do a NFB, you have to have pole and zero stuff. It is going to be hard to get all the pole and zero beyond the audible frequency range. This will cause phase shift, distortion and WILL affect the sound. The more trickery you put in, the more you change the sound, the more you distort the sound. I just don't see it can be better.
The more circuit you have, the more wires/traces the signal has to go through. The more chance of cross coupling you get. Case in point, I designed a guitar amp and I prototyped it. I did very careful layout and wiring. But when I start changing the design and put in more stuff, I had to bandage it in. Finally I build the new amp with the exact same circuit, it sounded different and much better. This is because I optimized the layout of the new amp with the newly added in circuit. The two circuit are the same, but the sound is very different. The old one always have a strange type of sound that I can never get rid of. the new amp has none of the strange sound.
Just because it's McIntosh, don't assume they know how to layout, seen too many bad layout in my life already. The more complicated it is, the more you can have mistake and problem. The best layout is to have input to the first stage, to LTP, to the power tubes, to the OTP, to the ouput jack all in one straight line. That does not necessary give you the best looking layout visually. Mc I saw in the video has all the components nicely line up visually to show off it's a tube amp, but the signal path has to be bent. I assume the input line has to route to the preamp tube at the back to the front, then the signal goes sideways before going to the power tubes and OPT at the back. You created a big loop for crosstalk.
The more circuit you have, the more wires/traces the signal has to go through. The more chance of cross coupling you get. Case in point, I designed a guitar amp and I prototyped it. I did very careful layout and wiring. But when I start changing the design and put in more stuff, I had to bandage it in. Finally I build the new amp with the exact same circuit, it sounded different and much better. This is because I optimized the layout of the new amp with the newly added in circuit. The two circuit are the same, but the sound is very different. The old one always have a strange type of sound that I can never get rid of. the new amp has none of the strange sound.
Just because it's McIntosh, don't assume they know how to layout, seen too many bad layout in my life already. The more complicated it is, the more you can have mistake and problem. The best layout is to have input to the first stage, to LTP, to the power tubes, to the OTP, to the ouput jack all in one straight line. That does not necessary give you the best looking layout visually. Mc I saw in the video has all the components nicely line up visually to show off it's a tube amp, but the signal path has to be bent. I assume the input line has to route to the preamp tube at the back to the front, then the signal goes sideways before going to the power tubes and OPT at the back. You created a big loop for crosstalk.
Last edited:
I basically am laying out my preamp in this fashion. Since it is two channel, there are two parallel lines with the power supply on one side of the chassis. Wires run from the back RCA jacks to the volume control which is on the front. From there it is a straight line back to the output jack. What do you think of this layout? It is the best I could come up with that balances pragmatics and optimal performance. The design has no feedback.
Ha ha, we think alike!!! I truly believe less is more. Put a cover over the tubes, then you don't have to worry about how it looks!!!
I am green in audiophile, but I can tell you in guitar amp, just because it's famous and sell a lot does not mean their amps are good engineered. Case in point, Fender Super Sonic, it has so many design problems that people question whether they know what they are doing. Like using opamp to drive relays!!!!
Last edited:
That's one of the reasons that it's useful to study control theory and understand feedback- it prevents severe misconceptions like this.
Yes, sideways signals are a huge and yet unrecognized problem.
For those that are not familiar with Alan0354's work, here is an example for your reference. Perhaps McIntosh can learn a thing or two...
An externally hosted image should be here but it was not working when we last tested it.
Last edited:
McIntosh amps are basically laid out in a line. It is much easier to produce a clean circuit in this configuration. I do find it hard to believe running a signal sideways will create a much of any problem, unless you have a general poor layout scheme. This will create other problems as well.
Read the Crowhurst article I mentioned. The Mac amps were a design lesson- the transformers were an integral part of it, but there's a lot of other gold in that mine for anyone who actually wants to learn something.
The latest QM theory shows that all is just classical mechanics with interference coming from adjacent universes. So you need to make sure that your "neighbors" are not playing annoying music. 🙂
I've heard that meditation can best get the message across. Something about coherence. So this could be the next big audiophool, I mean audiophile, enhancement.
--------------------------------------------------
The local Neg. Fdbks typically will reduce the apparent phase shifts from enclosed circuitry, so are usually helpful for the global N Fdbk stability. Its more a question of where gain is best spent.
,
I've heard that meditation can best get the message across. Something about coherence. So this could be the next big audiophool, I mean audiophile, enhancement.
--------------------------------------------------
The local Neg. Fdbks typically will reduce the apparent phase shifts from enclosed circuitry, so are usually helpful for the global N Fdbk stability. Its more a question of where gain is best spent.
,
Last edited:
McIntosh's older gear was well made and the*newer solid-state gear is also designed quite well, although, the prices are forbidding.*There is, however, little that could be considered particularly special about them. If it was not for the aesthetic appeal they'd probably not sell, nor, hold any perceived resale value.
Conrad Johnson's tube gear is mediocre at best, literally being nothing more than standard college book examples of triodes wrapped up in pricy champagne chassis. They prey on the unknowing and uninformed.
Conrad Johnson's tube gear is mediocre at best, literally being nothing more than standard college book examples of triodes wrapped up in pricy champagne chassis. They prey on the unknowing and uninformed.
I would have to agree with both comments..
Yes, it doen's look good. BUT this is a channel switching, the distortion channel has 4 stage gain to get the liquid high gain sound, with adjustable power output from 2W to 33W. ABSOLUTELY NO OSCILLATION. It is so high gain that the first OPT emitted enough EM to feedback to the guitar to cause oscillation. I found the cause and change the OPT to one enclosed on the side and solve the problem.
I am absolutely proud of this layout. The layout works one time through without any adjustment at all. It works like a champ and sounds great. I design two of this type of guitar amp recently and they all work one time through.
So please don't cross forum to smear me.
It just looks crowded, more than "sloppily" laid out. If it works and sounds good, so be it. If it were being marketed, another approach towards construction would be necessary to allow for more efficient production.
We have our history in the other forum, you definitely not showing my picture for good looks. I stand by my design and layout. 4 stages of gain of average 40 each is A LOT OF GAIN by any standard and that is before the power amp. I design RF circuits and layout RF boards with the same technique and theory.
RF is a different animal. You really need to be meticulous about layout there, especially in the GHz range up near the limits of copper. You can get away with much more in audio.
It's a small chassis for all the features. The adjustable power take up a lot of space as you can see the heat sink and all. Circuit is need to track the -ve bias when scaling the HT. Then muting circuit is needed to prevent the popping sound when channel switching. Because of the high gain, filaments of the front end are rectified DC. This is an existing chassis with big cutouts, A lot of space at the top left is wasted, The chassis is only 18" wide, there's not much space to work with. The signal goes from input at the left bottom to the right, then go straight back to the power stage. The power stage and filter cap are all one point ground close to the upper right to ensure no ground current go onto the chassis. The front end using the chassis as ground plane as the chassis has very little current run through it.
I have no intention to market it. I put in my 30 years as an engineer, unless opportunity just drop on my lap ( like the day pigs can fly!!!). This is just a hobby. Yes, if pigs does fly, I definitely do it in pcb form. There is no way to reproduce it.
The concept works the same, just less critical. EM radiation happens the same at audio frequency as RF, wire loops still pick up EM at audio frequency, just not as efficient.
As messy as it looks, I try my best to design signal path point to point particularly at the input. The wires that tied as a bundle are all DC wires like HT, DC filament etc. Layout is everything. Luckily, guitar amp mainly work up to 5KHz, so you do have an option putting small caps to the ground to cut the high frequency. Audiophile works to 100KHz, it's a different animal, I don't think you can have this kind of gain in this small chassis with audiophile.
Last edited:
- Status
- Not open for further replies.