Can you actually wind the MAC opt traffos?
I know Dennis Hoyer and we talk every couple weeks but the Mc30s AND Mc60s are C core. I wasn't aware the Mc75 and Mc275 were..I have two bad Mc225 OPT that need to be rewound.
Mike
I know Dennis Hoyer and we talk every couple weeks but the Mc30s AND Mc60s are C core. I wasn't aware the Mc75 and Mc275 were..I have two bad Mc225 OPT that need to be rewound.
Mike
There are a couple things here..What most people don't realize about the Mcintosh unity coupled output circuit is it's not a really a voltage amplifier in the output stage..This is why the driver has to be bootstrapped where you are putting about 340v PP of audio on the grids of the 6550s. The output stage is acting more as a cathode follower and it amplifies current as pulls its energy from the power supply and it actually swings the voltage from the grids..Notice when you pull the output tubes you can still hear your speakers when they are hooked up and it's not distorted but you have reduced bass and power of course..Also,if you measure the RMS voltage on G1 with reference ground, and then measure the plate with reference to ground,you actually see a reduction in the RMS voltage.
Current increases with signal and I realized this putting 6L6gcs in my Mc60s once. Static DC voltages for a 6L6gc and a 6550 run about 430vdc on the plate and -45dc on G1..At moderate levels it sounded fantastic but when I cranked up the volume,that's when my GE 6L6gcs literally melted a dent in the glass..I'm not kidding..LOL This is when I did a lot of technical investigation to find out why.
Last edited:
we did not push thru with that.....the scheme i proposed is wrong, sorry...
i have found a vendor doing those OPT's....Merren Audio - McIntosh
The McIntosh output stage has an effective mu of 2 due to the 50% cathode feedback. So it does do some voltage amplification, just far less than most output stages.
BTW, the driver does not have to be bootstrapped. I made a Unity-Coupled amp without a bootstrapped driver and it works really well. You just have to pay a lot of attention to the design of that driver.
As far as transformers go, I have used the Plitron Unity-Coupled transformers.
The big drawback is that they have a non-optimal turns ratio for 8 Ohm speakers. They are rated for 75W but it is difficult to get there because they have a 1K+1K : 5R impedance ratio. Nobody makes 5 Ohm speakers so you have to choose whether you are going to connect 4 Ohm or 8 Ohm speakers. If you choose 8 Ohm speakers, the shallow load line doesn't really allow you to get 75W out of them. I got 40W out into 8 Ohms.
They do have a really cool feature, though, a dedicated feedback winding. So you could apply feedback in some unusual ways.
And they have ridiculous bandwidth. I was sweeping the low end with small signals and saw no reduction in amplitude all the way down to 10Hz. I'm going to have to get my hands on another sig gen to find the point where it is 3 dB down.
The big drawback is that they have a non-optimal turns ratio for 8 Ohm speakers. They are rated for 75W but it is difficult to get there because they have a 1K+1K : 5R impedance ratio. Nobody makes 5 Ohm speakers so you have to choose whether you are going to connect 4 Ohm or 8 Ohm speakers. If you choose 8 Ohm speakers, the shallow load line doesn't really allow you to get 75W out of them. I got 40W out into 8 Ohms.
They do have a really cool feature, though, a dedicated feedback winding. So you could apply feedback in some unusual ways.
And they have ridiculous bandwidth. I was sweeping the low end with small signals and saw no reduction in amplitude all the way down to 10Hz. I'm going to have to get my hands on another sig gen to find the point where it is 3 dB down.
2SpreadSpectrum: MC275 final stage works with 100% local negative feedback and is not amplifying signal in any way, it is just feeding the signal coming from driver with current. Without anode-cathode 1:1 it is not true unity coupled, otherwise it is just normal stage with some feedback in cathode and is not working as unity.
Unity coupled means that you eliminate capacity between positive and negative wave in output transformer because they act in same paralleled way and you can't do that without unity coupling! 🙂 That is the reason why unity coupled transformer can go so high without problems...
Unity coupled means that you eliminate capacity between positive and negative wave in output transformer because they act in same paralleled way and you can't do that without unity coupling! 🙂 That is the reason why unity coupled transformer can go so high without problems...
You are correct.
You explained it very well to..Ive been going round and round with a friend of mine on this who believes there is a 2 to 1 voltage gain in the OPT and I keep telling him there is not.Thanks for the post..
You explained it very well to..Ive been going round and round with a friend of mine on this who believes there is a 2 to 1 voltage gain in the OPT and I keep telling him there is not.Thanks for the post..
I have not measured any voltage amplification in the output stage
It simply swings that voltage that exist on g1. As Wilks said it's 100% local NFB so I don't see how you could amplify that..In the case of the Mac to get the power and distortion numbers it has,it has to be bootstrapped..I'm sure the circuit could be redesigned but it is what is in this case. Dennis Hoyer actually knew Sidney Corderman and Frank McIntosh and he's really the go to guy on the Mac circuit..When you call McIntosh and you need a transformer rewound,he is the guy they send you to..Give him a call sometime..He's quite a guy to talk to.
It simply swings that voltage that exist on g1. As Wilks said it's 100% local NFB so I don't see how you could amplify that..In the case of the Mac to get the power and distortion numbers it has,it has to be bootstrapped..I'm sure the circuit could be redesigned but it is what is in this case. Dennis Hoyer actually knew Sidney Corderman and Frank McIntosh and he's really the go to guy on the Mac circuit..When you call McIntosh and you need a transformer rewound,he is the guy they send you to..Give him a call sometime..He's quite a guy to talk to.
Last edited:
WOW! What did you build there?
I get dizzy just looking at that..Lots of detail..I'll read it tho.
I get dizzy just looking at that..Lots of detail..I'll read it tho.
It is not 100% feedback. If it were, it would take a peak input voltage swing comparable to 2x the B+ to drive to clipping, like a cathode follower does. Cathode follower is 100% feedback. Unity-Coupled doesn't take that much drive.
It ends up taking about half the input voltage swing to drive to clipping that a cathode follower at the same B+ would take (big clue that it is 50% feedback).
Unity-Coupled doesn't refer to effective mu of the output configuration. It applies to the fact that the cathode and anode loads are tightly-coupled. You wouldn't be the first to make that mistake, though.
It ends up taking about half the input voltage swing to drive to clipping that a cathode follower at the same B+ would take (big clue that it is 50% feedback).
Unity-Coupled doesn't refer to effective mu of the output configuration. It applies to the fact that the cathode and anode loads are tightly-coupled. You wouldn't be the first to make that mistake, though.
MC275 driver stage is capable of doing approx. 160V RMS!! into KT's grid 1's. The same amount of voltage swing is on KT's anodes and same on cathodes. Final stage is not amplifying anything, just applying current to the signal and with cooperation of transformer transfers it into lower load.
It is indeed huge amount of driving, to make some current to that HUGE voltage swing that's why there are cathode followers to apply current to this swing to make it possible to drive final stage with acceptable distortion.
It is indeed huge amount of driving, to make some current to that HUGE voltage swing that's why there are cathode followers to apply current to this swing to make it possible to drive final stage with acceptable distortion.
Wiliks,
Please see the following link: https://ia802500.us.archive.org/16/items/am-1959-10/Audio-1959-Oct.pdf
Go to page 52 and you will find a detailed article by Norman Crowhurst on McIntosh Unity-Coupled amps. Pay close attention to the section on page 56 titled "50 Per Cent Feedback."
Please see the following link: https://ia802500.us.archive.org/16/items/am-1959-10/Audio-1959-Oct.pdf
Go to page 52 and you will find a detailed article by Norman Crowhurst on McIntosh Unity-Coupled amps. Pay close attention to the section on page 56 titled "50 Per Cent Feedback."
I can see, but in their common language to readers as: what may be called as 50 per cent feedback they mean it metaphorically as 50% is in anode and 50% in cathode I suppose?
But from electrical point of view it is absolute feedback very same as follower in my opinion, because everything that is conducted and "amplified" by anode winding, is immediately "decreased" by cathode winding. Which means it cancels each other and means no amplification.
I also have LT spice simulation for whole amp and the final stage is not amplifying anything. Probably depends on what version of MC it is because there are about 6 different mk versions of it.
But on the other hand there are many unity coupled amps made by MC and can differ, as they mentioned in the beginning of the article: the words "unity coupling" are applied
to more than one circuit.
But from electrical point of view it is absolute feedback very same as follower in my opinion, because everything that is conducted and "amplified" by anode winding, is immediately "decreased" by cathode winding. Which means it cancels each other and means no amplification.
I also have LT spice simulation for whole amp and the final stage is not amplifying anything. Probably depends on what version of MC it is because there are about 6 different mk versions of it.
But on the other hand there are many unity coupled amps made by MC and can differ, as they mentioned in the beginning of the article: the words "unity coupling" are applied
to more than one circuit.
Feedback is roughly 100% .... since output stage is unity gain...
Having split load in plate and cathode only limits the max voltage swing to half ....
Having split load in plate and cathode only limits the max voltage swing to half ....
What about converting Lockharts transformer to KT88 ?
http://www.tubebooks.org/Books/lockhart.pdf
Kim
Let me put it this way, if the output stage can be driven to clipping with less than 200Vrms drive, how can it be unity-gain? How can that be possible?
How can an output tube with a 450V+ supply be driven to saturation with only a little over half of the total output swing votage? The answer is: because it has a gain of two.
Since the load is split between the anode and cathode, the total output voltage swing is also split between the two, but only one side (the cathode side) gets summed with the input signal in anti-phase. If this is 100% feedback, how is the anode signal being fed back to the input? It isn't. We have output signal that is outside the feedback loop.
The term 'Unity-Coupled' does not refer to the total gain of the circuit, it refers to the very tight coupling between the bifilar-wound primaries.
Trust me, I have put a lot of thought into this and I have taken measurements on my own home-brew Unity-Coupled amp. It takes half of the input voltage to drive the output stage to clipping as it would to drive a cathode follower (a true example of 100% feedback) output stage to clipping. That right there alone should settle it.
But like I said, it is a common misconception to get confused by the 'Unity' in Unity-Coupling.
Last edited:
And that is exactly what you are missing, the fact that not all of the output signal is working to decrease the input signal, only half of it is. That is why this output stage is the very definition of 50% feedback.
Edit: Try this: imagine applying 10V to the grid of the output tube. That leads to a 10V (roughly) increase in the cathode voltage but also a 10V decrease in the anode voltage. The output stage swings a total of 20V across the entire primary half but only 10 of those Volts are fed back to the input signal via the cathode winding. Imagine a Unity-Couple output stage with a 450V B+. The output tube will saturate with around 210Vpk input because it is being squeezed between the two windings.
In a cathode follower working into the same primary impedance with 450V B+, there would only be roughly 10V of output swing for 10V input (again, what you would expect with 100% feedback). It would take over 400Vpk to drive it to saturation. Roughly twice as much voltage drive is required, which is what you would expect with twice the feedback.
Last edited:
No SpreadSpectrum, it does not have gain of two. Again...
to understand how the circuit works you need to understand how the output transformers are winded and how the winding works. When you wind transformer and want to wind lets say 1 volt, you make several turns, but when you make the same amount of turns but in opposite way, you cancel previous turns and result is zero voltage. What I want to say is that the unity coupling circuit is driving transformer only by that low voltage swing, not both together counted (anode+cathode) they act as one. Because the windings are coming from same direction to the transformer so voltage is basically the same but current goes in opposite way.
That means you need exactly about 200VRMS to drive output transformer and that is your half what you are talking about. It is not 400V. Also the output transformer has impedance ratio roughly 10:1 that is another advantage of unity coupling because you do not have many turns in primary and so have lower losses and also have higher chance to get halves of the primary to get perfectly matched. When you create standard PP amp you need to know Raa impedance of the transformer, let's say for pair of KT's that would be about 3600R. But you have two windings one in cathode and one in anode so you can split it in half and you get 1800Raa for cathode circuit and 1800Raa for anode circuit. But as this is working with great amount of feedback and huge amount of global feedback as well this strongly builds capability to drive lower impedances so you can nearly divide this again into half. Than you get 900Raa for anodes and 900Raa for cathodes. So basicaly what you need is bifilarly wound two symmetrical primaries of Raa=900R (beware, you need very good double or more isolated wire to do this, because there is 500VDC between the wires all the time, and to be very caution whne winding to not damage isolation). Now this is very close to 10:1 ratio if Rz=8R. And since this two waves are not counting together 1+1 but they are acting as parallel, thats also the reason why this circuit is sometimes called as "paralleled symmetric push-pull" circuit and they are creating in transformer core just half of the B+ voltage into swing. And trust me it works as I am decribing, I designed two different types of MC275 transformers from scratch, and I have built two different DIY versions of MC275 amplifier and they all works great 🙂
As cerrem said: thats why it is called unity gain amp, because the gain is unity = 1, the input and the output between two devices are the same level.
to understand how the circuit works you need to understand how the output transformers are winded and how the winding works. When you wind transformer and want to wind lets say 1 volt, you make several turns, but when you make the same amount of turns but in opposite way, you cancel previous turns and result is zero voltage. What I want to say is that the unity coupling circuit is driving transformer only by that low voltage swing, not both together counted (anode+cathode) they act as one. Because the windings are coming from same direction to the transformer so voltage is basically the same but current goes in opposite way.
That means you need exactly about 200VRMS to drive output transformer and that is your half what you are talking about. It is not 400V. Also the output transformer has impedance ratio roughly 10:1 that is another advantage of unity coupling because you do not have many turns in primary and so have lower losses and also have higher chance to get halves of the primary to get perfectly matched. When you create standard PP amp you need to know Raa impedance of the transformer, let's say for pair of KT's that would be about 3600R. But you have two windings one in cathode and one in anode so you can split it in half and you get 1800Raa for cathode circuit and 1800Raa for anode circuit. But as this is working with great amount of feedback and huge amount of global feedback as well this strongly builds capability to drive lower impedances so you can nearly divide this again into half. Than you get 900Raa for anodes and 900Raa for cathodes. So basicaly what you need is bifilarly wound two symmetrical primaries of Raa=900R (beware, you need very good double or more isolated wire to do this, because there is 500VDC between the wires all the time, and to be very caution whne winding to not damage isolation). Now this is very close to 10:1 ratio if Rz=8R. And since this two waves are not counting together 1+1 but they are acting as parallel, thats also the reason why this circuit is sometimes called as "paralleled symmetric push-pull" circuit and they are creating in transformer core just half of the B+ voltage into swing. And trust me it works as I am decribing, I designed two different types of MC275 transformers from scratch, and I have built two different DIY versions of MC275 amplifier and they all works great 🙂
As cerrem said: thats why it is called unity gain amp, because the gain is unity = 1, the input and the output between two devices are the same level.
Last edited:
- Status
- Not open for further replies.