Maby a better (and safer) solution is attenuation with resistors (24K/1K) followed by a OPA627 buffer between preamp board and amp board inside the Sugden.
The input resistor on the AMP printboard is 1K. Is this an option? With this attenuation in between both the pre-amp and amp part still see 1K.
Yes but you will not solve its way too high GAIN. That 47 Ohm to GND can not be. Your fixation on passive being not optimal is solved only by the root by changing the GAIN. If it is a good design there will be be a filter capacitor to filter out RF. That corner frequency wil change if you change the 1 kOhm. Just checked and there is indeed a 220 pF cap so take care with changing that resistor.
BTW your sources like a load of 10 kOhm or higher. Your integrated amplifier is 50 kOhm input impedance AFAIK. No trouble there unless you decide to go on with experiments and in-between stuff not making things better. Lapmiddelen in dutch.
Apparently Extreme_Boky has dealt with it before. Nothing bad can happen with that resistor clipped with micro crocodile clips specially bought to test this. Maybe you can ask him if it has been done in reality?!?! If I would be incertain I would like to make sure things will be effective and no detrimental effects. Personally I don't see a drawback except for clumsy things like clips falling off and causing shorts. Place them in deenergized situation and remove them in deenergized situation with pulled power plug.
BTW your sources like a load of 10 kOhm or higher. Your integrated amplifier is 50 kOhm input impedance AFAIK. No trouble there unless you decide to go on with experiments and in-between stuff not making things better. Lapmiddelen in dutch.
Apparently Extreme_Boky has dealt with it before. Nothing bad can happen with that resistor clipped with micro crocodile clips specially bought to test this. Maybe you can ask him if it has been done in reality?!?! If I would be incertain I would like to make sure things will be effective and no detrimental effects. Personally I don't see a drawback except for clumsy things like clips falling off and causing shorts. Place them in deenergized situation and remove them in deenergized situation with pulled power plug.
Last edited:
I think the attenuation with 1K and 47 ohm is good, because in the existing situation the transistor provides a low input resistance with 47 ohm to ground. So you have 1K + 2x 47 ohm as input impedance. Or am I wrong?
No the 47 Ohm is a series resistor to the 75 kOhm to GND. If the input impedance would be 1.047 kOhm then the 1 µF input cap would definitely not be sufficient. A 100 µF would be needed then. You have created a 1 kOhm load for the previous stage. Not all devices like 1 kOhm loads. Originally the A21Se specification says 50 kOhm input impedance so it is an easy load. I would keep it that way.
I won't repeat the cause of it all but I see it is a fight to accept it 🙂
My advice: don't do anything anymore and have personal assistance by an experienced person. If you live somewhere in the east of NL I can help out if needed.
I won't repeat the cause of it all but I see it is a fight to accept it 🙂
My advice: don't do anything anymore and have personal assistance by an experienced person. If you live somewhere in the east of NL I can help out if needed.
Last edited:
Existing input resistance is probably around 30k. Ideally, you would rejig the amplifier's feedback to reduce its gain to a sensible value. However, increasing feedback by the required amount could easily lead to stability problems, and I think spotting and solving them is way beyond you. If I were doing it for myself, I'd start by analysing the entire circuit in SPICE. I'd be prepared to bet that I'd then decide it was safer just to add resistive attenuation at the input. 110mV is stupidly sensitive.
For all that it's in a shiny new case, that circuit looks very old indeed. Probably a rehash of the original design from the 70s when 110mV sensitivity made slightly more sense.
For all that it's in a shiny new case, that circuit looks very old indeed. Probably a rehash of the original design from the 70s when 110mV sensitivity made slightly more sense.
I did a little digging into the theory of NPN transistors. I was wrong. For a 2SC2240 with an emitter resistance of 47 Ω and a typical gain of 200, the estimated input resistance of the base for AC voltages is about 9.4 kΩ. So with 1Kohm in series you get a generous 10Kohm as input resistance of the power amplifier printed circuit board.
Last edited:
@DouglasSelf talks a bit about attenuators and noise levels in his small signal book. Many modern opamps can easily drive lower resistance loads. Rather than use a high value attenuator, cut it back to something in the hundreds of ohms and drive it with a suitable opamp. No problem with input impedance or output impedance and the inherent noise level will be better than you can do with a typical higher value pot. He describes this far better than I can.
Since I don't know how the output of the preamp stage is constructed and don't want to risk a too heavy load, and also afraid of capacitive effects (and noise) of an attenuation with resistors, it seems better to me to keep my active attenuator between DAC and Sugden. I have already determined that it does not negatively affect the sound quality. In any case, thanks for thinking along!
What about drawing the preamp stage of your A21SE!?
We call it “thinking in solutions”. Don’t fear the unknown, embrace it.
We call it “thinking in solutions”. Don’t fear the unknown, embrace it.
Other thing occured to me...sometimes the idea needs to roost.
I have been doing lots of experiments with buffers and step up transformers. For simple reason, to amplify signal without negative feedback.
You need other scenario, to lower the signal, so why not to use line level trafo as step down?
I see main advantage that in step down, input impedance is hight, no issue for pre. Output impedance is low, no issue to drive anything. You can sellect how much you lower the signal, but maybe not as much as you need.
The only downside is added distortion. How much depends on line trafo, most of the time its beneficial. Like the interstage stepup trafo in M2 or Vfet amps. Or iron pre by ZM.
Just used other way around.
I never experimented this way, why would I? The goal is other way around, but you never know. These trafos are small, cheap and easy to work with, passive devices require no power. Think about it.
I have been doing lots of experiments with buffers and step up transformers. For simple reason, to amplify signal without negative feedback.
You need other scenario, to lower the signal, so why not to use line level trafo as step down?
I see main advantage that in step down, input impedance is hight, no issue for pre. Output impedance is low, no issue to drive anything. You can sellect how much you lower the signal, but maybe not as much as you need.
The only downside is added distortion. How much depends on line trafo, most of the time its beneficial. Like the interstage stepup trafo in M2 or Vfet amps. Or iron pre by ZM.
Just used other way around.
I never experimented this way, why would I? The goal is other way around, but you never know. These trafos are small, cheap and easy to work with, passive devices require no power. Think about it.
Before I started with the active attenuator I also thought about a signal transformer. A high quality transformer is expensive and they influence the sound. Since I am 100% satisfied with the current sound of the Meitner in combination with the Sugden, I chose an active attenuator of professional quality, super low noise level and a neutral character.
If your happy with the active solution please stay with it.
Now you have changed a lot(fase and distorsion), but you like it.🙂
I think the best solution was a resistor devider done the right way.(change no fase, freq resp. and distorsion)
Now you have changed a lot(fase and distorsion), but you like it.🙂
I think the best solution was a resistor devider done the right way.(change no fase, freq resp. and distorsion)
I think people put too much faith in the published circuit, but it only has a gain of 1k/47R, which does not seem anywhere close to 110mV.
There has to be a preamp stage in front, where gain modification would be trivial.