Hi all,
I am a little scared to post here as most seem very knowledgeable in regards to electronics and, well, to be honest I am quite the noob! So bear with me....
I am wondering how feasible it is to add a buffer (specifically RJM's B-Board) to the input of my LM3875 gainclone amp (built from Peter Daniel's kit). I have built up the amp boards with a passive switch and attenuator front end. Very simply: Switch->Pot->Amps. The reason I would like to add the buffer is that I have some source components (phono stage mainly) that don't really like the < 25k input load of the attenuator.
So, what I would ideally like to do is add the B-Boards between the switch and the attenuator. That way, at least in my mind, my source components would "see" the input load of the B-Boards instead of the attenuator and everyone would be happy.
Does this make sense? Do you think I could piggyback on the gainclone's PSU to power the B-Boards? I assume the transformer is waaaay to big for the B-Boards.
All info is greatly appreciated!
-Adam
I am a little scared to post here as most seem very knowledgeable in regards to electronics and, well, to be honest I am quite the noob! So bear with me....
I am wondering how feasible it is to add a buffer (specifically RJM's B-Board) to the input of my LM3875 gainclone amp (built from Peter Daniel's kit). I have built up the amp boards with a passive switch and attenuator front end. Very simply: Switch->Pot->Amps. The reason I would like to add the buffer is that I have some source components (phono stage mainly) that don't really like the < 25k input load of the attenuator.
So, what I would ideally like to do is add the B-Boards between the switch and the attenuator. That way, at least in my mind, my source components would "see" the input load of the B-Boards instead of the attenuator and everyone would be happy.
Does this make sense? Do you think I could piggyback on the gainclone's PSU to power the B-Boards? I assume the transformer is waaaay to big for the B-Boards.
All info is greatly appreciated!
-Adam
Source > cable > vol pot > cable > amplifier works IF each Source can drive the cable and the input of the receiver.
You are right to consider a Buffer for any Source that cannot drive the cable and receiver that follows.
BUT !!!!!!
the Buffer goes on the OUTPUT of the SOURCE that cannot drive what follows.
Do keep in mind that a vol pot is a very bad driver of cable and receiver simply because it has VERY LOW current capability to drive the capacitance at the high frequency end of the audio range.
You may need to add a Buffer after the vol pot so that your chosen vol pot can drive a cable and receiver.
Very short cables have very low capacitance, no cable has even lower capacitance.
ps, don't be afraid.
It's the folk that hide their limitations and deliberately mislead those that try to help that get jumped on.
You are right to consider a Buffer for any Source that cannot drive the cable and receiver that follows.
BUT !!!!!!
the Buffer goes on the OUTPUT of the SOURCE that cannot drive what follows.
Do keep in mind that a vol pot is a very bad driver of cable and receiver simply because it has VERY LOW current capability to drive the capacitance at the high frequency end of the audio range.
You may need to add a Buffer after the vol pot so that your chosen vol pot can drive a cable and receiver.
Very short cables have very low capacitance, no cable has even lower capacitance.
ps, don't be afraid.
It's the folk that hide their limitations and deliberately mislead those that try to help that get jumped on.
Last edited:
Thanks for the prompt reply Andrew!
For clarification (if it even matters) the switch and attenuator are built directly into the amp as per Peter Daniel's instructions. So there's not much in the way of cable after the attenuator -- maybe an inch or two. Still there is the impedance (the resistor on the input of the amp?) I guess. I don't think I will add a buffer between the attenuator and amp stage anyway, so let's assume that is hunky dory if we can. 🙂
Regarding placement of buffer in the chain: on the input of the amp (before the attenuator) vs the output of the source. If the buffer is placed on the input of the amp then the source will still have to drive the potentially difficult cable. If the buffer is placed on the output of the source then it doesn't need to worry about anything. Am I understanding that correctly? In both scenarios the issue of the source handling the input load of the amp is solved, right? I should add that my interconnects are all < 1m long and of the low-capacitance (microphone) variety.
The reason I ask is because logistically (not electronically -- I am no authority) it makes more sense to me to have the buffer in the amp because then it is available for all my sources. As opposed to needing a buffer for every source (in the case of adding them to each output).
Your input is appreciated 🙂
-Adam
For clarification (if it even matters) the switch and attenuator are built directly into the amp as per Peter Daniel's instructions. So there's not much in the way of cable after the attenuator -- maybe an inch or two. Still there is the impedance (the resistor on the input of the amp?) I guess. I don't think I will add a buffer between the attenuator and amp stage anyway, so let's assume that is hunky dory if we can. 🙂
Regarding placement of buffer in the chain: on the input of the amp (before the attenuator) vs the output of the source. If the buffer is placed on the input of the amp then the source will still have to drive the potentially difficult cable. If the buffer is placed on the output of the source then it doesn't need to worry about anything. Am I understanding that correctly? In both scenarios the issue of the source handling the input load of the amp is solved, right? I should add that my interconnects are all < 1m long and of the low-capacitance (microphone) variety.
The reason I ask is because logistically (not electronically -- I am no authority) it makes more sense to me to have the buffer in the amp because then it is available for all my sources. As opposed to needing a buffer for every source (in the case of adding them to each output).
Your input is appreciated 🙂
-Adam
A Buffer is needed when the Source cannot drive what follows it.
An amplifier and it's vol pot are effectively the receiver. They don't need a Buffer on the input. There is an exception, see below.
But your question does come back to :
What Sources do you have that cannot drive a 25k vol pot?
Most Sources can drive a 10k vol pot that is paralleled by some cable capacitance.
If your source/s cannot drive a big resistive load, then what can they do if they are presented with some paralleled capacitance?
Usually it's the paralleled capacitance that draws most current from a Source.
It's rarely the receiver resistance that stretches the source's current capability.
Exception:
An inverting receiver presents a low (or very low) impedance to the source. One can design a Source to drive this low input impedance, but it is often easier and simpler to "convert" the very low input impedance to a higher input impedance to suit the many sources that may have to drive the inverting stage. This can be done by adding a Buffer to the INPUT of the inverting receiver.
An amplifier and it's vol pot are effectively the receiver. They don't need a Buffer on the input. There is an exception, see below.
But your question does come back to :
What Sources do you have that cannot drive a 25k vol pot?
Most Sources can drive a 10k vol pot that is paralleled by some cable capacitance.
If your source/s cannot drive a big resistive load, then what can they do if they are presented with some paralleled capacitance?
Usually it's the paralleled capacitance that draws most current from a Source.
It's rarely the receiver resistance that stretches the source's current capability.
Exception:
An inverting receiver presents a low (or very low) impedance to the source. One can design a Source to drive this low input impedance, but it is often easier and simpler to "convert" the very low input impedance to a higher input impedance to suit the many sources that may have to drive the inverting stage. This can be done by adding a Buffer to the INPUT of the inverting receiver.
Last edited:
Hrm, well I suppose perhaps a buffer is not needed at all! I guess my intention was to mitigate potential problems-- not necessarily real ones.
To answer your question, I am mostly concerned about my valve phono stage. It has an output impedance of 4k. However, I also like to plug my phone/ipod in directly to the amp (the 25k attenuator in other words) as well. Do you think this is also okay?
Looks like you'll save me some time and money, Andrew 🙂
-Adam
To answer your question, I am mostly concerned about my valve phono stage. It has an output impedance of 4k. However, I also like to plug my phone/ipod in directly to the amp (the 25k attenuator in other words) as well. Do you think this is also okay?
Looks like you'll save me some time and money, Andrew 🙂
-Adam
Hrm, well I suppose perhaps a buffer is not needed at all! I guess my intention was to mitigate potential problems-- not necessarily real ones.
To answer your question, I am mostly concerned about my valve phono stage. It has an output impedance of 4k. However, I also like to plug my phone/ipod in directly to the amp (the 25k attenuator in other words) as well. Do you think this is also okay?
Looks like you'll save me some time and money, Andrew 🙂
-Adam
a valve stage with a 4k output impedance typically works well into a load impedance of 40k or more. ie: 1:10 or more. at 25k you are likely to experience some noticeable distortion from the valve stage being loaded too heavily. BTW are you sure about that 4k number?
Last edited:
PS: since for your purposes your valve stage is only a bit too high output impedance, the simplest solution may be to lower that impedance. often that's pretty easy. what is the valve stage design topology currently? a schematic would be nice.
a valve stage with a 4k output impedance typically works well into a load impedance of 40k or more. ie: 1:10 or more. at 25k you are likely to experience some noticeable distortion from the valve stage being loaded too heavily. BTW are you sure about that 4k number?
This is what I was afraid of, and why I originally started investigating adding a buffer in the chain somewhere. Would my original idea of adding the buffer before the attenuator provide the correct load on the phono stage? Thus reducing the potential for load-related distortion?
Regarding the phono stage, it is a Bottlehead Seduction I've built from a kit. Unfortunately I don't have the schematics on hand. Bottlehead publish it's output impedance as 4k (at least they used to). You can read about the topology here: Reduction Phono Preamplifier Kit Specs | Bottlehead Sorry I can't be of more help in clarifying.
-Adam
i followed your bottlehead link. i did not immediately spot any claim about the output impedance. but they say the two stages in each channel are 6922s (with which i have a lot of experience). since they say they amps deliver nearly 40db of gain, each of the two stages almost certainly are single ended common cathode. it's very easy to get the output impedance of a common cathode 6922 down to under 3k and with a few simple design tricks, something near half that is doable. all without a buffer. just to show how far you can go with enough effort, my own preamp has 4 parallel 6922s fed with a constant current source and has an output impedance of roughly 400 ohms (not kidding).
but since we don't have a schematic for your preamp, the path of least resistance is probably to add a buffer after your phono pre. for a very good tube like the 6922, the Pass Labs B1 (PCB and a full set of matched FETs is $40 total) or the Pass Labs Bride of Zen see: https://www.passdiy.com/project/preamplifiers/bride-of-zen or the MooseFET see: DIY Audio Projects Forum • MooseFET Preamplifier are quite suitable, cheap and easy to build.
but since we don't have a schematic for your preamp, the path of least resistance is probably to add a buffer after your phono pre. for a very good tube like the 6922, the Pass Labs B1 (PCB and a full set of matched FETs is $40 total) or the Pass Labs Bride of Zen see: https://www.passdiy.com/project/preamplifiers/bride-of-zen or the MooseFET see: DIY Audio Projects Forum • MooseFET Preamplifier are quite suitable, cheap and easy to build.
PS as a quick and dirty expedient you could simply add a 27k series resistor after the output cap in each channel of the phono preamp. you'll lose about 6db of signal but the phono pre will be happy. use a quiet resistor (the Dale RN60 carbon film available on tube depot is a reasonable compromise in terms of not affecting the phono pre's sonics). i would not use a metal film resistor in this spot but you may be OK with it.
Last edited:
Thanks for the great replies Steve.
Regarding the phono stage, I do intend on building their next up kit called the Eros (Eros Phono Preamplifier Kit Specs | Bottlehead) which they do state as having a 4k output impedance -- sorry I may have confused the two. When I build this kit perhaps I will experiment with adding resistors after the output caps. I will return to the forum for that help when the time comes 🙂
Regarding the buffers, I was looking at RJM's B-Board (http://phonoclone.com/diy-bb.html) simply based on form factor. I will check out your suggestions.
Do you think my idea of adding the buffer to the input of my amp (before the attenuator) is sound? Yes, I will lose the benefits of the buffer in regards to driving the cable. However, I will provide the phono stage with an input load it wants to see, correct?
(Along with the benefit of having the buffer there for other sources that may need it, rather than build multiple buffers)
Regarding the phono stage, I do intend on building their next up kit called the Eros (Eros Phono Preamplifier Kit Specs | Bottlehead) which they do state as having a 4k output impedance -- sorry I may have confused the two. When I build this kit perhaps I will experiment with adding resistors after the output caps. I will return to the forum for that help when the time comes 🙂
Regarding the buffers, I was looking at RJM's B-Board (http://phonoclone.com/diy-bb.html) simply based on form factor. I will check out your suggestions.
Do you think my idea of adding the buffer to the input of my amp (before the attenuator) is sound? Yes, I will lose the benefits of the buffer in regards to driving the cable. However, I will provide the phono stage with an input load it wants to see, correct?
(Along with the benefit of having the buffer there for other sources that may need it, rather than build multiple buffers)
Last edited:
Do you think my idea of adding the buffer to the input of my amp (before the attenuator) is sound? Yes, I will lose the benefits of the buffer in regards to driving the cable. However, I will provide the phono stage with an input load it wants to see, correct?
(Along with the benefit of having the buffer there for other sources that may need it, rather than build multiple buffers)
it depends.
is there a load R in the 47k range already in the phonostage output network (to help swamp cable capacitance)?
is it a short (less than 75cm) cable from the phonostage to the proposed buffer location?
if both answers are yes it will probably be OK.
without a buffer immediately after the common cathode valve stage, it can get dicey unless it's pretty close to a resistive load at least 10x the output impedance. YMMV
BTW the common solution of using a cathode follower buffer does not do well because capacitive loads need to both source and sink current with respect to the follower to work well -- simple cathode followers source current well but do not sink current well. mu followers and white cathode followers are much better at handling capacitive loads.
BTW#2 if this were my phonostage i would redesign it's last stage as a mu follower (gain and excellent load handling in one stage) -- but that would entail adding another 6922 or some other lower mu triode such as a 12au7 for the top half of each totem pole -- with the existing 6922 serving as the lower half. there was an excellent 6922 mu follower design posted by another forum member. if you're interested i'll dig up a link to it.
have you checked with other users of your buffer board to see if they have used it successfully with valve amps?
Last edited:
I really ought to dig up a schematic for these things so that I can answer your question a bit more accurately because I'm not exactly sure how things are implemented in the phono stage. On top of that I think I am going to build a new one in the very near future anyway. However I do believe the output stage is the same in both preamps, so it's worth investigating. I have the Seduction schematic at home -- I will dig it up and post it later.
Steve, you had mentioned adding resistance to the output of the phono to mitigate the loading problem. Alternatively, could I add resistance to the input of the amp? Perhaps to bring the input impedance north of 40k (resistor + attenuator). I'd lose some gain yeah, but perhaps it might solve any distortion issues.
Thanks again for taking time to help out this newbie.
-Adam
Steve, you had mentioned adding resistance to the output of the phono to mitigate the loading problem. Alternatively, could I add resistance to the input of the amp? Perhaps to bring the input impedance north of 40k (resistor + attenuator). I'd lose some gain yeah, but perhaps it might solve any distortion issues.
Thanks again for taking time to help out this newbie.
-Adam
resistance before the cable will better isolate the cable capacitance but resistance after the cable may be quieter if the pre can tolerate the cable capacitance. hard to predict in advance without schematics and knowledge of the cable's characteristics. so why not just try it one way or the other and see if you like the results.
here is another nice simple buffer: http://www.decdun.me.uk/gc/pedja.jfet.buffer.gif
here is another nice simple buffer: http://www.decdun.me.uk/gc/pedja.jfet.buffer.gif
as a shortcut that may help in lieu of a schematic:
1. what is each anode R value (one each is connected to pin 1 and pin 6)
2. what is the cathode bias? is it just an LED (pin 3 and pin 8).
3. is LED red (2volt) or another color?
4. what is the voltage of the plate supply?
1. what is each anode R value (one each is connected to pin 1 and pin 6)
2. what is the cathode bias? is it just an LED (pin 3 and pin 8).
3. is LED red (2volt) or another color?
4. what is the voltage of the plate supply?
Last edited:
- Status
- Not open for further replies.
- Home
- Amplifiers
- Chip Amps
- Adding buffer to Gainclone (LM3875) input