Jump to content

Recommended Posts

Posted

I have a leftover remote controlled 100k resistor shunt passive volume control lying around. It's not really matching my amplifier impedance as is so I'm thinking to put an buffer output stage on it. Want to do a really simple no gain buffer. found couple of candidate

https://www.ebay.com/itm/303707516390

https://www.aliexpress.com/item/4000824429551.html

https://www.aliexpress.com/item/4000824429551.html

I'm leaning towards the jfet buffer. What do you guys think. Or if anyone has any other idea of simple output buffer diy project. Another one I can think of is pass b1 but I think that uses 25k pot

Posted (edited)
5 hours ago, mloutfie said:

I have a leftover remote controlled 100k resistor shunt passive volume control lying around. It's not really matching my amplifier impedance as is so I'm thinking to put an buffer output stage on it. Want to do a really simple no gain buffer. found couple of candidate

https://www.ebay.com/itm/303707516390

https://www.aliexpress.com/item/4000824429551.html

https://www.aliexpress.com/item/4000824429551.html

I'm leaning towards the jfet buffer. What do you guys think. Or if anyone has any other idea of simple output buffer diy project. Another one I can think of is pass b1 but I think that uses 25k pot

 

I'm confoosed, Mahdie ... what does a remote-controlled vol ctrl have to do with an output (zero-gain) buffer?

 

Interestingly enough, I am also thinking of an output buffer - in my case, to inject a bit of H2 into the signal.  I was using Hugh Dean's "TLP" tube buffer in my system (on the mids & tweeters) - but recently removed it, because of the noise it produced.  Hugh said he could design a ss cct which would add some H2 ... not as much as a tube - but still some.  I had coffee with him last Wednesday and he drew it out for me on the spot!  :thumb:

 

(It's much simpler than the one @r3xposted.)

 

Andy

 

Edited by andyr
Posted
47 minutes ago, andyr said:

I'm confoosed, Mahdie ... what does a remote-controlled vol ctrl have to do with an output (zero-gain) buffer?

It's a 100k volume control so not really good impedance matching with my amps. That's why thinking to make a buffer the reduce the impedance that the amplifier going to see.

 

Care to share the diagram? Which jfet does it use?

  • Like 1
Posted
12 minutes ago, mloutfie said:

It's a 100k volume control so not really good impedance matching with my amps. That's why thinking to make a buffer the reduce the impedance that the amplifier going to see.

 

Care to share the diagram? Which jfet does it use?

 

Aah, OK - yes, that sorta makes sense.  :thumb:

 

The "SSTB" ("Solid State Tube Buffer"  :) ) uses a 2SK170.  I'll have to ask Hugh whether I can give you the cct diagram.

 

Andy

 

Posted
18 hours ago, r3x said:

here you go. Q1-4 are 2sc945 or equiv.

ElektorBufferOptimised.png

I'm not great at reading diagram but this is how I see it. is that V1 AC? Doesn't look right. V2 and V3 both DC +? and I assume one coupled with R5 is input? quite low input impedance of 23k? not sure that's quite suitable as buffer for 100k pot

Posted
2 hours ago, mloutfie said:

I'm not great at reading diagram but this is how I see it. is that V1 AC? Doesn't look right. V2 and V3 both DC +? and I assume one coupled with R5 is input? quite low input impedance of 23k? not sure that's quite suitable as buffer for 100k pot

c1 is your input cap. Impedance is north of 150k.

Posted

I've got spare B1 buffer pcbs if you want to try this ...

 

On "diyAudio" it was known as the Hypnotize or Mesmerize project and is basically a Salas shunt regulator driving a set of jfet buffers (2sk170/2sj74)  - I think one of the members still has parts kits available.

 

One of the main attributes of this more complicated project is that any simple buffer circuit works best with very clean power supplies, not just the usual 3 pin units like lm317/337, for example.

Posted
On 7/18/2021 at 2:49 PM, mloutfie said:

I'm not great at reading diagram but this is how I see it. is that V1 AC? Doesn't look right. V2 and V3 both DC +? and I assume one coupled with R5 is input? quite low input impedance of 23k? not sure that's quite suitable as buffer for 100k pot

Hi Mahdi

C3 in the schematic as I see it will form a low pass filter, rolling off audio high frequencies.  I would breadboard it first to ascertain what it does. 

But guaranteed it will add various flavors of reactance,  where there was little if any before.  If you really have to experiment with it,  get c3 out of the audio path, and instead turn Q2 into a capacitance multiplier current source  instead,  by grounding the capacitors negative end 

 

Much better though  is to come to terms with consumer line level and have your power amp with matching sensitivity to consumer line level. 

You can then  passive resistor attenuate in between, and avoid adding any reactance - that always totally upsets what might have been possible.  

 

Conventional passives cannot adequately individually tailor the rates of shunt and series resistance, which is exactly what you need. Typical figures you should see at mid volume  are 12k on shunts and 4k or less on series. The higher shunt resistance allows dynamics to be reproduced. Get these aspects right, and having them adjustable too via  the usual contact less means ...  and you are there.   

 

 

  

Posted
10 hours ago, stereo coffee said:

 

C3 in the schematic as I see it will form a low pass filter, rolling off audio high frequencies.  I would breadboard it first to ascertain what it does. 

But guaranteed it will add various flavors of reactance,  where there was little if any before.  If you really have to experiment with it,  get c3 out of the audio path, and instead turn

 

 

  

 

Err, no.  C3 is there to AC couple the base of Q2 to the audio signal.  Sure it's kind of in the audio path, but not in the way you imply.

Harmonic distortion is unmeasurable with my terratec pro-audio interface, it actually lowered THD compared to loopback.  IIRC it was something like 0.004% down to 0.003%.

You'd need some serious equipment to actually be able to characterise the circuit in real life.

ElectorBufferFreqResponse.png

ElectorBufferHD.png

Posted
2 hours ago, r3x said:

 

You'd need some serious equipment to actually be able to characterise the circuit in real life.

ElectorBufferFreqResponse.png

ElectorBufferHD.png

So its just a simulation, not a real-life actual circuit  ? 

Posted

Surprised you've not looked at transformers for this..... Sound be a simple, passive way to do impedance matching.

Posted (edited)
16 minutes ago, stereo coffee said:

So its just a simulation, not a real-life actual circuit  ? 

Did you read what i said or just look at the pictures?

 

I listened to it just last night.

Edited by r3x
  • Like 1
  • Haha 1
Posted
1 hour ago, r3x said:

Did you read what i said or just look at the pictures?

 

I listened to it just last night.

Well you have used your own serious equipment  

 

4 hours ago, r3x said:

 

You'd need some serious equipment to actually be able to characterise the circuit in real life.

ElectorBufferFreqResponse.png

ElectorBufferHD.png

 Glad you have real life,  partially characterised the circuit by listening to it last night, I am still none the wiser as to why it is needed in your system.  Could I guess that is  wrong choice of power amp sensitivity, now being compensated for, or is it something else ? 

 

What is your subjective opinion on how it presently sounds, and what do you attribute that to ?   

Posted
21 minutes ago, stereo coffee said:

Well you have used your own serious equipment  

 

 Glad you have real life,  partially characterised the circuit by listening to it last night, I am still none the wiser as to why it is needed in your system.  Could I guess that is  wrong choice of power amp sensitivity, now being compensated for, or is it something else ? 

 

What is your subjective opinion on how it presently sounds, and what do you attribute that to ?   

The buffer sits after a ladder attenuator, before a dividing network that splits the signal up into 2 high pass and 1 low pass outputs.  These passive dividing networks require constant impedance being seen both upstream and downstream.

 

The sound?  I can't separate it out from the other changes.  Combined with the upgraded filters the SQ jump was huge.  It's better than just a passive device, in my unique situation and probably better than a passive devices in most situations, but as always YMMV.  It was the simplest circuit (emitter follower) backed up with best linearization i'd yet seen.

 

I took the circuit, drew it out in microcap.  Swapped out the transistors for the best low noise transistors i had lots of, messed about with values and biases, implemented it and tried unsuccessfully to measure it.

  • Love 1
Posted
On 7/18/2021 at 3:16 PM, r3x said:

c1 is your input cap. Impedance is north of 150k.

Sorry I would need a bit of hand holding to understand that diagram. Back to my original question V2 and V3 both positive? And is that V1 the signal output I can see the input but I can see the output. Also for your other post already have a tube preamp so solid state for this project.

Posted
On 7/18/2021 at 3:50 PM, HdB said:

I've got spare B1 buffer pcbs if you want to try this ...

 

On "diyAudio" it was known as the Hypnotize or Mesmerize project and is basically a Salas shunt regulator driving a set of jfet buffers (2sk170/2sj74)  - I think one of the members still has parts kits available.

 

One of the main attributes of this more complicated project is that any simple buffer circuit works best with very clean power supplies, not just the usual 3 pin units like lm317/337, for example.

the one I'm not sure with a b1 project if 100k stepped attenuator would be suitable. cos that's the spare parts I have

3 hours ago, MattyW said:

Surprised you've not looked at transformers for this..... Sound be a simple, passive way to do impedance matching.

Would that work? I would have though s transformer will significantly drop the signal voltage coming from high to low impedance

Posted
27 minutes ago, mloutfie said:

Sorry I would need a bit of hand holding to understand that diagram. Back to my original question V2 and V3 both positive? And is that V1 the signal output I can see the input but I can see the output. Also for your other post already have a tube preamp so solid state for this project.

 

V2 +15V, V3 is -15V.  R5 is the load resistor.  Increase the R5, 23k to something like 220k or more (just to make sure that the output cap C2 is never charged up when you plug something in) and that is your output.  R9 is 47k.

 

You can simplify this circuit a lot, at the expense of performance.  Q2 can be replaced with a simple link between emitter and collector and Q3 & Q4 can be replaced with a resistor.  You'd be left with a simple emitter follower with slightly less gain and much more distortion.

 

https://en.wikipedia.org/wiki/Emitter_follower

 

Circuit implementations are common.  

Posted
23 minutes ago, r3x said:

Increase the R5, 23k to something like 220k or more (just to make sure that the output cap C2 is never charged up when you plug something in) and that is your output

I thought that's the input. So where is the input?

Posted
6 minutes ago, mloutfie said:

I thought that's the input. So where is the input?

V1 is the input.  Imagine there is an RCA socket there, or perhaps the wiper/switch on your pot/attenuator.

Posted (edited)
1 hour ago, mloutfie said:

the one I'm not sure with a b1 project if 100k stepped attenuator would be suitable. cos that's the spare parts I have

Would that work? I would have though s transformer will significantly drop the signal voltage coming from high to low impedance

 

Depends on the transformers used really. I've had success with various transformers though the best I've used for isolation are Western Electric 111C. Very transparent. Do something quite special to the sound though very hard to put a finger on exactly what it does. Adds a little body, a little additional spatial cues, removes digital jaggies..... Very nice overall.

Edited by MattyW
Posted
1 hour ago, r3x said:

 

V2 +15V, V3 is -15V.  R5 is the load resistor.  Increase the R5, 23k to something like 220k or more (just to make sure that the output cap C2 is never charged up when you plug something in) and that is your output.  R9 is 47k.

 

You can simplify this circuit a lot, at the expense of performance.  Q2 can be replaced with a simple link between emitter and collector and Q3 & Q4 can be replaced with a resistor.  You'd be left with a simple emitter follower with slightly less gain and much more distortion.

 

https://en.wikipedia.org/wiki/Emitter_follower

 

Circuit implementations are common.  

 

I've not made this, i've spent no time optimising it or even trying to make sure the results make much sense, but this is what  i mean...

SimpleBuffer.png

SimpleBufferFreqResp.png

Posted
On 7/17/2021 at 12:53 PM, mloutfie said:

'm leaning towards the jfet buffer. What do you guys think. Or if anyone has any other idea of simple output buffer diy project. Another one I can think of is pass b1 but I think that uses 25k pot

You can use any value pot for the B1. 

 


 

  • Like 1

  • Recently Browsing   0 members

    • No registered users viewing this page.

×
×
  • Create New...
To Top