MRupp said:
I've not tried any pots, as they don't track well enough for use in a balanced design. However based on my experience with them in other products, it seems that their deficiencies tend to be subtractive and not additive. This is infinitely preferable compared to the other way round. In other words, losing a bit of resolution with a pot is far better than adding grit, grain, and glare with an op-amp based volume control like the Crystal (Cirrus) and Burr-Brown electronic volume controls.
I've never tried the TKD's. They are difficult to source here in the US and seem quite expensive for what is offered. When I first did listening tests on the Shallcos I found them to be next to perfect, so I didn't spend a lot of time looking further. In addition, they are great to work with. The prices are reasonable, they will make any custom configuration we require, and have consistent deliveries.
Sigurd Ruschkow said:
If there is a "click" it could be a situation of either DC across the switch or else a manifestation of either make-before-break or break-before-make, depending on the circuit configuration. For most DIY'ers, this is not a problem. But for a commercial product it is a kiss of death. The attitude among normal customers is "If a $300 receiver doesn't make clicks, why does a $3,000 preamp do so?"
Sigurd Ruschkow said:
The problem with this schematic is that it will attenuate differential signal perfectly, but common-mode signals not at all. So as you turn the volume down, the CMRR of the overall circuit is reduced by the factor of your attenuator. If you circuit has 50 dB CMRR and you are listening at -40 dB, then you only have 10 dB CMRR left.
Sigurd Ruschkow said:
I have not listened to either one. Our latest design uses 240 resistors just for the volume control (60 steps times 4 sections for a balanced circuit). If we used $5 resistors the retail price of the unit would go up significantly. Plus they wouldn't physically fit in that design. As always, audio design is always a series of tradeoffs.
Hello Charles,
it is not DC voltage that causes the click sound, so it must be the ELMA rotary switches that is the cause. I can stand the clicks but they are annoying as it feels that the design is not finished. Personally, the clicks have to go.
The fixed series/variable shunt attenuator is not the perfect solution as you point out with the common mode issue. It is a compromise.
Still, with it I can have two fixed resistors of the TX2352 nude Vishay type, and use PRP resistors for the rest,
and have at most two TX2352s in series and one PRP resistor parallel to the signal path instead of maybe 200 PRP resistors.
Spontaneously, I would think that my solution has the sonical advantage.
Resistor size is a problem when using hole mounted types, but with SMD resistors it should be able to fit them all. I wonder if there are any audiophiles out there that would pay the extra cost of having an all-BulkMetalFoil-resistor attenuator with 240 resistors?
It surely would be interesting to have Stereophile test your new preamp with both types of attenuators.
Attached is a balanced series attenuator schematics.
Sigurd
it is not DC voltage that causes the click sound, so it must be the ELMA rotary switches that is the cause. I can stand the clicks but they are annoying as it feels that the design is not finished. Personally, the clicks have to go.
The fixed series/variable shunt attenuator is not the perfect solution as you point out with the common mode issue. It is a compromise.
Still, with it I can have two fixed resistors of the TX2352 nude Vishay type, and use PRP resistors for the rest,
and have at most two TX2352s in series and one PRP resistor parallel to the signal path instead of maybe 200 PRP resistors.
Spontaneously, I would think that my solution has the sonical advantage.
Resistor size is a problem when using hole mounted types, but with SMD resistors it should be able to fit them all. I wonder if there are any audiophiles out there that would pay the extra cost of having an all-BulkMetalFoil-resistor attenuator with 240 resistors?
It surely would be interesting to have Stereophile test your new preamp with both types of attenuators.
Attached is a balanced series attenuator schematics.
Sigurd
Attachments
Sigurd Ruschkow said:
Generally speaking, there can be two causes:
a) If the switch is MBB, then there will momentarily be two shunt resistors in parellel. This will reduce the gain at least 6 dB (depending on the step sizes of your attenuator design) during the transition. This can be perceived as a "click", especially if it occurs when the waveform is far from the zero crossing.
b) If the switch is BBM, then there will momentarily be a time when the gain increases to full volume with your attenuator scheme. This will also be perceived as a "click".
Sigurd Ruschkow said:
All designs involve compromise. That is the art of design.
Sigurd Ruschkow said:
Your conclusion seems logical. However, the only way to know for sure is with a listening test.
Sigurd Ruschkow said:
The magazines are not in the business of performing research into circuit design. Besides, I trust my own ears more than theirs. But in any case, our new circuit does not use an attenuator. Instead the incoming signal is fed straight to the gain stage, which has an adjustable gain. This was discussed in more detail in the other thread.
Charles Hansen said:
Hi Charles,
You are absolutely right here, and we used 84 in total, but this is a 6 channel attenuator, built with no imposed cost-constraints, which I understand will be marketed later on this year as part of an entire audio system.
As mentioned, the overall accuracy both in attenuating steps and across all 6 channels, together with the unvarying input impedance, was a satisfying result. This was mostly due to the fact that someone other than myself performed the resistor's calculations (!), and the availability of such awkward resistance values being made to order.
Using the fine-sounding TX2352's (and I must agree with Sigurd's enthusiasm for these 'naked' Vishays) but necessarily compromising with relays for reasons I won't go into, the sonic results were remarkably close to using 'conventional' (VSRJ) Vishays together with a 'manual' Shallco switch. This was my 'yardstick' sonically, being the most transparent attenuator configuration I have so far experienced.
The results surprised me a little, and delighted the speaker manufacturer who had commissioned this design.
Regards,
Hello Bob,
I sure am a fan on the the TX2352s!
A few years back I even used Bulk Metal Foil potentiometers for a line amp. The pots I also got from Texas Components. I always do business with Arbie Lopez - a very nice person do work with!
The line amp then became an all-BulkMetalFoil-LineAmp. The 1MEG input resistors I had specially made for me by TC; they put 4 pcs 250k in series. TC will also make me some 3 Ohm TX23252s
resistors by soldering several resistors in parallel.
Price for the pots has since then increased 100% (around 40USD / each now), and I am now looking for alternatives to the Bulk Metal Foil potentiometers.
I am also using a reed relay based stepped logarithmic attenuator
and I get click sounds with that one, too.
That loudspeaker company surely must be satiesfied with your work. Well done.
Is it click sound free?
Sigurd
I sure am a fan on the the TX2352s!
A few years back I even used Bulk Metal Foil potentiometers for a line amp. The pots I also got from Texas Components. I always do business with Arbie Lopez - a very nice person do work with!
The line amp then became an all-BulkMetalFoil-LineAmp. The 1MEG input resistors I had specially made for me by TC; they put 4 pcs 250k in series. TC will also make me some 3 Ohm TX23252s
resistors by soldering several resistors in parallel.
Price for the pots has since then increased 100% (around 40USD / each now), and I am now looking for alternatives to the Bulk Metal Foil potentiometers.
I am also using a reed relay based stepped logarithmic attenuator
and I get click sounds with that one, too.
That loudspeaker company surely must be satiesfied with your work. Well done.
Is it click sound free?
Sigurd
Bobken said:
Sigurd Ruschkow said:
Hi Sigurd,
Yes they are very pleased with this attenuator, and so am I. To have no cost constraints for any design like this, does make matters easier though.
I cannot go into much detail now, but there are no 'sonic' clicks/pops etc., although at times when all 24 relays happen to switch simultaneously (which only happens once in the 1dB steps between '0' attenuation & -63dB) the mechanical noise is a slight problem.
I needed to do a lot of experimenting in order to damp this down, and some relays are mechanically quieter than others, I discovered.
As for the sonics of these TX2352s, I have found them quite exceptional, and even slightly cleaner-sounding than the S102/VSRJs. I know that some users have criticised them for being maybe too open and clinical-sounding, but my experiences indicate that this is simply the result of their being so transparent that they reveal shortcomings elsewhere.
In my usual test-system I don't have any resistors *anywhere* which are not either bulk-foils, or a few Caddock MP930 (sonically superior to their MK132) for higher current areas, and a potential divider made with 2x TX2352s is just about as neutral and revealing as I believe it is possible to achieve, where resistors need to be used anyway. I am not a fan of using different makes of shunt and series resistors in one location, though, as this has never been to my satisfaction, and will always tend to colour the sonic results in my experience.
It is an unfortunate fact that in life one only gets what one pays for, and the cost of these bulk-foils is very high compared with average metal films. Also, the increase in performance is subject to the laws of diminishing returns and doesn't reflect the increase in cost, but where this outlay can be justified, I would use no other resistor by choice in audio circuits.
Regards,
The thread STAR could be something like this … 😉
I have been away of this thread for some time, I was quite busy at my job and in taking take care of my mother; she passed away a few weeks ago.
An externally hosted image should be here but it was not working when we last tested it.
I have been away of this thread for some time, I was quite busy at my job and in taking take care of my mother; she passed away a few weeks ago.
Hi Richard,
Thanks for sharing this schematic.
It is very sad to hear about your Mother, and I know what it is like to lose someone so dear, especially at a time like this.
I am sure that others will be thinking of you with sympathy.
Regards,
Thanks for sharing this schematic.
It is very sad to hear about your Mother, and I know what it is like to lose someone so dear, especially at a time like this.
I am sure that others will be thinking of you with sympathy.
Regards,
For the record, this is NOT exactly the CTC preamp circuit. The original CTC preamp circuit has NO loop feedback, and can operate with either single ended or balanced input and output, depending on the customer's willingness to pay for the added option of a balanced input.
John, however I do remember that you mentioned somewhere in this thread that there IS local feedback in the second stage, right?
I am just pointing out that this is a different design, but it is an interesting design, and should work as well.
Okay. It's official. I'm confused.
courage says, 'Local feedback in the second stage?'
You say, 'Yes, but at the input.'
Er...does 'input' refer to the first stage (J1-4), or to the input to the second stage (M1-4), meaning the Sources in the schematic above or the Sources & Gates if you happen to be using your older Gate & Source driven folded cascode (have you ever come up with a name for that topology)?
Or have the germs that invaded my respiratory tract also eaten my last brain cell?
Grey
courage says, 'Local feedback in the second stage?'
You say, 'Yes, but at the input.'
Er...does 'input' refer to the first stage (J1-4), or to the input to the second stage (M1-4), meaning the Sources in the schematic above or the Sources & Gates if you happen to be using your older Gate & Source driven folded cascode (have you ever come up with a name for that topology)?
Or have the germs that invaded my respiratory tract also eaten my last brain cell?
Grey
- Status
- Not open for further replies.