Andrew,
Who solder @ 184°C? This is way too low when you account for thermal mass of the PCB traces and components to solder. Been soldering for the past 30 years and I don't recall going lower 300°C and many time 340°C on the iron (I'm not saying it's the way to go but it's what I do). This allows to heat fast and do a quick solder without having to leave the iron on for prolonged time on the parts. Never overheated a single components, including ICs, never burned anything. I now use a Weller DW1 digitally controlled soldering station (which is the best one I've owned so far) and trying to solder parts using 63/37 under 200°C is not ideal and in my experience hard to do unless you're doing SMD type mounting.
Anyways, just my two cents, I know you have lots of experience so not arguing with you, just sharing my experiences.
Ciao!
Do
Dcb1 integration
Hello All
I require some advice regarding integrating the DCB1 into my evolving system. This is mainly due to my recently developed concerns about impedance matching...no doubt a result of spending considerable time researching the build process of my recently acquired Hypnotize kit from Tea Bag.
I'm currently using the BHL phono amp. It is said to have a fairly high output impedance which is between 2-3K...as I have read.
Although I already have the desire to replace it with the FSP, that will not happen immediately. This means I need to try and work with what I have to integrate the DCB1.
My underlying goal for a while has been to be able to experiment with the low output Denon 103 MC. A few years ago I acquired one and then in searching for a way to amplify its signal into my Audible Illusions MM phono stage I came across a ZYX CPP-1 for sale. Turns out it wasn't enough gain. In addition, I got tired of feeding the AI tubes.
I digress...besides wanting to be able to try the 103 I have a variety of other MM and MI carts I would still desire to use...along with misc digital sources. All of which I can reasonably facilitate by building the DCB1 into a preamp box that will allow a few sources and attenuation.
The attenuation seems the issue. I had bought some 25K stepped attenuators off of fleabay to experiment with because I originally thought I would build a standard B1 clone and that was the recommended value for the attenuators.
Now I face a dilemma as to where I should put the attenuator in the signal chain or if I need to make a different choice regarding the attenuator altogether.
I think more information about the system would be helpful.
I use a VPI HW-19 w/ a modified Rega RB250 as an analog source with my small selection of carts. Depending on the cart I either insert or remove the head amp. The head amp's input impedance is said to be 125 Ohms. I've searched everywhere for information regarding its output impedance...I found none.
Depending on the cart, the signal either goes through the head amp or is sent directly to the BHL phono amp. I thought I had come across a number for the input impedance of the BHL...it escapes me now. I have communicated with Jason of BHL a few times. I remember reading the info or him communicating that the phono amp was designed to have a high enough input impedance to allow it to be driven easily and a low enough output impedance to be able to "dive" under any following component. So no exact figure on input impedance...2-3K said to be output impedance.
Ideally the signal then goes into the attenuator at this point...according to what Salas had recommended early on in this thread...on its way to the DCB1?
However, I'm concerned that my 25K attenuator is not a good match for the output of the BHL phono amp. My understanding is that the buffer is supposed to be able to null any issues with impedance mismatching. Unfortunately it is not sitting between the phono amp and attenuator in this scenario. It is sitting between the attenuator and my gainclone amp...and perhaps rightly so?
My gainclone is built from the Peter Daniels premium kit. There is no coupling cap at the input currently. My understanding is its input impedance is set by the input shunting resistor which has a value of 22K. I have already acquired everything I would need to move onto my next amp project...a pair of F5s. Those also are not supposed to have any coupling in the signal path. Besides that, I have a pair of Quicksilver Horn Mono amps...which have not responded well to my system choices. I believe they require more efficient speakers. I also have various Rotel amps. For now, I am using and enjoying the gainclone and would like to continue to do so.
To wrap things up, the signal would be fed into a pair of Oris 200 horns which are open backed and loaded with Tang Band W8-1808s.
Until now a miniDSP has occupied the duties of the attenuator and DCB1. It allows me to supplement the bottom end of the full range horns with a powered subwoofer. I integrated a master volume control in the miniDSP when I built it from boards for attenuation control. I find it a bit noisy and want to try the attenuator/DCB1 in its stead while allowing the horns to role off and utilizing the on board crossover of the subwoofer to try and get it to blend into the bottom end.
So there it is. Not sure I can pull things off with what I'm working with. Guidance is greatly appreciated.
Kevin
Hello All
I require some advice regarding integrating the DCB1 into my evolving system. This is mainly due to my recently developed concerns about impedance matching...no doubt a result of spending considerable time researching the build process of my recently acquired Hypnotize kit from Tea Bag.
I'm currently using the BHL phono amp. It is said to have a fairly high output impedance which is between 2-3K...as I have read.
Although I already have the desire to replace it with the FSP, that will not happen immediately. This means I need to try and work with what I have to integrate the DCB1.
My underlying goal for a while has been to be able to experiment with the low output Denon 103 MC. A few years ago I acquired one and then in searching for a way to amplify its signal into my Audible Illusions MM phono stage I came across a ZYX CPP-1 for sale. Turns out it wasn't enough gain. In addition, I got tired of feeding the AI tubes.
I digress...besides wanting to be able to try the 103 I have a variety of other MM and MI carts I would still desire to use...along with misc digital sources. All of which I can reasonably facilitate by building the DCB1 into a preamp box that will allow a few sources and attenuation.
The attenuation seems the issue. I had bought some 25K stepped attenuators off of fleabay to experiment with because I originally thought I would build a standard B1 clone and that was the recommended value for the attenuators.
Now I face a dilemma as to where I should put the attenuator in the signal chain or if I need to make a different choice regarding the attenuator altogether.
I think more information about the system would be helpful.
I use a VPI HW-19 w/ a modified Rega RB250 as an analog source with my small selection of carts. Depending on the cart I either insert or remove the head amp. The head amp's input impedance is said to be 125 Ohms. I've searched everywhere for information regarding its output impedance...I found none.
Depending on the cart, the signal either goes through the head amp or is sent directly to the BHL phono amp. I thought I had come across a number for the input impedance of the BHL...it escapes me now. I have communicated with Jason of BHL a few times. I remember reading the info or him communicating that the phono amp was designed to have a high enough input impedance to allow it to be driven easily and a low enough output impedance to be able to "dive" under any following component. So no exact figure on input impedance...2-3K said to be output impedance.
Ideally the signal then goes into the attenuator at this point...according to what Salas had recommended early on in this thread...on its way to the DCB1?
However, I'm concerned that my 25K attenuator is not a good match for the output of the BHL phono amp. My understanding is that the buffer is supposed to be able to null any issues with impedance mismatching. Unfortunately it is not sitting between the phono amp and attenuator in this scenario. It is sitting between the attenuator and my gainclone amp...and perhaps rightly so?
My gainclone is built from the Peter Daniels premium kit. There is no coupling cap at the input currently. My understanding is its input impedance is set by the input shunting resistor which has a value of 22K. I have already acquired everything I would need to move onto my next amp project...a pair of F5s. Those also are not supposed to have any coupling in the signal path. Besides that, I have a pair of Quicksilver Horn Mono amps...which have not responded well to my system choices. I believe they require more efficient speakers. I also have various Rotel amps. For now, I am using and enjoying the gainclone and would like to continue to do so.
To wrap things up, the signal would be fed into a pair of Oris 200 horns which are open backed and loaded with Tang Band W8-1808s.
Until now a miniDSP has occupied the duties of the attenuator and DCB1. It allows me to supplement the bottom end of the full range horns with a powered subwoofer. I integrated a master volume control in the miniDSP when I built it from boards for attenuation control. I find it a bit noisy and want to try the attenuator/DCB1 in its stead while allowing the horns to role off and utilizing the on board crossover of the subwoofer to try and get it to blend into the bottom end.
So there it is. Not sure I can pull things off with what I'm working with. Guidance is greatly appreciated.
Kevin
Thanks very much for the quick response and guidance.
It's unfortunate that I'll have to get some other attenuator/pot and resistors to work with, but it's better to ask than make obvious mistakes.
Now I have to debate over what choices to make concerning these components. It is hard as a novice to know what you need and will do the job well and what is unnecessary and just overkill. I find a lot that the problem with the Internet can often be too many choices and opinions.
I previously bought the 25K stepped attenuators from Hong Kong on a whim because I had seen a few threads and an article on a DIY site where they were used.
This time I pretty much did the same thing and grabbed some "DACT" 50K attenuators off of eBay because I had again seen them mentioned in a thread. For some reason I can't paste the link. I'm already doubting my decision and can only think that to some point you get what you pay for. The Alps pots are probably the most universally mentioned, but they are not exactly cheap. I had also thought I would like to have dual mono controls as my old preamp has so that there was the ability to balance the channels easily. I guess I will find out and report back.
I'm not sure what to do about the need of the 680K input resistor as the stock PRP resistors seem to have been thought through well. Yet, I see others opt for much more expensive options. I guess I'll be looking for a supplier to find the PRP in the value I need and perhaps the 15-0-15 transformer as well.
It's unfortunate that I'll have to get some other attenuator/pot and resistors to work with, but it's better to ask than make obvious mistakes.
Now I have to debate over what choices to make concerning these components. It is hard as a novice to know what you need and will do the job well and what is unnecessary and just overkill. I find a lot that the problem with the Internet can often be too many choices and opinions.
I previously bought the 25K stepped attenuators from Hong Kong on a whim because I had seen a few threads and an article on a DIY site where they were used.
This time I pretty much did the same thing and grabbed some "DACT" 50K attenuators off of eBay because I had again seen them mentioned in a thread. For some reason I can't paste the link. I'm already doubting my decision and can only think that to some point you get what you pay for. The Alps pots are probably the most universally mentioned, but they are not exactly cheap. I had also thought I would like to have dual mono controls as my old preamp has so that there was the ability to balance the channels easily. I guess I will find out and report back.
I'm not sure what to do about the need of the 680K input resistor as the stock PRP resistors seem to have been thought through well. Yet, I see others opt for much more expensive options. I guess I'll be looking for a supplier to find the PRP in the value I need and perhaps the 15-0-15 transformer as well.
Last edited:
That 220k resistor sits in parallel to the pot's output. If not having >10x value than the pot itself it will change its log curve. So with 50k pot better use much bigger than 220K DCB1 input load resistor i.e. the 680k value I recommended. Original DACT is fine, no worries. DACT switching style cheap small ones are good enough sounding, they just don't last very long. Your 25K ones are good for most sources but when having a source with 3K output impedance like the one you mentioned, better use 50K pot. PRP and Dale 50ppm are good resistor ranges.
Hi Salas, chromentus,
Thanks for raising this I had the Sam question in my mind
Salas, what is the formula that you are using to correlate the
I put Resistance, pot Resistance and the adjustment of DCB1 resistors. I think it could answer the question on what pot are applicable for different set up kind of questions
Sent from my ONE A2003 using Tapatalk
Thanks for raising this I had the Sam question in my mind
Salas, what is the formula that you are using to correlate the
I put Resistance, pot Resistance and the adjustment of DCB1 resistors. I think it could answer the question on what pot are applicable for different set up kind of questions
Sent from my ONE A2003 using Tapatalk
As I wrote already its about not using less than 10x the pot value as input to ground DCB1 resistance. Use up to 20K pot with 220k input resistance for less interaction with the JFET input capacitance (minor deal, there is no gain to multiply it) but use higher pot value for high output impedance audio sources (proper loading is a bigger deal)
The minimum recommended load for a voltage Source is generally accepted as 5x the Source output impedance.
Many adopt 10x and it does no harm to be higher, even much higher.
I generally aim for at least 20x if I can arrange it.
The load impedance seen by the Source, set by the DCB1 is 220k.
The maximum Source impedance should be lower than 220k/5 and preferably <22k
The Source is the vol pot.
The output impedance of the vol pot varies depending on rotation setting. It varies from a low of zero ohms to a maximum of {vol pot value+the source output impedance} / 4
a 50k vol pot fed from a 200ohms source will have a maximum output impedance of {50000+200}/4 = 12550ohms.
applying that range of source:load factors one needs a load of roughly 63k to 251k
The standard DCB1 with a 220k suits a 50k vol pot. Part of the reason is that the vol pot has a very short built in trace link to the Buffer. There is very low and pre-defined parasitic capacitance on the vol pot.
In my view the standard DCB1 suits all vol pot impedances from 5k to 50k
Only 100k to 1M vol pots would need the DCB1 input to be set to a higher value, to prevent the log faking effect.
But if one wants to try the log faking effect then one can set the 220k to a lower value. No harm will be done.
You might like the effect that has on the SPL vs the rotation.
Many adopt 10x and it does no harm to be higher, even much higher.
I generally aim for at least 20x if I can arrange it.
The load impedance seen by the Source, set by the DCB1 is 220k.
The maximum Source impedance should be lower than 220k/5 and preferably <22k
The Source is the vol pot.
The output impedance of the vol pot varies depending on rotation setting. It varies from a low of zero ohms to a maximum of {vol pot value+the source output impedance} / 4
a 50k vol pot fed from a 200ohms source will have a maximum output impedance of {50000+200}/4 = 12550ohms.
applying that range of source:load factors one needs a load of roughly 63k to 251k
The standard DCB1 with a 220k suits a 50k vol pot. Part of the reason is that the vol pot has a very short built in trace link to the Buffer. There is very low and pre-defined parasitic capacitance on the vol pot.
In my view the standard DCB1 suits all vol pot impedances from 5k to 50k
Only 100k to 1M vol pots would need the DCB1 input to be set to a higher value, to prevent the log faking effect.
But if one wants to try the log faking effect then one can set the 220k to a lower value. No harm will be done.
You might like the effect that has on the SPL vs the rotation.
Last edited:
Interesting information...going to copy and paste into my notepad for reference.
Does this same formula apply in reverse to a load? For instance...if I decide to add a volume pot to the input of my gainclone?
The gainclone has an input load defined by a 22K resistor to ground. If I decided to add a 25K attenuator at the input (such as described in Peter Daniels build guide) would the resulting load be 25K + 22K/4?...or 11.75K?
Not sure I have it right. In Peter's guide he states that a combination of a 50K pot at the input in combination with the 22K input resistor results in a 15K load...hmmm.
Does this same formula apply in reverse to a load? For instance...if I decide to add a volume pot to the input of my gainclone?
The gainclone has an input load defined by a 22K resistor to ground. If I decided to add a 25K attenuator at the input (such as described in Peter Daniels build guide) would the resulting load be 25K + 22K/4?...or 11.75K?
Not sure I have it right. In Peter's guide he states that a combination of a 50K pot at the input in combination with the 22K input resistor results in a 15K load...hmmm.
Not quite.
You need to separate source and receiver and then analyse the pair.
A vol pot can be both a source and a receiver.
For the connection between vol pot and power amplifier the vol pot is the source.
The output impedance of this source loads the input impedance of the Power Amplifier.
Thus a 25k attenuator gives a maximum output impedance of ~6k and a minimum of 0k.
The receiver has an input impedance of 22k in parallel to whatever you have used as an input filter. Usually 220pF to 1nF.
Now join these together with a cable or some PCB traces.
What effect does one have on the other when signal is passed from source to receiver?
Analyse both cases 0k and 6k and maybe an intermediate value typical of normal listening level.
You need to separate source and receiver and then analyse the pair.
A vol pot can be both a source and a receiver.
For the connection between vol pot and power amplifier the vol pot is the source.
The output impedance of this source loads the input impedance of the Power Amplifier.
Thus a 25k attenuator gives a maximum output impedance of ~6k and a minimum of 0k.
The receiver has an input impedance of 22k in parallel to whatever you have used as an input filter. Usually 220pF to 1nF.
Now join these together with a cable or some PCB traces.
What effect does one have on the other when signal is passed from source to receiver?
Analyse both cases 0k and 6k and maybe an intermediate value typical of normal listening level.
Last edited:
OK...I'm not sure an exploration of input and output impedances of potentiometers/attenuators and what determines them has a real place in this thread except perhaps for those like myself that are trying to determine what options there are to place an attenuator in relation to the DCB1 in the signal chain if their source or load have unusually high or low mismatched impedance values.
I scratched my head immediately concerning the ~6K maximum output impedance stated for a 25K attenuator. I wanted to know why. The math is just way above my head. The only logic I was able to absorb is that the impedance on the output of the attenuator would be at its highest at the center of an attenuators rotational travel and at that point it's impedance should be about 1/4 it's total stated impedance value. And of course your statement that at the 0 degree rotational position the wiper would be grounded and the impedance would be
Regardless of my understanding the math, as you said I would have the attenuator acting as a source with a maximum output impedance value of ~6K driving the amp as a load that's input impedance is defined by a 22K resistor to ground.
I built the gainclone in its simplest form and as of yet there is no filtering capacitor at its input. I have had to be diligent about monitoring any sources that would generate additional DC.
So since the attenuator would be mounted in the amp with virtually no additional signal path to travel to its input and no additional capacitance other than what small amount may come from the attenuator itself are we simply just looking at the worst case scenario of the attenuators ~6K impedance as a source into the 22K load of the amplifier? If so, it doesn't seem like it would pass even the most basic impedance matching rule of thumb you stated as being 1:5.
The other question that occurred to me is that the attenuator, as you said, can act as both a source and load. I'm assuming it is actually both at all times as it is the nature of the device. Is its output impedance as a source affected by components that are driving it at the same time as a load? If so, does utilizing a buffer like the DCB1 in front of it prevent that from happening?
Edit: You already gave the example of the 50K pot plus 200 Ohm source...so sources do have an affect on attenuators output impedance and I suppose all we can hope to do is minimize that by adding the buffer in front of it so as to add as little additional impedance as possible and "buffer" the attenuator from a possible very high output impedance of a source further up the chain?
I scratched my head immediately concerning the ~6K maximum output impedance stated for a 25K attenuator. I wanted to know why. The math is just way above my head. The only logic I was able to absorb is that the impedance on the output of the attenuator would be at its highest at the center of an attenuators rotational travel and at that point it's impedance should be about 1/4 it's total stated impedance value. And of course your statement that at the 0 degree rotational position the wiper would be grounded and the impedance would be
Regardless of my understanding the math, as you said I would have the attenuator acting as a source with a maximum output impedance value of ~6K driving the amp as a load that's input impedance is defined by a 22K resistor to ground.
I built the gainclone in its simplest form and as of yet there is no filtering capacitor at its input. I have had to be diligent about monitoring any sources that would generate additional DC.
So since the attenuator would be mounted in the amp with virtually no additional signal path to travel to its input and no additional capacitance other than what small amount may come from the attenuator itself are we simply just looking at the worst case scenario of the attenuators ~6K impedance as a source into the 22K load of the amplifier? If so, it doesn't seem like it would pass even the most basic impedance matching rule of thumb you stated as being 1:5.
The other question that occurred to me is that the attenuator, as you said, can act as both a source and load. I'm assuming it is actually both at all times as it is the nature of the device. Is its output impedance as a source affected by components that are driving it at the same time as a load? If so, does utilizing a buffer like the DCB1 in front of it prevent that from happening?
Edit: You already gave the example of the 50K pot plus 200 Ohm source...so sources do have an affect on attenuators output impedance and I suppose all we can hope to do is minimize that by adding the buffer in front of it so as to add as little additional impedance as possible and "buffer" the attenuator from a possible very high output impedance of a source further up the chain?
Last edited:
Of course it on topic.
The DCB1 is a volume attenuator AND and Buffer.
Input and output impedance are important to both sections and to the Source that feeds the input and to the Receiver that is fed by the output.
I suggest you read that again and maybe some others that state the facts in a different manner. Sometimes one style of writing is not fully understood and when you read one or two other Authors on the same subject, the nessage is decyphered.
It is rare that a vol pot needs a Buffer in front of it.
Much more likely is that a SOURCE needs a Buffer AFTER it and before the cable that passes the signal to the next stage.
Last edited:
can we use BF862 / toshiba 2SK880GR
https://www.toshiba.com/taec/components2/Datasheet_Sync/200711/DST_2SK880-TDE_EN_6949.pdf
I just saw in other thread that reflector D circuit has got 2sk880GR.
But many say that using BF862 in B1 buffer sounds great. Since its a RF part its even said that stability is a big issue. What is the possibility to use the BF862 in the regulator circuit as it has even more transconductance than 2sk170 but can we use the toshiba 2sk889GR?
https://www.toshiba.com/taec/components2/Datasheet_Sync/200711/DST_2SK880-TDE_EN_6949.pdf
I just saw in other thread that reflector D circuit has got 2sk880GR.
But many say that using BF862 in B1 buffer sounds great. Since its a RF part its even said that stability is a big issue. What is the possibility to use the BF862 in the regulator circuit as it has even more transconductance than 2sk170 but can we use the toshiba 2sk889GR?
Dissipation is an issue too for a tiny package JFET in dcb1 because everything runs at idss level in a dcb1. BF862 averages as a high idss part, that brings even more heat to dissipate, also having to perform the frustrating process of sorting out many SMT parts for matching idss almost exactly because DC coupled means offset if not doing so. 2SK880GR would stay cooler in the PSUS area but its not that good for the audio quad because not enough current or transconductance will be available.
Im talking about using them in psu areas rather than at buffer. At buffer I agree the dissipation is quite higher. Im considering using BF862 at the PSU side.
There will be self dissipation issues even when used in the PSUs as I wrote. All run at IDSS there and four out of six have high enough voltage across themselves to overheat. If not chancing enough IDSS so to be readily self-destroyed, the long time reliability will surely suffer at least. Prefer GR range IDSS in Japanese compatibles. BF862 mostly show circa 15mA IDSS. Also prone to oscillation sometimes needing either gate resistors or mini inductors. Good FETS but not as drop ins without specific circuitry and PCB design.