Did I mention I also have this dac.
The mod list
1. replace the output cap with Vishay mtk1822 -> big improvement.
Yes, give the mkt1822 a try, it is very good.
2. up the voltage to 7.5V (change the resistor of the lm317)
and change Vref resistor to 1.5k and I/V resistor to 2.2k.
3. replace most of the decoupling cap, the most important is next to tda1543.
I used blackgate which I have on hand, others to rubycon zl.
This will give exceptional value dac for the money.
There are other ways to improve it of cause as discuss in above posts.
Other suggestion which I haven't done (I move on to the deq2496/tda1543)
1. replace the lm7805 to shunt lm431. maybe add a separate power for the 74hc00 and dir9001.
2. add input digital transformer for isolation
3. replace the 74hc00 to 74vhc00 or even better with ds8921 (not a direct replacement, google this)
The mod list
1. replace the output cap with Vishay mtk1822 -> big improvement.
Yes, give the mkt1822 a try, it is very good.
2. up the voltage to 7.5V (change the resistor of the lm317)
and change Vref resistor to 1.5k and I/V resistor to 2.2k.
3. replace most of the decoupling cap, the most important is next to tda1543.
I used blackgate which I have on hand, others to rubycon zl.
This will give exceptional value dac for the money.
There are other ways to improve it of cause as discuss in above posts.
Other suggestion which I haven't done (I move on to the deq2496/tda1543)
1. replace the lm7805 to shunt lm431. maybe add a separate power for the 74hc00 and dir9001.
2. add input digital transformer for isolation
3. replace the 74hc00 to 74vhc00 or even better with ds8921 (not a direct replacement, google this)
Attachments
I have the same 1543 card lying around and want to try it.
If ECs biasing method sounds better I think this would be the way to do it.
One question though: Is the biasing current 2,3mA total or for each output?
Lithium cells are 3V and are said to be the best for biasing tubes, so they should be useful here too.
Feeding the signal without caps directly into a FET SF or tube CF would take care of the impedance problem. Personally I think the output level of just above 0,5V rms is good, the common 2Vrms is always to high.
If ECs biasing method sounds better I think this would be the way to do it.
One question though: Is the biasing current 2,3mA total or for each output?
Lithium cells are 3V and are said to be the best for biasing tubes, so they should be useful here too.
Feeding the signal without caps directly into a FET SF or tube CF would take care of the impedance problem. Personally I think the output level of just above 0,5V rms is good, the common 2Vrms is always to high.
Attachments
2Vrms is a must have when you have a passive pre! 
I want to try the Fetishizator but with a 2SK170 instead of BF245A. I think it is an ambitious task for a layman like me.
Fetishizator
I want to try the Fetishizator but with a 2SK170 instead of BF245A. I think it is an ambitious task for a layman like me.
Fetishizator
Hey Tolu,
Is the sensitivity really that low in your amp? Usually a power-amp is around 1Vrms.
To fix this a little less NFB(if there is any)could maybe help.
Will also sim a gainstage with 2SK170 and 12dB gain to see if it will work. Probably you will need an added SF to bring down Zout. The high input capacitance of the 2SK170 could maybe be of some help to filter away HF.
John,
LEDs can not be used in the cathode circuit of a CF.
But if you want to build a gain stage this is a good idea. In this case you must compensate by adding 3,8V to your normal gridvoltage.
Don´t use blue, the red are the good ones.
I would use a 5842 or 6C45 with something like a 8:1 parafeed transformer to bring down Zout and gain.
Is the sensitivity really that low in your amp? Usually a power-amp is around 1Vrms.
To fix this a little less NFB(if there is any
)could maybe help.Will also sim a gainstage with 2SK170 and 12dB gain to see if it will work. Probably you will need an added SF to bring down Zout. The high input capacitance of the 2SK170 could maybe be of some help to filter away HF.
John,
LEDs can not be used in the cathode circuit of a CF.
But if you want to build a gain stage this is a good idea. In this case you must compensate by adding 3,8V to your normal gridvoltage.
Don´t use blue, the red are the good ones.
I would use a 5842 or 6C45 with something like a 8:1 parafeed transformer to bring down Zout and gain.
Tolu said:
I want to try the Fetishizator but with a 2SK170 instead of BF245A. I think it is an ambitious task for a layman like me.
Fetishizator
Here's a simple 2SK170 circuit that would do the job: http://myweb.tiscali.co.uk/nuukspot/decdun/gainclonepre2.html
Hi Lars
thanks a lot for your work.
What I don't understand is the 3V battery and R3. Why don't you go directly (with a series resistor) to the gate of the JFET? Why do you do I/V conversion first? Why do you take 2 JFETs? Wouldn't it be enough to use the Fetishizator but with 180 Ohm for gate and source resistor?
thanks a lot for your work.
What I don't understand is the 3V battery and R3. Why don't you go directly (with a series resistor) to the gate of the JFET? Why do you do I/V conversion first? Why do you take 2 JFETs? Wouldn't it be enough to use the Fetishizator but with 180 Ohm for gate and source resistor?
Hey Tolu,
You asked for a FET-buffer with 12dB gain(4x), capable of driving your 7kohm load. This is what is needed !
The "thing" on your sketch does not have any of these qualities. It has high gain, high distortion and high Zout. I´d say it is totally useless at these voltage-levels!
About the 3V and R3 it is the I/V and part of ECs novel design.
I asked about the sensitivity of your amp was but didn´t get an answer? Would rather see a sourcefollower instead of the above and higher sensitivity in the following amp. This to eliminate one gainstage in the chain......
You asked for a FET-buffer with 12dB gain(4x), capable of driving your 7kohm load. This is what is needed
!The "thing" on your sketch does not have any of these qualities. It has high gain, high distortion and high Zout. I´d say it is totally useless at these voltage-levels!
About the 3V and R3 it is the I/V and part of ECs novel design.
I asked about the sensitivity of your amp was but didn´t get an answer
? Would rather see a sourcefollower instead of the above and higher sensitivity in the following amp. This to eliminate one gainstage in the chain......
Tolu,
In the lightspeed thread Nelson Pass suggested a buffer to use with it. Not sure which page but a quick search would yield results. Also GeorgeHiFi recommends a tube buffer if you are a 'tube buff'.
Then you could drive just about anything.
Course you might need to replace the 100k pot in the lightspeed with a 5 turn or 10 turn 100k at that point
Uriah
In the lightspeed thread Nelson Pass suggested a buffer to use with it. Not sure which page but a quick search would yield results. Also GeorgeHiFi recommends a tube buffer if you are a 'tube buff'.
Then you could drive just about anything.
Course you might need to replace the 100k pot in the lightspeed with a 5 turn or 10 turn 100k at that point
Uriah
Hi Uriah
I recently tried a Yaqin Buffer with Mullard tubes. It was nice but the pure Lightspeed only with CDP was better (to my ears).
I think I will give Lars' schematic a trial. Or I try the JFET BOZ after the EC mod of the output of the TDA1543. Perhaps Lars' finds a trick to use only one JFET (->KISS)!
I recently tried a Yaqin Buffer with Mullard tubes. It was nice but the pure Lightspeed only with CDP was better (to my ears).
I think I will give Lars' schematic a trial. Or I try the JFET BOZ after the EC mod of the output of the TDA1543. Perhaps Lars' finds a trick to use only one JFET (->KISS)!
This is maybe not the best option as it does not have the best drive capability for your Lightspeed.
Check above, you must have a source-follower to buffer it.
Another option is to use: DAC with Sourcefollower, Lightspeed, Boz. But this way you will add another cap in the chain. Accoording to NP gain is ca 15x.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.