Some months ago, my friend gave me a cd player he doesn't use anymore and which has slightly modified analog output - Philips CD303 with TDA1540x2. So I removed allmost all components from output and stuck board with 2x4 AD844.
Last edited:
There is very little room to fit anything else without modifying it heavily and even couldn't put shield back. Measured power supply voltages and difference is more than 2 V (12.4 and -10.3)[emoji53]
I have a jumper to choose between ad844 paralled buffers output (4x100 ohm) or TZ outputs with OPA627.
I have a jumper to choose between ad844 paralled buffers output (4x100 ohm) or TZ outputs with OPA627.
I think the different positive and negative supply V were intended by Philips.
Your PCB looks nice.
Miksi
which configuration do you prefer in term of sound quality? opa627 or ad844 parallel buffers?
cheers
Hi guys,
My stacked AD844 IV is working nicely, but it gets hot (50+°C) after 10 or 20 minutes. How is yours? My supply rails are +-13.6V. Should I lower the voltage for less heat?
My stacked AD844 IV is working nicely, but it gets hot (50+°C) after 10 or 20 minutes. How is yours? My supply rails are +-13.6V. Should I lower the voltage for less heat?
Can one of you knowledgeable gentlemen please explain to me the difference between the AD844's TZ pin and the AD744's compensation pin?
Thanks...
Thanks...
Well, I've misscalculated opa627 amplification factor and it is quieter than AD844 buffers output so I'm not sure yet. On weekend I'd make correction and listen some more to decide.
That means you had the 627 running at "negative gain" if it was quieter than the 844's buffer.
Cheers George
The datasheet of each op amp has a simplified schematic which show where pin 5 is connected internally.
AD844 vs AD744
The AD744 is a fairly straight forward voltage op amp. Pin 5 is different in that it allows one to adjust the frequency response of the AD744 by applying a capacitor from Pin 5 to Pin 8. AD844 is a different animal all together. It was designed for video applications. It has a current mirror common base driver stage with Pin 5 access to the mirror just before the output buffer stage. This allows one to apply a no negative feedback resistor from Pin 5 to ground. In addition you can apply a capacitor across Pin 5 and ground to act as a first order filter. They are not interchangeable as I/V parts go. Don't know if that helps however the datasheets spell it all out. 😉
I think the closest would be AD846 which also has a current mirror followed by output buffer.
AD846
Yes. That probably is quite similar. The AD846 has a nice sound. I am kicking around the idea of using it in a MC phono preamp. The low input impedance of the - input terminal makes it ideal for this application running like an I/V application. Downside is the limited number of MC cartridges that would work with with a lowish 50 to 200 Ohm winding. I think I found one made by Denon that would work. Something to mess with in the future.... 😉
I am still wondering if the value of the capacitor at pin 2 to ground should be increased when we stack up multiple ad844. This capacitor forms a first order filter with the input impedance of the op amp. With 3 op amp the input impedance decrease to 1/3 of that with one op amp. as a result the cut off frequency will shift up. In order to maintain the originally intended cutoff frequency I thinh this capacity value should be increased.
The effect may be different on machines with or without a digital filter. Pedja designed for NOS TDA1541 i.e. no digital filter. If one use on a dac or cdp with a digital filter the effect may not be audible.
when I stack 3 ad844, the sound is on the bright side.
The effect may be different on machines with or without a digital filter. Pedja designed for NOS TDA1541 i.e. no digital filter. If one use on a dac or cdp with a digital filter the effect may not be audible.
when I stack 3 ad844, the sound is on the bright side.
George
It is logical to keep the i/v resistor and the cap at pin 5 to ground the same. But I have doubt about the one at pin 2 (input). Can you explain why this cap should be the same?
AD844 opinion...
Try it. There is no reason not to. I am not a fan of the NOS idea. I found a digital filter that works and it isn't a SAA7220. That is junk in every implementation I have ever seen. So try increasing the capacitor value by 3 times the original value. The correct way to measure it is with a CD with 19 Khz and 20 Khz tones. You want to see it starting to roll off about 1 db or so at 19 Khz using an oscilloscope. In truth... With NOS I'd be surprised if there is anything beyond 15 Khz. Even with the Sine/X compensation circuitry it will be tough to get data up there without at least 4X oversampling. Just not enough data.... This is just an opinion. YMMV... DIY is fun and educational! 😉
RF Hash
The problem may be ineffective filtering of the DAC products. One of the reasons I added the passive second order transformer based filter I shared earlier in the thread. 😀
The problem may be ineffective filtering of the DAC products. One of the reasons I added the passive second order transformer based filter I shared earlier in the thread. 😀
Combination of the cap and resistor form the HF filter in my case 2.7kohm with 470pf = -3db at125khz. The resistor set your gain.
Cheers George
Thanks Torchwood. It happens that I have a NOS dac. adding a DF will require a total reconstruction. I understand that correctly implemented DF is good.
I will try 3 time the original value.
George
The resistor and cap value that you mention are those at Tz pin.
I am talking about the cap at pin 2. This cap acts in parallel with input impedance of the opamp and form another filter. If this cap is two small relatively to the input impedance of the op amp then it would not have any filtering effect.
I will try 3 time the original value.
George
The resistor and cap value that you mention are those at Tz pin.
I am talking about the cap at pin 2. This cap acts in parallel with input impedance of the opamp and form another filter. If this cap is two small relatively to the input impedance of the op amp then it would not have any filtering effect.