Mjone,
>>As far as C105 goes larger value capacitors in this position introduce less internal noise
I was planning to keep the stock value of 1uf. Just to replace with film type capacitor.
>>If the I.C. dual op.amp you are using has low enough input bias current you can do >>without an input coupling capacitor entirely.
Stock opamp is njm4580 and I don't know what qualifies as low enough input bias. I was thinking to replace it with NE5532N, per earlier recommendation by Sreten.
So instead, you recommend SE5532 and to replace C105 with 100ohm resistor?
Or MC33078 and a jumper instead of C105?
What is the advantage of each of above mods?
As mentioned before, there are lots of stages in this mixer and around 10 coupling 47uf/16v electrolytic capacitors down the signal path. So I want to be sure it is a good idea to to the mod, having to change out all the coupling caps.
Would I be recalculating the new value of one cap and using that value for the rest of caps?
>>As far as C105 goes larger value capacitors in this position introduce less internal noise
I was planning to keep the stock value of 1uf. Just to replace with film type capacitor.
>>If the I.C. dual op.amp you are using has low enough input bias current you can do >>without an input coupling capacitor entirely.
Stock opamp is njm4580 and I don't know what qualifies as low enough input bias. I was thinking to replace it with NE5532N, per earlier recommendation by Sreten.
So instead, you recommend SE5532 and to replace C105 with 100ohm resistor?
Or MC33078 and a jumper instead of C105?
What is the advantage of each of above mods?
As mentioned before, there are lots of stages in this mixer and around 10 coupling 47uf/16v electrolytic capacitors down the signal path. So I want to be sure it is a good idea to to the mod, having to change out all the coupling caps.
Would I be recalculating the new value of one cap and using that value for the rest of caps?
I meant just the two series 47uF caps directly after the phono output. They feed a net 10k input resistor on the next stage,
so you have a rolloff of 1/(2x3.14x23.5uFx10k) or 0.7Hz. Changing to 4.7uF gives a 7Hz rolloff. Changing to 10uF gives 3hz.
Other parts of the circuit are similar. If other caps elsewhere feed 100k, then 0.47uF to 1uF is fine. If they feed 10k, then use 4.7uF to 10uF.
Last edited:
Rayma's reply covers your last question. My suggestions about C105 were made on the understanding that the board layout for your circuit is compact.
C105's value of 1 uF is to attenuate some of the low frequency before it reaches the amplification stage. It has a secondary purpose of preventing the bias current of the IC from flowing through the magnetic cartridge. (In purist terms you do not want d.c. currents on an a.c. source like the cartridge windings - some don't think the level to have any significant affect).
C105 is an electrolytic type and theoretically a film type will be better, however a polypropylene replacement will pose difficulties in mounting within the available space. If it has to be a film dielectric replacement then a tombstone package MKT would be more practical.
You could change that in the first instance and listen to the equipment over a few days to see whether or not it meets with your expectations.
If not, then remove C105 and bridge the connections with a 100R resistor. ( The resistor is to reduce bias current flowing into the windings of the magnetic cartridge.) Removing the capacitor is removing a source of noise - at the expense of some bias current flow. The changed capacitor values suggested by Rayma should do what is needed at the mixer stage and you would change these now.
If you installed a SOIC socket to allow a DIL component chip replacement you can compare SE5532 to a 33078 without any de-soldering.
If you want to try direct coupling of the cartridge to input you could short the 100R resistor with a wire on the component side of the board.
If you stay with the original chip you would first need to check out the input bias current level from the data sheet to see how it compares with the foregoing equivalents.
Modifying equipment is a process of correlating what one hears with what one knows from experience or technical sources. While one can learn from semiconductor manufacturers datasheets and application manuals it is often useful to look at audio equipment service manuals to see what professional designers have done.
Last edited:
mjona,
Yes, there is very little space on the phono board so big poly caps are out of question. However I can try small MKT.
Input bias current (Ib) from opamp datasheets:
mc33078 --300/750 (typical/max) nA
SE5532 -- same
utc mc4580L (existing phono opamp) -- 100/500 (typical/max) nA
NE5532 -- 200/800 nA
looks like the stock opamp has the lowest Ib! I will try what you suggest and completely bypass the input capacitor C105.
greatly appreciate the help and will report back.
Herman
Yes, there is very little space on the phono board so big poly caps are out of question. However I can try small MKT.
Input bias current (Ib) from opamp datasheets:
mc33078 --300/750 (typical/max) nA
SE5532 -- same
utc mc4580L (existing phono opamp) -- 100/500 (typical/max) nA
NE5532 -- 200/800 nA
looks like the stock opamp has the lowest Ib! I will try what you suggest and completely bypass the input capacitor C105.
greatly appreciate the help and will report back.
Herman
Hello Herman,
It was getting late at the time of my last message.
I looked at the 4580 datasheet this morning and came to pretty much the same conclusion as you. The manufacture of IC devices is a competitive business and I looked up the datasheet for LM833 another industry stalwart which has been around for as long as CD players. It was introduced by National Semiconductor which was taken over by Texas Instruments.
33078 is On Semiconductors version. In the (May 2002) datasheet for they say:
"The all NPN output stage exhibits no deadband crossover distortion, large output voltage swing, excellent phase and gain margins, low open loop high frequency output impedance and symmetrical source and sink AC frequency performance.
They provide an interesting "Representative Circuit Diagram".
My information on LM833 is so old now that the paper is dog-eared and yellow so I decided it was time to download an update from Texas Instruments.
In reading this I noticed that the LM833 "Functional Circuit Diagram" has changed at some time since the days of National Semiconductor.
While such diagrams leave out a lot of detail and differences are possible as compared with 33078, I cannot distinguish between the two circuit diagrams.
The LM833 datasheet is comprehensive and includes notes on RIAA stage design methods with graphs, formulae, and calculations. Even if you stick with the 4580 the datasheet would be useful for study or reference purposes.
Here one can buy LM833 from an Electronics shop whereas 33078 is an online or telephone order proposition for which I have to wait for mail delivery. I bought some 33078's this way recently - so I have learned something too.
this phono get rid of capacitors and uses inductors in the RIAA.
Van Den Hul - The GRAIL | Product
""Phonograph RIAA - equalization with coils only, no sound impairing capacitors in filter""
just interesting to mention it since we are talking about removing capacitors.
Van Den Hul - The GRAIL | Product
""Phonograph RIAA - equalization with coils only, no sound impairing capacitors in filter""
just interesting to mention it since we are talking about removing capacitors.
I believe if something works well enough one does not need to find novel solutions.
Cyril Bateman tested capacitors of various dielectrics and published his results in Electronics World in 2002. With film types his preference was for polystyrene but he could not distinguish that from polypropylene. Talking about the best examples the distortion percentages - the first four figures to the right of the decimal point were all zeros.
Bateman's articles included circuit details of his test apparatus so people could build it and select the best performing capacitors from manufacturers standard product ranges.
Some 60's preamps made use of an inductor as part of a low pass filter. From memory the Quad 33 was one. Later designs from Quad used active filters instead.
- Status
- Not open for further replies.