Hi,
I recently received an custom built (DIY) all-tube pre-amp. The line-stage works just GREAT but the phono-stage is not so good at all. It sounds to low, no dynamics and as if the tweeter was not working at all. Something is wrong and I suspect the design of the schematics to be off (wrong Cap/resistor vlaues probably?). Problem is, the person who built it cannot hear it as I do and could not fix it either.
Would anybody here be willing to run the schematics (which I have in PDF format) through spice or tubecad or whatever tool available, to check if all the parts values calculate correctly? This would be of tremendous help to me.
The schematics is very similar to the pre-amp described in the book "hoechst empfindlich" of Goetz Wilimzig.
2 E88CC in the phono-stage and 1 E280F (Triode mode) in the line-stage, with passive RIAA EQ.
I am an illiterate when it comes to electronics-design but like to build things. So I could try to fix it once somebody who knows better has determined the correct values or found a theorethic error.
Many thanks in advance for any help offered!
Dieter
I recently received an custom built (DIY) all-tube pre-amp. The line-stage works just GREAT but the phono-stage is not so good at all. It sounds to low, no dynamics and as if the tweeter was not working at all. Something is wrong and I suspect the design of the schematics to be off (wrong Cap/resistor vlaues probably?). Problem is, the person who built it cannot hear it as I do and could not fix it either.
Would anybody here be willing to run the schematics (which I have in PDF format) through spice or tubecad or whatever tool available, to check if all the parts values calculate correctly? This would be of tremendous help to me.
The schematics is very similar to the pre-amp described in the book "hoechst empfindlich" of Goetz Wilimzig.
2 E88CC in the phono-stage and 1 E280F (Triode mode) in the line-stage, with passive RIAA EQ.
I am an illiterate when it comes to electronics-design but like to build things. So I could try to fix it once somebody who knows better has determined the correct values or found a theorethic error.
Many thanks in advance for any help offered!
Dieter
Hi,
It is not a cartridge impedance problem. I tried various MM & MC/Step-Up set-up and compared the results to other (OP-Amp & Tube) phono stages.
This baby always came out way last.
But for the record: I tried a Shure MM with 90K load & 47K load. I tried various capacities with NO impact on the sound quality of this Pre.
Same (bad) results with a clearaudio Stradivari (MC) via a Silvercore 1:10 100K Step-Up which usually sounds marvelous, using 47K, 90K and NO load at the phono-input.
cheers, Dieter
It is not a cartridge impedance problem. I tried various MM & MC/Step-Up set-up and compared the results to other (OP-Amp & Tube) phono stages.
This baby always came out way last.
But for the record: I tried a Shure MM with 90K load & 47K load. I tried various capacities with NO impact on the sound quality of this Pre.
Same (bad) results with a clearaudio Stradivari (MC) via a Silvercore 1:10 100K Step-Up which usually sounds marvelous, using 47K, 90K and NO load at the phono-input.
cheers, Dieter
dikarner said:
Hi Dieter
Sorry for the trivial question, i didn't realise you were a seasoned vinyl-phile
A circuit seems essential then. Any idea what "NO load" means? There is obviously a resistor at the first tube's grid which determines the maximum value of the input resistance. I assume you change resistors in parallel to this internal resistor to set your desired value. The difference between 90k and 47k is usually quite subtle; what i had in mind was loading a MM cart (or the secondary of a transformer) with a 100ohms.
Schematics
I just now realized I could upload as well
. I hope this will suffice for anybody willing to do the math on the RIAA inverson:
There's one small error: the last resistor, connecting the Output to ground (2,2Meg) is missing on the circuit diagram.
The values are as published in the book. Actual, if different, values are provided in brackets. Major differences to the official version are marked in red.
The descriptions are in German:
müF --> uF
U1/U2 --> V1/V2
fehlt im Original --> does not exist in original schematics
BR
Dieter
I just now realized I could upload as well
. I hope this will suffice for anybody willing to do the math on the RIAA inverson:There's one small error: the last resistor, connecting the Output to ground (2,2Meg) is missing on the circuit diagram.
The values are as published in the book. Actual, if different, values are provided in brackets. Major differences to the official version are marked in red.
The descriptions are in German:
müF --> uF
U1/U2 --> V1/V2
fehlt im Original --> does not exist in original schematics
BR
Dieter
, but could you please be more precise?
Sorry, I wrote it in a hurry. I refer to the paper of Stanley P. Lipschitz titled "On RIAA Equalization Networks" (J.A.E.S. 1979 June). See section 8 Addendum where he describes the formula for passive eq networks. The relevant formulas are:
R1 * C1 = 2187 us
R2 * C2 = 109.05 us
R1 * C2 = 750 us
R2 * C1 = 318 us
R1 / R2 = 6.877
C1 / C2 = 2.916
R2 is R3 on your diagram, C1 is C4 and C2 is C5.
C4 = 3.0 nF and C5 = 1.0 nF give quite good values. The plate resistance ( cca. 16k according to my calculation) adds to R1.
R1 * C1 = 2187 us
R2 * C2 = 109.05 us
R1 * C2 = 750 us
R2 * C1 = 318 us
R1 / R2 = 6.877
C1 / C2 = 2.916
R2 is R3 on your diagram, C1 is C4 and C2 is C5.
C4 = 3.0 nF and C5 = 1.0 nF give quite good values. The plate resistance ( cca. 16k according to my calculation) adds to R1.
BTW I also use a passive RIAA circuit, but with 3 gain stages and the eq network is divided in 2 sections. Each gain stage is an SRPP. The gain is sufficient, noise is low and it is easier to calculate the values of the passive circuit. I described my circuit here:
http://www.diyaudio.com/forums/showthread.php?s=&threadid=119229
http://www.diyaudio.com/forums/showthread.php?s=&threadid=119229
Hi Dieter
What oshifis probably means is that some cap values seem an order of magnitude off the mark using the common formulas. Which does not necessarily mean they are wrong. I tried calculating the response of your riaa and it seems indeed wrong - 3db down at 2kHz and -6db at 5kHz and above. Of course it's best to simulate the riaa network but i'll only get time to do that tomorrow.
A set of safe values, with guaranteed response are the following:
R1 = 950k
R3 = 96k3 (95k3 can be retained with small error)
C4 = 3n3
C5 = 1n13
These assume that R8 is in place as shown (and that i haven't made a silly mistake).
The gain should probably be raised from the second stage. I haven't given this much thought.
Seems like i posted at the same time as oshifis
What oshifis probably means is that some cap values seem an order of magnitude off the mark using the common formulas. Which does not necessarily mean they are wrong. I tried calculating the response of your riaa and it seems indeed wrong - 3db down at 2kHz and -6db at 5kHz and above. Of course it's best to simulate the riaa network but i'll only get time to do that tomorrow.
A set of safe values, with guaranteed response are the following:
R1 = 950k
R3 = 96k3 (95k3 can be retained with small error)
C4 = 3n3
C5 = 1n13
These assume that R8 is in place as shown (and that i haven't made a silly mistake).
The gain should probably be raised from the second stage. I haven't given this much thought.
Seems like i posted at the same time as oshifis
Don't worry about plate resistance - it's too low to affect anything. And it's neither 16 nor 8k Your R1 is however wrong - there is 2M2 in parallel and that's why i suggest 950k.
Changing the gain of either stage won't affect riaa much. The first stage will have a negligible effect and the second none.
Your R1 is however wrong - there is 2M2 in parallel and that's why i suggest 950k.Changing the gain of either stage won't affect riaa much. The first stage will have a negligible effect and the second none.
A grid resistor is required, and as analog_sa wrote it affects the accuracy of the equalizer circuit. A trick which I use is that I put the coupling capacitor C1 before the RIAA network and the grid resistor right after C1 (plate - C1 - R14 to GND - R1 - ...). Then the grid resistor does not load the equalizer network and it can be 1M or less.
Oshifis
Not sure if i got the picture right but it makes absolutely no difference where you put the cap or the grid resistor - it is still in parallel with the series combination of R1 and whatever is the output impedance of the driving stage. There is one potential advantage in having C1 preceding the riaa though - you can use low voltage riaa caps.
Not sure if i got the picture right but it makes absolutely no difference where you put the cap or the grid resistor - it is still in parallel with the series combination of R1 and whatever is the output impedance of the driving stage. There is one potential advantage in having C1 preceding the riaa though - you can use low voltage riaa caps.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.