• WARNING: Tube/Valve amplifiers use potentially LETHAL HIGH VOLTAGES.
    Building, troubleshooting and testing of these amplifiers should only be
    performed by someone who is thoroughly familiar with
    the safety precautions around high voltages.

6SC7 Phono pre amp

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
I built this pre amp mule while ago, and recently dug it out, rebuilt the PS and added a second channel,,, Played a couple dozen albums and find it to be a little lacking in bass response with a TechnicsSL-D30/P28,,, it seems to respond better with the Dual 502/Shure M91E...
I want to try a different coupling cap, replacing the .0047 to .0068 or .01 or so to see if there is any boost in the bass response...

Question is,,, Will changing that cap change the EQ of the pre amp? How about if the 22M grid resistor is left in place?

Some searching turned up this post,,,
"I noted that in the Tronola articles, the coupling cap in an Eico phono section was changed from a 0.05 ceramic to a 0.1 µf film cap. This made sense and improved the LF response. So sometimes it is beneficial",

Para phrase from...POST #15
http://audiokarma.org/forums/index....or-values-in-tube-amp-preamp-rebuilds.744091/
Any thoughts from the designers out there!!
Thanks for any info,,,
 

Attachments

  • P9090002.JPG
    P9090002.JPG
    79.4 KB · Views: 798
  • P9110012.JPG
    P9110012.JPG
    81.3 KB · Views: 766
You have 22mR for your grid stoppers. The data sheet only shows a max of 10mR under certain conditions. Outside of that it isnt critical, so I suppose what you are using is ok, but 22meg...wow.

You say this is a phono preamp, however I am not seeing a RIAA circuit in your drawing. That would leave your bass response totally hosed.
 

PRR

Member
Joined 2003
Paid Member
........ Any thoughts from the designers out there!!...

There's not enough there, in several ways.

6SC7 is "high Mu" but not as high as 12AX7.

All 2-stage 12AX7 phono preamps have shortcomings. Worse with a lower-Mu tool.

The classic connection is P2 to K1 NFB, but the 6SC7's common cathode foils that.

NFB from P2 to G2 will not really cover the whole 40+dB needed to do the RIAA curve well.

Passive EQ may give a nice flat RIAA response but is likely to get hissy at the high end.

This design dates from the late 1950s, when speakers were all like +/-10dB lumps, when GE was trying to promote their low-output cartridges and needed a LOW-cost preamp to make them work "acceptably" (by the standards of the time).

Sow's Ear. Even if not aiming for a silk purse, there's not enough here to make a plain purse you would carry. (If you carry a purse...)

Also the supply voltage is rather low (cost-cut).
 
Thanks for the "plug", Kevin. :)

It is possible to get some "mileage" out of 6SC7s in the tweaked RCA setup. The common cathode connection is not an impediment to sharing a tube between channels as the grid leak biased 2nd gain block triodes. When sharing tubes between channels, the 6H2Π-EB (6n2p-ev), with its internal shield, is a prime candidate for 1st gain block duty. Those folks with large bank balances can consider an ECC808/6KX8 as the 1st gain position tube.

Don't give even passing thought to AC heating, with the types mentioned. :mad: :down:
 

Attachments

  • Full Tweaked RCA Phono Stage.jpg
    Full Tweaked RCA Phono Stage.jpg
    45.4 KB · Views: 879

PRR

Member
Joined 2003
Paid Member
There's not enough there.....

I'm going to walk this back.

The old topology will work. Those NFB EQ values are just not right.

At least according to the Idiot Assistant.

Simulation notes:

No 6SC7 model, 12AT7 is loosely similar. Most SPICE models do not do 22Meg gridleak bias with realism (although the tubes do work well this way); I added small batteries to fake what really happens. For real parts, these batts should be shorts.

(And you want your 47K resistor before my C1.)

I used Koren's reverse-RIAA because it was handy. I did not check it.

The EQ accuracy WILL depend on the tube. Popping a 12AX7 in for TU1 gives a *lower* overall gain(!) and some rise below 50Hz, though otherwise flat. Smaller change for the +/-20% spread of 6SC7 production.

In Real Life: Gridleak bias works, and tends to give about maximum gain; but the exact operating point is uncertain. This means a bit more spread of working gain. Also for a given B+ you won't get maximum output before distortion. Some of these variations will reduce with lots of B+ (overwhelm contact potentials which add/subtract from gridleak potential).

When doing gridleak you want *big* coupling caps, far more than you expect looking at 22Meg.

The THD spectrum seems to be high because of poor operating points and heavy loading, particularly above 1KHz. In that most music falls above 1K, this may not be a real problem.

The EQ lacks the 20Hz corner defined by IEC in 1976. I believe any modern phono should do this. Apparently other folks have other opinions. There is no simple way to add a 20Hz corner (grid impedances are not defined, filtering at input raises random 1/f, at output is load-sensitive).

This is essentially a re-power and re-cap of what you have on the board, with a few added parts for proper EQ, so I don't propose to beat my simulator further.
 

Attachments

  • 6SC7-phono.gif
    6SC7-phono.gif
    17.9 KB · Views: 662
Last edited:
Ok, I see where your schematic came from. It is a GE design.

First of all thanks for replying!!! I had this over at AK for a week and got nothing!!! and I questioned it not showing RIAA EQ...

It works fine as its built, very quiet at any volume,, good mids and highs, just lacking in bass as mentioned, which is teh point of the thread... so, it seems raising coupling cap value won't change the EQ, as it really isn't there!!!!
 
It's not for modern pick ups playing RIAA equalized recordings. Take a look in Analogue Source for a bunch of appropriately EQ'd designs.

Eli has an improved version of the RCA passively equalized design, you might want to check it out.

Thanks for replying,,, I have built two RCA phono preamps, they may actually be on this forum... First one is a copy of the original,,, 2nd uses 6SL7s,, which to me sounds better... I've seen the improved version schem also...

This project was to see how this simple pre would perform, and seems it is similar to an old Muntz TV,,, gets the job done with less parts!!!!

I appreciate the help, I will try replacing the coupling caps, as I'd like to tweak it as much as possible, without going to a "better" or different design... There's a little learnin for me here, and I appreciate your input and help in optimizing this old design...
 
There's not enough there, in several ways.

6SC7 is "high Mu" but not as high as 12AX7.

All 2-stage 12AX7 phono preamps have shortcomings. Worse with a lower-Mu tool.

The classic connection is P2 to K1 NFB, but the 6SC7's common cathode foils that.

NFB from P2 to G2 will not really cover the whole 40+dB needed to do the RIAA curve well.

Passive EQ may give a nice flat RIAA response but is likely to get hissy at the high end.

This design dates from the late 1950s, when speakers were all like +/-10dB lumps, when GE was trying to promote their low-output cartridges and needed a LOW-cost preamp to make them work "acceptably" (by the standards of the time).

Sow's Ear. Even if not aiming for a silk purse, there's not enough here to make a plain purse you would carry. (If you carry a purse...)

Also the supply voltage is rather low (cost-cut).

I agree with the gain assessment,,, 6SC7 is closer to 6SL7 in gain than 12AX7,,, however this build is an exercise in using it... Its a nice sounding tube, that is available for for next to nothing... and has a built in shield...

I did build a cathode follower for it, but found it had too much gain, even tho CF isn't supposed to add any...

I don't want to mess with the NFB circuit, however the B+ thru the amp does seem low, I haven't tried to raise it as I can't see where it would help bass response and may add some noise...

Again, I truly appreciate the time you gave this question, I realize its not optimum, but its a fun winter project!!
 
Thanks for the "plug", Kevin. :)

It is possible to get some "mileage" out of 6SC7s in the tweaked RCA setup. The common cathode connection is not an impediment to sharing a tube between channels as the grid leak biased 2nd gain block triodes. When sharing tubes between channels, the 6H2Π-EB (6n2p-ev), with its internal shield, is a prime candidate for 1st gain block duty. Those folks with large bank balances can consider an ECC808/6KX8 as the 1st gain position tube.

Don't give even passing thought to AC heating, with the types mentioned. :mad: :down:

Eli,,
Nice to hear from you again,,, you were one of the first to assist me when I retired and started this hobby!!! I do appreciate your improved design, and am familiar with it,,, however this project is more for me to see what simpler, if any improvements can be made to this vintage circuit... Its been on the bread board for at least a year, as it had an annoying hum at first... I dug it out a month or so ago, and rebuilt the PS with bridges and oil bypass caps which made it silent, hence the effort to proceed...
 
I'm going to walk this back.

The old topology will work. Those NFB EQ values are just not right.

At least according to the Idiot Assistant.

Simulation notes:

No 6SC7 model, 12AT7 is loosely similar. Most SPICE models do not do 22Meg gridleak bias with realism (although the tubes do work well this way); I added small batteries to fake what really happens. For real parts, these batts should be shorts.

(And you want your 47K resistor before my C1.)

I used Koren's reverse-RIAA because it was handy. I did not check it.

The EQ accuracy WILL depend on the tube. Popping a 12AX7 in for TU1 gives a *lower* overall gain(!) and some rise below 50Hz, though otherwise flat. Smaller change for the +/-20% spread of 6SC7 production.

In Real Life: Gridleak bias works, and tends to give about maximum gain; but the exact operating point is uncertain. This means a bit more spread of working gain. Also for a given B+ you won't get maximum output before distortion. Some of these variations will reduce with lots of B+ (overwhelm contact potentials which add/subtract from gridleak potential).

When doing gridleak you want *big* coupling caps, far more than you expect looking at 22Meg.

The THD spectrum seems to be high because of poor operating points and heavy loading, particularly above 1KHz. In that most music falls above 1K, this may not be a real problem.

The EQ lacks the 20Hz corner defined by IEC in 1976. I believe any modern phono should do this. Apparently other folks have other opinions. There is no simple way to add a 20Hz corner (grid impedances are not defined, filtering at input raises random 1/f, at output is load-sensitive).

This is essentially a re-power and re-cap of what you have on the board, with a few added parts for proper EQ, so I don't propose to beat my simulator further.

Thanks for the simulator work, I will study the schematic... I want to try raising coupling cap values first, to see how it improves bass response....
I did find quite a discrepancy in the noise level of the tubes... RCAs seem to be the quietest and best sounding so far... I can also adjust the B+ but if that is necessary, it would be better to just follow a different design... I'm sure GE designed for a specific cartridge back in the day, but its fun to experiment with old designs... and it may lead to following your new mods....

Thanks again to all that contributed,,, I don't expect miracles!!!
 
I just did a comparison to the 6SL7 RCA phono pre I built a while ago,,, the GE/6SC7 is no comparison.... I think I'll retain the nice PS and come up with another try!!
I could try the schem PRR posted but I have a few questions about it... I'd like to try something that could use the 6SC7s...

Also, I boosted the coupling caps incrementally up to 10X and it still was lacking bass,,, so the missing RIAA network is very necessary, as was posted here... maybe this GE offering was for cheap portable record players, as 78s also have a bunch of different EQs depending on who recorded them...
We learn from our mistakes! thanks again for all the input...
 
The GE setup relies on the inductance of the cartridge to "match" the RIAA curve. Give GE their due, as they were 1st in making mag. carts. available.

I'd like to try something that could use the 6SC7s.

I've already provided a phono preamp option. You could use a 6SC7 as the LTP in a Mullard circuit power amp, instead of a 12AX7/ECC83, provided you use full pentode mode "finals". High RP/low gm types, like the 6SC7, get into trouble attempting to drive anything resembling a substantial load.
 

PRR

Member
Joined 2003
Paid Member
The GE setup relies on the inductance of the cartridge to "match" the RIAA curve.....

Yes. Confused here because post #1 shows 47K input. Given that, there are not enough corners in the EQ to do RIAA. The old-school trick of R-loading the needle's inductance does work, if the needle maker commits to a specific value. GE did, today nobody does.

Now that we see the Fisher PR6, we see the same thing with R1 noted as 6.8K or per needle maker's recommendation. L(cart)+6.8K makes the 2KC corner. 120K makes the 1KC gain. 2700p+120K makes a 491CPS corner. 50Hz happens just because it runs out of gain.

Mis-done with 47K input, I would expect screaming highs, which is not what knockbill reports. The shy bass may be weak 6SC7 or poor biasing (DC checks would be first- are the tubes even running happy?).

PR6 booklet with notes:
 

Attachments

  • PR6-booklet.gif
    PR6-booklet.gif
    104 KB · Views: 550
The highs are a little pronounced, I wouldn't say screaming,,, As far as cartridge loading, seems the specs were accurate at the time it was designed,,, Pickering cart at 2500R or a GE at 6800R ( factory wired), were they the only Mag cartriges available at the time?... I really didn't give a thought to the timing until I saw this manual,,, As I have pair of Fisher 50C clones from the same era, but it has a lot more EQ controls...

FWIW,, 6SC7 voltages are well within specs... Would changing the input load resistor from the common 47K put this back to the RIAA specs, the amp is designed for? or maybe it will only be accurate with one those two vintage cartridges?
 

PRR

Member
Joined 2003
Paid Member
> I'd like to try something that could use the 6SC7s...

Use the 6SC7!! I only simmed with 12AT7 because I do not have a SC7 model; and to some degree tubes is tubes.

(However I am uncomfortable with the results I got. I suspect SC7 is not as hot-rod as AT7; also my AT7 model may be hotter than reality.)

You need that many R and C in the NFB EQ to define RIAA well. Anything with fewer parts 'cheats'. One cheat is that cartridge loading, no longer fashionable (data not available). Another is like PR6, let the "50Hz" corner happen where the tubes run out of gain (so low NFB in an important part of the spectrum, and wide variation with tubes).

Also you do not use my 1V grid batts with real vacuum tubes; the model does not have real electron clouds so does not simulate what really happens for 22Meg Rg.
 

PRR

Member
Joined 2003
Paid Member
For Shure M91E, 720mH, try 10K input resistor.

BTW, with this trick your turntable-preamp leads may be 30 feet, whereas for 47K we can't go over 3 feet. Also the extrapolated response runs far past the audio band. OTOH you have to change the resistor for every cartridge you use, and there is some rise in effective hiss.
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.