Preamp Idea

[coldcathode]

Hi all,

Working out a preamp idea, please look over this and comment, suggest, etc etc.

[coldcathode]

I had a go at the Power Supply

[anatech]

Hi coldcathode,
Before examining anything in detail I can see a couple things.

You might consider regulating your phono pramp section to prevent signals from the output stage from affecting your high gain stage. "Motorboating" may be an issue with this build otherwise.

Your "line in" should be terminated to ground with a resistor. Normally you would want to block any DC from external sources also.

The grid of your 6CG7 should be referenced to ground to avoid those big noises when switching sources. Where are you going to put your volume control?

Your output impedance should be as low as possible in order to drive the cable and input of the amplifier. Most amplifiers are around 47 K, although we are seeing units as low as 10 K these days.

Your power supply. A 5U4 to run three signal tubes?? Excessive I think. Consider something smaller like a 6CA4. This will also reduce the heater current draw. Your 12 V transformer is rated too low on current for one, try 1 ampere. It will run too hot as rated. Also, consider running your heaters AC, but biased up 30 ~ 50 VDC. If you really want to regulate your phono heaters, then you need a transformer rated for about 16 VDC. Your regulator may "drop out" on the lower part of the ripple on the raw DC supply with a 12.6 V transformer. Not what you want to hear! Understand that your regulator will get very warm. Use a heat sink and test run it to make sure you do have enough cooling.

-Chris

[GMuffley]

If you want to use cathode bypass caps, they need to be MUCH larger in value--at least 100 mfd.

[HollowState]

To somewhat repeat what the others have said:
A 5U4 is overkill. Better a smaller rectifier tube or perhaps nice quite schottky diodes. Decoupling the B+ between the two tubes with an R/C is necessary. Get a fixed resistor or volume control on the 6CG7 grid and a much larger cathode bypass capacitor if you want one there. I would not use an anode follower for the output either. Without running the numbers my gut says you will suffer high frequency response here.

Also:
The RIAA feedback as shown can't work because the 10uF cathode bypass capacitor puts it at AC ground shorting out the feedback signal. Is that a 4.7 meg resistor in the network? Seems quite large. And why the 47K between cathodes? I'd loose that as well as the other 10uF capacitor.

DC for the filaments is good for low hum but you do need a higher voltage to start with. (Higher current here would help too) You can't rectify and then regulate a 12 volt source and expect to still get 12 volts out under load. The regulator needs at least a few volts differential to function. And you need a much larger capacitor off the filament rectifiers. I like to use several thousand uF.

[coldcathode]

Anatech -

You might consider regulating your phono pramp section to prevent signals from the output stage from affecting your high gain stage. "Motorboating" may be an issue with this build otherwise

I considered that at the outset. One thing I still have difficulty with is determining the B+ current draw of the circuits. Any help on that will allow me to put and RC network for each section.

Your "line in" should be terminated to ground with a resistor. Normally you would want to block any DC from external sources also.

I am adding 47K from "line in" to ground and a coupling cap of .01uF Mylar, sound good??

The grid of your 6CG7 should be referenced to ground to avoid those big noises when switching sources. Where are you going to put your volume control?
Your output impedance should be as low as possible in order to drive the cable and input of the amplifier. Most amplifiers are around 47 K, although we are seeing units as low as 10 K these days.

added 470K from Grid to ground at the switch. Volume Control is at input of amp hence the need to adjust parallel resistance with 1M Pot and amp's input grid load.

Your power supply. A 5U4 to run three signal tubes?? Excessive I think. Consider something smaller like a 6CA4. This will also reduce the heater current draw. Your 12 V transformer is rated too low on current for one, try 1 ampere. It will run too hot as rated. Also, consider running your heaters AC, but biased up 30 ~ 50 VDC. If you really want to regulate your phono heaters, then you need a transformer rated for about 16 VDC. Your regulator may "drop out" on the lower part of the ripple on the raw DC supply with a 12.6 V transformer. Not what you want to hear! Understand that your regulator will get very warm. Use a heat sink and test run it to make sure you do have enough cooling.

5U4GB is OVERKILL, I am working with parts on/hand. I will revisit the B+ and maybe go with a 520 tranny and Diode Bridge. That tranny is a nice Gillfilan Bro's potted tranny. A lot of extra mA's but I will find uses on this chassis for it. I am thinking of adding another "Tape" stage with NARTB EQ. as well as cascading another gain stage with -6DB Low pass @ 150Hz for a mono block sub woofer amp. (2 x 829B's PP ??)

The filament supply for the ax7's- I will change to a 30V tranny I have laying around. Rectify that and drop it to about 17V, then use 1 LM7812 per tube and heat sink those, placing them outside the chassis on top in open air.

[coldcathode]

Originally posted by HollowState
To somewhat repeat what the others have said:
A 5U4 is overkill. Better a smaller rectifier tube or perhaps nice quite schottky diodes. Decoupling the B+ between the two tubes with an R/C is necessary. Get a fixed resistor or volume control on the 6CG7 grid and a much larger cathode bypass capacitor if you want one there. I would not use an anode follower for the output either. Without running the numbers my gut says you will suffer high frequency response here.

Also:
The RIAA feedback as shown can't work because the 10uF cathode bypass capacitor puts it at AC ground shorting out the feedback signal. Is that a 4.7 meg resistor in the network? Seems quite large. And why the 47K between cathodes? I'd loose that as well as the other 10uF capacitor.

DC for the filaments is good for low hum but you do need a higher voltage to start with. (Higher current here would help too) You can't rectify and then regulate a 12 volt source and expect to still get 12 volts out under load. The regulator needs at least a few volts differential to function. And you need a much larger capacitor off the filament rectifiers. I like to use several thousand uF.

Hollowstate:

As posted before I agree 5U4GB was overkill, I am revisiting the B+ source as I write this.

470K Grid to Ground added

Bypass Caps removed from Phono Section (were not in the circuit that I adapted from Dynaco PAS-2 not really sure what I was thinking)

As far as anode follower suffering high frequency loss I do not follow? The circuit info I got from RC-30 indicates that the loss would be well above typical audio frequencies. ie: > 25Khz

The 4.7M and the 47K between the Cathodes are in the Dynaco Circuit. I am assuming for a good reason (One that I certainly cannot figure out, LOL but the PAS2 was a pretty well respected circuit correct?)

Also as mentioned above the Filament supply will be revised. I am also considering using the 30V tranny and going all DC filaments, using a LM7805 with a 1V diode drop on the ground.

Any comments on ALL DC filaments.

BTW, will go with 10,000uF Caps after the regulator.

[kenpeter]

Cathode #3 impedance is so low, the "feedback" from plate #6
through the RC network is completely swamped out. Except at
the highest frequencys, where the caps quite simply ground out
plate #6. Rather than act as-if a legitimate negative feedback...

I'm also unsure what 47K does between two cathodes of low
impedance. Probably nothing.

--------------------------------------------------

I had a similar idea the other day in regard to a directly coupled
cathodyne. What if I fed back the plate to the preceeding stage
cathode? Could I lower the impedance at the plate (of the
cathodyne) to match the Cathode (of the cathodyne)???
In a word, No.

Tried in LTSpice for hours in vain, the impedances of plate V2
and Cathode V1 are just plain incompatible for such feedback
without additional buffering in between. All that happens is
Plate V2 grounds out though the feedback resisior to no good
effect.

I even tried CCSink with the "feedback" resistor doubling as
the effective "plate" resistor in the cathodyne totem, turned
back toward virtual ground as it were... That one didn't work
either. I'm not yet sure why.

[HollowState]

Originally posted by coldcathode
The 4.7M and the 47K between the Cathodes are in the Dynaco Circuit. I am assuming for a good reason (One that I certainly cannot figure out, LOL but the PAS2 was a pretty well respected circuit correct?)

IMO Dynaco (aka Hafler) was responsible for some of the more cheap and dirty designs I've seen. They were very popular because they were inexpensive. And they were great vehicles for modifacitions because they needed it.

Even so, I do own several pieces of Dynaco like the MK-III's, ST-70 & DH-200) but mostly for collector reasons. I also have some of fheir larger output transformers (440's & 451's) that I will build something with when I get the time and physic gasoline.

[anatech]

Hi coldcathode,

BTW, will go with 10,000uF Caps after the regulator.

Please don't do that. It's not necessary and without a reverse biased diode across your regulator (from in to out), it will probably die the first time you switch it off. It probably goes into current limit on power up too.

Your main storage goes before the regulator. Small decoupling capacitors go directly before and after the regulator. That's it. You can add RF bypass caps at each heater if you wish.

5U4GB is OVERKILL, I am working with parts on/hand.

Cool. Works for me.

Hi HollowState,

IMO Dynaco (aka Hafler) was responsible for some of the more cheap and dirty designs I've seen.

Without a doubt!

-Chris