Schade Common Gate (SCG) Preamp

It is an aesthetic thing for me not wanting magnetic resistors in the circuit. I readily admit I have never compared the sound of a magnetic resistor to one that is not magnetic.
In theory a magnetic conductor would be more susceptible to physical vibration than one that was non magnetic. It could also create vibration if sufficient current was passed through it. The magnitude of this effect would be miniscule hence the statement about phono stages. About 20 years ago I made a small SMD phono stage from a TI application note and chip. I stuck it inside the turntable base. I tried to make one small enough to fit inside the headshell, but that turned out to be the bridge too far.
 
I had thought it best to use a carbon film resistor for the gate stoppers even though I had usually gone for the old standby the carbon composition.

I checked a KIWAME I had bought for the TGV51 bleeder and it is magnetic, too.

I checked ones I had received from a TUBECAD kit and they were magnetic. Nothing special was claimed for them.

I have some RIKENs and they are not magnetic. Do they really sound better to those that say they do or is it the old expectation bias at work - they should sound better so they do. I know all about this phenomenon since I seem to enjoy being a victim of it. I have read that the gold plated leads are not really such a good idea.

At such low values could the carbon composition be the best choice/compromise? Just a rhetorical question.
 
I have found a (very) few cases where a carbon comp resistor will do better than than a carbon or metal film resistor at stopping oscillation. These have been in applications where high Gm vacuum tubes capable of operation into the UHF frequency range are used.

In theory there should be no difference in sound where no current flow is occurring (gate of a mosfet, or grid of a vacuum tube). Theory relies on "ideal" components that do not exist. If there is current flow through the resistor, the carbon comp can be noisier than a metal film.

My working SCG board has lots of Mouser budget grade metal film resistors in it, most of which are quite magnetic. The high power breadboard has some of them too along with some carbon film resistors. The stopper on the input fet is a 1K metal film. No issues there either. The combination does sound quite nice, now that the oscillation is fixed. I have since removed the snubber network across the OPT with no ill effects. The 1K 2W carbon film resistor in the snubber got quite hot and even started smoking when oscillation was present. That ceased with the addition of the ferrite covered carbon comp resistors on the gates or each of the 4 output fets.
 
  • Like
Reactions: 1 users
Thanks for that, Tubelab. I will leave the carbon composition resistors in the drawer. And stop obsessing over being attracted to magnets,

I am confident the circuit would sound good with almost anything. I know this is not important to the basic goodness of the sound.

Mot of the resistors I am using are DALE RNs. Hard to beat established quality.
 
  • Like
Reactions: 1 user

ra7

Member
Joined 2009
Paid Member
Yeah, at the time I thought it was another depletion mosfet but it is actually a chip to be used as a programmable CCS. So, I wouldn’t throw it in the gain position, even though I did :LOL:… it worked and sounded decent. But as the CCS part, it will work.

There is a wrinkle in that the output is taken from the source pin of the CCS FET, thereby looping in the AC load into the CCS correction loop. I imagine it will still work but watch the top end. I remember it was rolled off in the gain position with the 10M45S.
 
The original spec sheet did state that the 10M45S was a programable CCS "chip." The Block diagram on the old data sheet also indicated that it was a "chip" and mosfet pair. The current data sheet does not show this and some experiments with a curve tracer tend to reinforce the "it's really a mosfet" beliefs found on this and other forums. What's really inside remains a mystery. I should have taken one to the decap analysis lab when I worked at Motorola.
 
I have asked a version of this question before and I know there is not a definite answer but I continue to wonder and since ra7 and Tubelab have so much experience with these circuits I dare to ask again:

What would be the sonic effects of having the SCG as a buffer with no gain?

Just an opinion or conjecture would be appreciated.

I am using a xilica DSP with ZM's Singing Bush - my typical gain settings on the xilica are -14 dB input and -10 dB on the outputs. Total of -24 dB. Singing Bush is approx 18 dB per ZM.

An additional wrinkle is I will be using a CINEMAG input transformer for a balanced input from a balanced DAC which will add to the gain - at the moment everything is single ended.

From what I have read one does not want that much attenuation with a digital processor so that is why I am thinking follower. There are no large audible flaws but this cannot be ideal.

Sorry to go on so long and with the many details that for most are superfluous but without my reasoning would be even harder to understand!

All conjecture is welcomed.

Of course the only way to KNOW is to try it but if told it would ruin everything i will stick with half gain.
 
music soothes the savage beast
Joined 2004
Paid Member
In case there is anyone interested in speculating on this I did leave out that the question is based upon using the TDV circuit not the Singing Bush.
Use any of the M2X buffers, plenty to choose from.
SCG was designed as high gain. Asking for no gain seems counterproductive.
 
  • Like
Reactions: 1 user
Thanks for the guidance, especially to Ben Mah.

Obviously, I did not understand that aspect of things.

Would there be a way to use the proven elements of SCG and make it a "follower or common drain circuit"?

If a schematic could be provided I would be very grateful if this is not an absurd request.

And I realize it probably is.

I can always retain the gain and continue with the same levels of attenuation in the DSP.

My other question is could a DAC with a rated 5 volts output through the transformer drive a TDV/THF51?
 

ra7

Member
Joined 2009
Paid Member
Hi Rick, as others have said, you can reduce the gain of the SCG but making it unity gain would be not its best use. You could look to the B1 buffer if you already have enough gain to driver a VFET follower. The B1 would be just about perfect in this aspect because you could get all the gain you want and attenuate right before the power stage.

I am working on a power amp SCG circuit. Is that what you are asking about with "follower or common drain circuit"? In the meantime, you could try something like this. It is essentially a Zen Follower (TM) :D

My other question is could a DAC with a rated 5 volts output through the transformer drive a TDV/THF51?
If the transformer is stepping up the voltage (stepping down the current), the DAC would need a lot of drive (current) to ensure that it can drive the gate of the VFET. The other thing to watch for is that there is enough gain to deliver the power you need. Here is a quick table for reference:

OhmsPrmsVrmsVpp
8​
1​
2.8​
8.0​
8​
2​
4.0​
11.3​
8​
4​
5.7​
16.0​
8​
8​
8.0​
22.6​
8​
16​
11.3​
31.9​
8​
32​
16.0​
45.1​
8​
64​
22.6​
63.8​
 
  • Like
Reactions: 1 user
The transformer is 1:1 - used to convert a balanced output to single ended - to the SCG or to the amplifiers.

I have to admit I do not understand the table The transformers (CINEMAG bifilar) are nominally 600 Ohms. The DAC has a specified 100 Ohms output impedance.

I can always use the SCG as you intended it and retain the attenuation levels I have. I would use the 9 dB gain resistors.

My thought was that it would be better to be able to have less attenuation in the digital "domain".

I did not think the DAC could drive the amps but then I am skeptical that a B1 could either. I have plenty of Toshiba FETs.

I was thinking the power rails need to be where they are with SCG to drive the TDV amplifier? I am obviously unclear on how a buffer works.

I looked at the link circuit and that looks like a version of the MOFO. And the B+ strikes me as too low to replace what the SCG does - of course I am not too sure what is what!

I need something, to put it simply, to convert the DAC's balanced output which has been converted to single ended with a 1:1 transformer at its output to something that would drive a TDV using THF51s. If there is a follower circuit that would do the trick that would be great. Again, I would like to use the devices that you, ra7, have specified. In addition I have the power supplies.

To repeat and please forgive my repetition - for average loud listening approx 90db average I have the input to the DSP at -14 db and the outputs at an average of -10 dB - -24 dB total - The Singing Bush has gain of approx 18 dB per ZM.

I have gain to spare.

At the risk of driving everyone crazy my question is simply what would be required to allow the DAC to drive the TDVs using THF51 devices?

I greatly appreciate the help.
 
Last edited: