Paralleled CC Mic Pre

[rackmonkey]

I've got a power supply for an old project sitting around and wanted to play around with some circuits I haven't tried out before with some inexpensive tubes as well. The supply puts out 280v B+ as it is (fwb). So using that I pulled this together over a couple of lunch breaks and had some questions.

One of the circuits I'm curious about is one I just saved a jpeg of a generic example of some time back. I'm not 100% sure where I originally saw it,  but it looks like one of Broskie's. I didn't save the article it probably came from with it (I roll that way ;-P). It's labelled "Humbucking Common Cathode". I always thought this circuit was just called a paralleled input stage, but I'm probably overlooking some subtlety that differentiates it. What makes this humbucking beyond just the paralleled triodes? I know you halve the noise when you parallel tubes, and thus "buck hum" but is there something else going on here i'm not seeing? Or is it just another case of many names for the same basic circuit?

I figured this up for a 6CG7 with a 6C4 cathode follower, with a 1:10 input transformer. Total gain I came up with is just under 40dB with no OPT loss figured in yet. Fine with that, given the output Z (see below).

Questions I have are:

I've biased the 6CG7 at -5v. The curves I played with with a 280v B+ just sort of led me this way.  Average characteristics in the data sheet have between -8 and -11 at 250v, but the input voltage swing should still be adequate for a mic pre, I'd think. Any issues with bias at this voltage?

I'm also not running this tube very hard with this config so I have some leeway to change things.

I came up with just about 400 ohms output impedance for the circuit. Assuming that's correct, I'm thinking a 500:150 on the output would be the logical choice, but would you do something different?

I've messed this calculation up before, but it looks like the high end of frequency response for this thing would be well up into the RF range. Messing around with the grid-stopper resistor within reasonable values doesn't change things much. I'm concerned about susceptibility to parasitic oscillation without some kind of HF attenuation. Is it advisable to add in a separate HPF as insurance here, or is it not worth worrying about?

I didn't want to use NFB as I just want to hear this circuit without it since I'm just playing around. Also haven't added gain/volume control yet. Will do that if/when I build it.

Any other thoughts? Thanks in advance.

BT

 

[ruffrecords]

To me those two tubes are not paralleled. It looks like a cathode follower feeding a a common grid stage, something I tried a long time ago. The 200K input grid stopper is in series with the input so its noise simply adds to the input noise making the overall noise performance of this circuit very poor. The 6CG7 has a low gm so it does not need grid stoppers. I think the hum bucking comes from the pot divider biasing the second triode grid which injects a portion of the HT supply noise in anti-phase into the signal path. It looks like another version of Broskie's  Akaido circuit.

Edit: The cathode follower has 100% negative feedback so you do not need to dc bias to greater than the input signal. Even with a 1:10 input transformer and -20dBu input, the grid signal will be not more than 0dBu which is 0.775V rms which is just under 1.1V peak. So as long as you bias is at least 0.5V greater than this there will be not grid current even if it wan not a cathode follower. As it is a cathode follower, the grid/cathode signal will be 5% of the input signal (6CG7 mu-20) so there is even less to worry about.

Cheers

Ian

 

[moamps]

Isn't it just a LTP?

 

[AndreasSchwarz]

Isn't it just a LTP?

Yes, but with half the amplification as a single common cathode stage.

Regards Andreas

 

[ruffrecords]

Isn't it just a LTP?

Not really because the tail is short - this is not a joke.

https://www.ampbooks.com/mobile/amplifier-calculators/long-tailed-pair/

Cheers

ian

 

[moamps]

I tried the calc on site you linked for Rl1=0, Rl2=22k, Rk+Rt=1k2 and got the gain of 19.2dB for 6SN7.

 

[rackmonkey]

Thanks, Ian.  Very helpful insight, esp on the bias point.  I'm still curious to hear this thing, especially since it'll only take 30 minutes to build it up on the easy-bake board.  I'll remove the grid-stopper.

Any thoughts on the need for a LPF? I think the grid stopper was useless for hf attenuation because of the circuit's effect on Miller-effect capacitance, is it not? Despite the differences, the effect should be similar to that of any paralleled triodes, shouldn't it? Take cathode-to-plate capacitance and multiply it by the gain of the second triode. Add that to the grid-to-cathode capacitance. I'd have to figure this in to any lpf I might add after the first stage, correct? Then again, adding one after the first stage won't prevent any oscillation in that stage anyway. Would you even bother with this or not?

Any thoughts on an OPT, given about 400Z out?

Thanks again for the advice!

BT

 

[ruffrecords]

For the first stage you are going to have a relatively low source impedance (15K or so with a 150 ohm mic and 1:10 transformer) so Mr. Miller is not going to have much chance to affect HF performance. Second triode is grounded grid so Mr. Miller is switched off completely. Between stage 2 and stage 3 is where he will appear. But again the source impedance of stage 2 is relatively low so the effect will be small. Personally I would leave well alone.

The way you have the output cathode follower (CF) biased the quiescent current will be about 4mA. As I have said many times before, output impedance is no indicator of drive capability - that comes from the quiescent current. The peak output current from a CF equals the quiescent so the rms output current is just under 3mA. Into a 400 ohm load this is just 3.6mW which is about +5dBu. 2:1 is probably about the right ratio for the output transformer but you do not want to present the CF with a 400 ohm load - it wan't drive it. 2:1 load with 600 ohms looks like 2K4 to the CF which is a much more sensible load. 3mA rms into a 2K4 load is 21mW which is about +13dBu. Still not very high but possibly usable. If you want to achieve +20dBu inot a 600 ohm load with a CF you will need around 10mA of standing current.

Which is why I much prefer using an SRPP output stage. It is a true push pull circuit so with just 5mA of standing current you get 10mA peak output current; enough to drive +20dBu into 600 ohms.

Cheers

Ian

 

[rackmonkey]

Great advice. Thanks again, Ian. I asked about the use of a hpf after some experience w/parasitic oscillation in a guitar amp I designed. Once bitten. I'll give it a shot as is on that point.

SRPP would cure the anemic drive capability and I've built a few pres with those so no questions there except for phase inversion. SRPP inverts phase while I don't think this humbucking first stage does (correct me if I'm wrong there). If I even worry about it (and I wouldn't unless this thing has some appeal and I end up building a pre for use with it), my inclination would be the lazy fix of just swapping the input or output leads. Not sure if there's a better way .

Appreciate all the advice.

BT

 

[ruffrecords]

SRPP does invert but there is no need to worry as you are driving a transformer - just flip the leads.

Cheers

Ian