12AB5 (6V6) triode in the output of a mic pre?

Hi,

I am re-designing a mic pre I have due to the fact that I wanted to eliminate the totem pole follower output. The rest of the mic pre is single ended, no negative feedback and I wanted to keep that topology through out as a I prefer that sound. The mic pre has a 225vdc B+ and all tubes are heated in series using a 50.4 vdc H+. That gives 12.6 vdc to the heaters of each of the 4 tubes and allows a phantom power supply without need a separate supply.

The way I have it set up right now is 1:10 input tranny > 7062 (e180cc) common cathode no negative feedback stage > coupling cap > stepped attenuator > 7062 (e180cc) common cathode no negative feedback stage > coupling cap > 5:1 (15k:600) output tranny.  The former design uses 2 6072 stages and a third totem pole follower stage using a 12au7.

Although the 7062 is very close in mu to the 6072 and more current drive ability I cannot quite get the same output and voltage swing due to the beefy totem pole follower the original had. I want to add a third stage to better drive the output transformer. I plan to try a 12ab5, which is a lower voltage 6v6 with 12.6 volt heaters in a miniature Noval footprint. I plan to wire it as a triode in common cathode no feedback which should give 10 mu, 5000 gm, and 1965 ohms internal resistance. I believe this will be a nice fit but I only know enough to be dangerous. will it be too much for a mic pre output? Am I gonna blow up my converters with this thing?

I believe this will be a nice fit but I only know enough to be dangerous. will it be too much for a mic pre output? Am I gonna blow up my converters with this thing?

Click to expand...

I don't see why you should, presumably you have some sort of control in the audio path?

I made a no-feedback amp on similar lines some time back (attached) this should give you some ideas.

If your output tube is going into a transformer, then the maximum voltage there can be on the primary is about twice the B+ voltage less the cathode voltage.  This is peak to peak voltage so you divide by 2.8 28 to get the rms voltage.

This primary voltage is then divided by the OPT ratio to give you your output voltage.  This is only a ballpark check as it's a little more complicated than that and it will be less than this method gives you.  This should be enough to stop you worrying.

DaveP

What do you mean by a totem pole output stage??

The drive capability of a single ended triode stage depends on the quiescent current. The E180CC data sheet gives a maximum recommended design plate current of 8.5mA. If you set up the output stage for this current then this represents the peak ac current available so the rms current can be no more than 6mA. With a 5:1 output transformer this gives an available maximum 30mA rms current in the 600 ohm load which represents 18V rms which is +27dBu.

This demonstrates what is theoretically possible. Unfortunately, this represent an 18 x 5 x 2.828 = 254 votl swing at the plate which is a tad more than your supply has available. That is why DavePs suggestion of a transformer as the plate load is a good idea because you can then in theory swing up to twice the power rail.

Bottom line is I think it is mostly a case of correctly setting up the present output stage.

Cheers

Ian

P.S but running it this way with a 225V power supply will exceed the maximum allowed dissipation so don't do it. Data sheet says 150V plate voltage max with 8.5mA quiescent current  which is 1.275V. If you run the plate at 112.5V (half the supply) you will stay within the dissipation spec at 8.5mA. For a conventional single ended stage this means a plate resistor equal to 112.5/8.5 = 13K. This will dissipate as much as the tube so better use  a 5W part. You should still be able to drive +20dBu into a 600 ohm load with this and a 5:1 ac coupled transformer.

Ian

What do you mean by a totem pole output stage??

Click to expand...

That's what some people call a White Cathode Follower (WCF)

DaveP

Thank you for the quick replies.

As for the totem pole output. Just as Dave put it, yes, I am referring to a white cathode follower style output or more specifically in this case a 12au7 cathode folllower with the second triode acting as an active current sink. The original design sounded good but I think it’s just subjective preference here. With the white cathode follower I was hearing less 2nd order harmonics and a little extra high harmonics. A nice sparkle to the sound, but I am partial to the fully single ended no feedback approach and sound even though it’s higher overall thd. I like the extra 2nd order harmonics and the lower high order. So, I am going for a very purist approach.

My initial approach was to find a triode with a similar mu to the 6072 where I would get a similar amount of gain for the common cathode stages but that had lower plate resistance and more drive capability so the second triode could also drive the output transformer. I also have the constraint of 12.6 volt heaters and 200ma heater current since I would like to leave the power supply alone. I scoured many datasheets before I found the 7062/e180cc. As a bonus, this tube has quite a linear transfer curve and sounds very nice. Even better than the 6072. With my changes, the mic pre sounds very good but I run out of headroom before I can quite get the output I would like. That’s why I thought I’d try a third stage with a more powerful tube.

One thing I realized is I have the 7062’s biased for too much idle current. I need to decrease the cathode resistors a bit. Had them biased more like a 12au7. I need to see what I get after that change. I have a 15k plate load reistor for the output triode, I will change that to 13k as well. That should give me a bit more plate voltage. Also, I have a 4:1 (10k:600) I would like to use for the output transformer  to salvage a few db of output but when I swapped it in for the 5:1 I got a similar output and a bit of a slower sound and rolled off high end. This is also what drove me to want to try the 12ab5 to drive the output. Honestly Ideally, I’d like to use a 4:1 gapped carnhill tranny and dc couple the output 

Thanks,

Matt

This would be a very good time to master the tube curves and be less "dangerous"!

The 12AB5 data you quote shows the rp is ~2k triode wired, so a good starting point for the load resistor would be 3.9k or 4.7k wire wound 12W.  With a suitable capacitor (Parallel feed), you should be in the right area. 

If you wish to keep the same power supply, then you cannot draw any more current than the  WCF you already have.  This means that you will have to bias the tube with a higher voltage/resistor to reduce the current to the WCF level.

As you will not be using feedback, the output resistance will be the parallel combination of the tube rp and the plate resistor, say around 1.3k.  This means that with the 15k primary of the OPT you will form a voltage divider.

As Ian says, a better solution would be a gapped OPT that can handle the DC current in the primary.

Good luck
DaveP

I have the 7062’s biased closer to the center operating point now but with the first tube having a little more idle current for a little assymmetry in that first stage. It’s sounds great, better than the stock pre for my taste but I’m just still not quite getting the output and voltage swing I would like.

Ian mentioned the max plate voltage being 150 volts, but I don’t believe that is correct. Every datasheet I see for the 7062/e180cc has the max plate voltage at 275 volts. It gives a typical operating spec where it shows 150 volts plate voltage but it has 275 volts for max plate voltage. So, could I use a gapped output transformer and connect the plate of the second 7062 to the primary to get more voltage swing or will it still be beyond max dissipation for the tube?

Thanks

Matt

will it still be beyond max dissipation for the tube?

Click to expand...

This is under your control, you increase the cathode resistor until you get the current you want.

(Vp-Vk) x current = tube dissipation

DaveP

mbf90 said:

Ian mentioned the max plate voltage being 150 volts, but I don’t believe that is correct. Every datasheet I see for the 7062/e180cc has the max plate voltage at 275 volts. It gives a typical operating spec where it shows 150 volts plate voltage but it has 275 volts for max plate voltage. So, could I use a gapped output transformer and connect the plate of the second 7062 to the primary to get more voltage swing or will it still be beyond max dissipation for the tube?

Thanks

Matt

Click to expand...

That is not quite what I said. I explained that to increase the drive capability of the output stage you need to increase the cathode quiescent current. The data sheet shows the maximum recommended plate voltage for 8.5mA of quiescent cathode current is 150V and you really need about 8mA of cathode current to get the required drive. You can run the plate at 275V but you would need to reduce the plate current in order not to exceed the dissipation of the tube and hence the drive capability which is not what you want.

Changing the ratio of the transformer will give you more voltage at the output but less current. So it will improve the output level into light loads (like 10K) but not for a 600 ohm load. If you use a gapped transformer as the plate load you will be able to bet a larger voltage swing so to get the current drive you need on the secondary you will need a higher ratio transformer.

What is the bias current right now in the output stage? Exactly what level of signal do you want out of this stage and into what load?

Cheers

Ian

What is the