12AU7 Parallel Push Pull

[TIMBO]

Hey guys,
I have built Rob Robs micro bassman with good results and the 12AU7 sound pretty good.
Volume wise is not bad either, so...
12au7s are still cheap enough but OT's are getting up there in price.
Rob Robs micro bassman uses a 22.5k/8ohm OT.
If I was to use two 12au7 and parallel the triodes in push pull could I use an OT with a primary as low as 8K

[shooter]

should be reasonable, thou, I might use a 16ohm on the 8ohm tap just to fudge it closer to "normal". 

[kagliostro]

+ 1 for Shooter

Franco

[tubeswell]

What the others said. The lower you go for the load, the colder the biasing you need to avoid overheating the tube under signal conditions, which will rob output power.

[TIMBO]

Thanks guys.

[tubeswell]

FWIW even 16k is on the light side. Optimum Class A load line for 300V B+ is 33k (so for push-pull, 33k plate-to-plate, which will drop to 8k2 in Class B). The 1k4 Rk value is per triode (i.e. 700R for 2 triodes. 680R should be in the ballpark, no guarantees)


Edit: if running a 16k load @ 210V B+, also okay, but lower bias voltage (5.2V with 200R shared Rk) means more difficulty getting into Class B (because of sooner onset of grid current limiting). Note that if B+ = 300V, cooler bias (i.e. 13V with 850R shared Rk) is needed to stay under Pmax when signal starts wiggling around, but would be better for getting to Class B.

[TIMBO]

Thanks mate.
I'll be aiming for that 300v on the plates as per Robs original amp
I have good success with the 100v line transformers in the OT department

This one will give me more options to get a better primary impedance.
So looking at the chart it's possible to get a primary up to 32K

[tubeswell]

Sorry just re-read your post heading. So are you wanting to use 2 parallel 12AU7 on each side for push-pull? (I.e. 4 triodes altogether?)

[TIMBO]

Yep.
The 8k primary OT I mention was a more available OT than the 22.5k .
I was assuming that two valves each with their two triodes paralleled would need to have a plate load of around 11-12k
So 8k being too low or 16k a better match

[tubeswell]

Yep.
The 8k primary OT I mention was a more available OT than the 22.5k .
I was assuming that two valves each with their two triodes paralleled would need to have a plate load of around 11-12k
So 8k being too low or 16k a better match

Okay - so running 2 x 2 triodes at 300V through an 8k plate-to-plate reflected load on your OT is equivalent to the 33k loadline above. Each opposing pair would want a 680R shared cathode resistor

[TIMBO]

AWESOME!!!

[tubeswell]

Sorry What am I saying! Brain is missing a fuse today. Running 2 x 2 triodes in PP is equivalent to the 33k loadline per triode plot above, but your quad wants 16k plate to plate load for optimum Center bias. Use 680R shared Rk for each pair at 300v B+, and check that each triode biases at 2.75W.

You could use an 8k plate to plate load, (which is equivalent to the 16k load line plot above), but you want a cooler bias at 300V (2.25W per triode) so use 820R shared Rk per p-p pair to start with but maybe go next size up if bias is over 2.25W

[TIMBO]

All good mate!
I only throw in these curve balls because I"M confused....
Thank god for you guys.
 

[AmberB]

In the past, I've wondered about using line transformers for small output transformers.  This thread gives me the information that I had been wondering about.
Now I have to wonder how this would apply to 6SN7s...

[TIMBO]

Hey AmberB,
The local electronics store Altronics | Electronic components, AV, CCTV, Power & More
Have three 100v line transformers 15, 20 and 40w that have an the ability to reflect an 8k load and lower.
They have good frequency response and way cheaper than the real deal.

[TIMBO]

Hey guys,
This has gotten too confusing...
I will be using two valves only, with each valves triodes parallelled in a push pull circuit
 
And my speaker load - one speaker of 8ohms and one 4ohms can be 12ohms or 2.6ohms

[tubeswell]

Hey guys,
This has gotten too confusing...
I will be using two valves only, with each valves triodes parallelled in a push pull circuit
 
And my speaker load - one speaker of 8ohms and one 4ohms can be 12ohms or 2.6ohms

So what else is known? B+ = 300V still? OT impedance ratio? With 2 x 2 12AU7 in push pull at B+ = 300V, you want a plate-to-plate reflected load of 16k to make optimum output power (as I said previously). If so, I wouldn't use both speakers - it should be one or the other. If an 8 Ohm speaker, the OT impedance ratio wants to be 2,000:1 to make 16k. With a 4 Ohm speaker, 4,000:1. You're more likely to be able to find a 2,000:1 OT.


However, if you want to use your 8k-8R OT (1,000;1 impedance ratio) you need to bias the 12AU7s cooler to stay within Pmax (as I also said previously), but you'll make less output power. If you want to keep each triode biased to -13V, and have a single cathode resistor for all 4 12AU7s, you want a 430R cathode resistor with a 1W power rating. The next closest standard resistor is 470R (but you'll be a bit cooler still (so even less output power). Or you could try 390R and check to see what the Pdiss is once you have it running. You don't want it to go above 2.25W per 12AU7 triode (if you want the tubes to last longer)

[shooter]

if you want to use your 8k-8R OT (1,000;1 impedance ratio)

wouldn't a 16 ohm or 2 8's in series be an option?

[tubeswell]

if you want to use your 8k-8R OT (1,000;1 impedance ratio)

wouldn't a 16 ohm or 2 8's in series be an option?

Yeah - but I understand he's restricted to 1 x 8R speaker and/or 1 x 4R speaker

[TIMBO]

Hey guys,
Big thanks for all the hard work.
I'm revisiting this MINI BASSMAN
We didn't continue with this project due to a miss print on the board itself.
My partner in crime wasn't convinced that the sound produced was of a good quality.
We were intending to sell the boards as a kit and due to the cost to reprint the boards we didn't believe it was a viable venture.
SO I've got a heap boards to play with 
My power supply should hit approx. 300v
I was intending to use the two speakers but not a must do
I got plenty of AU7 valves this is the reason for the parallel pushpull
I'm keen to attenuate the output of the PI to reduce signal to the AU7 grid
I have done it here

and has worked well

[shooter]

I was intending to use the two speakers but not a must do

wire in series for 16 ohm, or use a single 16 ohm

I'm keen to attenuate the output of the PI to reduce signal to the AU7 grid

The MV should be fine for testing, it keeps all the pre "sound" just lowers amplitude (drive)
once it's up n run'n then if you don't have enough range you can tweak R's

[kagliostro]

A new PCB seems can be obtained on the cheap if you place your order in China and wait some time

Franco