Tried to add transformer to Pentaboost

[Esppse]

Hello,

I was trying to add a Mouser Xicon 42TM018 to the end of the pentaboost.

https://www.diystompboxes.com/smfforum/index.php?topic=70524.0

The result when running through the xfmr was very choked and tinny. I basically just when out of the Pentaboost into pin one, and out pin 4, then grounded pins 3 and 6 of the TM018.

Is there an impedance issue here, why the full audio signal is not passing through?

I also ran a Super Hard on boost in front of the Pentaboost, it drove the tube harder, but still the xfmr sounded tinny.

Are they supposed to sound this choked? It basically turned a humbucker into reversed phase single coils, that's how much low end was lost. Am I missing a part of the circuit to pull this off correctly? I was hoping to have manual control over transformer saturation.

[Rob Strand]

Is there an impedance issue here, why the full audio signal is not passing through?

More than likely.  The 100k (actually more like 38kohm) output impedance of the Pentaboost driving
a small 10kohm transformer is going cut the lows off  maybe even lose some output level.

[Esppse]

Hmm I thought so. I tried putting a 100k pot between the xfmr and output of the pentaboost and swept it but it didn't fix the issue. Any suggestions I can try?

[Rob Strand]

Hmm I thought so. I tried putting a 100k pot between the xfmr and output of the pentaboost and swept it but it didn't fix the issue. Any suggestions I can try?

The root cause is the inductance of the transformer loading down the source impedance.   When you add the 100k resistor it stops some of the loading at the Pentaboost output but then you get the effect when the 100k hits the transformer.

If you add a 100k like you have now then add another 10k across the transformer it will improve the bass a lot, maybe not enough, but the side effect is a very large loss of overall level.

If you want to stick with that transformer you would need to put a buffer between the pentaboost output and the transformer.   If you use an opamp buffer it's more than likely the output swing of the pentaboost is  going to cause the opamp to clip which will change the sound in a bad way.    So you will need to drop the output level using a divider before the opamp, maybe 100k in series and 220k to ground.   You will need an output cap on the opamp and also biasing for the opamp.     Another plan of attack is to use a BJT buffer DC coupled to the tube.  Use a high gain transistor (not a darlington) and maybe a lowish emitter resistor like 2k2 - that will save a lot of parts since the bias comes from the tube and no need for the divider.

EDIT:
Here's an example using a *MOSFET* buffer.  You can get a JFET or a BJT buffer to work (provided you are careful about bias points and choice of Vgs_off on the JFET.)

https://electronics.stackexchange.com/questions/174255/need-a-voltage-follower-buffer-for-a-tube-how-do-i-calculate-and-select-a-mosfe

[Esppse]

Thanks for the in depth explanation. I'll build the smallest buffer here, using a 2N5457. Hopefully it'll push the mismatch in the correct direction.

https://2.bp.blogspot.com/-ciSdhLUOXw8/VDAax2GGD-I/AAAAAAAAIog/jDJcvEog3ls/s1600/Buffers.png

[Rob Strand]

Thanks for the in depth explanation. I'll build the smallest buffer here, using a 2N5457. Hopefully it'll push the mismatch in the correct direction.

https://2.bp.blogspot.com/-ciSdhLUOXw8/VDAax2GGD-I/AAAAAAAAIog/jDJcvEog3ls/s1600/Buffers.png

It should work and will definitely help the response issue.   It could change the sound a bit if the JFET isn't biased right.  Might be worth checking the source voltage on the buffer.  If you need to raise the source voltage add resistor (large value like 2.2M to 10M ) from gate to +V.