6BM8 "one tube reverb"

[tubenit]

I have read quite a bit of info in ARCHIVES on the one tube reverb approach using a 6BM8 tube.  I've noted the difference between capacitor vs. transformer driven.  However, I don't understand a significant amount of the information around the 6BM8 designs in that thread.

So, in summary, for tube circuits, "usually" if you see a transformer coupled reverb driver, the tank input will be 8Ω. A cap coupled reverb driver will connect to a high Z tank input. And all tanks will have a high Z output.

My question is I already have Hoffman's 022921 reverb transformer and 4AB3C1B tank.  These are currently installed into my Tweed BluezMeister using a 12A_7 tube. 

I've never tried the 6BM8 "one tube reverb" idea.  I'm curious what it would sound like and if it has more "verb" then the 12AX7?

Can I use the same reverb transformer and same reverb pan and simply rewire for a 6BM8 tube as shown in the attached schematic?  OR …. would I need a different transformer or different tank?

With respect, Tubenit

[sluckey]

That tank will be fine, but the 6BM8 has a plate impedance of about 5600Ω so that 022921 reverb transformer is far from ideal. The 6BM8 would be much happier with Doug's Heyboer Output for Champ and Reverb units or any 5K or 6K primary SE output transformer. 5 watt size is fine.

[tubenit]

Great information!  THANKS, my friend.

With respect, Tubenit

[TIMBO]

Hi T, I just measured the transformer in an old Fi Sonic amp I have and it's more like 10k primary with a 8ohm load.

I have also used the 022921 seems to work ok.
 

[tubenit]

I have also used the 022921 seems to work ok.

Thanks, I appreciate and value your experience. Thank you for posting on this thread.

I am understanding this to mean you used the 22921 reverb trannie with the pentode section of the 6BM8?  Is that correct?

And if so …………… did it sound "better" to you then a one tube 12A_7 type?    Or just different?

I am trusting that Sluckey's advice about using a Champ type OT would be closer to ideal.  Having said that,  if using the 22921 with the 6BM8 pentode sound better then using it with a 12A_7, then it might be worth experimenting?

With respect, Tubenit

 

[TIMBO]

Yes, used the transformer on pentode side.
I don't think there is any one valve that sounds "better" than the other.
The difference between the pentode and triode driver is there is much reverb with the pentode.

But, Is the pentode basically acting as a triode?

 There is a difference in the reverb between transformer driven and capacitor coupled, kinda like the capacitor coupled.

[tubenit]

Good information!  I know you've done some incredibly innovative amps and experimented with some remarkably cool ideas, so I am grateful for you to take the time to share your experience.  Thanks!

With respect, Tubenit

[TIMBO]

No worries mate

[2deaf]

Valve Wizard gives a rule of thumb for the primary impedance on a centre-biased class A single-ended amp:  The plate voltage squared divided by the maximum plate dissipation.  You seem to use around 350V on your reverb drivers and the max dissipation above 250V is 5W.  350 squared is 122,500 divided by 5 is 24.5K.  The 022921 is right in there at 25K:8ohms.

If you measure the slope on one of those typical pentode lines, you could easily get the impression that the internal plate resistance is in the 20K range.

[tubenit]

That's great information!  Thank you!   

With respect, Tubenit

[PRR]

> get the impression that the internal plate resistance is in the 20K range.

Around the 300V op-point suggested, I get 200K.

You *never* "match* a pentode to a load.

And the reverb tank works best with a high-Z source.

[2deaf]

> get the impression that the internal plate resistance is in the 20K range.
Around the 300V op-point suggested, I get 200K.

I get 2K over by the knee and 2M down by cut-off.  But like I said, I get 20K for one of those typical pentode characteristic lines, which would be one out in the middle somewhere with a slope that is similar to a bunch of the other ones.  I only mentioned this because there was some confusion earlier between R_a and r_a.  Sometimes it's hard to tell what they mean on datasheets, especially non-English ones that use symbols other than the ones we are used to.  A rough idea of r_a taken from the slope of a characteristic line is frequently sufficient to determine what they meant.

You *never* "match* a pentode to a load.

I don't know if this still refers to me or not, but I have debunked the matching idea multiple times on this forum and various other ones (where I have just about as much clout as I have here).

And the reverb tank works best with a high-Z source.

This clearly doesn't refer to me.  Furthermore, I have no idea what you are talking about.  They work really good with a low-Z constant-current source. 

[PRR]

> I have debunked the matching idea multiple times...

Excellent!

> a low-Z constant-current source.

Puzzled.

[2deaf]

Excellent!

I think that you're the funniest guy on this forum.

Puzzled.

Well, that's probably because I don't understand what you said, which is frequently the case.

With the driver in this thread, the low impedance side of the transformer is what I consider to be the source that is driving the input coil on the tank.  The current delivered to the input coil is reasonably constant within the range of interest from 2 Ohms to 24 Ohms, which I consider to be a constant-current source.

Sand with a low output impedance and the capability to deliver constant current over the input impedance range that it is driving seems to work pretty good, also.   

[2deaf]

The 6BM8/022921 reverb driver can really drive a tank if you are willing to use enough current.  Node "B" on the power supply would be better than node "C" if you are going to use a lot of current.  If you have 358V from node "B", you will probably have around 354V on the 6BM8 plate and maybe a few volts on the cathode, so around 350V anode to cathode.  At the 5W maximum dissipation, that would be 14.3mA max. at idle.  I'm going to go 90% of that, so 12.9mA will be my target.

There are lots of ways to get an idle current of 12.9mA, but some ways are better than others.  The pentode uses more DC current than a triode for the same reverb drive because the pentode has screen current in addition to the drive current.  The higher the screen voltage the higher the screen current, so I don't like the 2.2K screen resistor.  Instead, I want to increase the screen resistor and decrease the cathode resistor so that more of the total current consumption is being used for the reverb drive.

Since more current is going through the first power supply dropping resistor, its value would need to be reduced to maintain the usual node voltages.  Probably something like half the original value.

[2deaf]

This one has a performance that is similar to the classic Fender 12AT7 strapped-triode driver with a bypassed 2.2K cathode resistor.