Serious design effort begins on mini-console project

[mashaffer]

OK, I see what you are saying. So the problem would be primarily when approaching grid current right? I haven't been able to test the situation at overload as of yet because the mic mixer I am using as a preamp is not capable of driving the power amp that hard. As it is my ears and the speakers are working pretty hard at maximum volume. I suppose I could use a RCA to 1/4" adapter and try plugging in to the mic inputs. I will have to dig out the manual for that thing and see what the gain is.

 

[Kegger]

Not so much grid current, as it's the output impedance of the driver stage feeding
into that input impedance of the output stage and it's ability to drive and be linear.
(the more it has to drive the more this comes into play as well)

It's nice if you can keep that ratio to say 10x or higher, so say you have that 220k
input impedance on the output section then shoot for 22k or less output impedance
on the driver section would be a good spot to be in, I'm not saying what you have is
horrible, just analyzing the circuit and how R18 looks to effect it.

 

[mashaffer]

OK, recalculate R18 for 22K total (i.e. R18 in series with rp || R2) and then reduce R4 to have the same ratio as currently with respect to the driver Zout correct?

 

[mashaffer]

Just had a thought. I could use the last section of the can cap for the cap on the 6CG7 B+ and put a discrete cap on the input of the filter string. In other words shift everything back one and put new smaller cap at the head of the line. Lots of possibilities. Only problem is I can't bring myself to stop listening to it long enough to take it back upstairs and sling more solder.

 

[Kegger]

OK, recalculate R18 for 22K total (i.e. R18 in series with rp || R2) and then reduce R4 to have the same ratio as currently with respect to the driver Zout correct?

Sounds like a plan to me.

 

[Kegger]

Just had a thought. I could use the last section of the can cap for the cap on the 6CG7 B+ and put a discrete cap on the input of the filter string. In other words shift everything back one and put new smaller cap at the head of the line. Lots of possibilities. Only problem is I can't bring myself to stop listening to it long enough to take it back upstairs and sling more solder.

I hear yu, tough to take a build down that works and "adjust" it.. ahh take yur time..

 

[mashaffer]

I took a look at this last night and it looks like in order to get over 5mA I will need to reduce the plate load also. This shouldn't be a problem though since my current load is about 8 times rp at 56k. Seems like 47k or even 33k should be OK.

I haven't done rigorous calculations yet but a quick simulation seemed to indicate that if I went down to 11k on R18 then a feedback resistor of 180k should put me at about the same feedback level.

 

[Kegger]

Using 11k R18 an fbk res of 180k also gets you a bit of extra voltage to the 6CG7.
(it gets voltage from the output tube side, an a bit of extra current with that to)
(agreed you can drop the plate load off some)

With that an 39k plate load and 1k cathode should net you just over 6ma an 200v
still on the plate, that looks like a pretty decent range to be in, probably what I'd
be doing in that setup about there. (that "should" make for a better driver)


-------------

I am sure you know all this, and I don't mean my posts to come across as I'm the
expert here and you know nothing, just bantering back n forth on the design here.

 

[mashaffer]

Thanks! will give it a try.

 

[mashaffer]

OK. Finally got around to trying the modifications. I changed the cathode resistors to 1K 2W MO (didn't have any MF in that value but I don't notice any increase in noise). For R18 I paralleled another 22.1K MF with the existing one (this was easier than replacing). I changed out the plate load resistors and put 39k MF in their place. Due to what I had on hand I also decided to parallel 200k MF resistors with the existing resistors which gives me about 120k total which should increase the FB by a couple of dB.

I am getting about 7mA on the 6CG7s now. When I add the preamp this will lower the B+ a little so that will drop some but it should be a pretty good operating point when all is said and done.

I have just finished listening to the "New World Symphony" and I don't notice any radical difference in sound or gain so I think that the FB level must be OK. Now listening to Dudley Moore's "Songs Without Words" album and the clarity and presence seems good.

All in all I don't think I broke anything.

Next step is to build proper vented enclosure for the drivers so that I can get a better idea of the tonal balance. I will probably put a passive line level HP at 40Hz between the mixer/preamp and the power amp so that I don't over work the little woofs. The final preamp will include the HP internally.

Appreciate the help Kegger.

 

[mashaffer]

Thought you might want to see it right side up.

With the braces still on...

Sorry for the grain in the following. There is only so much a consumer grade point and shoot digicam can do in low light.

 

[Kegger]

Pretty sweet!

 

[GloBug]

Nice project.

 

[mashaffer]

Did a little more work on a beta version of a ported enclosure. Ugly but effective...

The box is made out of scraps. The port is a chunk of sump pump tubing cut to size. I used heavy duty spray adhesive to attach 1" thick poly batting to most interior surfaces and glued it all together with gorilla glue. There were some holes that needed filling so I just shot in a bit of caulking into the holes.

The box (about 0.5 ft^3) is about twice the size of the small sealed enclosures I was using. The new cabinet is pretty solid down to 40Hz and sounds quite nice with music. With test tones I can hear some funny noise from the port but that might be due to the fast warble tones I am using. Still I am considering using a slot port in the final version. I also will be making the enclosure wider and less deep.

I will probably go ahead and glue together a mate to this one as I already cut out the pieces. Then, on to the preamp section.

Made a crude little video of the little driver in action. Of course the audio is about what you would expect from a cheap digital camera shoved right up next to the port with African drums running at full tilt.

http://www.youtube.com/watch?v=EsPOqpO6XS8

BTW, is there a way to embed a you tube video in a post?

 

[mashaffer]

In long term testing I have come to the conclusion that the margin of xmax is inadequate for a young person's system so I am looking at either building a subwoofer now or adding helper woofers. Over at DIYaudio I have been discussing the issue on the vendor board for the maker of my FR drivers...

http://www.diyaudio.com/forums/markaudio/194579-mini-console-based-chr-70-gen1.html

I have an 8" driver (salvaged) that could be used for a small TL sub.

It models quite well. The surround looks totally fine but even if that were a concern I have found very similar units available for not a lot over $50. However, I am currently leaning toward using integral helper woofers.

These two both look promising in ported enclosures at or slightly under 1ft^3.

http://www.madisoundspeakerstore.com/approx-8-woofers/aurasound-ns8-385-4a-8-woofer/

http://www.madisoundspeakerstore.co...rasound-ns6-255-4a-6-paper-cone-woofer-4-ohm/

Oddly enough the larger driver has a lower specified Vas and actually can be tuned lower (according to simulations) in the same size enclosure. The efficiency is a bit lower than the 6" in this situation but it appears that the lower efficiency is a better match to the FR. The output simulations indicate that the woofers would be able to hold up to over 100dB (each woofer) throughout their passband before becoming xmax limited.

The plan would be to put the FR in a small sealed sub-enclosure of about 5 liters and use a 2nd order passive LP on the woofer at about 100 - 110 Hz and maybe a simple blocking cap for the FR. Conveniently the bass control on the preamp will be almost exactly covering the same range as the woofer so a little bit of tweakability will be built in.

The added bass may present some microphonics issues but I am starting to mull over some decoupling options such as possibly using foam insulation board or sound blocking material to line the amplifier compartment and attaching the chassis to the insulating material with screws and insulating grommets to try to minimize the transfer of vibration to the chassis and cut down on airborne vibrations in the compartment. Another chance for a learning experience.

 

[mashaffer]

OK, so I built a passive crossover for a helper woofer. I plan to order a pair of 6.5" woofs on payday that sim well in the space that I have available and should match up nicely in level.

I tried the new crossover the with sub woofer shown above and even though the efficiency is notably lower the results were night and day improvement of just a ported box FR with solid fundamentals (though down in level a bit due to efficiency mismatch) down into the 40Hz or a bit below.

The FR is afforded quite a bit of protection with the 1st order HP and the sealed enclosure in comparison to the ported enclosure. Driver excursion is maybe 1/4 of what the ported box exhibits.

Crossover is at 100Hz with 2nd order on the woof and 1st order on the FR. I think we are on the right track here finally.

 

[mashaffer]

OK, Been a loooong time since I worked on this project. I have been having some health issues and haven't had the energy for it. With things dragging out I have been rethinking the front end.

All of the fancy tone controls, sub-woofer outputs etc. are very nice but probably unnecessary. So I am thinking it might be better to modify things a bit and get 'er done.

Here be my thinking...
1) I have been using that little rat shack mixer as a preamp and that layout almost gets enough volume out of it (not quite making it to clipping) and I doubt that it has all that much gain (just a two transistor gain stage with interstage feedback). Initially I thought of adding a single gain stage but I may not even need to do that. If I redesign the input stage of the PA to use a 12AY7 I should be able to get a bit more gain and I doubt that I would be giving up enough in sound quality to be noticeable. I would just use a passive volume control going into the 12AY7. Might need some more help from you Kegger.

2) As to the tone controls, consider that this will be used primarily for iPod and CD so the need for adjustment is probably little to none. Also the helper woofer is being crossed in right where a bass control would be. So it occurs to me that I can just use an L-Pad in either the FR or the Woofer circuit (which ever is loudest). I might even add a passive high cut tone control at the input to the power amp, maybe.

3) A filtered sub-woofer output also seems unnecessary as virtually all powered subs have speaker level input so all I need to do is have a set of connections on the back of the main console from the OPT.

Thoughts?

mike

P.S. Sorry for being gone so long. Missed you guys.

 

[cdk99]

Glad you bumped it, interesting reading. I'm a newb here and would have missed it. In a way this skirts being a tube boombox, which I would like to have myself. I travel and something portable but dulcet would be useful.

Would a tube boombox be lighter if it were SET? Since we're using FR drivers anyway...

 

[mashaffer]

This one is single ended but with big Pentodes. For a boombox I would consider SET 6EM7 or 6BM8 in PP to save weight on the output iron.

mike

 

[mashaffer]

OK, had a chance to diddle a liddle tonight just to see if I would be able to increase the gain somewhat with the switch to 12AY7 from 6CG7 in the VAS stage of the PA. I have done nothing to optimize a 12AY7 operating point nor have I even thought about what that might look like but given the higher rp and lower cathode current for the 12AY7 I ran a simulation with the load resistor on the now 12AY7 model from 39k to 100k, the cathode resistor from 1k to 1.6k and the feedback resistor from 118k to 270k. The DCV at the plate of the 12AY7 is very close to what it was with the 6CG7 and the output at the speaker increased by about 6dB.

So it seems like it might be a feasible approach. I just need to figure out the right bias point and feedback.

mike