Jump to content

Deep End DIY - My first speaker project


acg

Recommended Posts

With a low output impedance, and no clipping at the required power (which isn't a huge ask), then an (pre)amplifier will have no issue driving the passive crossovers, and multiple amplifier inputs.

 

IMO the next step - if it were required - is rather than a "particularly boutique" stereo preamp .....  would be one specifically designed with stereo input(s), and multiple individual outputs.    In those individual outputs, you could address each amplifier input separately for impedance and gain (assuming they looked different to the source), and include the passive filter elements, etc.

  • Like 1
Link to comment
Share on other sites



1 hour ago, davewantsmoore said:

 

 

IMO the next step - if it were required - is rather than a "particularly boutique" stereo preamp .....  would be one specifically designed with stereo input(s), and multiple individual outputs.    In those individual outputs, you could address each amplifier input separately for impedance and gain (assuming they looked different to the source), and include the passive filter elements, etc.

+1

 

I kind of assumed this initially which is why I mentioned a 6ch preamp in my post yesterday. 

 

A buffer at the input to the passive filters in the amps would effectively restore the input of the amps to being 'normal' again and can deal with volume control and source selection elsewhere. 

 

Could think of it like a phono stage design:

 

buffer/gain (input buffer) -> passive filter -> gain (power amp input stage) ->

Edited by hochopeper
  • Like 1
Link to comment
Share on other sites

12 minutes ago, hochopeper said:

+1

 

I kind of assumed this initially which is why I mentioned a 6ch preamp in my post yesterday. 

 

A buffer at the input to the passive filters in the amps would effectively restore the input of the amps to being 'normal' again and can deal with volume control and source selection elsewhere. 

 

Could think of it like a phono stage design:

 

buffer/gain (input buffer) -> passive filter -> gain (power amp input stage) ->

 

The electronic filters are connected directly to the grid of each input tube AND they operate after positive bias has been applied but before the negative bias, if this makes any sense (it is a complicated bias arrangement to in order to remove the capacitor at input?  Also, the coupling cap between input and power stages can be used as part of the electronic filter eg. a 0.022uF cap into the impedance provided by a half 6C33C will provide a high pass filter at 60Hz.  I am wedded to keeping the electronic crossovers where they are and would need some convincing to put them elsewhere.

 

What I would more readily consider is multiple inputs with switching at the input to the amplifier.  For instance having the dac and tuner simultaneously connected to the amplifier and using relay switching to choose my source.  This would mean I get to keep using the excellent software based digital volume control for the dac (assuming it drives the amplifiers adequately, which I am quite sure it will) and could just install a better VC in the tuner.  That would be the best of both worlds and also leaves open the option of plugging in a preamp instead of the tuner if I were to add more sources (and plugging the tuner into the preamp).

 

 

Like this...

59092c068237e_SoundSystem2.thumb.jpg.1e2aa2256293e98009490b76d001ae5f.jpg

 

I would just have to figure out the very best way do the input switching at the amplifier, preferably via remote, but a manual switch on the side of the amps would also be fine.  There is already an isolated 12V service voltage in the amplifier for tasks such as running relays and fans.

 

  • Like 1
Link to comment
Share on other sites

2 hours ago, davewantsmoore said:

With a low output impedance, and no clipping at the required power (which isn't a huge ask), then an (pre)amplifier will have no issue driving the passive crossovers, and multiple amplifier inputs.

 

IMO the next step - if it were required - is rather than a "particularly boutique" stereo preamp .....  would be one specifically designed with stereo input(s), and multiple individual outputs.    In those individual outputs, you could address each amplifier input separately for impedance and gain (assuming they looked different to the source), and include the passive filter elements, etc.

 

I do like this sort of idea, but as I mentioned earlier I am wedded to the idea of the electronic filters residing within the amplifiers.  My biggest hurdle has been finding a pre/buffer that will not alter what my best source, the dac, provides.  Dual input switching at the amplifier solves this immediate problem quite nicely, at least in theory:  no additional RCA/BNC connections, no volume control, no additional gain stage between the dac and amplifier, just a single switch of some sort.  That is a very attractive solution to me right now. 

Link to comment
Share on other sites

12 hours ago, hochopeper said:

What is between the passive crossover and the input of the amplifier? 

 

The passive crossovers sit after the BNC input jack and attach directly to the pins of the input tubes (6E5P, 6E6P).  They are in the middle of the complicated bias arrangement that removes the need for an input capacitor.

 

 

 

12 hours ago, hochopeper said:

The differences that you notice are likely all down to how that interaction happens right? It's not really about being 'transparent' is it? That's hard to define and ambiguos. It' s about driving the next stage without any freq response changes and not adding any noticeable harmonic distortion.

 

Yes, that sounds fundamentally correct.  In practice the ideal solution will ultimately be judged by what I hear, but it will certainly need to measure well first.  Transparency in this context means "sounds the same", or "when the pre-amp is added in between the dac and amplifiers that there is no change in sound".

Link to comment
Share on other sites



47 minutes ago, acg said:

 

The electronic filters are connected directly to the grid of each input tube AND they operate after positive bias has been applied but before the negative bias, if this makes any sense (it is a complicated bias arrangement to in order to remove the capacitor at input?  Also, the coupling cap between input and power stages can be used as part of the electronic filter eg. a 0.022uF cap into the impedance provided by a half 6C33C will provide a high pass filter at 60Hz.  I am wedded to keeping the electronic crossovers where they are and would need some convincing to put them elsewhere.

 

I wasn't suggesting to put them elsewhere.

 

What I was suggesting is that .....   you're going to have a "preamplifier", with stereo input(s).

 

You could make it with one single stereo output, which is capable of driving your amplifiers multiple inputs in parallel.....   or you could design multiple separate outputs from the preamplifier, each one with optimised gain structure, and with optimised buffering and parameters to drive each individual amplifier input.

 

 

Whether this is valuable, may depend on how different each of the amplifier inputs looks to the preamplifier.      If they are all very different, then paralleling them all together and driving them with one single output circuit from the preamplifier might be a poor choice.

 

OTOH, if they all look reasonably similar and stable to the preamp .... then perhaps a stereo output from the preamp driving all the amplifier inputs together will be ok..

 

 

 

Link to comment
Share on other sites

44 minutes ago, acg said:

I do like this sort of idea, but as I mentioned earlier I am wedded to the idea of the electronic filters residing within the amplifiers.

 

That's ok.    Maybe I'll try and rephrase  (all over the place today sorry).

 

The question is can you drive all those amplifier inputs in parallel from a stereo preamplifier .....    or would you be better to make a preamp which has stereo inputs, and then individually optimised outputs for each amplifier input.

Link to comment
Share on other sites

39 minutes ago, acg said:

"when the pre-amp is added in between the dac and amplifiers that there is no change in sound".

 

I think what Chris is trying to say here is ......   "unless of course, the sound changes for the better".

 

This would be the case if your DAC is unable to drive the input to your amplifier well.

Link to comment
Share on other sites



22 minutes ago, davewantsmoore said:

 

But doesn't your amplifier need a much bigger input signal than your DAC can provide?  (6 volts?)

 

I don't think so at this stage.  The amplifier will slip into Class A2 with about 6V input and the sound will go to soup...in my current room I am pretty sure that I will get away with the 1.5V input limitation, however if I extend the room to double its current size then I may be stretching things, perhaps.  Romy currently has a larger room than I will ever have and at 4.5V input needs a little more volume so I am just trying to plan ahead a little.

 

 

37 minutes ago, davewantsmoore said:

 

I wasn't suggesting to put them elsewhere.

 

What I was suggesting is that .....   you're going to have a "preamplifier", with stereo input(s).

 

You could make it with one single stereo output, which is capable of driving your amplifiers multiple inputs in parallel.....   or you could design multiple separate outputs from the preamplifier, each one with optimised gain structure, and with optimised buffering and parameters to drive each individual amplifier input.

 

 

Whether this is valuable, may depend on how different each of the amplifier inputs looks to the preamplifier.      If they are all very different, then paralleling them all together and driving them with one single output circuit from the preamplifier might be a poor choice.

 

OTOH, if they all look reasonably similar and stable to the preamp .... then perhaps a stereo output from the preamp driving all the amplifier inputs together will be ok..

 

 

 

 

Yes, I see now.  I guess the next step is to build an amplifier and graph its input impedance to see just how 'bad' it is or is not.  

 

Chris also called me on the phone before to explain what he was trying to say (instead of fighting to make a detailed post on his phone).  His idea about placing at the input of each amplifier an active buffer specially suited to drive the passive filters has a lot of merit.  I guess it all comes down to the specific problems that I will have at that time and I can see the potential for a fair bit of experimentation at that time.

 

The first question to answer is whether the dac, my primary source, adequately drive the amplifiers to sufficient volume and with sufficient precision?  The dac is probably two-thirds of my listening, or perhaps more, and is the "best" source I am likely to own.  In the meantime I will consider how to leave room inside the amp cases for an active buffer and how to implement its power supply.

 

I am still very drawn to the idea of dual inputs to the amplifier and will try to figure out the best way for that to be implemented.

Link to comment
Share on other sites

57 minutes ago, acg said:

The first question to answer is whether the dac, my primary source, adequately drive the amplifiers to sufficient volume and with sufficient precision?  The dac is probably two-thirds of my listening, or perhaps more, and is the "best" source I am likely to own.

 

If in any doubt, just put a buffer on the DAC outputs. If it's a valve DAC output, err on the side of caution, add a high-current solid state buffer.

Link to comment
Share on other sites

If we think about the passive filter design in isolation first.

 

The simplified topology of a phono amp that I mentioned this morning looks like this:
 
1.png.a1cd0be5f29a6e3056f510ba5a02f7e1.png
 
Where in your case:
At the moment the input buffer (A) is provided by the output stage of your DAC and any other source.
Passive filter network of R1, R2, C1, C2 would be your passive speaker crossover filter
A' would be the input stage of your power amplifier
+1 would be the output stage of your power amplifier
 
 
The variability in the interaction between the output stage of the source and the passive filter is the reason we're having this discussion. In my mind it's because you're missing this unity gain buffer stage. As @davewantsmoore pointed out also, ideally you'd have a separate buffer in front of each passive filter and that way you'd have 6x buffer input impedances in parallel rather than 6x passive filter impedances to drive in parallel.

 

 

Here are some options to represent the whole signal path in a simpler block diagram, I think these are all discussed above, with a few missing. 

 

1. is the 'base case'. I've intentionally broken out the DAC IC and IV to separate stages as that's what's happening and I don't want the 'simpler is better' crowd to think that this has a dramatically less complex signal path than the other options so I've emphasised that.

 

2. is the option for relays at to select inputs where you could bypass the buffer for one input and have other inputs go via a buffer/gain stage. 

 

3. The preamp option

 

4. I ran out of time to draw but would have multiple buffers, this would mean changing the amp chassis to have a separate input connector for each filter input, which would be annoying at this stage of the build process. So i wasn't too concerned about drawing it, although it should be there for completeness.

 

Capture.thumb.PNG.70322c1c243edc769e7f7d5635a9283b.PNG

 

 

 

 

Edited by hochopeper
  • Like 2
Link to comment
Share on other sites

11 hours ago, hochopeper said:

 

The variability in the interaction between the output stage of the source and the passive filter is the reason we're having this discussion. In my mind it's because you're missing this unity gain buffer stage. As @davewantsmoore pointed out also, ideally you'd have a separate buffer in front of each passive filter and that way you'd have 6x buffer input impedances in parallel rather than 6x passive filter impedances to drive in parallel.

 

 

Here are some options to represent the whole signal path in a simpler block diagram, I think these are all discussed above, with a few missing. 

 

1. is the 'base case'. I've intentionally broken out the DAC IC and IV to separate stages as that's what's happening and I don't want the 'simpler is better' crowd to think that this has a dramatically less complex signal path than the other options so I've emphasised that.

 

2. is the option for relays at to select inputs where you could bypass the buffer for one input and have other inputs go via a buffer/gain stage. 

 

3. The preamp option

 

4. I ran out of time to draw but would have multiple buffers, this would mean changing the amp chassis to have a separate input connector for each filter input, which would be annoying at this stage of the build process. So i wasn't too concerned about drawing it, although it should be there for completeness.

 

Capture.thumb.PNG.70322c1c243edc769e7f7d5635a9283b.PNG

 

 

 

 

Thanks Chris.  Thinking about this, and now that the steel is cut and ready for me to pick up, I have a few physical constraints at the amplifier end of things...have a look at the image below...

 

590a636888489_InputFilterRoom.thumb.jpg.2ccb1cf4fad63ecc31f5f8c2381b9d4e.jpg

 

You can see the back of the BNC adapter pointing up and the input tubes (x6) above and behind.  Beside are the two white boxes of two filter chokes and below them the aluminium box encased in steel where the filter for the tweeter reside.  The remainder of the filter components and the last component of the bias string will fit between the BNC and the various input tubes.  It is a pretty tight fit, especially given the size of the filter chokes...

 

590a64e3dde33_InputFilterRoom2.thumb.jpg.2544748a3b2a8ce4b30f9be2b0ace783.jpg

 

I can shift the pot next to the BNC input (it sets the DC Offset) and add one more BNC socket, but I don't physically have room for a third BNC unless I shift the filter chokes a little further away (which is possible).  Trouble with that option is that if I had 6 BNC inputs and 6 buffers that I would be relying on wire (aka antenna) to get from the furthest BNC to their respective input tube whereas at the moment it will all be component-to-component soldering (aka ratsnest).  I am sure that any input switching arrangement could be snuggled next to the big white choke or even tucked in immediately under the copper coloured shelf.  Then I would have to figure out if I need a DC offset pot for each BNC input...I'm not sure...but at this stage I doubt it.

 

Where does this leave us?  Well, 6 x inputs and buffers is I think out....not enough room.  2x BNC could be workable, with local input switching and perhaps even a gain stage on the second input, although at this point of the exercise I think I would prefer the single gain stage to be external to the amplifiers...mainly due to incorporating power supplies for the buffer into the amplifier PS and the difficulty that I will have tweaking it and the input filters if things gets too congested at that corner of the amplifier.  That leaves Option #2 or a variant thereof as my preferred solution.  If I ever get a third source then I will need the external pre/buffer which means that there is probably little difference between having the gain stage in the amplifier or in the preamp.  I think I will go with putting it in the preamp more so for simplicity of the amplifier build at this stage.

 

  

 

  • Like 1
Link to comment
Share on other sites



1 hour ago, acg said:

Where does this leave us?  Well, 6 x inputs and buffers is I think out....not enough room.  

 

 

I should also say that the thought of 6 sets of interconnects gives me heart murmurs.  After all, I have mounted the amplifier to the horn stack so that I don't have to see 6 sets of speaker cables!

  • Like 2
Link to comment
Share on other sites

Aw you guys.

 

 

Last weekend I spent doing a million things around the house but I did find some time to trim up some ancillary parts for the upcoming builds,,,namely the various rods were trimmed to the exact lengths required.  By the time they were all squared up it actually took a fair while to do all of these...

 

590bdec671e5f_RodsAnyone.jpg.6d44ec50b0e68b95a099ec398e16763d.jpg

 

Just some holes to tap in  the ends and in various parts of the square bar and they are ready for use.  The smaller 10mm aluminium rod are for the inside of the DSET amps and I will brush them and then add a clear coat.  The larger 30mm aluminium rods are will hold the horns to the horn stack and I am undecided if I will brush them and add a clear coat or if I will apply a black patina...not sure...will have to see what it all looks like when it is ready to go together.

 

 

 This morning I picked up some steel for the Horn Stacks and the Bass Cannons.  I have a few parts to test to make sure they fit as intended and I have only enough cut for the cannons to test 5 "tubes"...the rest will come later when I ok the prototypes.

 

590bdec3b7d05_CannonFodder.jpg.f1abbb4958940d3889e8b6ba895e9340.jpg

 

 

It is the Horn Stacks that have me excited though...

 

590bdec4b575c_HornStackBits.jpg.e30ec7d0908f29638a00ec193a5e845d.jpg

 

590bdec2c624d_TheBeanstalk.jpg.4ce4533ea8474c76896c858ffcc86a8b.jpg

 

To give you an idea of the scale of the speakers (next to our stunted Christmas Tree!)...

 

590bdec59cf80_JackbeNimble.thumb.jpg.c3d02990a881db232b6389b2eb89f80a.jpg

 

I'm 6'1" tall and the horns are bigger!!

 

 

 

 

 

 

 

 

  • Like 3
Link to comment
Share on other sites



So, three days straight of drilling steel and including the efforts of the previous weekend I'm up to about 35 hours on the tools to get the Bass Cannon prototypes, Horn Stacks and DSET chassis' ready for powdercoating (why am I doing three projects at once?).  I've now done most of the easy stuff and it is largely fiddly and tricky stuff from here but if I had to guess I reckon I am about one-quarter to one-third of the way to getting all that steel ready for powdercoating.  The welder is lined up for about a fortnight so hopefully I can have his stuff ready by then which gives me a semi-realistic target of handing everything over for coating at the end of the month.

 

Fingers crossed.  It would be nice to put the empty horn stacks in the room and maybe just install the RAAL tweeters and perhaps one or two other channels below to get some noise.  Maybe even just go mono at this stage.

Link to comment
Share on other sites

15 minutes ago, Upfront said:

What, no photos Anthony!?!?!

Sent from my SM-G900I using Tapatalk
 

 

Oh, bugger it...I just went downstairs and took some random shots for you Linc.  First a DSET floor...10mm steel...

 

Floor.jpg.e42456e9d4df2964ecb12e6812d47b5a.jpg

 

 

...and a pile of DSET panels and stuff, the floor sheets at the top.

 

Stuff.jpg.dc2bb35869866fb62388b3babe3718af.jpg

 


Actually in the background Linc you can see some damping tubes for the pneumatics dangling from the patio rook while the paint dries...that hammertone takes a good month to get hard.

 

  • Like 4
Link to comment
Share on other sites

Spot on regarding the hammerite. Noce stuff but painfully slow drying.
I'll be interested to see how these bass cannons come together.
Also, how will you attach the horn to the stand?

Sent from my SM-G900I using Tapatalk

Link to comment
Share on other sites

32 minutes ago, Upfront said:

Also, how will you attach the horn to the stand?

 

 

The horns will have a pair of 30mm diameter aluminium rods that screw into their rear (on 10mm threads).  See below for an illustration of the tweeter...

 

59110f97d6700_HornConnection1.jpg.210d3f286c54f8c344fc2dc495b4da1f.jpg

 

 

These rods will pass through a pair of sleeves in the frame and at the rear will be a third plate through which a 10mm bolt will pass and then screw into the back of each rod.  Twist the pair of bolts and the driver moves forwards and backwards.

 

59110f96da9b7_HornConnection2.jpg.b8f4ad05687e2619e19a368560b892ef.jpg 

 

 

Clear as mud!

 

 

  • Like 3
Link to comment
Share on other sites



  • Recently Browsing   0 members

    • No registered users viewing this page.




×
×
  • Create New...
To Top