SLOB project

Dedicated to those large boxes at one end of the room
Post Reply
Cressy Snr
Social Sevices have been notified
Posts: 8651
Joined: Wed May 30, 2007 12:25 am
Location: South Yorks.

Re: SLOB project

Post by Cressy Snr » Sat Dec 21, 2019 1:53 pm

Ray P wrote:
Sat Dec 21, 2019 1:38 pm

...The bit where I really was sleep-walking was the cap calculations. I used this cap calculator but I have no real idea why I put those values in to it;

https://www.v-cap.com/coupling-capacitor-calculator.php

The calculator notes talk about phase anomalies around the corner frequency and recommends a rule of thumb of using a corner frequency a tenth of your target frequency to avoid the anomalies and I think that's what adled my brain....
Hence my 6800uF output caps. Blowing up one of the smaller value caps with my knackered old arcing 13E1 on the breadboard, around this time last year gave me the chance to rethink and recalculate via that v-cap calculator, with the phase taken into account.

10,000uF (2Hz) to get decent phase down to 20Hz would have been ideal for my speakers, but I had none of those, and 6800uF gets me down to 2.9Hz (30Hz phase flat...ish) which will do. Yes...I know speakers aren’t resistors. Anyway; off topic, so as you were. :)
End of file found unexpectedly.

User avatar
ed
retired
Posts: 4034
Joined: Thu Jun 21, 2007 4:01 pm
Location: yorkshire
Contact:

Re: SLOB project

Post by ed » Sat Dec 21, 2019 3:01 pm

IslandPink wrote:
Sat Dec 21, 2019 12:27 pm
It would be a shame to use an active crossover into a super-transparent OTL amp and a FR driver - you're limiting your sound quality from the start.
Hi Mark,

in what way would that limit the SQ?

I only ask because it was the way I was going to go when I abandoned OB with the disappointing Beyma AMT. I was going to to feed 2 tpa3116 with an LR active per channel. This was what I intended with the resurrected Pass configuration also....

apologies for diverting Ray.

edit:
There was a complication with my method which involved a delay circuit for the TPA class Ds so that the mute could be applied at switch on to avoid the thump. It's more complicated because the same MO must be used at switch off as well.
There's nowhere you can be that isn't where you're meant to be

User avatar
Ray P
Social outcast
Posts: 3406
Joined: Thu Nov 22, 2007 5:18 pm
Location: Somerset

Re: SLOB project

Post by Ray P » Sat Dec 21, 2019 3:10 pm

ed wrote:
Sat Dec 21, 2019 3:01 pm
apologies for diverting Ray.
No need to apologise Ed, these are all learning opportunities and I'm happy to have started a thread that's getting people to pitch in.

User avatar
Ray P
Social outcast
Posts: 3406
Joined: Thu Nov 22, 2007 5:18 pm
Location: Somerset

Re: SLOB project

Post by Ray P » Sat Dec 21, 2019 5:30 pm

I've been working on some CAD drawings for the bass unit 'cradle' (i.e. the side panels, bass slot panels, etc) this afternoon.

I'm planning on the 'cradle' being machined from 18mm birch ply. I want the actual baffle to be removable so I'm on the lookout for some neat but discrete fixings to attach the baffle to the cradle - current thinking is to epoxy lenrgths of M6 studding into the ply and then use sleeve nuts through the baffle, something like these;

https://www.ebay.co.uk/itm/SLEEVE-NUTS- ... 1438.l2649

Any better ideas wil be welcomed.

User avatar
IslandPink
Oooo, Roberts Radio's sound nice
Posts: 7320
Joined: Tue May 29, 2007 7:01 pm
Location: Denbigh, N.Wales

Re: SLOB project

Post by IslandPink » Sat Dec 21, 2019 8:17 pm

ed wrote:
Sat Dec 21, 2019 3:01 pm

in what way would that limit the SQ?

I only ask because it was the way I was going to go when I abandoned OB with the disappointing Beyma AMT. I was going to to feed 2 tpa3116 with an LR active per channel. This was what I intended with the resurrected Pass configuration also....
Breaking the music up into little pieces then trying to put it back together correctly before it goes into the amps. Even at 24-bit. I haven't heard what I would call a serious system (SQ -wise ) where that was done in the mids or treble - but I have heard it work quite well below about 300Hz.
I haven't heard a Najda or other expensive solutions, though.
"The bass is the king of the instruments - it has no known natural predator" (Wobble)

User avatar
ed
retired
Posts: 4034
Joined: Thu Jun 21, 2007 4:01 pm
Location: yorkshire
Contact:

Re: SLOB project

Post by ed » Sun Dec 22, 2019 9:59 am

IslandPink wrote:
Sat Dec 21, 2019 8:17 pm
ed wrote:
Sat Dec 21, 2019 3:01 pm

in what way would that limit the SQ?

I only ask because it was the way I was going to go when I abandoned OB with the disappointing Beyma AMT. I was going to to feed 2 tpa3116 with an LR active per channel. This was what I intended with the resurrected Pass configuration also....
Breaking the music up into little pieces then trying to put it back together correctly before it goes into the amps. Even at 24-bit. I haven't heard what I would call a serious system (SQ -wise ) where that was done in the mids or treble - but I have heard it work quite well below about 300Hz.
I haven't heard a Najda or other expensive solutions, though.
ah, ok then, we may be at cross purposes here. In my scenario I don't see where I would be putting anything 'back together'. In effect Ray may be doing something different which I don't understand.
There's nowhere you can be that isn't where you're meant to be

User avatar
Ray P
Social outcast
Posts: 3406
Joined: Thu Nov 22, 2007 5:18 pm
Location: Somerset

Re: SLOB project

Post by Ray P » Sun Dec 22, 2019 3:24 pm

Mark, have I cottoned onto your suggested direction correctly; bandwidth limit the OTL amp via it's output cap so that the Fast-10S can run full range without any crossover? Then just integrate the bass unit to suit the roll-off of the Fast-10S - that sounds like it might just need a passive lo-pass filter, especially as I have the option of 'room correction' via HQPlayer convolution.

User avatar
Ali Tait
Social outcast
Posts: 3731
Joined: Fri Jun 08, 2007 8:10 pm
Location: Dunfermline

Re: SLOB project

Post by Ali Tait » Sun Dec 22, 2019 3:50 pm

My Bastanis run a 12” full range driver which is mechanically damped to cross over to an 18” woofer which is run with a Behringer with built in DSP, crossing at 200hz. Sounds very good to me.
Image[/img]

simon
Eternally single
Posts: 4172
Joined: Thu May 24, 2007 11:22 am
Location: People's Republic of South Yorkshire

Re: SLOB project

Post by simon » Sun Dec 22, 2019 4:06 pm

Me too! I also run my 208s directly connected to the GM70 amps with no crossover using the natural roll off of the 208s at around 100Hz. (I use a small cathode resistor bypass cap to roll the bass off just under 100Hz to keep the lower frequencies out of the OPTs but that's not an issue for you!)

Going from a (series) crossover to directly connecting the amps made as big an improvement as separate HT PS for the driver valves for me - very worthwhile.

User avatar
IslandPink
Oooo, Roberts Radio's sound nice
Posts: 7320
Joined: Tue May 29, 2007 7:01 pm
Location: Denbigh, N.Wales

Re: SLOB project

Post by IslandPink » Sun Dec 22, 2019 4:08 pm

Ray P wrote:
Sun Dec 22, 2019 3:24 pm
Mark, have I cottoned onto your suggested direction correctly; bandwidth limit the OTL amp via it's output cap so that the Fast-10S can run full range without any crossover? Then just integrate the bass unit to suit the roll-off of the Fast-10S - that sounds like it might just need a passive lo-pass filter, especially as I have the option of 'room correction' via HQPlayer convolution.
Yes.
You could do the bass LP passively, but getting it right could be expensive given you might need a few sizeable chokes. Having said that, you can probably use cored-chokes ( with laminations ) rather than big air-cores like the Quasar, since you may be crossing a bit lower.
One advantage of the slot-loaded arrangement is that the typical rise in the treble you get from a lot of bass drivers, which can require a 2nd-order low pass , is cancelled or subdued. It comes either from the cone edges or the dustcap, but either way, at frequencies over 500Hz, the slot chamber will act a bit like a mixing pot and cancel a lot of the HF.
But for getting the values, if you can get a DSP it will speed things up.
"The bass is the king of the instruments - it has no known natural predator" (Wobble)

Wolfgang
User
Posts: 165
Joined: Mon May 07, 2018 3:08 pm

Re: SLOB project

Post by Wolfgang » Sun Dec 22, 2019 4:29 pm

bandwidth limit the OTL amp via it's output cap so that the Fast-10S can run full range without any crossover? Then just integrate the bass unit to suit the roll-off of the Fast-10S - that sounds like it might just need a passive lo-pass filter, especially as I have the option of 'room correction' via HQPlayer convolution.
This will most likely not sound very good. I might be wrong as the individual room influence can accidentally correct things - soemtimes in a dramatic and positive way (or the opposite!) - and I haven't seen any measurements of Ray's listening room. But from experimenting with OB and active/passive x-overs for a while, including acoustic room treatment and DSP/EQ, I woudl say that it needs a steeper filter and also the cross over point needs to be precisely defined. The size of the baffle and to some degree its shape define where one has to set the cross over point (was it around 170Hz for this design?) . An amp with smaller output cap and a standard/typical shelving filter cannot deliver this precision. One needs also to consider the limits of the FR driver without any damping at higher SPLs which means that the x-over point needs to be reasonably high (or low).

I used less capacity for the output caps in my 6C33OTL because I use an active x-over at 70Hz in my sub. I measured the frequency response curve of the amp in order to see how it would all fit together considering my room measurements (really big room with one major peak at 42Hz/smaller at 84 and 168Hz in my sweet spot). Starting at around 100Hz with the peak at 84Hz it goes relatively flat to the cross over point at 70Hz where a 48dB filter is kicking in for the BLHs which I set before the DAC with the miniSHARC module (I2S connections).This way I need less DSP/EQ for the sweet spot and don't have to counteract the room influence with heavy filtering. Always a good idea.Every room has one or more frequencies(peaks) which cannot be corrected by any means except moving the listening position and one has to choose a position for the speakers and the sweet spot where these pos/neg peaks have the least influence on SQ.

The bass and the FR driver need phase adjustment (delay) in order to work seamlessly and sound "as one" in the sweet spot. No way around it. I don't know if all of this can be set in HQplayer? In my understanding from reading the manual it uses FIR filter settings from measurements with other programs and cannot be adjusted using a GUI with controls. But I am no HQplayer expert and there might be ways to work around this.

User avatar
Ali Tait
Social outcast
Posts: 3731
Joined: Fri Jun 08, 2007 8:10 pm
Location: Dunfermline

Re: SLOB project

Post by Ali Tait » Sun Dec 22, 2019 5:59 pm

All that is fine if you have a separate listening room, most of us don't have that luxury. You have to make the speakers work in a domestically acceptable position. Below is a measurement taken at Mark's of my Bastanis. A 12" driver is run full range and is mechanically doped to cross to a compression tweeter with a series cap. Original design used 1uf, this did not measure well so was upped to 2.5uf which gave a much better result. An 18" woofer is run actively with a Behringer INuke amp, crossing at 20hz. Final result-


Image
Image[/img]

chris661
Shed dweller
Posts: 2294
Joined: Mon Feb 22, 2010 7:29 am
Location: Sheffield

Re: SLOB project

Post by chris661 » Sun Dec 22, 2019 6:21 pm

Ali,
Is there any chance of a more zoomed-in graph? 20dB per division means quite a lot of the detail can be obscured.
Are you sure the crossover frequency was 20Hz? Sounds very low.

Cheers,
Chris

User avatar
Ray P
Social outcast
Posts: 3406
Joined: Thu Nov 22, 2007 5:18 pm
Location: Somerset

Re: SLOB project

Post by Ray P » Sun Dec 22, 2019 7:12 pm

Simple bit first; no, there s no GUI to play around with DSP settings in HQPlayer; you need to use other software to create a file with the required settings that HQPlayer then uses during playback. Of course the 'other' software packages will have GUI interfaces so while you can't adjust things on the fly with HQP I imagine that it's not too painful, if a little time consuming, to iterate through different settings to find the sweet spot. I also have a friend who is doing a degree course in sound and video engineering and he has offered to help with the process with some professional quality equipment.

My primary goal is personal sound quality satisfaction - if it measures well too so be it - and for that reason I would prefer to avoid using MiniDSP/Minisharc for anything other than experimentation/discovery. I've put effort into retaining the integrity of the data rate families (44.1KHz/48KHz) in my system so I don't want to have to use the sample rate converter to get everything into 48 or 96KHz PCM as required by MiniDSP (or worse to use ADC/DAC processing). Anyway, my replay is focussed on DSD in preference to PCM.

My original thinking was to use the Nelson Pass FET-based active crossover so that remains as my fall back position (after all the SQ of the Nelson Pass SLOB system was very well regarded and used a very similar active crossover) but I'm keen to explore whether this can be made to work passively and/or to remove any processing steps that can be achieved without detriment to my listening experience.

It's turning out to be an interesting journey.

User avatar
IslandPink
Oooo, Roberts Radio's sound nice
Posts: 7320
Joined: Tue May 29, 2007 7:01 pm
Location: Denbigh, N.Wales

Re: SLOB project

Post by IslandPink » Sun Dec 22, 2019 7:58 pm

chris661 wrote:
Sun Dec 22, 2019 6:21 pm
Ali,
Is there any chance of a more zoomed-in graph? 20dB per division means quite a lot of the detail can be obscured.
Are you sure the crossover frequency was 20Hz? Sounds very low.

Cheers,
Chris
I think he meant 200Hz. The plot is what I captured on the day ( a few off ) . It might be possible to open up the files again and do some better plots. I should have posted a few of these , to show where we started and where we got to. It was quite successful.
I will have a look after Christmas and post more on my thread. Ping me a reminder on my thread if I don't, by New Year.
"The bass is the king of the instruments - it has no known natural predator" (Wobble)

Post Reply