advertisement


Jim Rogers JR149 DIY build

Your choice is whether to make a replica or just use the lightweight cylindrical sealed cabinet idea with a modern and better driver pair.
The new Mission 770 is an example, measuring very well with better drivers than the original, but clearly let down by the front port output.
 
Your choice is whether to make a replica or just use the lightweight cylindrical sealed cabinet idea with a modern and better driver pair.

Good question... Today my prefrence is to use a modern and better driver pair. But still not really made the final decision yet.
 
I believe the 'modern version' would probably be composite rather than aluminium/lossy deadener, but following the same design mantra of a very light but stiff and well damped cabinet structure.

Modern natural composite sandwich structure would be perfect for this sort of thing.

If you want yo move away from the original design, but not concept.
 
I am leaning towards the Monacor SPH-145HQ / DT-19SU combination. Apart from anything else there is a crossover design avaviable. But it's this I really need some help with - some elements of it I understand and others I don't. I should say at this point I was, many years ago, an electronics engineer dealing with sophisticated control systems. The crossover looks like this.
Monacor_SPH-145HQ_Crossover by Robin Martin, on Flickr

For the DT-19 tweeter circuit:
L3, C3 and C4 form a 3rd order filter.
R3 and R4 compensate for the DT-19 being more efficient than the SPH-145
C5, L4 and R5 are some sort of compensation network? Zobel?

For the Bas/midrange SPG-145 circuit
L2 and C2 form a 2nd order filter. But what is R2 for?
And C1, L1 and R1 I don't understand at all.

Any insights the forum can provide would be much appreciated as I like to understand what I am doing and, I suspect, some of the components are for fine tuning that my ears wouldn't even notice... I like the mantra "keep it simple" and I don't want to incorporate anything that is unnecessary!

Cheers all
 
C1 in parallel with L1 form a notch filter at 1.6kHz [ equation is 1/(2pi x root(L1xC1)) ] and the 6R8 in parallel limits both the impedance rise in the notch zone and the extent (Q) of the notch in combination with the values L1 and C1. This is to flatten a peak in the bass/mid output centred at that frequency.
'C5, L4 and R5 are some sort of compensation network?' Correct. Possibly tweeter break up as it is centred around 17.6kHz, but it also mitigates against rising impedance in the inductance of the voice coil. The resistor limits the drop in impedance which would otherwise look like a short circuit at that frequency.
C2 and R2 is a Zobel network to flatten the rising impedance of the bass/mid voice coil.
Hopefully someone will be along with a fuller explanation.
 
(I think it's the other way round) C1/L1 fill in the dip at 1.6KHz - see Mr Marley's link to the measurements earlier in this thread.

C2 and R2 are part of the 2nd order filter on the bass/mid.


Some more pointers in the review:
"crossover

In view of the monitor requirements, the crossover is elaborately designed. Because of the excellent properties of the chassis, apart from the usual filtering and level adjustment of the tweeter, linearization only has to be done in two places, one of which is even optional. With the HQ, a second-order low-pass ensures a clean drop, and an additional blocking circuit reliably levels out the baffle step. The tweeter was given a third-order high-pass filter and a voltage divider for the optimum level. The second series capacitor is unusually large at 22 µF, but still has the absolutely correct value in that it brings the edge to the woofer into the correct shape and maintains the right phase position. As an option, the linearization of the tweeter by 17 kHz can be viewed using an absorption circuit. It expresses itself at most in a minimally fresher and freer high tone. If you are looking for a monitor that is perfectly linear up to more than 30 kHz, you can safely take this measure, because it will not do any harm. The sum of the branches results in a very linear curve with an absolutely minimal increase in the treble and an acoustic separation of just under 2 kHz."


The 17KHz notch filter seems to be optional.

I'm tempted by this project myself.
 
Where about are you located? I have an empty pair of the aluminum section, the one had been dropped and damaged all the wood, bent the bottom metal part that covers the crossovers and bent the metal rod inside. But the aluminum section seem ok.
 
Inductor/capacitor in parallel undergo a rise in impedance at resonance. In series they undergo a minimum impedance at resonance. Therefore if you put them (L and C) in parallel but in series with a driver, the rise in impedance surely forms a notch in the driver output?
 
(I think it's the other way round) C1/L1 fill in the dip at 1.6KHz - see Mr Marley's link to the measurements earlier in this thread.

C2 and R2 are part of the 2nd order filter on the bass/mid.


Some more pointers in the review:
"crossover

In view of the monitor requirements, the crossover is elaborately designed. Because of the excellent properties of the chassis, apart from the usual filtering and level adjustment of the tweeter, linearization only has to be done in two places, one of which is even optional. With the HQ, a second-order low-pass ensures a clean drop, and an additional blocking circuit reliably levels out the baffle step. The tweeter was given a third-order high-pass filter and a voltage divider for the optimum level. The second series capacitor is unusually large at 22 µF, but still has the absolutely correct value in that it brings the edge to the woofer into the correct shape and maintains the right phase position. As an option, the linearization of the tweeter by 17 kHz can be viewed using an absorption circuit. It expresses itself at most in a minimally fresher and freer high tone. If you are looking for a monitor that is perfectly linear up to more than 30 kHz, you can safely take this measure, because it will not do any harm. The sum of the branches results in a very linear curve with an absolutely minimal increase in the treble and an acoustic separation of just under 2 kHz."


The 17KHz notch filter seems to be optional.

I'm tempted by this project myself.
Thanks for that S-Man. Where did you find the review may I ask? I haven't been able to find it although that's probably due to the fact I guess the original was in German.
 
Thanks for that S-Man. Where did you find the review may I ask? I haven't been able to find it although that's probably due to the fact I guess the original was in German.
I did a Google translate on the German review.

Where did you find the crossover info?
 
To be honest I don't remember where the crossover circuit came from. It was either from one of the several links that have been added to this thread or I may have Googled it. Anyway I made sure I took a copy of the image....
 
Before I put my hand deep in my pocket to buy the drive units and crossover components I thought I had better check that I can actually make the metalwork for the cabinet. It's a bit rough and ready but it'll do as a proof of concept. The actual ones I will make that will end up in my lounge will be a bit more refined. The material is 1.5mm aluminium which was my best guess at what the original was made from.
prototype jr149 by Robin Martin, on Flickr
I think I can now confidently order the drivers :)

PS It doesn't look very symetrical in the pictrue but it is!
PPS The wooden structure on the right is mould arround which I bent the aly.
 
Superb job, looks like really clean bending. You have the potential to make something really special here!
 
Perhaps the Dolce?

https://wilmslowaudio.co.uk/stand-m...lce-dolce-kit-c/w-plain-mdf-flat-pack-cabinet

That Monacor SPH145HQ looks a good match with a ScanSpeak D2606-92200 but you'll have to design a crossover for it. A crossover made for the B110 and T27 will not work.

The ScanSpeak D2606-92200 tweeter is a better version of the Vifa D27TG-35-06.

PS Don't forget that the crossover may be costing as much as the drivers going by how expensive the LS3/5A crossover is. And if you need to voice it then you'll be spending even more money. Jeff Bagby of the Continuum II was only succesfull on the third build.

FWIW I was looking at the L5/19 at one time and Klang & Ton 2012/03 has a complete description of it. On ePray you can buy from Hong Kong the PCB for the 11 Ohm LS3/5A and it will fit the crossover perfectly.
 
Last edited:
With VituixCAD2 to model the crossover, REW to measure the drivers "raw" in the cabinet and a measurement microphone, its a lot easier these days to design a crossover
 
Before I put my hand deep in my pocket to buy the drive units and crossover components I thought I had better check that I can actually make the metalwork for the cabinet. It's a bit rough and ready but it'll do as a proof of concept. The actual ones I will make that will end up in my lounge will be a bit more refined. The material is 1.5mm aluminium which was my best guess at what the original was made from.
prototype jr149 by Robin Martin, on Flickr
I think I can now confidently order the drivers :)

PS It doesn't look very symetrical in the pictrue but it is!
PPS The wooden structure on the right is mould arround which I bent the aly.

Fantastic. Well done.
 
With VituixCAD2 to model the crossover, REW to measure the drivers "raw" in the cabinet and a measurement microphone, its a lot easier these days to design a crossover

I've just started playing with VituixCAD. It's a powerful tool, and not the easiest to learn (if you want to get the best out of it), but it's supposed to be capable of very accurate results.
 
With VituixCAD2 to model the crossover, REW to measure the drivers "raw" in the cabinet and a measurement microphone, its a lot easier these days to design a crossover
I gave up on Vituix. I suppose it might be good if you still have to decide on the box etc. but my mind had difficulty coping.

I'm using Holmimpulse to measure and Xsim to model. Found these easy to learn. Both are a lot easier to use than Vituix and/or REW and more than suffcient.
 


advertisement


Back
Top