Martin logan CLX mods

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Member
Joined 2005
Paid Member
The Martin Logan CLX (and the rest of their ESL line) can benefit from some selective upgrading.
Doing these mods will likely void your warranty, possibly reduce the resale value, and you are working on equipment which has AC input and high voltage supplies, so unless you know what you are doing and the speakers have been de-engerigized (AC and audio cables disconnected for a few days) then leave them alone.

Now that you've been advised and I've scared you all off here goes.

1. Remove the AC power cord from the speaker and leave them sit upowered for a couple of days to make sure they are dischaged.

2. Cover the FULL area of the speaker and cells on all sides with thick cardboard to avoid any damage while you work on them. tape up all edges. This is not negociable ! - a solder splash or drilling debris can damage your cells or finish!

Your preference and choice of wire, capacitors, and connectors may well differ as will which mods are of interest or fit your budget or skill set.

1. Speaker connectors - Martin Logan had a speaker connector made which looks sustanstial - I've had 2 break. I replaced them with WBT 730.12 platinum using the original Martin Logan plastic isolators. This requires using a drill with a 90 degree chuck and drilling out PART of the existing hole. Looks very good.

2. AC input connector - replace with an AC line filter -I used a Tyco C6EEJ1 or similar - you'll need to open the hole a little to get it to fit. I also like to use a 3.3 uf X2 rated cap such as Vishay BFC233820335 after the filter and put the whole assembly in a giant piece of UL approved heatshrink. The filter cap must be an X2 rated cap - no audiophile cap will do and would be unsafe!! This assembly must only use UL approved components and be installed in a safe and secure manner. Check with an ohmmeter for high impedance to ground for each of the Line to Line or line to Neutral wires after you finish and BEFORE connecting.

3. Wire. Martin logan used a UL approved TEW custom made copper cable for audio wire. I replaced it with DH labs Q 10 Teflon insulated silver plated OFC copper .

4. There are 4 Solen caps in the crossover board.
This is the biggest area for improvement. The planar resistors on the crossover boards are EASY TO DAMAGE and you will have a hard time to replace them without damaging the board so do this work on a old thick towel very carefully!!!! The original caps cost about $30 a speaker and the new ones will run about $1,000 total so this isn't inexpensive, but I don't know how to achive the same sonic improvement (and physically fits) any other way. I used Plumbers goop to hold the caps - much better than silicon, but don't make mistakes - it's also extremely much harder to remove once it sets.

The crossover caps are 2 x 22uf, 1 x 33 uf, 1 x 1.5 uf

22uf - replaced with a parrallel sandwich of 2 x 9 uf 100V MIT PPFXS , a 3 uf /100v MIT PPFXS, a 0.33uf /600v MIT RTX and a 01./600V MIT RTX. ( this totals 21 uf, but the MITs I had are a little over value and i selected mine to add up to the 22uf) The 2 x 9 uf and 0.01 uf caps fit where the Solens did and by using both PC boards you'll find you can put the 3 uf and the 0.33 uf with wire leads on the LF board without blocking the air flow to the planer resistors and when the 2 boards are joined the leads soldered to join the caps up and make the required value.

Note - The 9uf MIT PPFXS cap is the largest cap with the same width of the Solen caps which is why it was chosen - becuase it fits, takes the least space, and sounds good.

Don't block the airflow to the planer resistors.
I measured a 4 C rise with my mods on the resistor bodies - you really don't want the speakers on a carpeted floor blocking off the air flow intakes - make small granite platforms for them.

The 1.5 uf cap - use PPFXS 1.5 uf.

The 33 uf cap - replaced with a parrallel sandwich of 3 x 9 uf 100V MIT PPFXS , a 5 uf /100v MIT PPFXS, a 0.33uf /600v MIT RTX and a 01./600V MIT RTX. (this totals 32.34 uf, but the MITs I had are a little over value and I selected mine to add up to the 33uf) The 0.33 uf and 0.01 uf caps fit where the 33 uf Solen did . The rest of the PPFXS caps have to assembeled together off board and the leads brought back to the crossover board.

There are 2 ways to do this - I got rid of the components on the board which joins the cells to the transformer and mounted the 33uf's PPFXS caps there and hard wired the cells to the transformers - not as easy to service, but i prefer direct hard wired connections so I would have done it anyway and it left me with a place to mount the 33uf's PPFXS caps - otherwise you can keep the transformer to cell interface board stock and you'll have to mount the 33uf's PPFXS caps on a PC board and heatshrink the whole assembly and install it where it doesn't interfer - it must be insulated so it doesn't short against anything. (near the speaker input jacks there is some space).

5. The Fet turn on board (the lower PC board with heat sinks) has 4 of 0.33 uf/ 250 volt caps on it bypassing the IRF 640 Power fets - Replace with MIT PPFXS 0.33 uf.

6. The Fet turn on board also has 4 of 100uf/35v 85C Jamicon electrolytics - I prefer Panasonic FM or Nichicon HE series caps - both 105 C for long term reliability. Warning - this is a double sided board and unless you are very good at soldering , and have a good hot iron, solder wick, and lots of experience don't do it.
This is no place for an audio rated electrolytic cap like Silmic or Blackgate (which i like ) - it just does electronic housekeeping and it doesn't pass the audio signal that you hear.

7. The high voltge power supply board ( with 2 small black transformers on it) has about 10 of the 85C Jamicon electrolytics - again I prefer Panasonic FM or Nichicon HE both 105 C for long term reliability. Warning - this is a double sided board and unless you are very good at soldering , and have a good hot iron, solder wick, and lots of experience don't do it.

I also removed the LM 358's and socketed them - again years of soldering experience on boards is needed , a good hot soldering iron with solder wick is required - it is a double sided board and easy to lift tracks.

Next part ONLY FOR SKILLED MODDERS

I got rid of the transformers to cell interface board connections and the Molex connector between the cross over board . This makes it harder to service - you need to remove the crossover board, the inter-connect board, and the fet turn on board all together to service them, I did it several times durirng the mods as the updates progressed and it wasn't too hard. But some may not go this far ...


The end result - after a 75 hour burn-in - a much smoother, far more resolving sound, and superb detail. The structure of notes has higher intial impact and greater harmonic detail to the decay. Dynamics are improved as well. I couldn't go back to the original version.

pictures to follow .............................
 
Last edited:
Very interesting, and well explained.

My approach to crossovers is to just ditch the passive and use a top-of-the-line active solution.

Much simpler to implement, but it does lead to the question: How easy is it to bypass the CLX crossover elements?

From my prior ML experience doing that, it usually is pretty straightforward, but I've never peeked inside a CLX.
 
Member
Joined 2005
Paid Member
The crossover is all contained in 2 pc boards held together with long standoffs - One board is the low pass filter and the other the high pass circuit. I'll try and get the pictures posted soon along with some additional crossover details. The whole assembly could easily be bypassed. However there does seem to be a notch filter (the 1.5 uf cap is in series with a small inductor) on the high pass section I'll try and sketch them out this weekend.

The woofer filter is 6 dh per octave with an inductor and some series resistors which are switchable to adjust level by 1 and 2 db (although the switchs aren't mentioned in the manual)

The high pass filter is a multipole with capacitors, inductors, and planar resistors. While it was possible to upgrade the capacitors from metalized film to film and foil types the inductors have no values ( although I may have an inductance meter around here). I think it would be tough to replace the small guage air core inductors with a foil type - there's just not much space left to do it with-in the confines of the crossover assembly. The resistors are Ohmite 25 watt planars.

The positive and negative low and high pass outputs from the crossover go to a board with 2 sets of insulated back to back heatsinked IRF 640 Fets with the 0.33 uf bypass caps which don't allow the audio signal onto the transformers unless it senses the high voltage power supply are on - and the those supplies don't come on unless they sense an audio signal so the there is a small ribbon cable passing the audio sense and power supply signals back and forth.
 
Last edited:
The positive and negative low and high pass outputs from the crossover go to a board with 2 sets of insulated back to back heatsinked IRF 640 Fets with the 0.33 uf bypass caps which don't allow the audio signal onto the transformers unless it senses the high voltage power supply are on - and the those supplies don't come on unless they sense an audio signal so the there is a small ribbon cable passing the audio sense and power supply signals back and forth.

Are you saying that the high level audio signal from the power amplifier has to pass thru MOSFET switches before reaching the step-up transformers?
 
Hi all, (steve!),

I have just found this thread two years after it was posted by ticknpop, great stuff. As Bolserst knows i have been tweaking my ML Spires, to great effect. This post has a couple of things i can try on the boards on the spire, mainly the step up boards. Would you say that removing the input/output connectors and soldering direct gives a worthwhile sonic benefit? If so i will do this, although i understand the slight convenience issue, my only aim is better sonics. Would you also at the same time consider locating the transformer in an external shielded case, extending the wires back to the board?

Just a brief rundown of my mods so far:
Removed all electronics and xover to custom cabinet, re-wired using ptfe/silver plated copper 15awg between boards and panel, using a dielectric bias of 90v. Bass cabinet clad in granite/ceaserstone to improve panel to floor coupling, and reduce panel resonances. Spire spikes removed and replaced with Cerabase feet. Rebuild of the passive crossover with Clarity MR's, Mundorf foil inductors and Duelund Cast resistors.

So i'm looking at other area's that could be improved, so any suggestions are welcomed!
Regards,
Paul.
 
Member
Joined 2005
Paid Member
You have done a very good rebuild on the Spires with excellent parts. I hear very good things about the Dueland resistors and am looking to see if they will fit mechanically in the CLX crossover - the dual CLX crossover board assembly is too complex to mount externally and the bass driver is external on the CLX so there isn't the mechanical isolation to gain that you achieved with the Spire rebuild of Martin Logans that use crossovers inside the bass enclosure. I can't say if removing connectors and hard wiring the crossover cables will make mechanical sense for your rebuild, electrically(sonically) it's preffered, but I expect you will have a bigger improvement from your better crossover parts. I hope you matched the coil resistances as well as the inductance when you replaced the inductors as replacing them with the lowest resistance coils without matching the original resistance will affect the crossover negatively.

I'm not sure why you would remove the transformer - if it's an AC power transformer a shield isolating it from audio cables might make sense - careful not to interfer with transformer cooling. i shortened and shielded the AC input cable. I don't think shielding the audiostep up transformer will do much ( unless it's near the AC power transformer again be careful to allow for cooling). I have internal pictures of the updated CLX circuit boards after they were reassembled in the CLX base. I wish I had taken pictures of each of the 5 circuit boards before and after rebuilding before I reassembled them as it would help with the mod explanation.
 
Last edited:
So you say


Yes both the positive and negative audio signal for each of the bass and midrange / treble leads each goes through a heatsinked IRF 640 - there's a 0.33 film cap bypassing the fet.

This is the limiter....one for the bass one for mid-hi....the cap is new...but would work fine...an yes the limiter dose have a sound of it own...an now has the sound of a cap.
Be sure to get the best sounding 0.33 cap to you ,each diff cap well sound diff here.
 
I hope you matched the coil resistances as well as the inductance when you replaced the inductors as replacing them with the lowest resistance coils without matching the original resistance will affect the crossover negatively.

Hi & thanks for the reply.. Yes, i did consider matching the dcr as well as i could, the original ww coils were 18 & 20awg, i used 16, and notice no negative balance issues at all, they have pushed the Spire very high, an apt name now!

Re the SUT, i will most likely leave where i currently have it, and can re visit this in the future if i feel the need.

Enjoy your CLX, Some pics would be great to see what you have accomplished.

Happy New Year too!

Regards,
Paul.
 
Member
Joined 2005
Paid Member
I'll also document the crossover design. The bass panel crossover is 6 dB/oct with a choke with switch selectable series resistors to lower the bass panel output. The mid/high panel crossover is multipole ( maybe 18 or 24 db /oct ) with a notch filter in the top end - a 1.5 uf in series with a choke both in parrallel with the output to the HF torroidal step up transformer
 
Hoping someone can help me out ---- I had a flood in June and it submerged my ML CLX speakers in 3 ft of water -- the panels & cabinets seemed fine and I had all the boards cleaned up etc.. and after putting everything back together -- the bass panel is not getting any sound on both speakers -- curved panel is fine & all the wiring seems fine. Can someone point me to the instructions for removing the panels so I can check the connection from the cross over board to the panel to see if the connection might be the issue -- I don't see an easy way to remove the panels

thanks

Bart
 
Member
Joined 2005
Paid Member
Yes I'm still around, my friend bought his V caps and Dueland resistors , but hasn't opened his CLX yet. I changed subs to JL F 212.

Bartgilchrist - I spoke to you before , we are in the same town, yes I know how to disassemble them, it is unlikely to be just your connections , without seeing them I believe they could be saved, but you'll need clean and/or to rebuild the 5 PC boards.
 
Member
Joined 2005
Paid Member
It is more likely the power supply board for the LF section that needs repair/replacement. You must change all the electrolytics in the 5 boards or keep having failures every time one of the water damaged electrolytics fail. Martin Logan did not use the best electrolytics, I changed mine to 8,000 hr 105 C Panasonics rather than the cheap 85 C Jamicon's that Martin Logan used . probably $40 worth of parts...
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.