Skip Navigation.
Section 0

ChromaTalk Archives: December 2010

Re: Any news about the Chroma Enabler?

Go to first message in thread, November 2010

Randel Osborne [21030467]


Thanks for the questions about the Enabler. I have more photos and a description at:

Although I have shot some video, I do not yet have it available on the web.

I've taken deposits for 2 more Enablers, and am currently building them - I hope to have them finished in December. They are extremely labor intensive to manufacture, and take me several months to complete.

I'll post to the list when I put some video up, or have an update.

Jeffrey D. McEachin [21030073+]

Wow - that's pretty amazing! Nice job!

Would you mind sharing where you got the red-on-black LCD that's compatible w/ CC+?

David Gowin [21030611]

Just had to say, the Enabler is a beautiful work of art! Reminds me of the 2600 for some reason. Now I just had to get a CC and some cash together!

RJ Krohn [21030455]

ok, im only asking cause of this quote: "...get some cash together".

is the chroma enabler something that would potentially be for sale, or as a DIY project?


Looking very forward to teaming a Enable up with my CC+ and AT Chroma and CC+ (one day soon) Chroma Expander along with Matrix's IPAD editor. Should be quite the ultimate Chroma rig.

Also, Randel has been great to deal with in every way.

Merry Xmas/Seasons greetings

Odysséas Tsakas-Grigoriadis [21010242]

Yes, If you follow the link provided by Randel, you'll see at the bottom of the page that the Enabler is now for sale. As for the price, we should definitely "...get some cash together". Very nice work by the way, and an update with a demo video will be most definitely welcome!

Randel Osborne [21030467]

Hello, and thanks for all of the kind words about the Enabler project.

At your request, I've uploaded a short video of it in action here:

The MIDI output of the Enabler is feeding the CC+ input, and I'm adjusting parameters while I am playing, and while the Chroma is playing an arpeggiation.

Arun Majumdar [21030348]

Enabler Video (at last!)

Hey Randel,

If you ever do decide to offer plans/Kits/parts or anything to replicate your really absolutely superb and fantastic work, I am all ears and ready to dish out the $ needed to get one ... I have two Chroma's and the pressure kits and the CC+ etc... so this would make the studio setup much, much more usable :)

Please do let us know if it does become available!

Jeffrey Koepper [21030671]

Re: Enabler Video (at last!)

HI nice work, is there a way to have the enabler connected to the midi in of the chroma and also a midi sequencer connected to drive the chroma as well? The chroma only has one midi input.

Doug Terrebonne [21030114]

Re: Enabler Video (at last!)

You could just use a MIDI merge box but it would be nice if it had that capability built-in. While the Enabler looks wonderful I just don't know if it's $2500 better than my BCR2000... :)

Jeffrey D. McEachin [21030073+]

Re: Enabler Video (at last!)

Not only a nice demo of how the Enabler works, but also of how "Arpish" the Chroma sounds. Especially towards the end. ;)


Re: Enabler Video (at last!)

I have the BCR setup and can't wait for the Enabler : )


Re: Enabler Video (at last!)

On Dec 4, 2010, at 12:35 AM, Vco3 wrote:

HI nice work, is there a way to have the enabler connected to the midi in of the chroma and also a midi sequencer connected to drive the chroma as well? The chroma only has one midi input.

These are about $50 us

Matrix [21030220]

Re: Enabler Video (at last!)

I just bought the 4in from zzounds for $99 so I could use my iPad with my MIDI keyboard to control my racks. It works great.

Jesper Ödemark [21010135]

Re: Enabler Video (at last!)

VERY impressive!

I really need to fix that new front panel for my Doepfer Drehbank...

Jesper Ödemark [21010135]

Re: Enabler Video (at last!)

Doug Terrebonne skrev 2010-12-04 07:26:

You could just use a MIDI merge box but it would be nice if it had that capability built-in. While the Enabler looks wonderful I just don't know if it's $2500 better than my BCR2000... :)

Screw some cherry wood to the side of the BCR2000 and you're almost there. ;)

James Coplin [16330036+]

Re: Enabler Video (at last!)

There is also the free template for the Mackie C4 Pro as a cheaper option.

Arun Majumdar [21030348]

Enabler Video (at last!) / Okay - Found link!!!

Hi Randel,

I sent you a private note per your web link but since I sent a note on the public (dummy that I am sometimes) not knowing that you in fact do offer it, I'm updating the public site --- I've been out of touch for a bit but your video looks really great and I went to your site to see more pictures - they look great!!! Having the Prophet-T8 and the CS-80 and Chroma, I'm super excited about the Enabler since it provides that hands-on that is missing. Looking forward.

Jeffrey D. McEachin [21030073+]

At 07:21 PM 12/2/2010 -0500, Randel Osborne wrote:

The LCD is just the display, however. What makes it "compatible" with the CC+ are two microprocessor-based modules that turn the knob rotations into MIDI, and use the MIDI data to push the appropriate text messages to the LCD. Designing, manufacturing, and writing the firmware for those modules was a significant hurdle in the development of the Enabler.

Ah. Thank you for the clarification. Since the Enabler is out of my price range, it looks like I will have to proceed with my plan to frankenstein a serial LCD control board onto a red-on-black LCD. Really wish I could afford the Enabler, though.

Items on eBay

Chris Ryan [21030691]

Modulation Controller and Transformer: Item #270675118781, starting bid US$24.00, ends December 7. "Left hand modulation controller, matching transformer assembly and right hand blank panel."

Volume Pedal: Item #270675120022, starting bid US$25.00, ends December 7. "Looks unused. Has a 1/4" jack on the side and the cardboard sleeve is marked LSC-227-124." This is different than my pedal; perhaps it's an earlier model.

Pictures from the auctions:

David Gowin [21030611]

Hmmm, doesn't look like my pedal either. (And another Seller only shipping with the USA)

Odysséas Tsakas-Grigoriadis [21010242]

Sold for 24$ and 51$ respectively.

Also, anyone have seen this: [eBay item #5196a48dde]. Are these custom sounds you think? Compatible with the Chroma?

Picture from the auction:

Chris Ryan [21030691]

Looks like they're sounds for the Polaris.

Re: Expensive JL Cooper MIDI Interface

Go to first message in thread, November 2010

Chris Ryan [21030691]

On 2010-11-25, at 4:01 PM, David Clarke wrote:

It would be interesting to get a copy of the listed 'accessory order form' for the site, but I can't see that being that much of an extra draw such that someone would want to pay over $120 US for a JL Cooper MIDI interface:

This package sold for USD$152.50. Is the buyer on the list?

New album - Chroma related

Christian Kleine [21030210]

Hi list,

Please pardon the self promotion. My new album is out and it's full of analog synths/drum machines/spring reverbs... E.g. the piece "human response" is made almost only with a Rhodes Chroma.

Other synths used on this album: GRP A6, Xpander, Synthex, AKS and a few others.

For the opening track, there's a small video:

The album is self-released as mp3/flac for the time being and available here:

Illusion by Christian Kleine

All tracks can be prelistened on this page.


Filter Chip CEM3350 ???

Peter-Jan Kleevens [21010014]

Hi all,

my Chroma (sn 21010014) has been out of my studio for a long time now due to an extensive and complete overhaul at the Service Connection in Holland. Now the end of this long repair session is coming in sight but there is still a problem with getting a very essential part. It is part of the VCF filter on one of the voiceboards: a Curtis chip type CEM3350. Apparently this chip is very hard to get according to the people at the repairshop. Any advice would be of great help. Thanks and best regards.

Heinz Weierhorst [21010276]

Hi Peter-Jan.

There is one source I know: Touched by Sound. They sell a used CEM3350 and will see ? 99.00 for it !!!!! So it's your decision.

Voice Board and Pedal Unit on ebay

Michael Zacherl [21030253]

just in case you didn't come across that:

[eBay items #300503565966 and #300503568864]

Pictures from the auctions:

Jesper Ödemark [21010135]

voice board - new Ebay record?

201 euros for a board that isn't guaranteed to be working. The prices are going u-u-u-up!

Paul DeRocco [21030230]

Re: voice board - new Ebay record?

Doesn't sound so outlandish to me, given what the Curtis chips fetch by themselves.

Jesper Ödemark [21010135]

Re: voice board - new Ebay record?

But in reality this could be a board with broken ICs. These used to be a lot cheaper when taken out of a Chroma but they're apparently going up, just like the Chroma.

Patrik Gudmundsäter [21010240]

Re: voice board - new Ebay record?

Thats Cheap :)

Help! my chroma has sufferered from PSU malfunction....

Daniel Benoit [21030109]

First a little info about my Chroma. It as been upgraded with a new switching power supply 2 years ago, a CC+ was also installed in the same period.

My last upgrade, the Poly pressure kit, arrived a few weeks ago. On the same day it arrived, I powered up the Chroma, but for a reason only explanable from one of Murphy's laws, the PSU went bad. (Switching noise coming out of the case and the leds, on the front panel, going on and off.) After removing and checking the PSU, I replaced it the day after.

The Chroma went back to life, but didn't see any "tunable' voices. It instantly showed 12345678 on the display.

At this time I suspected that there was a bad IC on the I/O board, instead of trying to find the bad one, I decided to replace all of them (except for the DA converters)

I unsoldered all the ICs, put some good IC sockets in place, found all the replacement ICs and powered up the chroma with still the same problem...

For now:

  • The Chroma will power up and show 12345678 on the display
  • All the switches are working, I can call any program or edit parameters,
  • All voltages are OK on all PSU outputs.
  • I can force a Set split 33, but no voice wil be working

The keyboard scan doesn't seem to work (I dont see any flicker on the alpha digit went playing on the keyboard) and the "set split" note selection doesn't work.

I have check the data buss with my scope, but can't figure out if there is something wrong. Could there be something wrong on the CC+ side?

I would need some help a this point, Any advice could would be appeciated

Luca Sasdelli [21010226]

Hi Daniel,

I can't find your name within my SPSU kit records. Which SPSU did you install? Maybe was it purchased with another name?

First of all, please check all voltages: -12V, +12V, +5V digital and +5V analog (this one should read exactly 5.05V).

The Err 12345678 could show a DAC misalignment. Anyway, just check voltages and then post readings on the mailing list.

Paul DeRocco [21030230]

Could also be on the channel motherboard.

There's really no substitute for looking at what's going on, rather than just replacing everything. Even a simple voltmeter can identify audio signals that have unexplained DC offsets on them, such as blown op-amps.

That said, in my own units I've had the most problems with electrolytic caps drying out and RCA CMOS chips (especially analog switches) blowing up.

Heinz Weierhorst [21010276]

Hi Daniel.

Is it right - you replaced all ICs on I/O board? Are you sure there are no "killed" vias ? So the best way is to make copies of the I/O board schematics and with an Ohm-meter or a beeper check every connection and mark it with a color pen at the schematics.

Daniel Benoit [21030109]


The PSU upgrade was built DIY, from the original PSU article, the +5 psu was from a 'industrial' (PowerOne) swithing power supply. I have been using this "brand" for a lot of projects whitout problems

Voltage are:

  • +5 analog = 5.05v
  • +5 digital =5.16v
  • +12 analog = +12.06v
  • -12 analog = -12.06v

Before deciding to swap all the ICs, I did a lot of "probing" using a scope and a volt meter, (for about about 2 days...) . I found that (Z55) was note giving the correct output, It was the first IC a replaced, then went on with all the 74ls138(I/O strobe decoder) and Z28, Z29, Z30(miscellaneous I/O). then every thing else...

As I said initialy, I don't think the keboard scan circuit is completly working. When I use " set split", I'm not able to set the split note value form the keyboard. The keyboard scan circuit seems to work on the decoder side ( up to Z36 and Z37 input pins). Pin 34 of ths Z31 (keyboard scan CPU) generates a pulse for every new note event But I can't see a (KINT) signal and the (KMASK) is allways "low"

Should I be able to see the (KINT) even of there are no "active voice board", What should the "KINT" look like? I can trace it to J9-14 connector, but form there it goes in the CC+ board.

When a look to B0 to B7 on my scope, B7 is the only ont that as a realy different "wave shape" I will try to poste some pictures of the scope output

Heinz Weierhorst [21010276]

Hi Daniel.

Z49 B and C are a R/S flip-flop where Z31 pin 34 sets !KINT to Low. When the Main Computer has fetched note and velocity values it's reseted to Hi by !RDVEL. !KMASK should be only active during Autotune and Cassette operations. It also acts as reset signal but removes the keyboard while tuning by holding !KINT Hi. So may be there is something wrong.

Paul DeRocco [21030230]

Your main problem is something that is common to all channel boards. I'd yank out all the channel boards, and then use Set Split 31 to run the special auto-tune that doesn't disable the boards. While it's "tuning", you can monitor the digital and control voltage lines on the channel motherboard to make sure they're all working. If that doesn't find anything, you can put the boards back, use Set Split 31, and see if reasonable looking audio signals are multiplexed back to the I/O board.

Heinz Weierhorst [21010276]

Hi all.

Mr. DeRocco may be right. It's good practise to observe the Zerocrossing detector output with a scope to make sure there are any audio signals. Another trick is to select a preset with square waves - # 17 bank 1 for example and look to TP1 and TP2 on the voiceboards. There must be a high frequency for a short time and a then low frequency. That will show that the DAC and S&H bank is working. When the frequency on TP1 and TP2 is appr. 10-11 Hz after autotuning the board fails to tune for some reasons - may be a dead Oscillator, Filter, Amp or some of the Cmos multiplexers. With Setsplit 33 one must hear sound - but untuned! Not to forget: The Autotune latch on CMB and the !OSCEN latch should also be scoped. At least: While Autotuning the leds should flash 16 times!

John Leimseider [21030434++]

Has anyone found a fix for the tuning on voice boards that are just a little more out of tune than they should be? I recently fixed a bunch of boards, and 2 or 3 of them are almost tuning right, all the filter stuff seems correct and tracking, but one oscillator just doesn't track well enough. Any ideas?

Randel Osborne [21030467]


You can try replacing the op-amps and caps that are part of the sample-and-hold circuit on the voice cards. If either of these parts are marginal, the oscillator's pitch will not track as expected.

Heinz Weierhorst [21010276]

Hi John.

Do all voices pass autotune or did you use Setsplit 31? I'm asking that 'cause I did have a DVB with a failed Cmos switch just after OSC B. The computer wasn't able to "listen" to that OSC during autotune so the OSC remains untuned i.e with wrong scale and offset but did follow the keyboard. Another reason may be the charge pump IC (most RC4151). You can replace it by KA331, LM331 or XR4151 whereas the 331s have slightly better specs. A dirty adjustment pot or the tempcos are also possible. See Dave Clarks statement and solution.

Luca Sasdelli [21010226]

Hi John,

there should be some drift in thermal compensating resistors. Normally I simply apply a 39kohm resistor in parallel of each one to obtain perfect 50% duty cycle and tuning.

John Leimseider [21030434++]

I already changed the S/H stuff, the oscillator, the comparator, and the tempco... It passes auto tune. The pulse width is perfect. The tuning sounds bad, maybe 10 cents as compared to the normal 3 or 4 cent tolerance...

Heinz Weierhorst [21010276]

Hi John.

Try this: Remove R7 and R8 pins leading to th S&H bank. Connect them to the opposite S&H so that OCS A is fed by S&H B and OCS B is fed by S&H A. So if the fault remains the same it must be something on Oscillator B.

Paul DeRocco [21030230]

From: John Leimseider

I already changed the S/H stuff, the oscillator, the comparator, and the tempco... It passes auto tune. The pulse width is perfect. The tuning sounds bad, maybe 10 cents as compared to the normal 3 or 4 cent tolerance...

Did you clean all the flux off the board after your changes? Leakage current can really screw up an oscillator or a S&H circuit.

IO board - Chroma repair

Sean Vindahl [21010078]

Hi guys,

So after a small break I've begun to work on my Chroma again.

Earlier I had installed the CC+ upgrade and the Chroma worked fine for a couple of days. Then suddenly it began to switch itself off, and I therefore bought the PSU upgrade. Had some tiny problems with the new PSU, but located the problem.

But now I'm having this problem.

When i turn on the Chroma nothing switches on? No leds ... nothing? If I turn off the Chroma and then on again, some random leds a on. Turn it off and then on once again, different leds are on and some of the display lights up with random unreadable things?

Since the CC+ worked fine after installation and the new PSU has the right output values, Im thinking that I have some problems with the IO board... or am I way off?

Any hardcore Chroma users that could lead me in the right direction for location the problem?

Heinz Weierhorst [21010276]

Hi Sean.

First at all check the Reset circiut is working and is connected to its destinations.

New EQ

Heinz Weierhorst [21010276]

Hi all.

Here's my new EQ for the Chroma.

The reason for this is that my Chroma has stereo outputs and an external input. So I need two additional jacks one for input and the other for a width control. So the two XLR outputs were removed 'cause I don't like to drill any new holes. Now all outputs 0-3 are balanced, the mono output is balanced too but with a so called impedance balanced circuit (a poor mans version).

The output drivers could be AD SSM2142 , TI DRV134 or the better THAT 1646. Those cross coupled amps behave like a tranformer so one can use them as balanced or unbalanced outputs.

The quad vca is a AD SSM2164 which has volt to dB controls.

The disadvantage is you have to remove the pots from the old EQ board, the advantage is you have two CEM3360 as replacement parts.

So if anybody here knows where to get those pots let me know.

I will order the PCB this year but I don't think it will arrive soon. Because of holidays (xmas and new year) many companies have a brake until the next year.

Luca Sasdelli [21010226]

Hi Heinz,

great job indeed, especially about the VCAs stuff. A question: why didn't you include a headphone amp? Maybe because of the jack hole? :)

Heinz Weierhorst [21010276]

Hi Luca.

Thank you for the tip. I've added connector for supply and signal for a phone amp.

pro-actively replacing dried out caps?

Brian McCully [21030361]

Quite possibly this topic has been covered in the past, but I haven't run across it in the archives yet. I've enlisted a local tech to help install the spsu, cc+ and poly AT kits. He also suggested for an instrument of its age, that I should consider replacing all of the electrolytic caps because they dry out over time.

Is this 'all cap replacement' something that needs to be done, should be done or only in a 'fix as needed' type situation? He said if he did the work it would be ~$4 a cap, which would cover parts (caps are ~$1 each, pro audio spec'd(?)) and labor (soldering mostly). I looked at the parts list and it seems like an extensive and expensive endeavor. Are there any caps that are best replaced immediately - i.e. have a higher priority or tend to break down more than others? I'd appreciate any/all thoughts on this topic, thanks.

Paul Hackett-Evans [21010094]

Hi Brian,

I sold my Chroma last year (and yes! I DO regret it And I'd like to visit it in its new home sometime!) but when it was being restored, Roy Paynter replaced all the capacitors in mine. It sounded wonderful when it was finished. Of course, Roy did more than just the caps but I'm sure they contributed to how fantastic it sounded.

I can only imagine that if components such as caps do age and change over time, then the instrument cant be working at its best if it has old and deteriorated parts inside.

I'd suggest it's worthwhile doing, unless you are currently happy with the way your Chroma performs.

Paul DeRocco [21030230]

It may not be worth replacing all the power supply bypass caps, because even if a few dry out, they're in parallel with a whole bunch more that are probably still okay.

However, it's not a bad idea to replace DC blocking caps that pass a signal. The most important are C13 and C14 on the dual channel boards, because if they go hi-Z the filter won't self-oscillate and the board will fail auto-tune. When I resurrected my Chroma and Expander, I found five out of sixteen boards that had this problem, one of them on both filters, so I just replaced all 32 caps. Took less than half an hour. I also replaced the giant cap in the power supply, but that's a whole other thing.

Pressure sensor success

Paul DeRocco [21030230]

I received the CPS yesterday, and installed it tonight. It works perfectly.

One thing I did that wasn't in the "quick and dirty" installation instructions was to remove all the front felts, so that the key "bottom" is completely controlled by the pressure sensor itself. If you don't do this, then the pressure output becomes very sensitive to the vertical alignment of the sensor, because it doesn't take much to shift the pressure from the sensor to the front felt. Relying only on the sensor to stop the key results in maximum sensitivity, but I found I had to lower the key dip a smidgin, because the bar isn't perfectly straight, and a few keys around middle C were clicking on the front piece of wood. All in all, it feels good now.

Does anyone have any reason to suspect that removing the front felts is a Bad Idea?

Brian McCully [21030361]

Pressure Sensor Success, et al.

Just got back from working with a tech on the big upgrade with the three kits - SPSU, CC+ and pressure sensor kits. All three took about 6+ hours which was way beyond our original time estimate, but the tech gave me a break at the end for billing in regards to the unforeseen difficulties.

I am glad I worked with a tech because he had the tools.

I didn't try to address Paul's posting here regarding removing the front felts, because I didn't have the Chroma open over the weekend, as we had to stop before the holidays and I had to take it home, then bring it back in today, due to the tech's shop being too small to leave it overnight.

We didn't remove any key felts and the action seems fine, so far. There was a slight problem with closing the front panel due to the crossbar and polyAT board being higher inside (after the kit installation), but it still does close up (noted below in 'part 3').

Here's my summary of the installation 'events':

  1. The CC+ kit went in smoothly and quickly. Yay! no more PS hum! The cables needed to be zip tied to get them clear of drooping on the keys, but otherwise all went well. I was happy to find that a bank of patches was loaded in already when it was powered up - which saved time rather than having to do a sys-ex dump, etc. to get an init patch while on the tech bench to test it.
  2. The SPSU wasn't quite as quick to install. As a 'drop-in', in regards to lining up the older screw holes through the bottom, it was very tight up against the circuit boards that line the back of the keyboard. So we re-drilled the holes to place it more central in the old PS area, and I needed to fetch some new metric screws from the hardware store, as no screws came with the kit and the old ones were not securing it well enough.
  3. The Pressure sensor required the most work. We should have read the "Quick and Dirty" and Luca's write-up -first-, because we basically lined everything up as per the 'initial' instructions on this site, removed the sticky tape, mounted it, and found it was backwards (after noticing the ribbon connector wouldn't reach). At that point we re-read all the other instructions and surmised what we had done wrong (hmmm - when is a C a B?), but I needed to pack up the Chroma, to bring it back after Xmas weekend.

Meanwhile, the tech filled and glued the crossbar's end drill holes with dowel, because remounting it was slightly offset (i.e. reverse relative to the original mount) and needed re-drilling only on the ends (luckily) for re-alignment sake. The two center brackets aligned fine.

Upon returning, the install went relatively quickly, but the sensor MUX board wasn't quite clearing the front panel when the panel was screwed back in place, so the tech filed away a bit of the front panel inside lip, to allow a board component (that was leaning up into the panel) to be free and clear. In an email exchange over the weekend, Chris Borman sent me a few more pictures of the installation and orientation.

One thing I should have done -before- the PolyAT install was to measure the key depth. Luckily I have a second Chroma and was able to measure it after the fact, and make the adjustments accordingly today. Since the new 'felt plus sensor board' is thicker than the old felt, the cross-bar must be adjusted higher. The tech actually added two little shims under the cross-bar end braces just to add a little more height, to make up for the additional thickness difference (he used two short pieces of a paint stirring stick - just the right thickness).

After closing it up, the Chroma powered up fine. The only glitch was that the D key that's an octave above middle C was intermittently playing. I changed patches and it seemed to work fine, so I'm not sure what's going on, as all the keys around it were fine and the key contact looks fine. I'll see if it continues being flaky after I spend some time with it. The polyAT worked across all of the keys but I haven't had a chance to assess how evenly things respond across the keyboard. It is very cool though!

That's all for now - I still need to test out the new MIDI functions, etc.

Jesper Ödemark [21010135]

Re: Pressure Sensor Success, et al.

Brian McCully skrev 2010-12-28 23:01:

--- removed the sticky tape, mounted it, and found it was backwards (after noticing the ribbon connector wouldn't reach).

*bummer* I think I quadruple-checked that, being scared to make that error. It _feels_ as if the mux should sit better on the other side. Great that you sorted it out...

--- but the sensor MUX board wasn't quite clearing the front panel when the panel was screwed back in place, so the tech filed away a bit of the front panel inside lip, to allow a board component (that was leaning up into the panel) to be free and clear.

I think we've all experienced that tightness. I remember thinking "wouldn't an angled connector had been great now...

--- It is very cool though!

Agreed and congratulations! I'm eagerly awaiting my replacement CPS since my first was so flakey. The CC+ and the new PSU works like a charm though.

Keep those Chromas singing during 2011, everyone! :D

Luca Sasdelli [21010226]

Re: Pressure Sensor Success, et al.

Hi Brian,

no screws with the SPSU kit?

Every kit is shipped with a small transparent bag with two screws and a short silicon spiral for earth cables tidy-up. Maybe didn't you see it within the package?

Strange the matter of re-drilling the fixing holes; I never heard of such as a misplacement before. Anyway, to move it towards back panel it's enough to use the two ex-transformer rear holes instead of the two front ones. If possible, I'd like to have some details about it (e.g. measures of existing room between back panel and keyboard contact boards).

Anyway, congrats for your upgraded Chroma!

BTW: while you're working on it, I suggest to buy and fit the (cheap) UART on CC+, together with the latest firmware, so that if and when you'll choose to add the alphanumeric display, it will be only a matter of external items.

Thank you for the feedback!

Brian McCully [21030361]

SPSU placement, kit parts

Hi Luca,

There certainly may have been a separate bag of extras with the kit. The various kit contents were all left in their respective boxes intact, and brought to the tech shop, but could have been lost somehow. I did end up with one different front panel screw - which is a problem when you run out of time on a tech bench and have to pack it up to take it home. I looked through my boxes of leftovers, and there was no separate bag.

Yes we did consider the extra set of holes for mounting. I can't remember the underside of the new PS wood block, but I thought there were 4 holes (or maybe it was just the transformer holes in the former PS area plus some other holes...?). And also I now remember that there was a thick grey cable pinched under the PS when we removed it, which I think fed one of the audio XLR's - I haven't tested that connection yet. Anyhow the PS was going to be too tight up against the keyboard contact boards (literally touching), so we moved it. It didn't take too long to re-drill. I will take a picture of where the PS sits, next time I open the Chroma. I can also estimate where the former holes were from the underside, to provide further feedback.

It is so nice to power up the Chroma without the big 60 cycle hum !!! And the keyboard weighs less without that hunk of transformer.

I haven't read about the alphanumeric display yet. Sounds interesting - I will look into it, thanks!

Luca Sasdelli [21010226]

Re: SPSU placement, kit parts

Hi Brian,

the alphanumeric display is IMHO a very useful add-on; it shows with meaningful words what's going on while editing parameters, leaving free from checking the parameters chart almost each time. There has been no definite project about how to mount the display: it is Chroma-powered and it requires only three wires; I've mounted my one in a RS-components plastic box that has the exact size in height, and has enough room for both display unit and MIDI ports in a single box, using the MIDI hardware serial cable for all. Not terrific looking but practical. The RS code is 219-346.

Italian to English Translation

Chris Ryan [21030691]

Hi all, Odysséas [21010242] recently sent me scans of an article on the Chroma from the 1980s Italian magazine Strumenti Musicali. As I did for the Synthesizer Magazin article, which was translated from the German by Christian Kleine [21030210], I'd like to post this in English as well as the original Italian. Is there anyone on the list who could take this on? The article is just a couple of pages long.

Werner Schöenenberger [21010114]

Hallo Chris,

I do not know Italian, but if you do not find anyone on the list, I will ask a friend of mine to do the job. Just let me know.

Luca Sasdelli [21010226]

Hello all,

I've missed the original request from Chris, but I'm available to do the job. Please send me the scans.

Enrico Dibennardo [21030494]

... and we can split it in two parts, Luca, if you're busy.


Heinz Weierhorst [21010276]

Hi all.

Here a discription of the CEM3350 what I believe it's working. Note all components red marked in the schematic are always present in the Chroma. All OTAs are LM13700.

What is new is SW 2. If it is set to the right position we got a BP response. Together with SW1 to right postion we got an Allpass.

The circiut is not adapted to the Chroma yet. Frequency control is set to somewhat 0.5V/Octave. My problem at the moment is the Q-control. When I'm pulling the FIX input to ground (no pos. feedback) the filter behaves as described.

Because I have no audio sweeper I did't know what is the Q when stepping trough from 0-6 with Q-control.

My only choice is the pink noise source as a audio source and measure the increase of output level at a fixed frequency i.e. 32. But - the source isn't bad but not the best.

So what I need is more measurements from different Chromas to have an average.

Ok - the next problem.

The LM13700 needs current sources for control. Thus the expo converter is hooked up around a PNP array. Hard to find and all controls have the opposite direction.

But the good news. The filter is breadboarded and on my bench. It oscillates from 16kHz to appr.12 Hz (better as in simulation). Because I have no quad PNP array I've glued two 2SA798 together. In other words - the filter works.

Inside the CEM 3350

The CEM 3350 Dual State Variable Filter is a very unique design. I will explain it as I suppose it will work, so when I'm wrong please correct me.

The filter core itself is nothing special: two OTAs (A&B) in series but with full (pos.) feedback. A very, very simple model is shown in [Fig. 1]. Two variable resistors in series with active feedback. In reality an OTA is a bipolar current in, bipolar current out and unipolar current controlled device or a two quadrant multiplier. Please note there are no buffers or any resistors, tha's new!

Next OTA C comes in. It has full (neg.) feedback and acts as a shunt to ground. Thus the three OTAs could be seen as a T-pad. [Fig. 2]

Now we connect two capacitors to the OTAs, one to the output of OTA A (BP) and the other to OTA B (LP) output [Fig. 3]. We can do that because OTAs have a very high out impedance. A small signal is fed to the FIX input. When we vary the control inputs Iabc currents of OTA A&B equal, we vary the frequency. Thsu we have a current controlled filter with very high Q.

If the input is connected as shown in [Fig. 4] we got a HP and BP response, as it is done in the Chroma.

With a current to OTA C we can control Q of the filter by lowering the feedback. But—here is a great difference to other filters! Unlike "normal" filters where the current for Q is constant for a given Q, in the CEM3350 the current for Q must track with the frequency control currents (T-pad!). So if all three currents are equal we have a filter with a flat response and no "Q peak." When we lower OTA C current by two in relation to the frequency current we get a peak of 6dB or a Q of two, when we lower it by four we get a peak of 12 dB and so on. The pasband level is not affected as Q increases when we use the FIX input, that's new too! Most other filters suffer from the so-called "Q-loss." When we use the VAR input passband gets lower as Q increases because with lower current the gain of OTA C gets lower. So we can decide what we want: No Q-loss, Q-loss or something in between, also new!

Conclusion: The CE 3350 is a very clever and versatile design. There is a minimal count of external components (without Q-enhancement). It seems to be the first or second filter without buffers created by Doug Curtis. Maybe the CEM 3328 was first or the CEM 3350, I don't know.

An indicator for bufferless filters are the relatively high values for the capacitors, 22nF for the CEM3350, 33nF for the CEM 3328. The CEM 3320 has 300pF only!

Paul DeRocco [21030230]

What's your goal? To design a small circuit board that can be plugged into a socket in place of a 3350? Or to design an updated Dual Voice Board that doesn't use 3350s? If the latter, you'd better clone the 3360s too.

I think you're going to have trouble with the mismatched PNPs in the expo converter. You could hand-match them, but that's only practical if you're making a pile of them. This could be fixed by using one pair for Q1 and Q4, and a second for Q2 and Q3 with a single trim on their bases. The trim could be a small R to ground, and a selected R either to V+ or V-.

If I were replacing the entire board, I'd design transconductors from scratch, and not use any of the off-the-shelf OTA chips, since they tend to be a little noisy and have CV feedthru issues. You could do better with modern matched pairs and FET op-amps.

Heinz Weierhorst [21010276]

Hi Paul.

Thank for your hints.

Yes I'm thinking about a PCB for replacing a fault CEM3350 but it will not be easy to do cause all components should be SMDs to save area.

The quad PNP Array should be a THAT 320. So the two 2SA798 were only for testing knowing it's not the best solution. Cloning the CEM3360 could work with Mike Irwin linear VCA showed a several yers ago on EDN.

But- the gain core of the SSM2164 is total different from the CEM3360 seems to be a kind of Gilbert cell.

No problem when connected to an inverting Opamp, but what's about Filter modulation?

In this case the current out of one SSM2164 is tied to a 1.87k tempco resistor.

What is the output compliance? The datasheeet shows nothing. Please tell me more about the transconductors. Something like the early filter designs with the CA3096? DeFranco or so if I remember right?

In the meantime I've tested the BP and AP mode. The AP sounds better than I expected.