Skip Navigation.
Section 0

ChromaTalk Archives: September 2003

My Chroma is not happy

Luca Pilla [21010218]

Hello from Italy!

I played my Chroma today with JL Cooper Chromaface and I've discoverd some problems:

  • 4 voice boards are dead. I tried to change their positions, but nothing. I tried Split 30: the sound of silence for those cards. Another voice board seems to have a trouble with VCA, the second or first oscillator is always open...
  • I checked the batteries, because my Chroma has lost every memory. I changed the batteries and I saw some oxidated contacts near the batteries. Nothing! My Chroma is without memory.
  • I have two keys not working

Any suggestion?

Thanks
Luca

ps: I'v got the original tape for voices, the original manual, pedals and flight case. The Chroma seems in good conditions.

David Clarke [21030085++]

  • 4 voice boards are dead. I tried to change their positions, but nothing. I tried Split 30: the sound of silence for those cards. Another voice board seems to have a trouble with VCA, the second or first oscillator is always open...

Luca - if you know of anyone familiar with debugging electronics, they could probably fix the voice cards for you. In general, the parts on the card are all still available - and the schematics provide sufficient information to narrow down most problems.

  • I checked the batteries, because my Chroma has lost every memory. I changed the batteries and I saw some oxidated contacts near the batteries. Nothing! My Chroma is without memory.

I have seen several cases where the circuitry responsible for switching the batteries in to the memory chips has failed. Again, the schematics provide a good place to start.

With the power off to the Chroma (but with the batteries installed) you should be able to measure a voltage on the RAM chips. If you don't have any voltage there, it should be a simple matter of tracing that voltage back to the batteries and to see where the signal is broken. If you like, we can provide some suggestions as to where to look and what to look for (assuming you have access to a voltmeter).

One straightforward item to check before going further, is whether or not the write-protect switch on the back of the Chroma is enabled or not. If memory is write protected then you won't be able to save new sounds (potentially making it appear that memory is not working).

  • I have two keys not working

This is often a problem of the key contacts themselves not making proper connection. If you have the keyboard open, locate the contacts at the rear of the bad keys. If you can make the chroma sound properly by pressing on those contacts manually (with your fingers), then a light cleaning/position adjustment may be all that is required.

Good Luck,

David

P.S. If you haven't done so already, it would be nice to have your Chroma listed in the Chroma registry.

Luca Pilla [21010218]

Many Thanks for your reply, David

David Clarke wrote:

Luca - if you know of anyone familiar with debugging electronics, they could probably fix the voice cards for you. In general, the parts on the card are all still available - and the schematics provide sufficient information to narrow down most problems.

I have the schematics, a tektroniks 475 oscilloscope and a voltmeter, but no experience in debugging eletronics.

If you like, we can provide some suggestions as to where to look and what to look for (assuming you have access to a voltmeter).

I will be happy to know what I have to do, and yes, I have the schematics

One straightforward item to check before going further, is whether or not the write-protect switch on the back of the Chroma is enabled or not.

Not enable, of course

If you have the keyboard open, locate the contacts at the rear of the bad keys. If you can make the chroma sound properly by pressing on those contacts manually (with your fingers), then a light cleaning/position adjustment may be all that is required.

thanks for your advices, I will try in the next day

Let me know what and how to do to fix my chroma, please
best regards
Luca

Luca Sasdelli [21010226]

Luca Pilla wrote:

Hello from Italy!

Hello from Italy too :-)

  • 4 voice boards are dead. I tried to change their positions, but nothing. I tried Split 30: the sound of silence for those cards. Another voice board seems to have a trouble with VCA, the second or first oscillator is always open...

Have you tried to swap'em over the motherboard? Does the failed board number change accordingly? Four voice boards is a big number for this machine; I suspect something about the analog section, e.g. an incorrect trim on the charge pump circuitry. This could be roughly verified by acting on the trimmers, setting them to a given position (e.g. beginning with full ccw, then increasing by a 1/4 of turn each test) and following each set with a [set split] 50 (system reset).

If the failed boards are still the same four pieces, even by swapping them across the mainboard, try to leave just one at a time in circuit, together with the other working four cards. In this way, you can avoid any kind of interference between failed cards.

Sometimes a voice card is muted due to a double OA broken. Just replace them with TL072 dual-OAs, cheap and easy to find.

In a short time, I'll replace the original PSU of my Chroma with the switched one, as the famous "nuked power" article mod states. During that period, if you're still in trouble, I can try to check your cards on my units.

I reside in Bologna.

Ciao
Luca

Luca Pilla [21010218]

Ciao Luca from Luca!

Luca Sasdelli wrote:

Have you tried to swap'em over the motherboard? Does the failed board number change accordingly?

Just do it.

Four voice boards is a big number for this machine; I suspect something about the analog section, e.g. an incorrect trim on the charge pump circuitry.

Where is this trimmer?

Sometimes a voice card is muted due to a double OA broken. Just replace them with TL072 dual-OAs, cheap and easy to find.

Good advice, Luca. how can I check this OA?

David Clarke [21030085++]

Re: My Chroma is not happy (Memory Problems)

If you like, we can provide some suggestions as to where to look and what to look for (assuming you have access to a voltmeter).

Luca,

In the schematic package, locate the sheet entitled "Computer Board Schematic" (page 6-3).

In the upper left-hand corner you should see a section entitled CMOS RAM POWER SWITCHING. This is the section that controls the backup voltage for the RAM, and the area that we'll want to look at for memory problems.

I have attached a picture of the CPU board and battery locations - and have marked three main test points, marked "A", "B" and "C"

A is the anode of CR1, and should always have a voltage equal to the total battery voltage (approx. 3 volts).

When the main Chroma power is turned off, test point B (the cathode of CR1) should have the voltage that would be presented to the RAM chips.

Test point C is the CMOS RAM voltage, as seen at one of the RAM chips. This should always be the same value as the one measured at point B (if it is not, then there's a bad connection on the board itself).

So:

  1. Measure voltage at A.

    If this voltage doesn't equal 3v (or so), then the previoius battery corrosion may have damaged the battery holder, or the battery holder may just not be making a good connection with the batteries. Measure the voltage at different points in the holder, to determine why you're not seeing the 3v at the output.

  2. Assuming 3v was seen at point A, measure the voltage at point B.

    If you don't see voltage here (2.4->3v), replace CR1.

  3. If voltage was seen at point B, measure the voltage at point C.

    If the voltages at B and C don't match, look for bad connections/solder joints between point B and C.

If the proper voltage is always present at point C, but the Chroma still isn't behaving as you would expect, please provide a description as to what it is doing/isn't doing that you believe is incorrect (i.e., how does the memory problem manifest itself).

Luca Pilla [21010218]

Re: My Chroma is not happy (Memory Problems)

David Clarke wrote:

  1. Measure voltage at A.

Point A, Chroma On 3.1 , Chroma Off 3.1

  1. Assuming 3v was seen at point A, measure the voltage at point B.

Chroma On 4,2, Chroma Off 3

  1. If voltage was seen at point B, measure the voltage at point C.

That's the problem! Chroma On 2,9 Volt on every memory chip, Chroma Off nothing!!

I suspect some bad connections, I have attached a jpg of this side. As you an see, there are some points of corrosions.

Meanwhile, I have fix the keyboard, and you are right: only contacts to clean.

Sandro Traversi [21010217]

Re: My Chroma is not happy (Memory Problems)

I take a look at your picture. It recalls me a similar problem I had on computer board. I may suggest you to remove all that oxide from pcb. Clean clean clean as much as you can. I used isopropylic spirit and compressed air Be also aware all that oxide can also lay under some IC around. It needed to desolde some IC , clean and solde IC on new sockets.

Luca Pilla [21010218]

Re: My Chroma is not happy (Memory Problems)

Yes, there is a bad connection. It's the trace near the number 229. Can I fix it with a cable from + of battery and the point at the end of the trace?

David Clarke [21030085++]

Re: My Chroma is not happy (Memory Problems)

Yes, there is a bad connection. It's the trace near the number 229. Can I fix it with a cable from + of battery and the point at the end of the trace?

You can fix it with a wire/cable - but it will need to go to a slightly different location.

Referencing the attached schematic excerpt (schm.jpg), your proposed connection would connect points "A" and "B."

Based on your measurements, there's actually a problem between points "B" and "C" (i.e., there's a break in that connection - due to the battery damage).

The following picture shows an alternate connection, which should just repair the problem area (B->C) (you can put a wire between the two red squares).

Before making the change, I strongly support Sandro Traversi's suggestion to clean everything off as much as possible. I would pay particular attention to the solder via's identified in yellow in the fix_me.jpg picture [below]. In order to ensure continued operation of the board, you should make sure that these all make proper connection through the board.

Chroma on eBay

Chris Ryan [21030691]

Chroma 21030200 (not in the registry [it is now]) is for sale on eBay. [Item 2555548193]

A mod Chroma on Ebay.de

Sandro Traversi [21010217]

URL removed (item 2556180085). Here are the pictures.

David Clarke [21030085++]

Translating from the German, and looking closely at the picture, the knob addition appears to be the sound-bank expansion mod. Anybody know what the four switch-like things are for?

Sandro Traversi [21010217]

It seems each switch enables a 50 sound expansion - so 4 switch X 50 sounds = 200 sounds (wow) !

Chris Ryan [21030691]

This was a mod offered by Ken Ypparilla/Chroma Cult. "This modification will expand your program storage to 200 sounds accessible as four banks of 50. The banks are selected with two toggle switches that you mount on the right cheek block of the Chroma or on the back panel of the Expander."

More info: Chroma Cult 1986.10.02: Memory Expander.

I'm not sure why there are four toggle switches on this Chroma: two switches, as mentioned in the Chroma Cult blurb above, would presumably provide access to four banks of 50, e.g. off/off, off/on, on/off, on/on. The large knob [see above] seems to be labelled "SOUND BANK," though the second word isn't clearly legible. I'm not sure what the smaller knob/button is for--there's a label above it I can't make out; or whether the circle to the left of the large knob is just a mounting screw or something.

See Chroma with Memory Expander on eBay.de (September 2004) for another spotting of this keyboard, with a clearer picture of the mod.

Fixing voice cards

Luca Pilla [21010218]

Hello!

As you know, I'm trying to fix the four voice cards (with a big help from Luca Sasdelli). Tonight I checked the voltage and waveform of Tp1 and TP2. I can see the right voice card play the timbre of choice, with relative waveform changing in pitch, with a Volt/Div of 50m on my scope, but I need a 2 Volt/div for the square waveform of wrong cards. It seems that the wrong cards have more output in Volt in respect of right card. Any advice?

Sandro Traversi [21010217]

May you explain a form about "How to test a Dual Channel Board using an oscilloscope" ? Not a "tractatus", just your knowledge. It'd be really interesting also to all mail-list.

Luca Sasdelli [21010226]

By my very personal point of view, I believe that the adjustment done by ear is quicker and more accurate than the one with the scope. It should be mentioned somewhere in the manual, but it's enough to connect a headphone between a voice card test point and the ground, play some keys until the pitch is quite low and audible, and then adjust the relevant trimmer to obtain the sound as "hollow" as possible.

The reason is simple: the square wave (50% duty-cycle) comprises only odd harmonics, therefore the optimal setup for the charge-pump oscillators is achieved when all even harmonics aren't heard anymore during adjustment.

P.S.: during the adjustment, the tone frequency varies, and it should be ignored, keeping mind only to the harmonics.

Sandro Traversi [21010217]

Answering Luca Pilla on this list I posted some conclusion about fixing Dual Channel Boards that I hope would be interesting to all mail-list - otherwise trash this mail !

Repairing my DCBs - as an absolute electronics beginners- I have reached a little personal method

  1. from Z19 and Z20 to VCO ICs , I check in and out voltages and compare to correct voltages appeared on full functioning DCBs whereas I find badfunctioning , I swap it with a new chip (plus socket ) - often I got really good results
  2. from VCO section to VCA out it's possibile to "hear " - in several DBC points and using a headphone jack as a probe - the sound from DCB (obvoiausly in Set Split 31 mode ) so as you can check at which section (VCO mix - VCF in - VCF out - VCA ...) sound still goes or interrupts - whereas I find signal interruptions , I just swap the suspected IC - even with good results
  3. as far my experience I found CEM as very stable and hardly subject to be broken
  4. so acting I have fixed 5 DCB on 6 broken
  5. new DCB problems are now arising and I still got 3 boards to be fixed ...

Luca Pilla [21010218]

May you explain a form about "How to test a Dual Channel Board using an oscilloscope" ? Not a "tractatus", just your knowledge.

Hi Sandro

I'm not a techniciam, as you know. I use my oscilloscope (Tecktronics 475, dual channel 200 MHz) to view the waveform and to misure the amplitude in Volt. I thin that a good software could replace an oscilloscope to do what I do,

See Fixing voice card (October 2003)

Chroma on eBay

Rob Vandivier [21030396]

Thinking about selling my Chroma. audio present only at direct outs, havent tested XLR outs yet. Anyone intrested before I eBay?

Freq./Pitch Tracking of Chroma

David Clarke [21030085++]

Has anybody out there has hooked up a digital tuner to the output of their Chroma and looked to see how well/bad it tracks up and down the keyboard?

If so - any conclusions (i.e., good tracking, bad, seems to error always low on low keys, high on high keys, etc.?)

Don Tillman

I suppose you'd want to mute all but one of the voice cards to measure this.

I should point out that the VCO's used in the Chroma are of a unique design. Most synths use a VCO with a sawtooth core and derive other waveforms from the sawtooth. One problem with a sawtooth core is that the reset time is finite, and will be a larger fraction of the waveform at higher frequencies. So higher frequencies will run flat. Sometimes this effect is compensated for to some degree.

The Chroma VCO is also a sawtooth core, but it is the only synth VCO that uses a "charge pump" to reset the sawtooth. The sawtooth reset takes a longer amount of time, but the charge removed from the capacitor is constant, so the high frequencies are unaffected.

This is very cool. I'm guessing that the designers recognized that a polyphonic synth of the quality of the Chroma needs more accurate tuning.

David Clarke [21030085++]

All - based on some testing done here I've become concerned that the firmware may be making some incorrect assumptions during auto-tune and incorrectly tuning the oscillators.

Specifically, I have concerns about the way the firmware determines the Reference DAC (RDAC) value used to set/scale the Volts/Octave output from the main system Digital to Analog converter.

To determine if there is really a problem (possibly requiring a firmware update), I need your help to perform some measurements on your Chroma.

To do the measurements, all you really need (in addition to the Chroma) is an external frequency measurement device (something like a Boss guitar chromatic tuner should do fine).

Test instructions:

  1. Set up a scratch patch (most easily achieved by ensuring that the EDIT A and EDIT B LEDs in the EDIT MODE section of the panel are on, and then holding down the PARAM SELECT button while pressing each of the numbered buttons (1-50) once.)
  2. Edit parameter 1, to change from a value of 0 (16-channel mode) to a value of 1 (one of the dual-channel modes).
  3. Store this patch in one of the patch locations.
  4. Press SET SPLIT 7 seven (7) times (7 board numbers should appear in the display, preceded by an "E")
  5. Press SET SPLIT 27 (to mute all B channels)

The steps above should only leave only one channel, on one voice board sounding.

  1. Press AUTO TUNE
  2. Press the lowest note on the keyboard recognized by the digital tuner.
  3. Adjust the "TUNE" slider on the chroma so that the external tuner indicates that the Chroma is tuned.
  4. Press the highest note on the keyboard recognized by the external tuner and note the output.

You can repeat steps 6-9 as many times as necessary to feel confident that you see a pattern.

What I'm interested to know is, what pattern is seen? Specifically, does the tuner show that the high note being played is regularly too high, too low, or just in tune?

If we do find a problem, we can then discuss the best way to address it.

Chroma Items on Ebay

David Clarke [21030085++]

Looks like there's a Chroma there (S/N 21030256): [Item 2559735875]

As well as single Chroma voice card (by a different seller): [Item 2559629731]

Sandro Traversi [21010217]

Re: More Chroma Items on Ebay (Expander)

[Item 2560368536]

David Clarke [21030085++]

Re: More Chroma Items on Ebay (Expander)

It looks like this is another, previously un-registered board (Expander S/N, 0104).

more compact cabinet?

Michael Zacherl [21030253]

Hi,

I'm aware that just to think about it may be considered as a big sacrilege, but however:

For the time beeing I'm just asking if anyone ever considered putting, or even did put the guts of a Chroma into a more compact box without it's keyboard and built-in MIDI interface?

19" is too small as far I can tell from knowing its inside, any other considerations?

David Clarke [21030085++]

The voice cards take up quite a bit of space. The next biggest thing is the power supply - and then behind the front panel there's 3 reasonablly sized circuit boards.

If you want to keep the front-panel intact, then you're probably not going to be able to get too awfully much smaller than the Expander.

If you don't want any front-panel, then you could probably drop down a bit smaller (especially if you were to use a smaller, switching power supply). The 8 voice boards are still probably going to be responsible for taking up most of the real estate.

Michael Zacherl [21030253]

David,

thank for replying - I think it's better to leave it as it is and restore the original case and save this rare beast. ;-)