-
Repair attempt TC-Electronic Nova System: DSP Audio (guitar)
Posted by obiwanjacobi on 24 Jan, 2015 10:59 -
Hi,
I have bought a TC-Electronic Nova System from someone who accidentally connected (touched) the left output to 230v mains (oops!).
So I thought this would be a nice learning project and perhaps I end up with a working unit.
When the unit is powered up, the LCD display shows the initialization message but hangs after that.
So I figured not all is lost, the processor still seems to work...
I opened the unit up and measured the supply voltages. Seems all ok: +3.3v, +5v (for digital) and +/- 15v (for analog).
Then I tried to find the functions for the chips that are used on the main board. Here's the list:
1x 39VF040 Flash ROM (Software v1.3)
1x (AMIC) A42L0616S-50F 1M x16 CMOS Dynamic RAM (with EDO Page Mode)
3x (CYPRESS) CG7130 ATT (More RAM?)
1x (Motorola) DSPB56362 AG120 DPS Audio Processor (used as main processor?)
1x CS8414 CSZ 96kHz Digital Audio Receiver (?)
1x AKM AK4620 BVF 24 BIT 192KHZ AUDIO CODEC (with IPGA) (ADC/DAC)
3x 74LV00 4x 2-input NAND gates.
6x NE5532AP Dual low-noise op-amp. (Texas Instruments)
1x HCT4052 Dual 4-channel analog multiplexer/demultiplexer
1x 74HC139 Dual 2-to-4 line decoder/demultiplexer
2x DG4410 Quad SPST CMOS Analog Switches
Because the (left) output was hit with 230v (connector is located on the main board) I figured that I would work back from that. It is possible that some of the analog circuits are fried - although no caps are blown or other physical marks/smells- I think I have to focus in on the DAC portion of the circuit. I think the processor hangs because it is waiting for some hardware signal that is not coming (because of a faulty component). My guess would be that either the CS8414 (audio receiver) or the AK4620 (DAC/ADC) chips are likely candidates. But I want to confirm that in some way.
The question is basically:
- Is my reasoning sound?
- How do check if these chips are still operating correctly? (is there some signal I could check?)
I do have a multimeter and an old analog 20MHz scope - but no logic analyzer. (Do I need to get one?)
I know I may be a bit in over my head, but still would like to give this a try.
Digital part of main board:
Analog part of main board:
-
Self test menu:
1.Power on while pressing "DELAY" + BOOST".
2.Loop the midi in and midi out.
3.Follow the instructions as shown in the display.
4.Power OFF and ON.
Also there is service note:
If you have a product freezing or rebooting intermittently, especially when performing a drop-test,
you must pay attention to the flatcable.
-
The unit does not enter the test mode. The same initialization message is displayed (Nova System / TC Electronic).
I checked connectivity of the flat-cable, and all wires connect fine.
Would be very odd if this would have been the fault...
But thanx for the suggestion.
-
I turned the unit on for a while and felt the chips with my finger for hi temps (don't have a fancy IR camera like Dave) and both the DSPB56362 processor and the CS8414-audio receiver are getting warm really fast; after a few seconds they're is almost too hot to touch. The AK4620 is getting a little warm but not hot - but I would think a DAC/ADC should not get warm at all.
Is it normal for a DSP processor chip to get really warm/hot (without heatsink!)? Or is it a sure sign it has faulted. Strange how that LCD display is getting data - I can see no other processors on the main board. (there is (probably) a small processor/MCU on the top board with the buttons and display - would that be responsible for the LCD activity?)
Top left what looks like a small MCU (I have refitted the print and did not look at the part number).
The cheapest DSP is about $15,= and the cheapest CS8414 is about $12,= (havent looked up the ADC/DAC yet). Not too bad - but not cheap either...
Do I have a good chance replacing these chips will fix the problem? What else could I try? -
Check R166 ( between TP153 and TP20 ), and R167, they are 0R0 protection resistors between AGND, DGND and PGND.
Top left cpu is ATmega168, and yes, he is driving the LCD. -
Will do - but that will not explain these chips getting so hot.
Yes, the Atmega fits the bill
Thanx! -
Found R166 (it's ok) but can't find R167 - can you give a hint of its location?
BTW: how do you know these things? Do you have a schematic? -
Found R124, R168 and R169
-
Ok, I have contacted the tc-electronic distributor and ordered a new main board. The price of the (3 digital) chips alone was more than half that of new board. So it will be an easy fix.
-
Bummer! There was an error in the price and it now more than double the first quote.
This is not gonna happen after all. 
Now its back to the decision whether or not to buy the individual chips and push forward...
Suggestions? -
1) Did you check how many Ampere the board is sucking up?
2) Can you desolder the chips suspected brocken? You should than check how the current at the power supply will than change. -
No I only measured the supply voltages. Because two chips were getting really hot, I did not bother to look further.
I am (still) not sure if I am willing to cough up the money for the new chips. But your suggestion is a good one and I can do that before buying anything.
Thanx! -
A bit off topic maybe but I am planning to get a Nova system and modify it for 4-cable method. From your pictures i can make a qualified guess that the drive unit is on a separate board. Is that so? and could you take some pictures of that maybe since you have already disassembled it?
-
That sounds like an interesting mod. I would be interested in hearing the details. There is enough room to drill holes for 1 or 2 more jack connectors.
It is correct that the drive input comes in on its separate board and -I haven't decoded the chips there- I suspect it does contain the drive circuit. The tc site boasts with having an analog drive circuit - so.
The white 'cable'/connector on the right goes to the main board and comes in near some analog multiplexers -if I remember correctly. The (empty) connector on the bottom left comes from the top board that contains the display and buttons. No idea what that is for to be honest.
Drive input board.
Here's an overview shot. The main board on the bottom is normally positioned on top of the power supply (left) and the drive board (right).
Hope it helps.
Marc -
Oh and I have measured the current for the +3.3V and +5V rails:
With faulty chips on Main Board:
+3.3V 0,227A
+5V 0,145A
Not blazingly high to my surprise...
I have successfully removed the DSP package with hot air and am now turning my attention to the Audio Receiver chip. After that I think I also remove the ADC/DAC chip and measure current again - to see what the difference is. -
Thank you so very much. Time to go hunting for a used one. IMO the Nova system would be be the perfect companion for a lots of amps having a fx loop including the new BluGuitar amp1. And as it seems from your pics and description the mod should not be to hard to make. I would believe that The larger connector is for the analog signals and the smaller empty one is for data transfer for parameter change. Maybe I2C or SPI.
One more question, I suspect the Line input jack is a breaking type since it disconnects the analog unit as i understand. Can you confirm that? -
Yes, it looks like there are three NC contacts that open when you insert a jack. But that should not matter as long as the drive board keeps working - haven't reverse-engineered that though.
-
Thank you kindly for your info. I'll keep you posted as work progresses
-
+3.3V 0,227A
+5V 0,145A
Do you have the power supply specification? I assume you already checked against that... -
I do not have any specs for the supply. But it does not seem to mind the load at all. It's a switching power supply.
I am currently waiting on some SMD soldering supplies to properly clean up the board now that I have desoldered the two chips that got really hot. I also have to restore one or two SMD C's that I knocked out of place while desoldering the chips.
When I got the board clean I will measure the current again.
-
I was researching where I could buy the chips that were defective. I noticed the DSP chips has on-board ROM and RAM.
So although there is a separate Flash ROM chip on the main board, how can I be sure there is not also some programming in the DSP chip?
So if I pop in a new one - it will not have that programming and the unit will not work - even though the chips are then fixed...
What to do? Abandon the repair or chance it that all the programming is in the external Flash ROM chip?
The three chips that need replacing (at least) will run about $50,= total. For me that is a bit much to gamble with. I could first only buy the DSP (15,= and up) and see if the processor runs again...
Suggestions? -
After careful consideration I have decided to spent another $15,= and try to fit a new DSP chip. Prepped the PCB for its arrival.
-
Hi guys, sorry for a possible noob hijack of this thread. I have a similar weird problem with a used nova system that worked fine for the first few days but after leaving it on for a few hours it locked up and whatever effects were on at the time stayed on and working. It refused to do anything else. When I powered it off and back on, it booted as normal and worked properly for a few seconds then the same brick thing happened again. I eventually worker out how to get to the factory reset before it bricked, and this did the trick for another day or two before the problem returned. I can boot into the fronttest by holding the boost and delay, and everything works fine. Weirdly, it doesn't freeze while doing this so maybe it's just the DSP that's overheating or something if it's left on. I'm trying to get my money back and get a quote on a repair but I'm definitely considering opening the thing up because I didn't pay a fortune for it. I'd also love to mod it for the 4 cable method. If any of ye have had similar problems any help would be appreciated. Cheers
-
Self test menu:
1.Power on while pressing "DELAY" + BOOST".
2.Loop the midi in and midi out.
3.Follow the instructions as shown in the display.
4.Power OFF and ON.
Also there is service note:
If you have a product freezing or rebooting intermittently, especially when performing a drop-test,
you must pay attention to the flatcable.
I'm resurrecting this thread as I recently got a used Nova System that I am troubleshooting. My unit will power up and works but freezes after a couple of button presses. It still passes audio, the DSP remains working and the signal is effected but the UI stays frozen until I power cycle. I will reseat the flat cable at bot ends, but I am interesting in your "service note", is that something that came from TC? Is there any documentation that goes with it (service manuals? service bulletins?) Thanks. -
I suspect the smal ATmega in the upper left corner drives that part - since the DSP seems to continue to work.
Check that out, see if there is any change between working state and when it hangs.
Just reseating might not be enough. I would beep-out every wire in that flat cable - just to make absolutely sure.
I have not found any service manuals. tc-electronics is notoriously secretive.
Perhaps PM Zarhi? He seems to know more.
(I abandoned my repair after I found out that also the ROM (on the DSP board) was fried too. I still have the unit though.) -
I have tried reseating the large flat cable with power on, this seems to reset the entire system as if I have cycled the power cable.
I have also run the self-test (delay and boost held down while powering up) and I could go all the way through, no issue with any buttons, LEDs or encoders. The unit still freezes when powered up.
The ATMega would be nice fix, surely it is programmed though? I can attempt to desolder and dump it but I imagine it's code protected.
Would a dump of the main socketed ROM chip help you out with your unit? Mine has a v1.2 sticker. -
I just went through this thread with the help of Google translate - funny to see the same names come up again and again in the DIY guitar electronics communities, Dirk_Hendrik seems to be everywhere
http://www.gitaarnet.nl/forum/showthread.php?167760-Nova-System-Repair&s=307ac744598859f01a90bc0cd256cb47 -
I tried to immediately run the factory reset routine on power up. It ran, reset and now everything appears to work. I'm thinking there may have been corrupt preset data stored that was causing a lock up on footswitch presses. I have heard of this only working temporarily, so I will probably keep the screws out for the next few days. If a factory reset fixes it again I might look at swapping whatever memory holds preset data. If any measurement from a working unit would be helpful then let me know.
-
Yeah, sure - if its not too much trouble.
Just post it here as an attachment.
Thanx!
-
Just getting to this. I dumped it with a Tl866 and PLCC32 adapter but I'm not totally certain it's a good copy. Looking at the hex I don't see many recognizable strings, usually I see a copyright notice somewhere.
GNU strings does find text like this:
>DIMD IpmueR iecdev<
> - sySmet - <
>P -ser teB /kna - <
Which looks like mangled text for the menus, so text may be stored in some unusual way (?)
-
Looks good, I could find preset names in the first block. (2 blocks: FF-data1-FF-data2-FF)
The menu strings are in the ATmega, I think - you will not find them in the DSP ROM.
Thanx. -
You are right, each letter of the preset title has some extra characters in between.
This might explain some of the freezing issues reported online, saving to flash may cause corrupted data which locks up the unit when it's reloaded. My unit recovered by a full factory reset. I have seen accounts of swapping the flash chips between units to revive them too. -
"A bit off topic maybe but I am planning to get a Nova system and modify it for 4-cable method. From your pictures i can make a qualified guess that the drive unit is on a separate board. Is that so? and could you take some pictures of that maybe since you have already disassembled it?"
Hello guys. I would really like to know if this mod. could be done else I'll have to spend some bucks to see myself
-
FYI: I have an almost complete reverse-engineered schematic of the front panel board. Shoot me message if you're interested.
