Attempting to fix Theta DS Pro Gen II

Not true DIY but as we do not have a repair / refurbish forum, figured this would be acceptable.

OJ sent me his Theta DS Pro Gen II as he was decluttering for his move. The one channel was not working and I thought I might take a stab at fixing it.

I never appreciated just how SOLID these things were. I have worked on a few sonic frontiers dacs and still own the transdac and although they are well built, the Theta is a tank.

some pics of the boards





First thing that I needed to do was replace the power rocker switch as the paddle had snapped off previously and I could not get it to toggle any longer - I just grabbed an NK switch I had in my parts bin and soldered it in to get started.

once powered up , it was obvious that the right channel was not in a good place - it had 14V of DC sitting on the output

My first task was to try and find a schematic - not surprising, nothing is available , so my only option is to try and do a basic reverse engineer on of the layout of the IV stage and see if I might be able to identify the point at which things go wrong.

From a brief look at the components, it looks like the signal comes from the dac - goes to a logic chip and then onto 2 serial opamps before an output current buffer ( not sure what the motorola chip is - it has been sanded down to prevent identification)




I do have one question for the digitally savvy folk on the forum



anyone know what the B W stands for - google has not been my friend and I feel that I should know this but cannot think what it represents .


My fear is that the vintage "can buffers" are the source of the problem



I have not looked yet, but suspect they might be a challenge to find / replace. If this is the case, I might end up replacing the entire I/V stage

more to come ..dB

 

think the penny finally dropped - think this might represent the various clock paths

Bit Clock, Word Select and L/R clock

but then again, maybe not

..dB

 

On my Gen III it was one of the buffers that stopped working when one channel went dead. Fortunately we found a replacement from a spare I/V board that was salvaged from dead Gen III. You can still find these buffers online, but they are expensive and beware of fakes. If you can find a another Gen II I/V board that might be your best bet.

 

Start by scoping out the IC outputs in the IV stage’s faulty channel. It’s prolly stuck to +VDC.

 

Hi, dB. This is a guess, but B and W may correspond to the bit and word clock signals respectively. This hunch is based on Mike's previous comments on the BWD transfer protocol. [This post] may be relevant. So the header you showed may include the bit and work clock signals and the digital data for the left and right stereo channels. I think @Scott Kramer may be a good resource here since he's done so much cool mad scientist stuff on the digital side with the Schiit gear.

 

Thanks that was useful, also makes good sense why there is a logic chip after the dac..dB

 

Adding @baldr to this thread, he might be interested or have some suggestions.

 

https://www.diyaudio.com/forums/digital-source/34457-buf03-equivalents.html

 

You can also roll a PCB adapter if you find an opamp with similar specs but not the same package

 

You could transplant the left (working) IC in the right path to see if that's actually the problem

 

Interesting! At first glance agree it's probably a shorted cap or transistor in the output stage, since one channel is just fine.

Checked over the pics... there's dual mono ADxxxx AnalogDevices dac chips there, B W L R but no D labeled on the digital board-- so it must be LD & RD (left Data, right Data) normally (or i2s today) there’s just a single (D)ata stream, with the (W)ord clock alternating L/R selecting the correct channel out of the data stream. So here it’s doing it beforehand to prep for the dual mono dac chips (TLDR; this suggests a single channel has the possibility of going out on the digital board).


First thing I’d do is look up those ADxxxx chips and get the pinout (or eyeball it). Put an oscilloscope on each of the digital pins and verify BWD on each chip. If all good you’ve ruled out the digital board.


Tap around on the analog board (digi too) while playing music. Look for cold solder joints. Then I guess work your way with a multimeter (start at the dac chip analog side) slowly work your way right looking for a bad (shorted) opamp, cap, transistor (look up pinouts on chips wherever you can)… at least there's a mirror circuit to compare. You may have to pull a few components out of circuit to find the short.

However, if one of B,W,D is missing at either of those dac chips… (high probability the major components on the digital board are fine, since a channel works, re: the motorola DSPs / firmware / input). You could get lucky and it’s one of the shit-ton old school 74xx TTL logic chips.

Start right at the dac chip digital side again, and work the other way, make sure the BWLR (labeled) is active (rule out the grey hookup wires). Flip the digital board over and look for a cold solder joints, particularly on the 74xx chips. -- anyway, it could easily be a bad 74xx chip not passing D(R) thru, check for bad caps.

Good luck!


That Digital board is funny, it's like baldr went back in time and had to build a Gungnir Multibit using the equivalent parts of the time ;D

 

Lot of tantalums on those boards. They can fail by shorting. I’d look there, but I’m an amateur.

 

The good thing about stereo is that you can compare left and right, doing spot checks with an ohmmeter (with power off, of course).

 

Curious, what happened with this?

 

I bought a new IV board which I was going to retrofit. Then after much poking about figured it might just be one of the AD chips that was damaged and needing replaced. Then looked at the major mess I could make if I did this on the fly, so shelved the project till I could clear my desk and have this as the only focus.

So at this point there are 2 paths.

Replace an opamp and everything is back to normal OR mod it slightly and bypass the IV stage (which I suspect was responsible for much of the magic and thus trying to avoid this)

More to come
.. dB

 

This is the most difficult thing in any project
The table filled with the remains and corpses of old projects depresses me

Then accumulate millions of different unidentified boxes, which are probably already other forms of life

 

Anyonw knows what the algorithm selection does ?



Very good sounding unit. Bass is similarly good as the Yggdrasil and less fat. This one's even got Mike's sig: