Akai DPS24 ADAT expansion card woes: no audio

[OneWorld]

I have some external effects which I want to use as insert effects, which use ADAT (in SPDIF - stereo) mode, I have set up the RME for ADAT to work as stereo, I had the red lights showing on the ADAT output, etc but try as I may I simply cannot get get the audio into the RME, in this insert effects scenario, but the ADAT works in other setups. It has to do with the patching or clocking I am sure.

It may just be sloppy terminology but I just wanted to check... You can only run ADAT format signals in stereo if you are using a 192kHz sample rate. So do you have all the RME inputs and outputs set to SPDIF or ADAT? It should be SPDIF. Are you using the RME as the audio clock master and is everything else set to external audio clock?

That might be the issue then, as the external equipment does not have the 192kHz sample rate. It is not such a big issue as I am using the analogue instead

[Coz14]

Hi Anthony - I think I might be able to help you out with your ADAT expansion card. I am in a similar situation as you which is how I found this post. I have a DSP24 which I bought in 2004 and still own. I have just upgraded my PC based DAW (Reaper) and wanted to use the DPS24 as a front end (Pre-amp and ADC) to send up to 16 tracks of ADAT to my RME HDSP 9652 soundcard.

Currently I am able to send 8 tracks via the DPS24's internal ADAT port.
Back in ~2012 I found/bought the ADAT extension board. The board worked for a little while but it was a bit flakey. Whenever I sent audio out on ADAT 9/10 the gain would increases in steps until the audio was distorting. Eventually out of frustration I hot swapped the board (removing and reinserting it while the DPS was still powered on), it stopped passing audio from that point. I fear/suspect that I may have damaged the Xilinx Spartan FPGA at the heart the board.

Recently, I have had renewed interest in seeing if I can get this card working again (or at least understanding/verifying what is wrong with it). This is what led me here. I saw your request for help, and I am willing to test your expansion card. If your are still interested in help, please reach out.

Steve

[Anthony gilbert]

Hi Steve,

Thanks for sharing your experiences. Very interesting read. Funny you mentioned the Xilinx Spartan IC, which I replaced. Of course it needs programming, so no board detection at all when I did this.

I'm not able to send the expansion card over until the new year, but I would love to see if you have any luck on your DPS24.

Send me a PM if you can for details.

Cheers,

Anthony

[Coz14]

Hi Anthony - I am unable to send you a PM. I have not been given PM privileges yet, due to my lack of posts on this forum.

I just went back and re-read this thread and I have some questions. 1) Did your expansion card ever work properly? 2) When did you replace the FPGA, was this before or after you noticed the LEDs in the ADAT TX ports flashing/extinguishing? 3) Do you still have the original FPGA? As you have stated these IC's need to be programmed with the appropriate firmware. At this stage further troubleshooting is pointless without the original FPGA on the board.

Steve

[Sam Spoons]

Hi Anthony - I am unable to send you a PM. I have not been given PM privileges yet, due to my lack of posts on this forum.

Steve

One more approved post and your future posts will be auto-approved along with your PM privileges being activated.

[Coz14]

Anthony - I did want to pass on some more information. I don't know if you have any of the schematics for the DPS and/or expansion boards. Many years ago one of the guys on the DPS World forum posted the DPS24 schematics, board layouts, service manuals and software files. I checked and it appears that all this content is still available and can be found here:
http://steve.sideroadsband.com/DPS24/

To see the entire ADAT path within the DSP24 you will need the following four schematics:
"IB_24ADT_PC_ADAT" - this is the adat expansion board schematic.
"PC Mother" - this shows the connectors and wiring of the DSP expansion bay where the ADAT, SMPTE & SCSI card connect.
"PC CPU 2" - This shows the Internal Master FPGA (IC2) and where the Master/Slave FPGA communications originate. It also shows how the Internal FPGA is connected to the Internal ADAT TX/RX Ports.
"PC CPU 1" - This shows how the Master/Slave FPGA communications are sent from the CPU Board to the PC Mother via J201-J203 then ultimately onto the expansion card.

Using the schematics we can now work off of some of your observations. For instance the Adat LEDs that are flashing then turning off (As stated in a previous post, this is not normal. The TX LEDs on my expansion board always stay solidly lit, mimicking the TX internal port). We can see that the LEDs within the adat TX ports are simply driven from the 5v that the DPS supplies to the expansion board. If they are flashing and turning off, it would suggest that the DPS is shutting down that 5v line. This can be verified by probing the Sync connector with a multimeter on pins 3,4 or 5 (5v) and pin 1 (ground) at bootup.

Why would the LEDs flash and shut off? There could be many reasons (take my explanation below with a grain of salt. I am not an expert and I am learning this stuff as I go). Again looking at the schematic there are SPI communication signals going between the Master FPGA (Internal) and Slave FPGA (Expansion) these are denoted by MOSI and MISO (Master In Slave Out, Master Out Slave In). If there is a communication problem at boot up this could be throwing an error causing the 5v line to shut down. Also someone else in this thread stated that the expansion card may be good and the problem could be within the DPS, this is also a valid possibility.

The other day I removed my ADAT card and powered it up on the bench using a 5v power supply feeding Pins 1 (5v) & 14 (GND). I was able to verify that the board power was working find (Adat ports were powered, the 3.3v and 2.5v voltage regulators were working, the 50MHz clock was working). My next step is to figure out a way to instrument the board to see if IC communication is present and if digital audio is getting passed to the board.

I hope this helps.

[Anthony gilbert]

Coz,

Many thanks for your detailed appraisal of the fault - and my apologies for taking so long to reply (a very busy start to 2022).

I have performed only basic tests on the expansion board, but it does appear to have the 5v, 3v3 and 2.5v rails working. I also have the schematics.

I haven't yet had chance to hook up my data analyser to look at the SPI messaging. I'm not convinced it will uncover anything that I can compare to what a working solution - perhaps some basic handshaking, but given the main DPS24 menus acknowledge the ADAT expansion, I'm guessing this initially works.

Testing my ADAT expansion on another DPS24 still appears to be the most logical step. The FPGA is on a socket, so my original programmed version could be copied/tested on your expansion card - that in itself sounds like it could be useful for your situation. I shall try to send a private message with my email now...

[Anthony gilbert]

Keen to further investigate the provision of power to the ADAT expansion card, so here's an update:

The Expansion ADAT GP1F32T (TOSLINK Transmitter) LEDs switch off after initial boot-up, which seems contrary to what other people are experiencing. Is this a result of the expansion card itself being powered down, post-boot?

In short, it appears not.

For one, the schematics suggest the DPS24 doesn't have a modular set of 5v supply rails for option cards. Next, I have added a breakout to the 9-pin ADAT sync D-Sub. This proves the ongoing existence of the 5v rail, as well as some periodic serial-out on pin-9. Interestingly, this serial-out data first passes through an AND gate with a PWRGDB as the enable. PWRGDB (Power Good) appears to work as a general enabler across several external facing I/O including the MIDI jack. The use of "Power Good" is used in other Akai products, and seems to rule out the idea of high power consumption (failing component) shutting options off.

So, we have 5v power, component tested 3v3 and 2v5 regulators, expansion board system recognition (handshake) on boot, PWRGDB enabled, and the FPGA pin 139 generating serial streams to the 9-pin ADAT SYNC D-Sub.

The expansion board on the face of it appears healthy, but with the transmitter LEDs still going off after c. 5 seconds from boot. They are controlled directly from the FPGA pins 117, 118. Puzzling, but more investigation to follow.

P.S, made an error in a previous reply when I called the socketed IC the FPGA. It is merely a generic array logic IC.

[Anthony gilbert]

As a quick update to this tread. I have decided with a heavy heart to finally part company with my DPS24. I have put the advert in SOS readers ads, and at a price which reflects the issue with the ADAT expansion card. Everything else is pretty damn good. I'm sure I will miss it!

[ajay_m]

The ADAT protocol runs at clock frequencies well within the capability of any half decent digital scope to analyse (like my Siglent SDS1202 200MHz scope) and I would have been tempted to rent one for a few days to troubleshoot (as they are somewhat north of £300 to purchase).
[or see if there's a grey-haired electronics enthusiast like myself in your local area!. ]
Scopes like this can decode a number of protocols but in any case their triggering options are sufficiently powerful compared to analogue scopes that you can generally get a good idea of what's going on.
[because if nothing else you can set it for a one-shot trigger and it has indefinite persistence on the display with zoom and scroll.]

They are generally a much better proposition than logic analysers because they tell you whether there's noise on the signal, whether the rise and fall times are good, whether there is ringing on the transitions, whether the logic high and low levels are within tolerance etc.

Albeit you would probably want to probe the signal by using an optical receiver such as the devices on eBay for around £20. (I'm assuming you only have optical out, not familiar with the DPS24).

Otherwise without proper test equipment it's frustrating and futile, unfortunately, there are things like this where a DVM is just not enough.

If it's any consolation, a scope like the Siglent would have cost the equivalent of a decent family car a few decades back, and if you have an interest in electronics design it becomes an invaluable tool.

[andysimm]

As a quick update to this tread. I have decided with a heavy heart to finally part company with my DPS24. I have put the advert in SOS readers ads, and at a price which reflects the issue with the ADAT expansion card. Everything else is pretty damn good. I'm sure I will miss it!

Hi have you sold the DPS24 now? I would be interested if you still have it. Email me on andy_simmons@yahoo.co.uk Thanks in advance!