| Rapid |
Hi everybody!
I've searched but couldn't find any mods for my behringer :( . Has anybody upgraded it with better components, like OP amps...? I'm using digital input (optical) and analog out.
All mods are welcome :)
Cheers,
Mattias |
|
|
| Rapid |
| Come on guys :) Could somebody point me in the right direction? A link or idea would be great! I've heard people modify the older 8024. |
|
|
| Bat |
Rapid,
What do you think of the DEQ2496? I've been looking at the Ultramatch Pro 2496, was wondering about the quality of the D/A conversion. Probably the same DAC as the DEQ2496. I'm thinking of getting one & then modding it heavily - it's cheap enough, if I screw up, I can throw it away & buy another. I was thinking of maybe using transformer outputs instead of op-amps...
...but the basic DAC has to be good enough to work with.
Sorry, I know this doesn't answer the question you asked - but maybe we can start a trend?
Bat |
|
|
| Rapid |
Hello Bat!
Sure, I'll copy my reply from another message board (audioasylum) because the question was almost the same :) Hope you don't mind:
All signal is passed through the 2496 but I only choose to equalize the bass range because I haven't succeeded in eliminating the peaks in the midrange perfect. That's why I'd use acoustical damping instead.
It's a pretty big difference if I put some pillows on the floor infront of the speakers. It smoothens out the response above approx. 1Khz
This is how things are connected:
CD/transport -> Digital optical inputs on the behringer (it's either optical or AES/EBU (?) )
Analog phono (XLR availible) outputs -> (phono->RCA converter) -> active filter
I'm using analog out with the built-in DAC. To me it sounds good but since I've heard people complain about the Sss sounds (and some don't), I'm not sure if my demands are lower than "audiophiles". Some people claim they can hear difference in cables. I can't but I hear a small difference when my friend connects his $160 PS Audio DAC to his pioneer CD player.
If I could choose wether to buy it again or not, I'd buy it. Maybe you can borrow it from a shop and try.
There's digital output from the DEQ2496 so you can run your favorite DAC instead of the built-in.
Cheers,
Mattias |
|
|
| Bat |
Thanks for that, Rapid.
I'm pretty much decided now to get the Ultramatch Pro. As far as I can see, there are no unblanaced outputs, so I'm planning to add a couple of Lundahl line transformers, maybe bypassing the output op-amps, depending what I find when I open it up.
Might get the DEQ2496 as well tho... |
|
|
| ChuckT |
Just to revive this thread.:smash:
I just got this gear, the boards inside are very compact, not much room to mod. |
|
|
| Kuei Yang Wang |
Konnichiwa,
| quote: | Originally posted by ChuckT
I just got this gear, the boards inside are very compact, not much room to mod. |
Well, the Op-Amp's for the analogue stages are on the board with the connectors on the back. You can upgrade these and also of course you could use a lot of the present space to fit input and output transformers completely replacing the active analog stages....
The PSU should also be easily improved....
Sayonara |
|
|
| ChuckT |
Holy, that was a fast reply.
Here's the dac part. The ak4393 output travel over the computer cable entering the opamp in the connector board. |
|
|
| ChuckT |
Here is the output stage. Cheapo opamp at work here, will replace them with either AD8066 , AD8620 or LM6172 .
May be replace the whole output stage when I find the time and the right circuit, or may just a single transformer. |
|
|
| lucpes |
| Rubycon ZA for local decoupling / Rubycon ZL for the 'bigger' PS caps would fit like a charm. |
|
|
| ChuckT |
Hi KYW,
what would be a good output trans? I have got a UTC W-786 600 to 600 trans on hand, thinking about using as a balance cable to single conversion . |
|
|
| stef1777 |
| quote: | Originally posted by ChuckT
Here is the output stage. Cheapo opamp at work here, will replace them with either AD8066 , AD8620 or LM6172 . |
Do not work.
You need +/- 15V op-amp. |
|
|
| Kuei Yang Wang |
Konnichiwa,
| quote: | Originally posted by ChuckT
what would be a good output trans? |
Depends upon the circuit you are going to use. There are quite a few good 1:1 types, but they are not really suited to directly interface with the DAC.
| quote: | Originally posted by ChuckT
I have got a UTC W-786 600 to 600 trans on hand, thinking about using as a balance cable to single conversion . |
Should be fine for that.
Sayonara |
|
|
| Kuei Yang Wang |
Konnichiwa,
| quote: | Originally posted by ChuckT
Here is the output stage. Cheapo opamp at work here, will replace them with either AD8066 , AD8620 or LM6172 . |
Any of the above will do fine. I'd say AD8620 on the analogue inpoit side, AD8066 or LM6172 on the output side.
| quote: | Originally posted by stef1777
Do not work. You need +/- 15V op-amp. |
Actually, the supply in the Behringer was +/-12V, last time I looked (inside my own), so au contraire, does work just fine....
Sayonara |
|
|
| stef1777 |
| quote: | [i]Actually, the supply in the Behringer was +/-12V, last time I looked (inside my own), so au contraire, does work just fine....
Sayonara [/B] |
I'm very surprised.
Mine is +/- 15V. It use JRC LM7815A et LM7915A in the switched power supply (It has 17,7V before the regulators). It has +15V on the red wire, -15V on the second wire of the power connector.
I used OPA2604AU. |
|
|
| ChuckT |
Thanks KYW.:)
Strange, it seems there are different configure out there for this gear. Mine uses LM317 & 7805. I found only those 2 regulators on board, maybe there are smaller ones that I've missed.
Here's the pic. |
|
|
| stef1777 |
These are for the digital board power only.
In fact, the +/- 15V regulators are in the switched power supply. Have a look. It's wrote on the PCB (IC1: 7815 and IC2: 7915).
Power connector:
1 Red = +15V
2 = -15V
3 = 9V
4 = 15V ground
5 = another ground
6 = 5V
7 = 3V
But the best if you want to be sure is to test on the op-amp pins... |
|
|
| ChuckT |
Your are right, I see the 7915/7815.
And thanks for the power connector info, very useful.
One thing is bugging me after a bit of probing on the output circuit. Why are there ceramic caps directly couple to the balance out, it is directly in the signal path!!!:xeye:
Surely this can't be good!! |
|
|
| tiroth |
| If they are connected to ground, it is to remove switching noise. |
|
|
| ChuckT |
Yes, they are connected to ground and the values on the board looks like is 100p.
Here's a pic. I am now starting to mod the PS. In goes the Blackgates. |
|
|
| stef1777 |
Be careful with the power supply. This is a switched power supply.
I will suggest you at start to mod only the two 470UF and the two 10UF before and after the 15V regulators.
- Replace C4 and C5 by 1000UF 35V with low ESR / hight ripple current (as Rubicon ZL). Will fit but you will have to push a little D1 and D2.
- Replace C1 and C2 by 47UF 25V tantalum (use good quality model as AVX Tap).
Write for reference, all voltage before and after regulators before tweaking. Write voltage with the digital card connected and without the digital card connected.
But the most important is to replace opamps. |
|
|
| ChuckT |
I just found out that is only two 0.1uf decoupling caps for the four opamps. All the ceramics next to the opamps are for filter circuit, the 2 decoupling caps are located much further away.
The opamp will be replaced by lm6172, but I will definitely replace the whole output stage later when I have more time. |
|
|
| stef1777 |
| Please put picture here if you identified coupling / decoupling caps on the analog board. This will help others. |
|
|
| ChuckT |
| Refer to my opamp pic, the pair of decoupling ceramic are just located in the top left of the pic where one is hidden under the black ground wire and another next to it. |
|
|
| stef1777 |
I have the DEQ with a complete 'new caps' power supply for a week and all is fine. :dead:
For the caps on the analog board, I waiting for an Black Gate PK source.
We have a thread there but in French. You can have a look on the pictures.
http://www.homecinema-fr.com/forum/...4761&highlight= |
|
|
| ChuckT |
Hi Stef1777,
may I recommend not the use the PK grade blackgate, I have use those before and find them not very effective and not worth the price. I would probably just use a higher value ZL which also fits. Just my opinion.
Also, I have tested some the ecaps on the output board and none of those are from the +/-15v rail. So I believe they are just for the relay.
Do you know why it has 4 dual opamp for the output, the ak4393 datasheet shows that output stage uses only 2 for balance out.
Could it be possible that using lower gain will bypass 2 opamp? |
|
|
| stef1777 |
| quote: | Originally posted by ChuckT
may I recommend not the use the PK grade blackgate, I have use those before and find them not very effective and not worth the price. I would probably just use a higher value ZL which also fits. Just my opinion. |
Actualy as I don't realy know the use of the 7 ecaps close to the in/out, I haven't moved to replace them. The PK was not really expensive. Rubycon ZA seems better than ZL. 6.3mm caps can fit on the analog board. Elna (all type) do not fit execpt if we change value or voltage.
| quote: |
Also, I have tested some the ecaps on the output board and none of those are from the +/-15v rail. So I believe they are just for the relay.
Do you know why it has 4 dual opamp for the output, the ak4393 datasheet shows that output stage uses only 2 for balance out.
Could it be possible that using lower gain will bypass 2 opamp? |
I don't know sure but you have 2 stages. I changed the 7 opamps... but at start I was looking to put AOP2134 on input and AOP2604 on output. As not sure where are the right one, I changed all of them will the same opamp.
But be careful, Behringer's engineers are vicious.
The PCB is really FRAGILE. |
|
|
| ChuckT |
Ok, more probing shows that the ecaps next to the relay are coupling caps for the 1st opamp stage. There are 4, but I only trace the 2 to the opamp, the other two not sure.
So changing this should make a difference.
I put in Blackgate-N for the opamp coupling and cerafine for the other 2. |
|
|
| ChuckT |
| I have also change the opamp to lm6172. As the decoupling is fairly sad on this area, I have put in some cap that fits on the +/- pins. |
|
|
| stef1777 |
| HOT! for the last picture...;) |
|
|
| stef1777 |
| quote: | Originally posted by ChuckT
Ok, more probing shows that the ecaps next to the relay are coupling caps for the 1st opamp stage. There are 4, but I only trace the 2 to the opamp, the other two not sure.
So changing this should make a difference.
I put in Blackgate-N for the opamp coupling and cerafine for the other 2. |
Sorry, but can you specify wich one are the coupling caps on the picture? What's the value of the new ecaps?
I'v hires pictures of the board without the 7 opamps if this can help to analyse the circuit.
Thanks. |
|
|
| ChuckT |
I believe the cap no is show in the pic, they are the 4 cap next to the relay. I believe they are 47uf, I use 33uf.
BTW, have you try the auto eq, does it stop by itself? It seems to go on forever and I just stop it. I couldn't get a very flat reponse!:( |
|
|
| stef1777 |
| quote: | Originally posted by ChuckT
I believe the cap no is show in the pic, they are the 4 cap next to the relay. I believe they are 47uf, I use 33uf.
BTW, have you try the auto eq, does it stop by itself? It seems to go on forever and I just stop it. I couldn't get a very flat reponse!:( |
I'ts never flat or if you try, you will break the music.
Read a little more the Manual. It's wrote inside.
Just push the button after 2mm, and click on DONE (next screen). Put delta max at 3 dB. The best is to switch to double mono mode. You can also use a destination curve as audio X-Curve. Great!
This not the right thread. You can open one "Use of DEQ23496" thread if you need. |
|
|
| ChuckT |
Right. Thanks for the reply.
Back to the mod, I have added BG FK 100uf to the +/-15v supplies beneath the digital broad connector. I believe I have done as much as I can/want with the output stage. The next upgrade should be a new outstage!!:eek:
Well, maybe one more when I have time. There are various filtering smd caps in the opamp circuit that IMO are quite critical to the sound. The small values (100pf and 390pf, please correct me if I wrong) are likely to be NPO grade (unfortunately can not confirm) and no need to be change, however, the 2.2nf are critical and unlikely to be NPO (due to cost). So changing them to NPO smd should make an improvement. I have also found some pps film smd caps from Panasonic in Farnell that might fit.
Also, it may be possible to replace the 2 smd 0.1uf decoupling for the 4 output opamp to Blackgate NX-Q 0.47uf/50v. |
|
|
| stef1777 |
Sorry, I've have no time to look at the DEQ right now. I'm leaving for my summer vacations by the end of the week.
I would again suggest you to publish pictures with red circles around of the different identified caps and your remplacement suggestion wrote on it. This is the best to understand and follow.
Concerning smd caps, Wima propose a Samples program. Have a look.
http://www.wima.com/navig/smdsp.htm |
|
|
| deerhunt |
It seems nothing has happened about this thread for a while...but I'll try anyway.
I've opened my deq2496 and (think I have) identified how the DAC op runs in the flat-cable to the op-board.
AK4393 has balanced op on pin 23,22(L) and 21,20(R). The op is 2,4Vpp so I believe it's suitable to be connected to a Zapfilter2 analogue op-stage:
http://www.lcaudio.dk/index.php?page=6
Attached picture shows DAC and flat-cable. I got the following results (DAC - cable):
23-7
22-8
21-10
20-11
Does anyone have a schematics where this can be confirmed? I would be very grateful for a schematics!
Maybe someone has done this already (completely new op-stage) and can share some useful or not so useful information
:) |
|
|
| Kuei Yang Wang |
Konnichiwa,
| quote: | Originally posted by deerhunt
It seems nothing has happened about this thread for a while...but I'll try anyway. |
Sorry, I did not get around to doing any work on my DEQ...
In the meantime I have confirmed that the analogue stages gegernally match in terms of layout the ones in the DCX Crossover (screenscrapes from acrobat attached).
Input Circuit to ADC (this with the takeoff of the digital feed):
Output Circuit after DAC:
The +12dbu/+10dbv/3VRMS=0dbfs switch option links in there somewhere, not sure yet where. Ignore all the component numbers, they have no bearing on the DEQ, only the principle design is the same.
For those wanting SE Outputs at lower levels it is easy to just cut out a whole stage, remove the XLR Sockets and fit RCA ones.
Also, the various DC blocking capacitors should be possible to be simply eliminated and replaced by a wire, only one is really required, namely between the Input Stage first stage and the phasesplitter driving the ADC (marked C6 on that particular schematic). You may have to add some offset nulling in the analog stage after the DAC, but that is not a big deal.
| quote: | Originally posted by deerhunt
AK4393 has balanced op on pin 23,22(L) and 21,20(R). The op is 2,4Vpp so I believe it's suitable to be connected to a Zapfilter2 analogue op-stage |
Yes, if you can find the space to fit all neccesary bits that can be used. Depending upon your needs you could use much simpler circuitry.
| quote: | Originally posted by deerhunt
Does anyone have a schematics where this can be confirmed? I would be very grateful for a schematics! |
Me too please, it makes work ever so much easier. I have been able to find the DCX2496 Schematics, but not the DEQ one. As said however, the basic structure between the analog stages of the DEQ and DCX matches, so the DCX schematic helps a little.
Sayonara |
|
|
| Kuei Yang Wang |
Konnichiwa,
Armed with the Schematics from the DCX and some datasheets here are some thoughts of what I feel would likely make for the best I/O stages for the DEQ (and by extension the DCX). From a "purists" viewpoint I rather take a single high quality transformer instead of 3 - 4 Op-Amp's, tons of capacitors and so on. So, can we roll transformers in there easily?
First, if we consider the output from the DAC, 4.8V Peak-Peak (differential) with a 2.5V DC offset. The Output clearly can drive 1K2 differential, as example circuits show a pair of 620R Resistor to ground.
If we have the DEQ in +12dbu level setting we get nomainally 3V Output Voltage, which is 8.5V.
So, if we where to take a commercially available "universal" line transformer of the 2K5:10K Type offering a level handeling of at least +13dbu @ 50Hz and connect the 2K5 side across the output of the DAC we will have an output of actually 3.4V for full scale and we will be able to drive a load of 5K or higher. No need for anything more. The transformer will form a 2nd order lowpass.
A suitable type would be the Stevens & Billington TX896 (ideally with some minor mods). The transformer is rated with an 0.5db bandwidth of better than 45KHz, good enough for 96KHz converters, with +12dbu @ 40Hz level handling we have just the right level without having to use handgrenade sized transformers. DCR is around 200R per winding, so in stepup mode on the analogue output the output impedance will be around 1k2 resistive, not outstanding, but non too bad either.
If purchased theorugh the normal distribution channels they will probably run at around 75 - 85 US $ each (sorry the poor US $ Exchange rate really louses things up for non europeans) not cheap but if you count a bunch of non-polar Black Gates, SMD Op-Amp Adapters for OP627 etc. etc the costs ratchet up quickly too....
These will be a little hard to mount in the rather low hight of the case. The flying lead version simply bolted to the bottom of the case is taller than the case allows, placing them on their sides is difficult. Using the PCB version on a dedicated PCB might work well. Otherwise using a piece of suitable angled Aluminum should make the mounting easy enough, using the flying lead version. In either case there is enough space between the powersupply and the main DSP PCB to place them there, in the DCX it would be a little tighter a squeeze.
If we also need and option for +22db the transformer should also allow a 1:6 Stepup, but that is likely to cause interesting happenings, for once, the permissable load on the output would be very low. Hence the option would be not very usefull for studio users, but just fine for domestic use.
Now on to the analogue inputs. What can we do here? We need 4.9V Peak-Peak (differential) with a 2.5V DC offset to drive the ADC to clipping, but in the case of the Behringer gear a 10db Pad is in front of the ADC. If we remove the shunt resistor of the Pad we have a load onto whatever drives the ADC of nominal 4.6KOhm in parallel with a 600R/1.5nF Zobel.
If we where to use the same transformer as on the Output, but reverse (setdown mode) our source would see a load of around 18K in parallel with 375pF. Our transformer again will operate as lowpass, basically anything should be pretty peachy.
I suspect we could actually remove the RC lowpass on the DAC or at least scale it to lower C-Values if we want to do more SMD messing about than strictly needed. In that case, the Transformers own DCR will amount to around 300R effective feeding the inpouts of the ADC.
Thus removing all the resistors of the L-Pad and the Capacitor in parallel to the input and soldering the wires directly to the pins of the IC should work. Then add a sensible value (say 470pF polystyrene or Silver Mica) across the input and we should have an effective input impedance of around 17K nominal, in parallel with 120pF, a load that will work fine with virtually any active output stage from source gear and even most passive preamps. Full Scale input voltage would be 3.4V RMS.
There you go, universally balanced or unbalanced inputs and outputs for the DEQ2496 all fully passive, with levels appropriate to HiFi Gear (not Studio). The outputs have the same levels balanced & unbalanced and don't really give sack jit if you use them balanced or single ended....
Sayonara |
|
|
| deerhunt |
Thanks for your detailed answers Kuei Yang Wang (T.L.).
I did have the Zapfilter laying around so this evening I've spent putting it into the DEQ2496.
Smoke is still inside the components (all electronics components contains smoke to function - if the smoke gets out the unit brakes down - 1st class of electronic engineering):D
I can post a few pictures after the holiday.
Sounds good so far, I can make a small conclusion later.
Good Night |
|
|
| deerhunt |
Now I have som photos and information that might be useful for those who are going to put an solder iron into this box.
If you are going to keep the original o/p contact, DON'T try to bend the leads going down on the circuit board after cutting. I did that and then it broke deep into the plastic, not possible to solder on. So, better to cut in two places - see red marking. And a Neutrik contact doesn't fit in the original hole...
I slitted the band cable with a razorblade, cutting lead 7,8,10 and 11.
The Zapfilter fits into the box. I haven't found place for power supply board and transformer so they will be placed in a separate box. The 5W on the power board becomes rather hot, so if putting all in the DEQ2496 box - venting holes has to be made (I guess).
So what about the sound?
A way to describe the change in sound is to go from pixeled sound to a more organic flow. The sound became "darker".
Did I loose some energy (and resolution) in the upper register?
I made some adjustments in the mid-high range on the DEQ, going from -2dB to +1,5dB.
Well, no doubt that there is more dynamics, more extension in the lower frequencies.
Around voices a small tendency of "ringing" before now disappeared. I can play at higher sound levels, keeping sonic appearence.
In conclusion, its definintivly worth the hours and money, not saying that Zapfilter is the ultimate solution. There are as Kuei Yang Wang writes others. |
|
|
| george a |
Hi All
Where can one download the DCX2496 schematics from?
Thanks in advance |
|
|
| Geoff4 |
Hi
Ive just started modding my DEQ 2496. The first thing was to
try to tame the SMPS. I used an active gyrator type filter which
attenuated the line noise somewhat. Next Ive fed the clock inverters with a relatively clean +5V using an LT1763. Also
replaced the ADC & Dac decoupling and voltage reference caps
with Oscons. Also the o/p caps have been replaced. I use the
AES/EBU input from my Pioneer DV565. I can do a direct comparison with my switched pre-amp and the modded DEQ
sounds just a lttle bit veiled compared to the bypass. Next
is to do something about the o/p stage which is currently the
NJM 4580 op amps. Since Behringer wont release a schematic it
looks like Ill have to do my own o/p stage.
Geoff |
|
|
| Kuei Yang Wang |
Konnichiwa,
| quote: | Originally posted by Geoff4
Since Behringer wont release a schematic it
looks like Ill have to do my own o/p stage. |
You can do that. Or if you can read a principle schematic and translate this into what is on a PCB you can mod the original. I can post a principle schematic of the circuit in the DEQ, both analogue input & output, if there interest. Note it will not show component designations but values, which should match the DEQ but may be slightly different.
Sayonara |
|
|
| leo van doorn |
Dear Kuei Yang Wang,
Thanks for the suggestions on i/o using transformers, I will try to implement these in my dcx.
What would you suggest as the road to take for the power supply ? Accu/battery based ?
Thanks for sharing your ideas !
Leo |
|
|
| Kuei Yang Wang |
Konnichiwa,
| quote: | Originally posted by leo van doorn
What would you suggest as the road to take for the power supply ? Accu/battery based ? |
Hmmm, switched mode supplies are a lot better, POTENTIALLY, than their reputation. I'd probably fix the existing SMPS, which is not too difficult if you know what you are doing and have a fast and sensitive 'scope.
DIY Supplies, batteries are always a PITA. I'd probably make a raw 20V + 20V + 9V supply that is choke input and uses multiple layers of RF filtering plus schottky diodes and feed that into the SMPS PCB and upgrade the reg chips there.
If I get the time anytime soon (unlikely) I'll mod my own DEQ.
In mind I have an abridged version of the original analogue stage (I&O) with consumer levels and SE (RCA) jacks (fitted instead of XLR's), the remaining Op-Amp's changed to AD8066 and the "liberated" op-amp's used as DC servos to allow direct-coupling of Inputs and outputs. If I get any suitable size Transformers in time from S&B (it's on their list to develop) I'll use these of course.
Then quiet down the SMPS and upgrade all the digital supplies.
Sayonara |
|
|
| Geoff4 |
Hi Kuieu
I for one would be interested in any schematics you can come up with. It would take far too much time to reverse engineer the
original design thats why I was thinking of doing my own.
Ive delevoped my own discrete opamp specially for this kind of
application but it is quite big.
I managed to filter the main power rails to the op amps
quite well. I originally tried LR/C networks but this didnt work as well as I hoped due to the inductors interwinding capitance I
think.
I think its worth perserving with this device because even in its current state I like what it does the sound of my system in my room.
Geoff |
|
|
| Kuei Yang Wang |
Konnichiwa,
| quote: | Originally posted by Geoff4
I for one would be interested in any schematics you can come up with. |
Analogue Input section:
Analogue Output section:
Sayonara |
|
|
| stef1777 |
| The schematics posted below are for the DCX not the DEQ. |
|
|
| Kuei Yang Wang |
Konnichiwa,
| quote: | Originally posted by stef1777
The schematics posted below are for the DCX not the DEQ. |
Note that I said the anaologue I/O for the DEQ used IN PRINCIPLE the same circuitry. I did a rudimentary trace on my own unit to confirm. I noted that component designations would not line up.
However, for people wanting simple mods it should be non too hard to trace the DAC signals especially to the first summing mp and equally the ADC inputs to the Balancing Dual Op-Amp.
The changing the relevant resistors to the "right" types should be childs play.
Sayonara |
|
|
| R Dijk |
Just a nightthought about 100 step resistor nettwork using binary counter. putting two resistor-netts in paralell and bin counter selecting pot lets you take "one half" step at the time.
anyone tryed somthing similar ? |
|
|
| ryssen |
| This question may not belong in this thread,but when you take the signal out of the digital output is it the processed by the eq or is it a unprocessed signal? |
|
|
| deerhunt |
A picture says more than...
Hope it helps.
brgs |
|
|
| ryssen |
| Is it okay to use LM317 and LM337 to adjust the analog power to +- 12v to use AD8620?For output and input |
|
|
| stef1777 |
| quote: | Originally posted by ryssen
Is it okay to use LM317 and LM337 to adjust the analog power to +- 12v to use AD8620?For output and input |
You can may be use 12V fixed version to have less components to add. Space is counted.
You can may be replace the two caps before reg with better and bigger version, en replace the caps at out with tantalium caps.
I personaly used 2 x 1000UF 35V and 2 x 10UF 25V tantalium. |
|
|
| ryssen |
| Another question,is it allright to short the input e-lyts and the ones inbeetwen the ic´s? |
|
|
| OLES |
Hi
what does it do this amplifier - TL072 ? |
|
|
| ChuckT |
For the pure passive output stage KYW mentioned, I found this app note in Jensen trans with circuit for CS4396.
http://www.jensen-transformers.com/as/as093.pdf
http://www.jensen-transformers.com/...ts/11ssp8ma.pdf
It uses the very expensive JT-11ssp-8ma "Repeat Coil" trans for voltage out differential dac like the CS4396 which is almost the same as the AKM4393 used in the deq/dcx2496.
Now I don't know is the difference between a split winding "repeat coil" and other 600:600 trans. I have a UTC W-786 600:600 and wonder if this is also suitable for the Jensen output stage. Thanks. |
|
|
| ryssen |
| Could I remove the digital input transformer,sins it is 1:1 and take the digital direct to dac input? |
|
|
| stef1777 |
15V linear PSU added to the DEQ to serve the op-amps in replacement of the original PSU.
4 cables to cut and it's ready. Large difference show on the scope. |
|
|
| stef1777 |
Top = new PSU
Botton = original PSU
 |
|
|
| ryssen |
| Do you have a schema of your mod? |
|
|
| stef1777 |
| quote: | Originally posted by ryssen
Do you have a schema of your mod? |
Not realy but it's not difficult.
First build a small 15V PSU using LM317 and LM337. I use a 2x15V 15VA toroid. Fix it a the best place depending of the PSU size. Be sure to isolate the LMxxx and the PCB from the DEQ box. You will not need a large cooler for the LMxxx.
On the power cable cut the first RED cable (+15V), the second black cable (-15V), the fourth black cable (analog GND). Connect the + and - 15V from the DSP board to the new PSU. Connect the new PSU GND to the GND cable coming from the old PSU and the GND cable going to the DSP board.
That's all.
With the new PSU, you reduce the noice by x10, have less heat on the old PSU, and more power for the analog and digital circuits.
 |
|
|
| stef1777 |
The last tweak.
New op-amps for the DEQ2496.
Input: 3 x BB OPA2134 (on the left).
Output: first I/V stage from AKM using four BB OP627 soldered on home made adapter and final stage using two BB OP2604.
The small adapters are only 5 millimetres tall.
The coupling caps have also be replaced with Blackgate PK serie.
Great mod. No more noise and great sound. |
|
|
| rossl |
I opened up my DEQ2496 just to look. I am surprised to see a DSP board that doesn't have a ground plane. The decoupling on this PCB is terrible. I wondered why I had an occasional lockup where I had to power off and back on. Now I know!
I had some tantalum caps laying around and I soldered them near the SHARC chips. Right on top of the small ceramic caps.
I may add some more capacitance in the power supply cable instead of modding in a new power supply. (I'm using digital in and out, not the analog outs) |
|
|
| rossl |
| quote: | Originally posted by ryssen
Could I remove the digital input transformer,sins it is 1:1 and take the digital direct to dac input? |
Why would you want to do that? The transformer provides ground isolation. It is a barrier to digital noise, EMI, and ground loops. You want to keep that noise away from the DAC. |
|
|
| rossl |
From post #17 in this thread:
Power connector:
1 Red = +15V
2 = -15V
3 = 9V
4 = 15V ground
5 = another ground
6 = 5V
7 = 3V
Does anyone know what the 9V is used for? Is it the relays? |
|
|
| stef1777 |
| quote: | Originally posted by rossl
I opened up my DEQ2496 just to look. I am surprised to see a DSP board that doesn't have a ground plane. The decoupling on this PCB is terrible. I wondered why I had an occasional lockup where I had to power off and back on. Now I know!
I had some tantalum caps laying around and I soldered them near the SHARC chips. Right on top of the small ceramic caps.
I may add some more capacitance in the power supply cable instead of modding in a new power supply. (I'm using digital in and out, not the analog outs) |
Your power problem comes from a bad contact on the power connector (quality of the contact inside). I add the same problem. Check the 5V pins first.
Coupling caps are ok on the DSP board but I you want to tweak, you can unsolder the curent 10UF on the board and replace them with Elna RJH or OSCON. |
|
|
| seoman |
| quote: | | Your power problem comes from a bad contact on the power connector |
I seccond that!!!!!!
Remove the glue dot off the connector en wiglle it.
That wil solve ít.
I had the DEQ for 3 days and it started switching on-off.
When removing the power supply, in search for bad kontakts, the problem was fixed.
I never have had a power failure again.
But I also agree on the bad PCB layout.
I only took a closer look on the analog part and i think that can be done much better.
Regards Simon |
|
|
| stef1777 |
| quote: | Originally posted by rossl
From post #17 in this thread:
Does anyone know what the 9V is used for? Is it the relays? |
Don't know.
The right Power connector pinout:
1 Red = +15V
2 = -15V
3 = 9V
4 = analog ground
5 = digital ground
6 = 5V
7 = 3V
Same on the DCX power supply (but not all used). Mod the power supply on the DCX is really a good thing. The PSU is really too small on the DCX. |
|
|
| rossl |
| quote: | Originally posted by stef1777
Don't know.
The right Power connector pinout:
1 Red = +15V
2 = -15V
3 = 9V
4 = analog ground
5 = digital ground
6 = 5V
7 = 3V
Same on the DCX power supply (but not all used). Mod the power supply on the DCX is really a good thing. The PSU is really too small on the DCX. |
Thanks. I will work on the connectors.
I think the 9V may be for the display light. |
|
|
| ChuckT |
| I believe the 9V goes to the regulator LM317 that powers the dac and other chips, so this supply is very important. But somebody please double check. |
|
|
| ryssen |
| Is there a schema for the powersupply for DEQ2496? |
|
|
| reins |
Hi,
has anyone ever thought of putting a remote volume control into the DEQ?
As any needed voltage is available, this with a new output stage would be a interesting feature, eliminating the pre-amp. Would the power supply deliver enough current to drive the additional units?
Thanks
Stephan |
|
|
| stef1777 |
| quote: | Originally posted by ryssen
Is there a schema for the powersupply for DEQ2496? |
No available.
| quote: | Originally posted by reins
Hi,
has anyone ever thought of putting a remote volume control into the DEQ?
As any needed voltage is available, this with a new output stage would be a interesting feature, eliminating the pre-amp. Would the power supply deliver enough current to drive the additional units?
Thanks
Stephan |
I don't think that the stock PSU can handle more circuit. |
|
|
| seoman |
I'm thinking of turning my UC into a preamp.
it should be simple.
There are 3 different inputs and i want to expand them with some relais.
Ofcource there has to be a motorpot at the analog outputs.
The thing is that with midi you can completely controll all functions so with a couple of buttons and a controller it would be easy to load a memory preset and switsh some relais.
Right now I use all 3 inputs with 2 different corrections.
2 (cd comes in optically and the pc by spdif on the xlr) have the same room correction.
For surround i use the analog inputs to adjust the center and subwoofer channels of my processor.
I was very dissapointed by the sound of both the center and the subwoofer but after roomcorrection they fit much better with my other speakers. (even though they are not corrected anymore)
I'm so pleased with it that i'm thinking to sell my processor (rotel rsp 1066) and buy me a extra deq2496 and a dvd player with multi outputs and make a multichannel preamp myself with the DEQs.
The thing is that the deq2496 has more potentials to be a good pre amp than all the **** you can buy below 2000$.
When it comes to signal handeling that is. I don't need the video switching!
The motorpot i can rip out of an old sony surround processor ESD type (its six channels). The relais will be the Nais TQ2 they are very good and can even handle the digital signals! maybe even video!
It should be easy :rolleyes: .
If this plan becomes more concrete i'll post it ofcource!
But the plan has just started.
Regards Simon |
|
|
| lsisaxon |
| Has anyone tried OPA2228 instead of OPA2604? The THD is supposed to be an order of magnitude lower than OPA2604. |
|
|
| FLECOM |
i was just wondering about output transformers, instead of the $200 dac output transformers that were linked to two pages ago couldent you use these sowter transformers and still get a balanced line out?
http://www.sowter.co.uk/specs/3603.htm |
|
|
| OLES |
I greet
what is on regulator Vout LM317? |
|
|
| courage |
LM317T front view, from left to right:
1= ADJ
2= Vout
3= Vin |
|
|
| janneman |
Hi,
I have designed a form-fit-function replacement output board for the DCX2496, an article on it has been published by AudioXpress (August issue). You may want to take a look at that.
Jan Didden
Edit: Just realised this thread is on the DEQ not the DCX. Sorry.... |
|
|
| courage |
Jan,
No problem. As I am also looking into possibilities to use the DCX with my Maggies, the information may still be of benefit to me and hopefully others.
Thanks for mentioning. |
|
|
| ChuckT |
Haven't touch my deq2496 for a while as I was working on my harddisk media player (TVIX) with I2S to TDA1541 (yes, that setup sounds really good, if someone can tell me how to interface I2S with deq, I will be very happy...)
Anyway, last night, I tap the ak4393 direct out to a tamura 300:600 tranny, it sounds really good, only got one channel working so far...
But the PS, the 9V is really really important, it feeds the 7805 that provides all chips that uses 5V, this includes the dac of cause.
And the 9V doesn't seems to go thru any inductor to reduce noise, so should be a good idea to add a coil and more caps for that 9v. The 7805 gets really hot with means its drawing quite a bit of current.
The LM317 will take the 5V and output 2.5V, so should be for the dsp chips. I found that there is a ceramic directly at the LM327 output, a real no no for this type of regulator, so I took it out. |
|
|
| mbcouple |
Instead of improving the supply to the 7805, you would be better of removing the 7805's from the board, and using outboard discrete regulators like jung, super-jung or similar boards. Selectronic sells a replacement regulator kit at an affordable price, or you can do your own.
Hans |
|
|
| tubesguy |
Hans -
Do you have a cite to the Selectronic kit that you mentioned? Apparently my google skills are not up to snuff. TIA- Pat |
|
|
| ChuckT |
Ouch, that regulator is expensive.
Providing seperate very low noise for each chip doesn't cost that much. Eg, some like the LT1027, LT1763 will probably do as good if not better. |
|
|
| FLECOM |
So I drifted over to the Slim Devices website, and they are coming out with a new device http://www.slimdevices.com/pi_transporter.html called the transporter... which they claim uses the "Miracle DAC" - AK4396
ok, so, I have played with AKM dacs before, since my behringer has an AK4393... so i head over to AKM's website... and it looks like a really nice chip... and it looks like its pin-per-pin compatable with the 4393, 4394, and 4395!
but the best part, it needs NO output caps! none at all!!!
I am definately going to see about trying to find some of these IC's and try swapping out the DAC in my DEQ2496 and making a new, capacitor-less, LPF and see what happens :) |
|
|
| Nature B0y |
I've got a SRC2496, which supposedly uses the same 4393 DAC, and I was looking to do that DAC swap as well, or at least bumping up to the 4395. Have you found a source for small quantities of the AKM DACs? The only place I've found so far is DIP international in the UK, and their minimum quantity seems to be 50... :(
One thing to think about is that those filter caps *are* in your signal path, even though there's no DC blocking caps required. Obviously, it's easy enough to get high quality caps in there, but I was thinking of going passive with a transformer, say a Lundahl line-level or somesuch. |
|
|
| FLECOM |
well, they are not really directly in the signal path dc blocking, just part of the filter... so some nice caps and you should be good to go... or you can just omit the filter entirely I guess...
I have not been able to source them in small quantities yet... working on it :) |
|
|
| wastl |
densen2002, the file has gone. May I ask for a repost (or email)?
Thanks! |
|
|
| janneman |
| quote: | Originally posted by FLECOM
[snip]but the best part, it needs NO output caps! none at all!!!
[snip] |
Neither does the '93, if you use the balanced to se converter you showed and which is shown in the data sheet. Both the '93 and the '96 need coupling caps if used balanced (their outputs sit on 2.5V DC).
Jan Didden |
|
|
| kepa1 |
| quote: | Originally posted by FLECOM
i was just wondering about output transformers, instead of the $200 dac output transformers that were linked to two pages ago couldent you use these sowter transformers and still get a balanced line out?
http://www.sowter.co.uk/specs/3603.htm |
Here is what Sowter answered when asked about it:
Hello Alain
3603 would be ideal provided the output impedance of the DAC is 1000
ohms or less. The load should be as high as possible so your high
impedance amp will be ideal.
http://www.sowter.co.uk/acatalog/SO...NCERS_2.html#a9
Best regards
Brian Sowter
SOWTER AUDIO TRANSFORMERS (E A Sowter Ltd)
This transformer will handle the balanced/unbalanced conversion as well. That must be a good candidate but I definitely don't have the knowledge to really recommend it;) |
|
|
| Spartacus |
The higher priced 5069 just below it will provide lower distortion in the bass.
Dan |
|
|
| FLECOM |
| quote: | Originally posted by janneman
Neither does the '93, if you use the balanced to se converter you showed and which is shown in the data sheet. Both the '93 and the '96 need coupling caps if used balanced (their outputs sit on 2.5V DC).
Jan Didden |
according to the datasheet the 96 dosent need any dc blocking on the output at all... |
|
|
| Spartacus |
| quote: | Originally posted by FLECOM
according to the datasheet the 96 dosent need any dc blocking on the output at all... |
It will need a cap if you use only one of the outputs from the chip e.g. the + OR the -, in singled ended fashion.
Dan |
|
|
|
