| InfiniteGain |
Hi All,
I read on a few threads around this site that I could replace the cheap and nasty NJM2114D opamps with many other varieties including the OPA627 which would require mounting 2 on a Brown Dog as they are single opamps not dual.
I have a batch of APO627AU and the appropriate Brown Dogs, got them on the boards last night after some really tricky soldering. I of course, observed where PIN 1 was and ensure no shorts between the tiny pins of the opamp chips.
I had removed the NJM2114Ds from my Arcam Alpha 7 (under the name JRC2114D) and soldered in DIL sockets to make experimenting with opamps mcuh easier. Thing is, it is the OPA627 that everybody raves about that I really want to hear.
I put in the BrownDog mounted OPA627s and quickly got a burning smell. The right channel OPA627 pair burnt out. Removing it and turning up the volume until I can hear the transformer hum over my speakers indicated that I had a very low signla output from the other channel, but that it was working.
Either the gain is off compared to the NJM, which for recommended "drop in replacements" seems a bit off or (more likely) something else is taking the opamps way out of their safe operating limits. Probably the latter, but looking at the datasheets the tolerances, working voltages etc... all look fairly similar to the NJM2114D.
Has anybody ever done mods to the Arcam Alpha 7 CD player, or know offhand why this would occur?
Help greatly appreciated. :confused:
Infinitegain |
|
|
| pinkmouse |
| Most likely oscillation causing overheating. These chips are NOT drop in replacements, they are much faster. You need to look at extra rail decoupling at the minimum. |
|
|
| InfiniteGain |
Thanks. I wouldn't know without the schematic - I just trusted the info I saw.
After hunting around the web for the schematic I thought I would actually ask Arcam and now have the schematics for the Alpha 7,8,9 and associated DAC upgrades - in fact they sent me the full service manual for the series. :D |
|
|
| analog_sa |
Infinitegain
If i were you i'd double check that soldering job. Oscillation killing a 627 may be possible but not very likely. What is the supply voltage? The 627s have 4v less max supply than the opamps they replaced. |
|
|
| InfiniteGain |
Hmmm.... the schematic confuses me. 4 diagrams for one board. Physically I only find 2 NJM2114D opamps on the board, but the schematic has 6!! 4 in the diagram and 2 on their own with different labelling but with the voltage defined as +/-18V. This is on the limit of the OPA627s - the supply is regulated.
I attach the schemat for the DAC and output section for the more experienced to look at as I am a newbie to this. My father who is now retired, used to work for Raytheon and so is a bit of an electronic wiz and I can get him to look at it and explain it all closer to Christmas. In the meantime, pointers would be good as I think I will have to rip out quite alot of the ouptut stage from the player to use the OPA627s to good effect. |
|
|
| InfiniteGain |
| I just want to apologize if there is too much missing from the schematic, but I can only upload a max of 1000 pixels in any direction and any smaller and the component values are lost! :rolleyes: |
|
|
| AndrewT |
Hi,
I think you need to check the pin supply voltage to ensure you don't exceed the spec. Is 627 @+-18V absolute max or recommended max?
Definitely add pin to pin decoupling and if ground is nearby also add pin to ground decoupling. Using the very shortest routes you can manage, even using the back of the PCB if needed.
Does the 8 pin 627 need any correction when operating at low gain or adjustment pin loading?
By the way, big change in roll off philosophy from the earlier Delta 70.2 which used three pole Bessel whereas the Alpha uses (almost) 2 pole Butterworth. |
|
|
| poynton |
[QUOTE]Originally posted by InfiniteGain
[B]Hmmm.... the schematic confuses me. 4 diagrams for one board. Physically I only find 2 NJM2114D opamps on the board, but the schematic has 6!! 4 in the diagram and 2 on their own with different labelling but with the voltage defined as +/-18V. This is on the limit of the OPA627s - the supply is regulated.
Hi,
The schematic shows only 2 op amp chips. Don't forget they are dual op-amps - 2 per chip. The 2 blocks on the lower right are only to show power supply and are not opamps.
I too would recheck the soldering - it sounds as though there may be a short on the working channel. As for the other ...!!!!!!?
Andy |
|
|
| JonHarrison |
Hi 'Infinitegain',
If you email me the service manual I may be able to help with your problems.
Could you post a close up picture of the 627's mounted on the brown dog adapters? I'm curious how you have connected them up given that there must be a fair amount of wiring to match two single opamps to a dual op-amp pinout.
Regards,
Jon |
|
|
| InfiniteGain |
JonHarrison,
I could send you the full schematics but you have your preferences set not to be emailable!!! No need to worry about wiring two single op-amps together, that's taken care of by the Brown Dog PCB board care of Cimarron Technologies.
Unfortunately, the girlfriend has my digital camera at her work residence - but maybe in a few days I could post a close up of some soldered chips. For the PCB, look here:
http://cimarrontechnology.com/index...WPROD&ProdID=24
Thanks.
InfiniteGain |
|
|
| JonHarrison |
Hi 'InfiniteGain',
>I could send you the full schematics but you have your preferences set not to be emailable!
I chose that option when I joined but have since enabled it. However, given that I can't even email myself it obviously hasn't changed anything. I don't understand why this might be but I've emailed the administrator. I'd post my email address but I don't want to be spammed!
The browndog adapter looks just the job. If the connection of the op-amps is ok then I'm afraid nothing obvious springs to mind at the moment. I was wondering if current from the supply is being driven into the output of the op-amps ! Perhaps they are having problems with the DC blocking capacitors (10uF C4 and C104).
I'll post another message if I have any other ideas or my email setting gets sorted out.
Cheers,
Jon |
|
|
| InfiniteGain |
It transpires that it wasn't on this website that I saw the OPA627s as drop in replacements. It was here:
http://www.soundodyssey.com/AudioParts/OpAmps.html
...where some opamps are listed (including mine) and then 3 replacement upgrades touted, including the two single OPA627 chips mounted on the Brown Dog adaptors. Specifically on the OPA627 link on the same site, which can be seen here:
http://www.soundodyssey.com/AudioParts/OPA627.html
.. it mentions:| quote: | | At Sound Odyssey we offer the OPA627 chips mounted on an adapter module that is a drop in replacement of the existing Dual OP-Amp chip. |
.. and this is why I believed these were drop in replacements. I think the Sound Odyssey website made a sweeping generalization in that statement which could result in some damaged CD players.
However, to be fair to Sound Odyssey, I am aware that there are many DIY audio websites selling the same product without giving cautionary advice (except about electrical work in general).
As my first outing into DIY audio, I trust in walking a path others have walked before me and follow their advice/direction - it was unexpected for a "drop in" replacement product to not be a "drop in" replacement. I chose this upgrade as it was theoretically a no-brainer and required litle electrical engineering knowledge.
InfiniteGain |
|
|
| JonHarrison |
Hi InfiniteGain,
I've got my email settings sorted. Send me a private message with your email address and I can help you further.
Jon |
|
|
| guytou |
| Hi infinite gain , did you solve your problem ?anyway you don't need opamps with this chip , because the PCM 1710 already has an opamp AND an analog filter on the chip .What you could do (say what you have to do )would be to hook a good plastic cap of 10 microfarads at vout R and v out L , and go to your RCA.this way you get a clean sound and even check your dac functioning or not . |
|
|
| poynton |
| It's interesting to note that the BurrBrown datasheet shows a dc blocking capacitor prior to the low pass filter - missing on the Arcam schematic. The Arcam schematic also seems to show an 'option' for audio output immediately after the dac. I would assume that the 'missing' capacitor is an extra rolloff filter to supplement the internal filter, when the extra stages are not fitted. |
|
|
| InfiniteGain |
I can drive the outputs directly?
Is this wise? Surely Arcam put the analog output stage there to filter some noise or bandwidth which had an adverse effect, otherwise why pay the extra cost of the added components?
As for the DC blocking capacitor, I agree on this. I also think the resistance in the feedback line in the Arcam may be too high, as well as the resistance to the ground line. Thoughts please? The OPA627 datasheet suggests something lower where they provide a butterworth configuration.
Thanks,
InfiniteGain |
|
|
| InfiniteGain |
Just to keep any followers of this thread in the loop....
JonHarrison has sent me some ideas I will be looking at over the weekend when I have time. I will probably then have to obtain some components or possibly even make a separate daughterboard with a new o/p stage. This might take a little while but....
I will keep you posted. |
|
|
| Jay |
| quote: | Originally posted by InfiniteGain
I can drive the outputs directly? Is this wise? Surely Arcam put the analog output stage there to filter some noise or bandwidth which had an adverse effect, otherwise why pay the extra cost of the added components?
|
Yes, I think that's wise. At least to check if the DAC is alrite (as guytou has mentioned). While doing so you can listen or compare the sound result with or without the Opamp LPF.
The reason for extra component is probably to cater for uncertainties in consumer's setups. You know what pre-amp this opamp is driving, and you can fix something so that you can bypass this opamp
Why bypass? Well, for short signal path. You should have a better (upgrade!) power supply in CDP. With less noise from upgraded power supply, you don't have too many reasons to install extra components. At least you can low-pass at very high frequency.
It is also difficult to know at what frequency you should construct your LPF, or with what degree. Listening test will tell you. And that's why you have to try bypassing the op-amp. May be you need RC filter (it's the missing cap poynton has mentioned).
Arcam also doesn't know if you need that mute circuit or not. With the mute circuit, I guess a little gain is required from the op-amp. If you don't need the mute circuit, you don't need extra gain, and maybe you don't need the whole op-amp. Who knows? |
|
|
| Elso Kwak |
| quote: | Originally posted by Jay
Arcam also doesn't know if you need that mute circuit or not. With the mute circuit, I guess a little gain is required from the op-amp. If you don't need the mute circuit, you don't need extra gain, and maybe you don't need the whole op-amp. Who knows? |
confused?:rolleyes: |
|
|
| Jay |
| yeah elso. confused. what do you think? |
|
|
| Jay |
| hello elso, are you confused with the extra gain issue? Well don't be because I'm not confused :D |
|
|
| guytou |
That's it , Jay , I could'nt say it so clear .
Please , Mr. Elso , could you explain to us the exact purpose of these output stages , since this DAC chip already complies itself with the Red Book (2 Volt out/about 1k. Z) ?
Many thanks in advance . |
|
|
| InfiniteGain |
| quote: | | The reason for extra component is probably to cater for uncertainties in consumer's setups. You know what pre-amp this opamp is driving, and you can fix something so that you can bypass this opamp |
Hmmm.... Well I am confused. I interpret the last few postings as being advised to use the DAC directly because the LPF is just for making the CD player more compatible with pre-amps as the DAC might not be able to drive such a variety.
I would think using the DAC directly unwise if it reduces the CD players ability to drive a wide variety of pre-amps - I do not use Arcam power amps (think Chord electronics here please!!!). If it cannot drive a pre-amp properly, I could end with all sorts of problems reducing sound quality or worse: maybe some type of damaging oscillation. Also note, bypassing the filters or modifying my player to match my amp is not good as my amp is liable to change and I want to produce something that will sound better than the stock unit regardless of what I hook it up to.
I also think it odd that I am being given this advice and then the last post asked for Elso to clarify the exact purpose of the LPFs. If you are not sure, how can you advise bypassing them as the right solution?
I do not think Arcam are daft. Two things are important from their perspective - the first is to make the CD player compatible with a wide range of pre-amps, which I think is a design consideration which should remain. The second is that for all we know the DAC will ouptut out of band noise which is harmonic with in-band (audible) noise and if the LPF is removed it might really mess things up sonically. I need to find out before progressing - but the DAC specs themseleves are not great.
From my discussions with Jon Harrision, I am more likely to either use the DAC upgrade socket of the CD player to build a dedicated DAC and construct ideal filters for use with the OPA627s, or just build something to handle the raw data stream directly from the transport. The Arcam is on a PCB where I cannot re-route the circuit path to all situations people may tout as a the solution, but the actual solution may need it. Therefore it is probably better to re-direct the signal currently going to the DAC to a bespoke PCB and take control in my own design from that point.
Prior to building a DAC the best I can hope for is removing/changing of components to get the opamps to a working state. Jon suggests C5 and C105 be removed, which I agree with and the filter cap that is hinted at by Arcam is nowhere near a suitable value. On this I agree with Jon and I will probably give this a go on Sunday.
InfiniteGain |
|
|
| Elso Kwak |
you guys have ever heard of Intermodulation Distortion, aliasing distortion or whatever?? LOL :rolleyes: :clown:
(look at her <<!) |
|
|
| InfiniteGain |
| quote: | | ever heard of Intermodulation Distortion, aliasing distortion or whatever?? |
My point exactly, just more eloquently put Elso!!! :D
I suppose I sounded a little flustered above but I am not :cool: |
|
|
| guytou |
"you guys have ever heard of Intermodulation Distortion, aliasing distortion or whatever"
Maybe anything more consistent/pertinent in the context ?????
infinitegain : you seem rather involved in the DIY thing , but it appears you experimented a desillusion , so I understand your position . But do you want to stop or go on ?
My proposition is a "go on" proposition.But if you don't feel , you have to let back sometime(excuse my poor english):nobody has the pure truth , but yourself .
If it helps , I had to modify an Alpha 7 : the sound was so annoying with this kind of "ready to use"chips , that I tapped a TDA 1543 directly from the decoder in EIAJ . Better ? Worse ? Same thing ? the owner is happy . |
|
|
| guytou |
| 1bit/delta/sigma are convenient for brands , not for customers . |
|
|
| InfiniteGain |
Hi Guytou,
RE:| quote: | | infinitegain : you seem rather involved in the DIY thing , but it appears you experimented a desillusion , so I understand your position . But do you want to stop or go on ? |
Not quite - i have read and understand a lot of theory (and seeing results from other modders of other electronic things, not audio) but I have little in the way of practice - I do not have the 'schematic' intuition that comes from that practice -> being able to quickly identify potential problems. The example in this case, is that I understand (or can confidently reason) why Arcam use the filters, but I have not got the experience to know how to adjust them for another chip. This is something I hope to get over time. An analogy would be that there are experts in Formula 1 who understand alot of the science and reasoning for certain things, but that doesn't make them a competent mechanic.
I think in the second half of what I quoted you saying you are trying to say I was misled by the website who sold the opamps. I'd agree with that.
I was also trying to say that I do not want to tamper with the player in a way where I may risk damage to the player. I would rather design something separate (PC simulation) and give that a try, or confidently change/remove existing components. One way or another, I will mod the player. I do have a backup in the form of a nice Meridian 506.20 which is currently my main player.
Currently, I do not have any schematic/simulation software and given my experience I was hoping somebody could just look at the circuit and confidently point out what is preventing the opamps from working. I think Jon Harrison did this via email, but I will be trying out his suggestions on Sunday.
Hopefully a good software suite will allow me to experiment safely without risk of changing things on the actual circuit board and watching components burn out (as the OPAs did!). Problem is, I run Linux and have not found a decent package that works well on it yet.
Thanks,
InfiniteGain |
|
|
| Jay |
| quote: | Originally posted by InfiniteGain
I also think it odd that I am being given this advice and then the last post asked for Elso to clarify the exact purpose of the LPFs. If
you are not sure, how can you advise bypassing them as the right solution?
|
No, I didn't ask elso about the purpose of the LPF. I didn't say that
bypassing them is the right solution either. I just advise to try and
listen. Because it is possible that you like it without the op-amp.
And it will take you less than a minute to try it. No extra component.
I don't have technical approach to audio. I know electronics only a
little. My approach is by doing it and listening. I use my limited
electronic knowledge here just to know that you can safely bypass the
op-amp, and that you have the possibility to have a preferable sound.
Well, you said it that compatibility with various amps is important
to you. Then obviously we have different perspective. I mod my stuffs
every now and then :)
| quote: | Originally posted by InfiniteGain
I do not think Arcam are daft. Two things are important from their perspective - the first is to make the CD player compatible with a wide
range of pre-amps, which I think is a design consideration which should remain. The second is that for all we know the DAC will ouptut out
of band noise which is harmonic with in-band (audible) noise and if the LPF is removed it might really mess things up sonically.
|
That's it. You could control the level of those noise and eat medicine
only when needed, and with the right dosage. If you mod a lot of CDPs
and review the result, you must know what you can get.
| quote: | Originally posted by InfiniteGain
but the DAC specs themseleves are not great.
|
The problem is you're not bypassing the DAC. Not even the built-in
output amplifier.
Oh, I know that this DAC is not special. As far as I can remember, it
is Arcam, right? Hmmm... I think Arcam uses Philips DACs??? But from
my experience with BB oversampling DAC, I can tweak them to sound
better (to my ears) than the non-os TDA DACs. And unfortunately,
this is by trying to remove the unnecessary cures to non-existent
deseases.
| quote: | Originally posted by InfiniteGain
Prior to building a DAC the best I can hope for is removing/changing of components to get the opamps to a working state. Jon suggests C5 and
C105 be removed |
I don't know how to put the OPA627 to a working state by removing C5.
IME exchanging opamp, it is only the supply voltage that matter.
Well I know about the speed of the 627, but never found in practice.
Especially you're upgrading to 627, not from 627.
How about C3/C103? :) Oh, for this kind of application, I prefer
a paper type capacitor. What I mean here is the "silver-metallic foil"
type. It's capable of producing the most enjoyable sound.
To tell you the truth, OPA627 is not special soundwise hehehe ;) |
|
|
| guytou |
Yes , I think I understand . Maybe I gave you the "go on/lazy " proposition , it's good to try to fix the problem .
Iin case of , and for going at the end of the idea , you would be led to resolder a DIP-8 socket and see what happens with a 5532 .
Otherwise (and to reply at your aforementioned question ), I can say that a PCM17...(don't remember)tapped directly as I said works NO WORSE than in stock state .
success , |
|
|
| InfiniteGain |
Hi Jay,
Thanks for the comments. I won't write too much as I have written War and Peace as it is ;) However, apologies if I took something you said out of context.
I understand what you mean - I might try the DAC directly. BTW: it's not a TDA type its a PCM1710U - 11 year old design and still going!! But it's not a good DAC.
| quote: | | The problem is you're not bypassing the DAC. Not even the built-in output amplifier. |
Well, I was thinking of using the offboard DAC adapter designed for the Alpha 7 to Alpha 8 upgrade. I could use my own DAC and o/p stage from there. I was thinking of the PCM1792 (two of them) in mono-mode supporting XLR outputs and 132db SNR as on the datasheet.
You may not like the sound of the OPA627s but I only wanted to try them given so many people rave about them and even Douglas Self's website says it is a good opamp for audio - and given he hates subjectivism it means it measures well too. I will say the descriptions of "warm and valve-like" does not sit well with me, but I find that term is often confused with rich and 3D-like, which I do like. I just want a transparent, detailed but very 3-D sounding opamp which improves the player.
Guytou:| quote: | | you would be led to resolder a DIP-8 socket and see what happens with a 5532 |
I had already put the DIP-8 sockets in. 5532 are pretty cheap, I could give them a go. They are virtually guaranteed to be compatible.
InfiniteGain |
|
|
| carlosfm |
| quote: | Originally posted by InfiniteGain
I put in the BrownDog mounted OPA627s and quickly got a burning smell. |
So fast? :D
Something went wrong, but I must tell you: if the PCB of the 7 is similar to the 7SE, it's one of the worse layouts I've seen on any CDP, at whatever price, down the scale.
For instance, on the analog stage, the only (and I mean the only) PSU decoupling caps are some 10cm from one channel, ~5cm from the other channel.
With such a miserable PSU decoupling only the likes of JRC/NJM2114, NE5532, etc. work. And they just 'work'.
The OPA627 on sockets and adapters will oscillate and kick like a mad cow.
I'm not sure if it's enough to smoke and burn them, but they will probably get so hot that they self destroy.
If I remember btw the player had +/-15V for the analog stage, regulated by LM317/337 regs, without Cadj. caps (!).
So the problem was not excessively high PSU voltage.
Conclusion: without some decent (re)work, the 7(SE) is a weapon. :D |
|
|
| InfiniteGain |
Hi Carlos,
Yep, that slow smoking/burning (no flame) sounds like it.
Eveything you say about the layout is true.
Well I am up and about to hop on a plane - to be back Sunday to try a few things, but I tested a couple of crucial PSU resistors the the other day - two of them were supposed to be 10 ohm devices, but they look browned and are only outputting about 5.5 ohms, so they need replacing and I will have to order some.
InfiniteGain |
|
|
|
