Hotrodding the UCD modules

[LTK]

Again, given a UDC 400AD with both CRD pads popoluated -

Is there any harm replacing the AD 8620 with an LM 4562 & leaving both CRD's in place? Will the unit run?

Also on the above UCD module, does anyone know the board id # for the surface mount opamp supply caps? I am thinking about adding bypass caps to those SMD's.

Thanks in advance.

I'd personally take them off, if you don't want to mess about straight away just try the LM4562's without them, see if you like these new op-amps first.
If you do like the LM4562's then its worth fitting the CRD's in the right place

I don't know the board numbers for the UCD400 but if you can show us a picture we can show you which caps are the op-amps decoupling

 

[Stevenacnj]

T.

These are not my modules, but do show the bottom of a UCD 400. Notice that the CRD's are not populated on these. I do have Sonicaps installed the same way:

An externally hosted image should be here but it was not working when we last tested it.

Hope you can see clearly enough to tell me which are the surface mount opamp supply caps.

Thanks in advance.

 

[LTK]

T.

These are not my modules, but do show the bottom of a UCD 400. Notice that the CRD's are not populated on these. I do have Sonicaps installed the same way:

An externally hosted image should be here but it was not working when we last tested it.

Hope you can see clearly enough to tell me which are the surface mount opamp supply caps.

Thanks in advance.

I'm afraid I can't see that picture very well, a close up would have been better
The decoupling caps are easy to find though for a dual op-amp, all you have to do is look at pin4 of the op-amp which is for the negative supply and pin8 which is the positive supply, closely follow the track from those pins and they go straight to the decoupling caps so follow from pin4 first, make a note of the cap it goes to and do the same for pin8.
If you fit any polar bypass caps be SURE to fit them the right way around Tants WILL explode if you fit them the wrong way.
Please check and double check

As example, here is a picture of the underside of a UCD180 I found, you can see the caps which go to pin4 and pin8 of the AD8620.
The UCD400 probably looks different but you can still follow the track from pin4 and pin8

 

[Stevenacnj]

t.

Thanks for that helpful info.

Question: you mentioned about the tantalum cap orientation. If installing tants, does the plus side of the cap go towards the opamp leg?

Thanks in advance.

 

[LTK]

It depends on the voltage rail you fit them, on the positive rail the positive pin goes to the op-amps pin8, for the negative rail the negative pin goes to pin4

Another thing with Tants, never run them close to their max voltage! the main reason I've not used them in my UCD's yet is because I don't really trust them, if running on 12v supplys I'd make sure their rated max voltage is much higher than that.
Again please be careful, you've been warned if any doubts leave em out.
I could be wrong but I think tants can go short circuit if they blow which wouldn't do your supplys much good

 

[anatech]

Hi Stevenacnj,
One thought. Tantalums have a tendancy to turn into a flaming blob, or short. Consider using a 1 uF ceramic for decoupling or a film cap. I bet a 1 uF film would do nicely but I'm lean towards a ceramic. Then polarity does not matter.

-Chris

 

[deandob]

T, Chris

Funny you should say that about tantalums, I had a tantalum blow in one of my amps yesterday, shorting the power supply. It actually caught on fire, and luckly as I was listening to it I could hear the amp developing a hum that got worse over a minute and I switched it off. I was running the tantalum a couple of volts below its rating. At least it was simple to replace.

regards,
Dean

 

[Robert F]

There are some extraordinary statements made in context of which rail to fit the CRD.

I acknowledge the reason advanced by most as to why a fitting a CRD to a certain rail on an opamp might make sense from a PSRR perspective.BUT I also explained many months ago on this forum good technical speculation on why it would be worth trying connecting a CRD to the negative rail on opamps which 'require' a CRD on the positive rail for the conventional reason.Turning the NPN devices on irrespective of PSRR may pay sonic dividends. Like Ric says maybe we need to listen here.

Oh BTW connecting the CRD to the wrong rail will not damage anything...

 

[Stevenacnj]

Thanks everyone for the most helpful info.

It would be easy for me to simply leave the two factory installed CRD's in place, when/if, I replace the AD8620 with the LM4562.



What exactly do you think will be the sound quality differences between the two opamps.


Thanks in advance.


PS: After reading the above, I will not be using tants to bypass anything. Will look for a film cap bypass solution.

 

[deandob]

Robert,

There are good reasons for both (better PSRR vs better linearity with only the PNP o/p transitor active), you will have to listen to determine what works for you. I have tried both ways and can't tell much of a difference (and I have a revealing system).

Regards,
Dean

 

[anatech]

Hi Dean,
They are easier to find and replace when they destroy themselves. BTW, I am starting to see some surface mount electrolytics do the short thing, no flaming blobs. Much harder to find. If you place a soldering pencil tip near the defective caps you will smell the electrolyte immediately. Nice eh? New product.


Hi Robert F,
I do have a concern when people start sticking devices in, possibly reversed. Unless you are a technician, please don't go playing around with these things.

Also, different op amps can be designed in ways that current sources do not help. The problem is that someone will always say there is a positive difference. Especially when they have spent money to have it done. It does take a brave person the say "I don't hear a difference" or "it sounds worse". No one wants to appear to have poor hearing or taste. That's the only thing that allows snake oil to exist.

-Chris

 

[Robert F]

Anatech,

You are a moderator on a DIY forum and you are advising people
(or is it just me?) not to do something as simple as installing a CRD unless they are a tech? PLEASE...

Yes there could be damage from installing the CRD the wrong way 'round but not by installing it properly to either rail.

What I object to is the fact that people assume that the ONLY way to install a CRD is to connect it to the rail where the op-amp has the worst PSRR. Oh that and also the assumption that listening is not important in making ones choices.

 

[anatech]

Hi Robert F,
I am a moderator, but first I am a member and have made the same comments from the day I first joined. Before that, and continuing now, I am an audio technician. I'd like to think I'm pretty good at what I do.

Over the years, I've seen more than my share of destroyed equipment done in the name of "mods" to improve performance. Sometimes it's as basic as pulling plated through hole connections out. Other times people have ruined equipment by replacing all the resistors with metal film. Or caps (got one now like that).

Yes there could be damage from installing the CRD the wrong way 'round but not by installing it properly to either rail.

Most certainly.

Oh that and also the assumption that listening is not important in making ones choices.

If you ever read my posts on this subject, you will notice that I am strongly for a balance between measurements and listening. Just listening leads to designs that may sound bad and have a tendency to fail electrically. If something is designed through measurements and good engineering only, they may sound bad, but will normally last. Only when good engineering practice, measurements and listening are all considered will you have a reliable, good sounding device. Unless you have a good understanding of what you are doing, some procedures will almost guarantee a damaged device. That is my point though long experience looking at repairs.

You may have some training in servicing or engineering (I don't know), but most people do not. The most dangerous person is one who thinks it's simple to swap these things. Sometimes it is, many times there are other factors to consider. Unless you have the many years of experience, you are foolish to hack away at a device unless it does not matter if it survives the procedure. I would have a greatly reduced problem with this if you were attempting these things on a breadboard or evaluation board.

What I am worried about is the individual who, after reading a post decides to "improve" the performance of their equipment and ends up destroying or causing damage beyond what they can fix or afford to pay for in repairs. I've seen that many times.

So what have I said that you have a problem with?

-Chris

 

[Bgt]

replaced the AD8620's on my UCD400's with LM 4562 chips, left the CRD on the minus rail, swapped it over to the + rail, no difference for me.

 

[Stevenacnj]

BGT

Couple of questions:

1. Did you try the LM4562 with both CRD's in place (on + & - rails)?

2. What was the difference in sound quality between the AD8620 & LM4562 on your UCD 400?

Thanks in advance.

 

[Bgt]

I tried them separately on the + and -, not together.
Difference for me is it sounds more direct, open(vocals). Not a big difference but present.

 

[Ric Schultz]

Don't think it would work well at all to try a current source on both rails....the one current source would just draw its current from the other current source. you would just be adding the sound of current sources to the output...not good....the whole idea of a current source is to force one side of the output stage of the op amp into Class A. So when you use it on one rail only the current has to be suppled through the op amp.

 

[LTK]

I tried them separately on the + and -, not together.
Difference for me is it sounds more direct, open(vocals). Not a big difference but present.

I thought the LM4562's didn't have that hollow midrange I found with AD8620.
All in all I found this op-amp to add less of a signature compared to the AD8620 or the original 5532, it didn't sound as impressive at first listen but after a few hours its the best I've tried

I've around a dozen different amps here and the UCD is still one of my favourites

 

[Stevenacnj]

Thanks for the help so far. My last question on this topic.

Again, I have UCD 400AD's with both crd's factory populated.

Question:

1. Are my current factory installed crd's on the + & - rail of the AD8620?

Thanks in advance.

 

[Bgt]

- rail