Head-Fi.org › Forums › Misc.-Category Forums › DIY (Do-It-Yourself) Discussions › Some LME49600 implementations
New Posts  All Forums:Forum Nav:

Some LME49600 implementations - Page 13

post #181 of 292

Well, with the caveat that it might not be working 100%, it sound as good or just a little bit better than my PPAv1 (OPA627, Glassman Buffers).

I could not hear any side effects if it is indeed oscillating (I can't think of any other reason for it to get so hot...).

No, I don't have an oscilloscope (I sold my Philips PM3260E years ago).


Originally Posted by Misterrogers View Post

Subjectively - how does it sound?
post #182 of 292
Thanks mate. Looking forward to building one.
post #183 of 292
Thread Starter 

Mine runs pretty hot too, so I'm concerned that it might be oscillating. I couldn't see anything with the 60MHz scope, and the RMAA results seemed reasonable, and it sounded OK and it didn't destroy itself so I turned my attention to the other board I had to build . Now I'm thinking break the loop either by lifting a pin or cutting a track and see if it runs cooler. With hindsight it would have been a good idea to have had a small res between the output of the 49990 and the 49600.


I've been a bit busy with my PC, it developed a software fault in the browser (Chrome) and I'm also messing with the Raspberry Pi which arrived a few days ago. I'll try this pin lifting experiment soon, but maybe not till tomorrow.


As far as the protection circuit is concerned, you didn't forget that the transistor footprint has a problem?


I'll try to get a simulation up to check the response time of the servo and see if can be tweaked a bit. It's always going to be a compromise between response time and overall frequency response.


post #184 of 292

I think my voltage divider for the protection circuit is the problem.

The voltage on one half was not what I expected...I have not had

time to troubleshoot it further.

I took the 10uF caps that formed the filter for the chopper op amp

back out but the settling time did not improve.


I mounted Q3 upside down and sideways...you can see it in the pictures.

post #185 of 292
Thread Starter 

I had another look with the scope, I couldn't see anything. 


Looking at the voltage drop across the 10R in the PSU filters, its about 850 mV, which means ~85mA (+/- 5%). The calculated current with the BW pin connected to Vee (as it is in my case) is 80mA, not taking the worst case values from the datasheet. I've got an NEC relay in there that draws 10mA more than the Axicom, I had about 20 that I bought to build a switched attenuator. 


The hottest things are the 49990s, but they're not finger-pull-away hot which probably means they're no hotter than 50 degrees C, and they take a while to get there, there's not a lot of cooling available to them.


It would be nice to get a 500M scope on there, I wonder if the university would give me 5 minutes.




Edit:- corrected slip of the pen from 10% to 5%

Edited by wakibaki - 7/3/12 at 11:36am
post #186 of 292

I'm glad mine behaves the same as yours...at least I built it right biggrin.gif

I have my LME49600 in wide band mode too.


I'll try and get a total current draw measurement when I can, might not

be until Thursday or Friday

post #187 of 292
Thread Starter 

After a long walk with the dogs in the rain, always a good time for thinking, I came back with a better analysis of the current requirement. The circuit is arranged so that the power requirement is symmetrical. Considering just the positive rail we get:-



Device Amps Number Total
LME49600 0.015 2 0.03
LME49990 0.009 2 0.018
LT1150 0.0008 2 0.0016
LM2901 0.0008 1 0.0008
LM317 resistors 0.008333333 2 0.016667
Relay 0.02 1 0.02


This corresponds very closely with what I am seeing.




Using the Axicom relay you should expect 10mA less.

Edited by wakibaki - 7/3/12 at 11:47am
post #188 of 292

I'm getting 97 mA with the amp cold.

Both rails the same.

post #189 of 292
Thread Starter 

That certainly seems a bit high. With the Axicom relay and OPA211 you'd expect ~82mA (87 - 10 + 5.6).


I'm fairly convinced mine isn't oscillating. I'll hit it with a square wave and capacitive loading and see how it likes that. It'll keep me from messing with this raspberry pi, which is driving me mad. 


You could try lifting pin 2 on one of the 49600s. If the current drops that's a pretty good indication that it's oscillating.



post #190 of 292
Thread Starter 

Scope traces 1kHz square wave, amp adjusted for a gain of 1:-


With scope load only (1M)...




With 16R...




With 16R || 1u...




With 1u





post #191 of 292

I added some extra bypassing right across the pins of the LMEs.

Didn't make any difference to the heat or current draw.

Still sounds great, though!



post #192 of 292
Thread Starter 

It's both puzzling and troubling that the amp is getting so hot though.


Another simple experiment you could try is to reduce the buffer BW by removing the 0R resistor.


Hopefully somebody else will complete one soon...



post #193 of 292

Just for reference, "The Wire" runs at +- 15 volts and the LME49600 in wide band.


I wonder how hot they run?

post #194 of 292

Here is some of my simulating:




For an input offset of 100 mV...

Point A = 2.1 volts

Point B = 100 mV

For a signal of 1 Vpp at 1kHz...

VF1 = 1.4849 Vrms

Gain = 2.1


Frequency response of the servo:

Swept from .001 Hz to 1 Hz...


Servo Responce.jpg


Does all that jive with your design math?

post #195 of 292
Originally Posted by Avro_Arrow View Post

Just for reference, "The Wire" runs at +- 15 volts and the LME49600 in wide band.

I wonder how hot they run?

Cool as a cucumber. redface.gif

Seriously, they run extremely cool. Cool enough that I never felt the need to measure it. The PS regulators are the warmest parts and even they are relatively cool.
New Posts  All Forums:Forum Nav:
  Return Home
Head-Fi.org › Forums › Misc.-Category Forums › DIY (Do-It-Yourself) Discussions › Some LME49600 implementations