 
|
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
|
|
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
98.221.125.138
I sure hope you can help me here. I am running into oscillation issues with my Dahlquist DQ LP1 low pass crossover. Apparently the OPA2601 op amps I recently put in are not compensated. Unfortunately I love the way these op amps sound so I'd like to fix the problem.
There are 5 of these op amps in the unit. I've read about using small capacitors to aleviate this issue, but I need to check with experts to ensure I get it right. The articles say one should put the capactitor on each of the +15 volt and -15volt power rails to ground.
1) wanted to check my interpritation (it's not one cap between these 2 pins right? Gosh that would be easier)
2) I have a large supply of capacitors on my work bench. There are formulas to calculate the proper value but honestely they are beyond my abilities. Does using a 0.01uf cap sound like a good value?
3) Do I really have to solder in 10 capactitors (2 for each opamp) to tame the oscillation? Could I get away with using one on each leg of the power trace on the circuit board, or 3 inches away at the power supply? (I am trying to simplify things and reduce the number of capacitors needed space can be tight)
Thank you for looking and I appreciate any thoughts you can share.
Let me share something back for any DQ LP1 owners: I found out that modifying one of the resistors allows you to modify gain. I realized, after years of trying to figure out why I don't have good bass, that the gain on the LP1 couldn't match the gain of the amps I use. Spoke someone who worked at Dahlquist with these units and was taught that changing this resistor will increase gain by 3db. Works wonderfully. Let me know if I can help. The reason I am not listing it here is that I am not next to the schematic and I don't remember the resistor number for sure, not trying to hide something. If anyone is interested, just let me know.
Thanks,
John
Follow Ups:
When I replaced the NE5532 in my Revox B226 with some LM4562, I really had to try different decoupling schemes before it sounded good. The B226 doesn't have ANY local decoupling on the opamps, so it was like starting from scratch... thank god there was some ground planes just next to them.
I ended up with 0,1uF MLCC from V- to ground, and another 0,1uF MLCC between V+ and V-.
A cap on each supply rail didn't seem to work as good in this case; experimentation with bigger EL caps paralled with the 0,1uF made the sound slow and flat and was immediately unsoldered.
I don't know about OPA2601, perhaps it's less fussy, but I'm afraid "general ruel" doesn't really apply to all opamps...
This was posted in the speaker forum. Do a search with this title "Dahlquist DQ-LP1 troubleshooting" in the speaker forum.
This guy used the OPA2604's also.
Thanks Cougar.
First thing you need to do is replace C16 and C16, and then put a 0.1µF in parallel with each.
Edits: 06/17/16 06/20/16
Sounds like a great idea. Thanks for seeing that one, it is obvious now that you've pointed it out.
John
I was trying to look it up and it doesn't come up. I have heard of the OPA2604 but not the OPA2601.
Here is a copy of the schematic. I should note that I followed many others' advice on "upgrades" to this unit - that's how I chose this op amp. (IOW, I didn't just pick any old op amp)
While there are numerous "upgrades" to this circuit, none are any different than what Frank Van Alistine recommended in his article decades ago, and that many others have implemented. Upgrade capactitors to better brand, replace resistors, I have put in sockets to make op amp changes easier.
BTW, if oscillation occurs, would the offending op amp get hot to the touch? I've seen strange things happen to this circuit with cold solder joints, I've often wondered if I may have overstressed one of the op amp pins inserting and pulling several times.
Thank you each for your help, I sure do appreciate it. John
I didn't think of modifying the subject line when I posted the schematic. This post is just to let you know the schematic has been posted. John
Did you use a scope to see the oscillation or or ringing or does it feel really hot to the touch? Also brightness in the sound could be another sign of oscillation.The reason I ask is that your question asking does an opamp get hot when oscillating.
So from I understand from your reply these exact opamp have been used in this exact circuit (crossover) without any problems? the only thing I see that I would change or try is changing the 47 ohm resistors. I would change them to 100 ohm to see if it stops the oscillation (that's if it's really oscillating). This would be quick, simple, and cheap to do.
Also, I have seen sockets cause opamps to oscillate due to the extra capacitance on the socket. Once the socket was removed, the opamp stopped oscillating. This was in a high speed circuit and "shouldn't" cause this in this circuit but you know how electronics go! anything can happen. Wouldn't hurt to try removing and see if that is the cause if all else fails.
Hope this helps!
Edits: 06/17/16
I get a condition like feedback but the volume does not increase. Just goes distorted with a pulsating beat or literally oscillation.
Since it does not literally feedback and get louder no fuses blow no speakers are damaged.the best way to describe it is an oscillating pulsating distortion.
You asked a good question. I hope I'm interpreting this correctly. Since it is intermittent I do not have a scope handy. I wondered if a faulty opamps might get hot in hope of nailing down the culprit.
It is not hard to turn one around thinking you have it in right and I have seen this thru the years on many circuits that use OP amps..The LP1 is a fine crossover and I still have mine but it is very simple in nature and even if you get them in backwards,you will not blow the unit..It isn't unusual for different manufacturers to orient their ICs in a different way to make you think you are putting them in correctly..Look at them very carefully.If that doesn't do it,you may have to compensate for the new ICs with a cap change or two.
"For every complex problem there is an answer that is clear, simple, and wrong" H. L. Mencken
Edits: 06/18/16
output pin of the opamp. You may have to put a 100 ohm resistor on the output of the opamp to isolate it from the circuit that comes after directly connected to that opamp's output pin.Sometimes you can't just drop in a different opamp and it works perfectly.
Try the decoupling caps on the +/- pins .1uf-.01uf. if that doesn't work then add the 100 ohm resistor. Sometimes a compensation cap is need but that should be the last resort.
It would be nice to see the schematic to help us. Is the opamp the same gain as the original? also can it drive high capacitance load? Also haveyou done anything else to the surrounding circuit that was designed for the original opamp?
Edits: 06/17/16
I find it **really** hard to believe that any competent designer would lay out a circuit board with no decoupling caps on the op amps.
OP, can you post a schematic and maybe more important, a photo of the board? More information will help us help you more effectively.
That's power supply bypass (a good idea), not compensation.
Compensation is deliberately reducing the -3dB point of the opamp to make the opamp stable (in your case, to eliminate oscillation). Your opamp will have terminals where the compensation capacitor goes. I don't know your circuit, and I don't have enough engineering know-how to tell you what value to use.
Recommend getting a decent book on electronics. Any modern book should explain compensation and how to select a value for your circuit.
Yes, you need 10 caps. One to each power rail to ground for each opamp. Keep in mind there may be other reasons for the oscillations. Changing opamps isnt always a good idea if you dont know the possible tradeoffs involved.
Thanks. Good point about other reasons. I'm going to try something and replace all 5 opamps. Maybe all the in out plugging and unplugging for listening tests damaged one of them. Maybe I'll get lucky and all will be OK with out having to bypass with 10 caps. I appreciate your help.
Post a Followup:
| FAQ |
Post a Message! |
Forgot Password? |
|
||||||||||||||
|
||||||||||||||
This post is made possible by the generous support of people like you and our sponsors:
Top of Page ]
[ Contact Us ]
[ Support/Wish List ]
[ Copyright © 2025, Audio Asylum, all rights reserved ]
