Just looking for some user experience on bypassing the power pins on the LM6171
I believe the CarlosFM method is 22-100uF electrolytic with 0.1uF ceramic as close as possible (which I've used happily with other opamps - thanks Carlos!). Datasheet shows a configuration with a 2.2uF tantalum with 0.01uF for the LM6171.
What has worked well for you with this chip?
I believe the CarlosFM method is 22-100uF electrolytic with 0.1uF ceramic as close as possible (which I've used happily with other opamps - thanks Carlos!). Datasheet shows a configuration with a 2.2uF tantalum with 0.01uF for the LM6171.
What has worked well for you with this chip?
Thanks all! Good to know it's not hypersensitive - you just need to treat it properly. 
I'll probably start with the method I know, and experiment from there.
I was going to go without socket (which I don't like to do, but wanted to avoid the extra capacitance). Maybe I'll give it a try with some good machined sockets first, keep the caps as tight as possible, and see if I can keep it out of oscillation.
And I definitely won't forget the 100-330nF from V+ to V-
I'll probably start with the method I know, and experiment from there.
I was going to go without socket (which I don't like to do, but wanted to avoid the extra capacitance). Maybe I'll give it a try with some good machined sockets first, keep the caps as tight as possible, and see if I can keep it out of oscillation.
And I definitely won't forget the 100-330nF from V+ to V-
old thread but can you elaborate? i don't understand BG N. do you mean 4.7 resistor across it? i was told something about a 4.7k resistor bypassing it battery but was across the in and out i think. Are you using this bypass method as a way of doing a line in?
You can't get zero offset. Even the LMP2021, which I designed, by the way, has -400 nV of systematic offset.
According to the data sheet (http://www.ti.com/lit/ds/symlink/lm6171.pdf), the LM6171 has 8 mV worst case input offset voltage. If that isn't good enough for your application, you need to do something about it.
Tom
I am impressed. i was told by a friend it can be used discrete as an in and planning on using it but haven't had the time to pay with it, I have about 100 of em and used them for different things. i remember in the white pages it said something about when discrete using tantanum parts?
I don't understand your question. I'd love to help, but the above sentence makes no sense to me at all.
If you're looking to use the LM6171 in a circuit, I suggest reading the LM6171 data sheet - in particular the application section towards the back of the data sheet. It'll tell you all the tips and tricks you need to know to use the product successfully.
Tom
yes sir, i did read them in detail several times and know the chip fairly well but i am new to audio amp designs and some of the lingo and terminology I have to learn and other things I need to understand. In the white pages it says battery can be bypassed by 1uf on pos before and after and .1uf on neg before and after the chip (if i am not mistaken). At this point is it basically the same a fet input array where i then proceed to negative offset and vas section? Also one of the engineers who is helping me with this project at one point said something (if i am not mistaken) about bootstrapping a 4.7 resistor across the ins and out and it would also allow you to bypass battery to use it discrete and be biased A was he incorrect and meant capacitor.
What would the outcome be if i used a higher value bipolar cap?
Sorry i have so many questions, i read as much as i can but don't have a formal test rig yet where i can see what happens when you do certain things and just picking up the terminology. My career is in finance and understand numbers and do math with the best of them and I can pick things up quickly but a lot of the white pages were written for EE so sometimes i end up reading more about what a parameter is. I have been accused of overthinking things in the past, i just like total comprehension.
Would you mind answering these few questions for me?
What would the outcome be if i used a higher value bipolar cap?
Sorry i have so many questions, i read as much as i can but don't have a formal test rig yet where i can see what happens when you do certain things and just picking up the terminology. My career is in finance and understand numbers and do math with the best of them and I can pick things up quickly but a lot of the white pages were written for EE so sometimes i end up reading more about what a parameter is. I have been accused of overthinking things in the past, i just like total comprehension.
Would you mind answering these few questions for me?
I still don't get your question. "Bootstrap a 4.7 resistor"? 4.7 Ω? 4.7 kΩ? Bootstrap where?Why? Can you draw a schematic of what you intend to do?
There's no special sauce required to use the LM6171. It's a high-speed op-amp (200 MHz if I recall correctly), so some thought needs to be put into the decoupling to ensure a low supply impedance beyond 200 MHz.
Tom
There's no special sauce required to use the LM6171. It's a high-speed op-amp (200 MHz if I recall correctly), so some thought needs to be put into the decoupling to ensure a low supply impedance beyond 200 MHz.
Tom
as Tom hinted above, page 18 of the datasheet has info that should be helpful.
http://www.ti.com/general/docs/lit/getliterature.tsp?genericPartNumber=LM6171&fileType=pdf
be warned again, this is a high speed part, so you have to be careful about bypassing AND other things to have the most success with high speed parts.
so study the datasheet.
are you sure this is the right opamp for your application and experience?
good luck,
mlloyd1
http://www.ti.com/general/docs/lit/getliterature.tsp?genericPartNumber=LM6171&fileType=pdf
be warned again, this is a high speed part, so you have to be careful about bypassing AND other things to have the most success with high speed parts.
so study the datasheet.
are you sure this is the right opamp for your application and experience?
good luck,
mlloyd1
Last edited:
I believe he said a 4.7k ohm resistor across the vin and vout and was in one of his moods where he said it's in my notes when he picked up and left the country. He's back temporarily and hoping he finds them but the guys like a mad scientist - great at what he does but all over the place. The idea was to use the lm as an input stage and run it discrete with an a bias. I don't have a proper test rig and need to figure out what to buy still. I guess I can measure the MHz of a preamp out and work into it. The manual however said something totally different and this guy does things unconventionally so that's where I get messed up.
Experience? Hahahah I always get in way over my head but work **** out eventually. I could do anything from start a fire get zapped, shoot a nail thru my hand spinltet metal and get it stuck in my eyes but I'll get it eventually. Too stupid to quit. I have literally no experience with using ICs in an amp. I know what I want my output devices to be. I know what I normally driver them with. If I am trying to run it discrete with no power other than what's coming from the line in and what other parts should I be bypassing?
Rather than starting with "must use this IC", I suggest figuring out what you want to build and what the specs of that project should be. Then select parts that fit.
If you're building a preamp, something like an LME49710 or OPA1611 would usually be better choices due to their low THD, high loop gain, and high PSRR.
Tom
If you're building a preamp, something like an LME49710 or OPA1611 would usually be better choices due to their low THD, high loop gain, and high PSRR.
Tom
Thanks. No this isn't for preamps. I did order some of the lmes for more power and seemingly easier design with an IC front end and driver stage but being and end of life cycle chip and a new line it's more for educational purposes. The LM6171 was picked because of the high sensitivity of each driver. The lowest sensitivity is 89 on the subs I chose but they may be changed out because the speaker supplier hooks me up and is involved and is strongly recommending another one over the labs - the new ones will be a lot higher. My 10s are 98, 5.25 are 99 and tweeters are 117 I think (or 107) I have to check I swear i have a warehouse of drivers here. I rather have enough power to push each driver to max based on the various applications at least for this design. I know sensitive speakers need faster amps to do the drivers justic hence my preoccupation on speed. All drivers are actively crossed requiring a mono block per driver which I intend to keep inside the speaker as they are already large enough. I have several ways to cool the amps. So heat will not be an issue.
negative, not drive a speaker or even an output device or driver for that matter. Operate as a line in on a power amp discretely without having to use a step down, allowing you to then go to negative offset, vas then output transistors, line a normal amp.
i am currently doing everything discrete, and not only from a build time with adding a dc servo plus the addition of the LM will make a cleaner input state and be faster then the mspsa42 i am currently using - a dc servo circuit to compensate for negative offset would be nice and speed things up and i hate pots after the new replace front end i would plan on using the the bc 559 and 549 array solid array and then drivers and then LFETs.
I am sorry i don't know all the industry jargon yet to fully articulate myself, i hope this makes it easier to understand.
i am currently doing everything discrete, and not only from a build time with adding a dc servo plus the addition of the LM will make a cleaner input state and be faster then the mspsa42 i am currently using - a dc servo circuit to compensate for negative offset would be nice and speed things up and i hate pots after the new replace front end i would plan on using the the bc 559 and 549 array solid array and then drivers and then LFETs.
I am sorry i don't know all the industry jargon yet to fully articulate myself, i hope this makes it easier to understand.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.