Owen,
I have a very strange problem.
I am testing "the wire" connected to my buffalo + legato, as the D1 is not ready yet. Now if I connect only one channel, everything is OK. When I connect both I got a DC offset on the input of the buffer and on the output of the headphone amp. There is no dc offset with only one channel or with the circuit disconnected. I cannot figure out what's wrong.
I get around 40 mV on output, always on the right channel.
Any idea ? Any test I can do ? i also tried to power the legato and the buffer with the same power supply, but it did not do any difference.
D.
I have a very strange problem.
I am testing "the wire" connected to my buffalo + legato, as the D1 is not ready yet. Now if I connect only one channel, everything is OK. When I connect both I got a DC offset on the input of the buffer and on the output of the headphone amp. There is no dc offset with only one channel or with the circuit disconnected. I cannot figure out what's wrong.
I get around 40 mV on output, always on the right channel.
Any idea ? Any test I can do ? i also tried to power the legato and the buffer with the same power supply, but it did not do any difference.
D.
do you get the same result if you try the second side by itself ie no offset? have you tried both channels individually? are you sure both channels of legato have no offset at the output? and by that you would need to check with reference to the other channels ground and across the poles. have you biased the input of legato to avcc/2?
i hope you get a response, i havent heard boo from him in over a week now i think about it. I couldnt wait any longer and moved forward, guess hes busy with life stuff. everything is perfect with the new amp to amplify the 1khz sine wave; (a cmoy on steroids with shunt regulated supply) except now i'm finding none of my meters are accurate enough right down in the very low mVrms AC voltage measurements. all are fine for everything else i have ever had to do, but i have never had to measure fractions of a millivolt rms ac down very beginning of the range, as even 1mVrms out of <18 makes quite a profound difference t the tc results. so i'm organizing a loan of another better meter, because not a chance am i buying yet another meter when the 3 i have already, 2 bought for this matching and no other reason are perfectly fine for everything else i have ever had to do. i was planning on getting a nice used benchtop meter in a month or so for final tweaking of the f5x, but i'll take my time with that, keeping an eye out for a deal. vds matching is done and settig the operating point for each fet is no problem, but measuring vgs in low mVrms is too inaccurate
sorry for the ot guys.
i hope you get a response, i havent heard boo from him in over a week now i think about it. I couldnt wait any longer and moved forward, guess hes busy with life stuff. everything is perfect with the new amp to amplify the 1khz sine wave; (a cmoy on steroids with shunt regulated supply) except now i'm finding none of my meters are accurate enough right down in the very low mVrms AC voltage measurements. all are fine for everything else i have ever had to do, but i have never had to measure fractions of a millivolt rms ac down very beginning of the range, as even 1mVrms out of <18 makes quite a profound difference t the tc results. so i'm organizing a loan of another better meter, because not a chance am i buying yet another meter when the 3 i have already, 2 bought for this matching and no other reason are perfectly fine for everything else i have ever had to do. i was planning on getting a nice used benchtop meter in a month or so for final tweaking of the f5x, but i'll take my time with that, keeping an eye out for a deal. vds matching is done and settig the operating point for each fet is no problem, but measuring vgs in low mVrms is too inaccurate
sorry for the ot guys.
Last edited:
Yep,
One side no problem, and no offset on output or input. Which side does not matter.
After one side is connected, I connect the second one. Note that there is no offset on the second side before I connect it.
Connect the ground: no problem
Connect the + bias goes up to 20 mV on one side only, connect the - bias goes up to 40 mV.
The Legato is biased correctly. There is the on-board buffer also on.
I played with the grounding, as I thought the problem was there, but I could not find a configuration that fix it
Thanks,
Davide
One side no problem, and no offset on output or input. Which side does not matter.
After one side is connected, I connect the second one. Note that there is no offset on the second side before I connect it.
Connect the ground: no problem
Connect the + bias goes up to 20 mV on one side only, connect the - bias goes up to 40 mV.
The Legato is biased correctly. There is the on-board buffer also on.
I played with the grounding, as I thought the problem was there, but I could not find a configuration that fix it
Thanks,
Davide
yeah that was another thought of mine, that it could be a ground loop, are you able to listen to it with some crappy headphones? ibuds or something. i have heard of these buffers having offset issues in some builds, but it seems you are getting offset at the input also? the offset has to be coming from before the opamps somewhere, because these things have exceedingly low input and output offset. its not a large cct, have you tried going through point by point to find where the first appearance of offset is?
tis a strange problem i agree
tis a strange problem i agree
Last edited:
Hi,
I'd live to have a pair of the PCB.
Please tell me how much it cost.
Of course, how one would be able to calculate the cost of the midnight oil that you've burned.
Please reply to tuantran287@yahoo.com.
Anh thanks again.
Tuan Tran
Thanks
I'd live to have a pair of the PCB.
Please tell me how much it cost.
Of course, how one would be able to calculate the cost of the midnight oil that you've burned.
Please reply to tuantran287@yahoo.com.
Anh thanks again.
Tuan Tran
Thanks
Pins are correct, I am not using connectors at the moment.
Legato has output paps, there is a resistor to ground after the cap, in my case 220k. Do you suggest to add a series resistor ? something like 1k?
Thanks,
Davide
Hi Davide,
That is indeed a very odd problem you're having. If we're going to get to the bottom of this, you'll need to check a few things for me:
1. When everything is connected together, is the offset measurable at the input of The Wire?
2. Are you measuring the output offset from one channel to ground, or between the two channels? It's easy to get the pins mixed up.
3. Have you tried removing the output balanced to SE buffer on the Legato? If not, try removing R17, 18, 19 and 20 from the Legato to take the buffer out of the picture completely. It may have gone south and could be reacting strangely.
4. Does the offset go away when you load the output of The Wire? Try a 20-100 ohm resistors and see what happens.
5. I think you already answered this, but do you measure any offset anywhere when the two aren't connected? On the output of the Legato, on the input of The Wire, or on the output of The Wire?
6. Try removing all ground references between the Legato and The Wire. Just run the +/- phase from each channel and nothing else.
The Wire has a gain of 1, and is DC coupled, so if there's any DC on the output of the Legato then it will make it's way to the output of The Wire. Hopefully we can get to the bottom of this so you can start enjoying your amp!
I've used mine with six different balanced sources, and have not seen any DC offset or stability issues at all:
1. NTD1 DAC
2. M-Audio Fast Track Pro USB sound card
3. Behringer Ultra Curve Pro
4. Pass Aleph-Ono phono preamp
5. Pass BOSOZ preamp
6. Analog Devices AD1955 Reference DAC board
Regards,
Owen
That is indeed a very odd problem you're having. If we're going to get to the bottom of this, you'll need to check a few things for me:
1. When everything is connected together, is the offset measurable at the input of The Wire?
2. Are you measuring the output offset from one channel to ground, or between the two channels? It's easy to get the pins mixed up.
3. Have you tried removing the output balanced to SE buffer on the Legato? If not, try removing R17, 18, 19 and 20 from the Legato to take the buffer out of the picture completely. It may have gone south and could be reacting strangely.
4. Does the offset go away when you load the output of The Wire? Try a 20-100 ohm resistors and see what happens.
5. I think you already answered this, but do you measure any offset anywhere when the two aren't connected? On the output of the Legato, on the input of The Wire, or on the output of The Wire?
6. Try removing all ground references between the Legato and The Wire. Just run the +/- phase from each channel and nothing else.
The Wire has a gain of 1, and is DC coupled, so if there's any DC on the output of the Legato then it will make it's way to the output of The Wire. Hopefully we can get to the bottom of this so you can start enjoying your amp!
I've used mine with six different balanced sources, and have not seen any DC offset or stability issues at all:
1. NTD1 DAC
2. M-Audio Fast Track Pro USB sound card
3. Behringer Ultra Curve Pro
4. Pass Aleph-Ono phono preamp
5. Pass BOSOZ preamp
6. Analog Devices AD1955 Reference DAC board
Regards,
Owen
i was referring to the output of The Wire. See item number 4 from the post above by OPC.
and i wish i had ordered one
you cannot buy one, there are none left, but i feel pretty sure there will be another run at some point, as there have been quite a few people wanting one since the gb closed.
Hey guys, some of the parts are out of stock at mouser.
I assume that these should have no issue as C22, C24:
EEU-FR1E122L Panasonic Electronic Components Aluminum Electrolytic Capacitors - Leaded
and these should have no issue as J4, J5
39-31-0028 Molex Headers & Wire Housings
Also, if I am going to use those conn headers, where do I get the cables for them?
Perhaps I should just skip them.. hrm..
I assume that these should have no issue as C22, C24:
EEU-FR1E122L Panasonic Electronic Components Aluminum Electrolytic Capacitors - Leaded
and these should have no issue as J4, J5
39-31-0028 Molex Headers & Wire Housings
Also, if I am going to use those conn headers, where do I get the cables for them?
Perhaps I should just skip them.. hrm..
hint, when looking at headers etc that have a mating product to go with the, click on the features tab and you will find a list of parts that people who bought that part also bought, in this case 39-01-2025 Molex Headers & Wire Housings is the mating product. digikey also has a similar feature and usually lists the mating products down the left side of the product page.
myself i would just skip them. the caps look fine, except they are a 5mm leadspaced part where the nichicon is 7.5
myself i would just skip them. the caps look fine, except they are a 5mm leadspaced part where the nichicon is 7.5
Hi nullstring,
Everything you have there is correct. If you want to go the connector route, then you'll need the board mount Molex connector, the cable mount Molex connector, and two pins for the cable mount connector.
If you're mounting everything in a chassis and never plan to take it apart, then don't bother with the connectors. They just add cost, height, and a single solder joint is always the preferred method.
Those caps will do just fine, and all the PTH caps on the boards are 5mm leadspacing and 10mm case size.
I think qusp was imagining the NTD1 in his head when he said 7.5mm.
qusp:
I'll send you a note in the morning about the fet matching stuff. I was out of town the last week and a bit, so I have a lot to catch up on.
Regards,
Owen
Everything you have there is correct. If you want to go the connector route, then you'll need the board mount Molex connector, the cable mount Molex connector, and two pins for the cable mount connector.
If you're mounting everything in a chassis and never plan to take it apart, then don't bother with the connectors. They just add cost, height, and a single solder joint is always the preferred method.
Those caps will do just fine, and all the PTH caps on the boards are 5mm leadspacing and 10mm case size.
I think qusp was imagining the NTD1 in his head when he said 7.5mm.
qusp:
I'll send you a note in the morning about the fet matching stuff. I was out of town the last week and a bit, so I have a lot to catch up on.
Regards,
Owen
haha youre right. for some reason i had 7.5 on the brain.
as far as the emails you can skim the first couple mostly, as i've solved the ground offset, exactly 45vdc supply voltage and sine amplitude problems. only need you to look over the pics to see if there are any glaring flaws, all appears to work fine and vds measureents check out, just the low mVrms measurement for tc wasnt possible with my meter; which has also been solved providing the HP voltmeter i bought on ebay is fully functional.
as far as the emails you can skim the first couple mostly, as i've solved the ground offset, exactly 45vdc supply voltage and sine amplitude problems. only need you to look over the pics to see if there are any glaring flaws, all appears to work fine and vds measureents check out, just the low mVrms measurement for tc wasnt possible with my meter; which has also been solved providing the HP voltmeter i bought on ebay is fully functional.
Owen,
Thanks you for the help. Here are the answer of some of your questions. I am in duty travel now, and I will spend only 24 hours at home before taking off again. I think that if I touch the soldering iron my wife will kill me, probably using the soldering iron itself. So I can make more tests only in two weeks or so.
Ok, down to business:
1) The offset is at the input and at the output. I guess cause and effect.
2) The offset is from one channel to ground. I have no connectors, only dealing with wires, so no risk of confusion.
3) This I will do next. first I will try to power it down and then to disconnect. Thinking now, the legato has output caps, so, unless one is defective, it cannot be his fault. On the other end "the wire" and the balliste are kind of in parallel, so I think the balliste is doing something odd.
4) This I also will try. Up to now I connected only to the F5 input, but 101K is not a big load. The offset is there connected or disconnected. By the way I connected it to my denon amp, and it works, but in that case I had not check the bias.
5) when the legato and "the wire" are not connected, no offset. I also tested the circuit with a signal generator and it works with no offset.
also, if a single channel is connected I do not get offset. This is really strange.
6) I tried as my first guess was that it could be a grounding problem. So I tried mostly all the combination of grounding, including no connection.
I'll let you know.
Best Regards,
Davide
I got me Headers & Wire Housings 2 CKT RCPT HOUSING !!!
Is just housing but i got me wires and wire crimps from old PC PSU ! Anyway, this Single Wire Crimping should be the right one ...
BTW, i got me Wire up and running, now looking for encloasure, and right POT for volime controll ...