WHAMMY parts kit and complete kits are now available for pre-order. We are advertising "mid November" for shipping. It might happen sooner. We're just waiting for a single part to be added and they will be shipping.
WHAMMY Headphone Amplifier – diyAudio Store
WHAMMY Headphone Amplifier – diyAudio Store
fault
The voltage of R17 is not stable, the first time I measured my multimeter was set to 20v,and it was -0.06 and I remeasured last night and this morning on 200mv and it varies between -1.3 and -2.9.
I thought remeasured when you said that there was a very low voltage on R17
The voltage of R17 is not stable, the first time I measured my multimeter was set to 20v,and it was -0.06 and I remeasured last night and this morning on 200mv and it varies between -1.3 and -2.9.
I thought remeasured when you said that there was a very low voltage on R17
The voltage on R17 should be rock steady and very very low. Not to worry, lets see what we can come up with 🙂
Leaving the circuit connected as it is measure the voltage across R5. With an LM833 fitted you should see around 30 millivolts (which I think will be negative in value). It should be steady. Any variation here will reflect in an amplified change at the opamp output. See whether this voltage across R5 is stable or not.
Lets be absolutely sure on this by doing a further easy test. Now link out R5 (short it out with a bit of wire) so that pin 5 of the opamp connects to ground directly. That will force a true zero volts onto the input.
The opamp output (R17) should be steady and at a very low voltage.
Leaving the circuit connected as it is measure the voltage across R5. With an LM833 fitted you should see around 30 millivolts (which I think will be negative in value). It should be steady. Any variation here will reflect in an amplified change at the opamp output. See whether this voltage across R5 is stable or not.
Lets be absolutely sure on this by doing a further easy test. Now link out R5 (short it out with a bit of wire) so that pin 5 of the opamp connects to ground directly. That will force a true zero volts onto the input.
The opamp output (R17) should be steady and at a very low voltage.
the voltage of R5 is not stable -15 to -42 mv
The voltage of r17 is not very steady it varies between -5 and -9 mv (short-circuited R5)
The R5 voltage should be steady, however your readings of -15 to -42 mv are in the right sort of area (but should be steady).
Could that variation be down to measuring technique and stray pickup from you holding the probes? It is possible.
R17 voltage with R5 shorted. Although it should be steady the absolute values and range of variation are very low. There may not be a problem in other words.
If we accept that the R17 voltage is around -5 to -9mv then when the circuit is restored back to original configuration (with R8 once again connected back to the main headphone output) then the same -5 to -9mv should now be seen at the headphone output. That is so because the whole output stage is then wrapped up or enclosed within the feedback loop.
So I think we have to now try that. Leaving R5 shorted connect R8 back to its original configuration and then check that the DC offset remains at -5 to -9mv.
So the circuit goes back to looking like this.
The theory of what we are doing is rock solid 🙂 but there is obviously something strange going on with your particular build. We are not out of options though but do the above first...
If we get nowhere doing that then the next (and easy) step is for you to remove the opamp from the socket and also remove the short from R5. Now power up and check the voltage on pin 5 and pin 7 of the empty socket.
What are they?
Next you would link pin 7 of the empty socket to ground and measure the DC offset at the headphone output. The voltage should be low but not zero but must be steady.
Could that variation be down to measuring technique and stray pickup from you holding the probes? It is possible.
R17 voltage with R5 shorted. Although it should be steady the absolute values and range of variation are very low. There may not be a problem in other words.
If we accept that the R17 voltage is around -5 to -9mv then when the circuit is restored back to original configuration (with R8 once again connected back to the main headphone output) then the same -5 to -9mv should now be seen at the headphone output. That is so because the whole output stage is then wrapped up or enclosed within the feedback loop.
So I think we have to now try that. Leaving R5 shorted connect R8 back to its original configuration and then check that the DC offset remains at -5 to -9mv.
So the circuit goes back to looking like this.
The theory of what we are doing is rock solid 🙂 but there is obviously something strange going on with your particular build. We are not out of options though but do the above first...
If we get nowhere doing that then the next (and easy) step is for you to remove the opamp from the socket and also remove the short from R5. Now power up and check the voltage on pin 5 and pin 7 of the empty socket.
What are they?
Next you would link pin 7 of the empty socket to ground and measure the DC offset at the headphone output. The voltage should be low but not zero but must be steady.
Attachments
voltage
R5 short-cicuited :
Jack : R +25.1
L - 2.0
ground to :
R32 : +25.1 or /-6.5 to -8.5 \ depending on how I position
R29 : +25.5 or \-6.5 to -8.5 / the tip of the multimeter
But to the jack never change.
R22 : - 2.5
R16 : - 2.0
No OP amp No R5 short-cicuited :
N° 5 -15 to +13 mv
N° 7 -55 to -59 mv
N°7 to ground :
R32 -75
R29 -75
R22 -51
R16 -52
Jack : R -75 L -50
R5 short-cicuited :
Jack : R +25.1
L - 2.0
ground to :
R32 : +25.1 or /-6.5 to -8.5 \ depending on how I position
R29 : +25.5 or \-6.5 to -8.5 / the tip of the multimeter
But to the jack never change.
R22 : - 2.5
R16 : - 2.0
No OP amp No R5 short-cicuited :
N° 5 -15 to +13 mv
N° 7 -55 to -59 mv
N°7 to ground :
R32 -75
R29 -75
R22 -51
R16 -52
Jack : R -75 L -50
I'm getting a bit confused here 😉 there seems to be a lot going on here tbh 🙂 As always the devil is in the detail.
Lets go through them one at a time.
Now I thought the right channel was good and had a low offset. Is 25.1 actually millivolts? and not volts. You are using the same units for both here.
I'm guessing you mean -2 volts and +25.1 millivolts.
Any flux on the board can cause less than perfect connection with the probes. You sometimes have to press quite hard to get a perfect connection.
That shows an anomaly. Look at the circuit. If R5 is shorted then you should have 0.00volts on pin 5 because the other end of R5 is ground. That needs investigating.
Begin by measuring it again (with R5 shorted) and checking that you see identical voltages on each end of R5.
If you have the same -13 to +15 millivolt then something is amiss. A small and constant voltage at an earth point could be possibly due to small volt drops within the copper print of the board caused by current flow in the output stage... it is explainable.
If we think that is what is happening for that then use the grounded end of R5 as the main ground reference point for all measurements.
The last set of results I assume are all millivolts, -75mv, -51mv and so on.
If so then they look OK as the output stage is just free running and self biased. If they are steady then those are OK at this point.
I'm struggling to know what to suggest to you now tbh.
Lets go through them one at a time.
Now I thought the right channel was good and had a low offset. Is 25.1 actually millivolts? and not volts. You are using the same units for both here.
I'm guessing you mean -2 volts and +25.1 millivolts.
Any flux on the board can cause less than perfect connection with the probes. You sometimes have to press quite hard to get a perfect connection.
That shows an anomaly. Look at the circuit. If R5 is shorted then you should have 0.00volts on pin 5 because the other end of R5 is ground. That needs investigating.
Begin by measuring it again (with R5 shorted) and checking that you see identical voltages on each end of R5.
If you have the same -13 to +15 millivolt then something is amiss. A small and constant voltage at an earth point could be possibly due to small volt drops within the copper print of the board caused by current flow in the output stage... it is explainable.
If we think that is what is happening for that then use the grounded end of R5 as the main ground reference point for all measurements.
The last set of results I assume are all millivolts, -75mv, -51mv and so on.
If so then they look OK as the output stage is just free running and self biased. If they are steady then those are OK at this point.
I'm struggling to know what to suggest to you now tbh.
Attachments
I have just heard from the warehouse the kits are now ready to ship, so they are now no longer pre-orders. All orders that were taken overnight should ship tomorrow.
Assembling one kit right now and realized I had forgot to order the correct output jack for the chassis. >.>
Is this Neutrik the one used in the kits?
Is this Neutrik the one used in the kits?
No but that is close and is out of stock.
You need this one:
Neutrik NMJ6HF-S Switched 1/4" TRS Jack | Full Compass Systems
Has loops instead of pins for soldering to.
Some Neutiks also don't isolate from the chassis.
Full Compass has the correct ones in stock (at least for a while). May be a run on them when this post hits.
You need this one:
Neutrik NMJ6HF-S Switched 1/4" TRS Jack | Full Compass Systems
Has loops instead of pins for soldering to.
Some Neutiks also don't isolate from the chassis.
Full Compass has the correct ones in stock (at least for a while). May be a run on them when this post hits.
1/ no all my measurements are in mv
2/I repeated the measurements several times :
R5 short circuited
jack L -6.3 mv
R+25.1 mv
R32 +24.9 mv
R29 +25.3 mv
R22 -2.2 mv
R16 -1.8 mv
OPamp
1_0.08 v 8_+14.82 v
2_0.00 v 7_-0.15 +0.12 v
3_0.00 v 6_-0.02 v
4_-14.96 v 5_-0.01 v
no OPamp
at the end of R5 -0.6 mv and -0.6 mv
hole n°5 opamp -0.7 mv
with OPamp it's the same voltage
without R5 short-sircuited
R32 +25.1 } I didn't change anything, I don't know why ? for right channel
R29 +25.5 }
R22 ~~
R16 ~~
2/I repeated the measurements several times :
R5 short circuited
jack L -6.3 mv
R+25.1 mv
R32 +24.9 mv
R29 +25.3 mv
R22 -2.2 mv
R16 -1.8 mv
OPamp
1_0.08 v 8_+14.82 v
2_0.00 v 7_-0.15 +0.12 v
3_0.00 v 6_-0.02 v
4_-14.96 v 5_-0.01 v
no OPamp
at the end of R5 -0.6 mv and -0.6 mv
hole n°5 opamp -0.7 mv
with OPamp it's the same voltage
without R5 short-sircuited
R32 +25.1 } I didn't change anything, I don't know why ? for right channel
R29 +25.5 }
R22 ~~
R16 ~~
Well those actually look OK and you have no high and wandering DC offset like you had earlier on.
The offset on the channel with R5 shorted is lower (better) than the other channel simply because of the way the LM833 type opamp works.
(The technical reason why is because of the the unequal resistances seen by the two inputs. A short on pin 5 and the 4k75 feeding pin 2 are much more 'equal' than 100k and 4k75)
So you have low offset with R5 shorted. The obvious thing to do now is remove the short and see what the offset does. It should actually go to a similar value to the other channel.
As the fault seems to have changed a little do you think there could be any problem with soldering and bad connections etc?
The offset on the channel with R5 shorted is lower (better) than the other channel simply because of the way the LM833 type opamp works.
(The technical reason why is because of the the unequal resistances seen by the two inputs. A short on pin 5 and the 4k75 feeding pin 2 are much more 'equal' than 100k and 4k75)
So you have low offset with R5 shorted. The obvious thing to do now is remove the short and see what the offset does. It should actually go to a similar value to the other channel.
As the fault seems to have changed a little do you think there could be any problem with soldering and bad connections etc?
I have already tried to redo a few welds, maybe that's what changed something. But despite all the fluctuations are still there. I will try to do them all like that, there will be no more ambiguity.
I take this opportunity to ask you at what temperature is it better to solder? me for example I set my soldering iron to 400 c ° celsius with this one Fer a Souder avec Controle de Temperature 50W 450degC - Audiophonics
you have to put the flux on the tip of the iron ?
I would like to put all the chances on my side now.
Thank you Mooly for all the time spent helping me.
Last edited:
no need
you'll never use amp outputting in that peak area, in between normal and lowered rails
except if you use it on speakers, but even in that case - that's highest THD area
you'll never use amp outputting in that peak area, in between normal and lowered rails
except if you use it on speakers, but even in that case - that's highest THD area
As mighty Zen Mod says. 
Basically the 10K+10K dividers give lots and lots and lots of bias, more than ever needed, regardless of rail voltage, so that there is always enough for the optocoupler to do it’s job and get things set perfectly.
Basically the 10K+10K dividers give lots and lots and lots of bias, more than ever needed, regardless of rail voltage, so that there is always enough for the optocoupler to do it’s job and get things set perfectly.
let's just phrase it little different
Wayne made string of 4pcs of 10K resistors, where we have most upper and most lower one (both) covering range(s) of, say, 15V-4V (15V rail, 4V gate voltage), so 11V
current is around 1.1mA
between gates we have approx (4+4)V, with two 10K resistors nested in between gates
current through these two is 8V/20K , so around 400uA
difference to 1.1mA (flowing through first mentioned 10Ks ) is flowing through opto transistor , so around 700uA
what's clever in that is - more current through biasing impedance string, it is tougher in high swing domain (think about caps between gates loosing charge) , so Wayne practically with two additional resistors nested between gates increased overall current, which is resulting in faster recovery time .......... in ZM layman terms - biasing is tougher
hope Pa did got full schematic with his Tesla, with Wayne around, who needs Official Service
🙂
Wayne made string of 4pcs of 10K resistors, where we have most upper and most lower one (both) covering range(s) of, say, 15V-4V (15V rail, 4V gate voltage), so 11V
current is around 1.1mA
between gates we have approx (4+4)V, with two 10K resistors nested in between gates
current through these two is 8V/20K , so around 400uA
difference to 1.1mA (flowing through first mentioned 10Ks ) is flowing through opto transistor , so around 700uA
what's clever in that is - more current through biasing impedance string, it is tougher in high swing domain (think about caps between gates loosing charge) , so Wayne practically with two additional resistors nested between gates increased overall current, which is resulting in faster recovery time .......... in ZM layman terms - biasing is tougher
hope Pa did got full schematic with his Tesla, with Wayne around, who needs Official Service
🙂