HThank you, I'll see if digikey/mouser or partsconnexion stocks it. But I still don't get how to switch between three or four inputs using a switch and not a rotary.
Yep, have the same question what about these Connex Rotary Switch
which one will do it for 4 stereo inputs?
which one will do it for 4 stereo inputs?
The two deck 4 way switch is the one you want unless you specifically want to switch the ground connections as well. Simpler and easier wiring to just switch the signal.
Like a lot of threads on here there are multiple threads covering the same topics over and over again.
There is also a MESMORIZE thread and the DCB1 as well as the B1 which all cover all these issues.
There is also a MESMORIZE thread and the DCB1 as well as the B1 which all cover all these issues.
Output voltage off
Ive just powered a hypnotize and have some voltage issues
rectified DC is 18.5 and -18.6
Drop across CCRs is 1.89 and 2.04
Vout is 10.1 and -13.1 ?
offset 6.2 and 6.3 mv
relay clicks temp is ok everywhere
Any thoughts on what to investigate regarding the large difference in Vout would be much appreciated.
Im running a 24 CT transformer but out of a variac so I can vary line voltage above and below normal, this doesn't appear to change the Vout imbalance
Fritz
Ive just powered a hypnotize and have some voltage issues
rectified DC is 18.5 and -18.6
Drop across CCRs is 1.89 and 2.04
Vout is 10.1 and -13.1 ?
offset 6.2 and 6.3 mv
relay clicks temp is ok everywhere
Any thoughts on what to investigate regarding the large difference in Vout would be much appreciated.
Im running a 24 CT transformer but out of a variac so I can vary line voltage above and below normal, this doesn't appear to change the Vout imbalance
Fritz
Have a look at the gate voltages on the big FETs, then compare Vg against Vs and Vd. These are not matched so you might be seeing a big difference in the Vgs voltages of two components.
The output offset is good so you have no real issues.
Improving the matching of the +/- rails wont affect your offset.
The output offset is good so you have no real issues.
Improving the matching of the +/- rails wont affect your offset.
and also check the CCS current just in case one Shunt FET cannot shunt all the extra current to ground.
Once you know that Kat & my checks are OK, then you should check the voltages in the sensor. If the sensor has the wrong voltage in any one of the 4 legs of the measuring bridge then the output voltage will be wrong.
Once you know that Kat & my checks are OK, then you should check the voltages in the sensor. If the sensor has the wrong voltage in any one of the 4 legs of the measuring bridge then the output voltage will be wrong.
Thanks Kat and Andrew this is helping.
I'm getting to the limits of my understanding at this point so take into account I may misinterpret what you say.
I downloaded the data sheets for 9240 and 240 according to the sheets pin-outs are identical for the 2 devises ie G-D-S facing plastic side
Ill call the FET pair closest to the AC end pair A and the pair closest to output B
Pair A
Vg -13.1 and 13.2
Vs -16.6 and 16.7
Vd -13.1 and 10.1?
Pair B
Vg -9.8 and 6.6 ?
Vs -13.1 and 10.1
Vd 0 and o
After measuring the resistance across the tandem resistors CCS? I get .0585 amp on negative rail and .0552 amp positive rail. So about 6% different.
The corresponding resistance measures 34.06 and 34.72 or about 2%.
So Im guessing the the mismatch is in the A Pair were equal gate and source voltages are giving very different drain voltages which then feed the gates at pair B differently?
Ive got extra FEts should I swap first P240 to see if I get a better match or as you mention it may be of little or no consequence?
What does determine my output offset? The 4 matched K170s? Im fine with 6 mv just notice its at the upper end of the target range.
One note: I had a burnt LED in one of the triplet LED groups, a yellow that didnt make it through the build I replaced it with a yellow from another set that had a similar total voltage but haven't checked the balance of the triplets.
I do measure similar V drops across both triplets but maybe I should yank them and check their totals? I get 5.46 and 5.48 volts across the triplet LED arrays.
Fritz
I'm getting to the limits of my understanding at this point so take into account I may misinterpret what you say.
I downloaded the data sheets for 9240 and 240 according to the sheets pin-outs are identical for the 2 devises ie G-D-S facing plastic side
Ill call the FET pair closest to the AC end pair A and the pair closest to output B
Pair A
Vg -13.1 and 13.2
Vs -16.6 and 16.7
Vd -13.1 and 10.1?
Pair B
Vg -9.8 and 6.6 ?
Vs -13.1 and 10.1
Vd 0 and o
After measuring the resistance across the tandem resistors CCS? I get .0585 amp on negative rail and .0552 amp positive rail. So about 6% different.
The corresponding resistance measures 34.06 and 34.72 or about 2%.
So Im guessing the the mismatch is in the A Pair were equal gate and source voltages are giving very different drain voltages which then feed the gates at pair B differently?
Ive got extra FEts should I swap first P240 to see if I get a better match or as you mention it may be of little or no consequence?
What does determine my output offset? The 4 matched K170s? Im fine with 6 mv just notice its at the upper end of the target range.
One note: I had a burnt LED in one of the triplet LED groups, a yellow that didnt make it through the build I replaced it with a yellow from another set that had a similar total voltage but haven't checked the balance of the triplets.
I do measure similar V drops across both triplets but maybe I should yank them and check their totals? I get 5.46 and 5.48 volts across the triplet LED arrays.
Fritz
You don't need matching between the +ve devices and the -ve devices. In many FET N & P devices, N to P matching is virtually impossible, eg. irf9140 and irf240 make a better pair, than 9240 & 240. But even then, the better pair do not match.
Similarly matching and/or precise selection of LED Vf is not required. You have 5.46V and 5.48V for the two groups. That would give a 20mV difference in the +ve and -ve output voltages. That 20mV "error" is completely swamped by other +ve and -ve mismatches in semiconductors.
For a 10Vdc output your Vs on pair A @ ~16.6Vdc is a bit low. I wonder what transformer you have used? Did you measure the open circuit voltage of your transformer before adopting it for the build? How much lower will it go if mains voltage drops to minimum limit? I have ~+-22Vdc as my input voltage from a 15+15Vac transformer. A 12+12Vac transformer kills the performance of the CCS when main's voltage goes low.
The +ve voltages look about right.
16.7Vdc input to 16.7Vs.
Vgs ~-3.5V, OK.
Vds ~-6.6V, low but OK.
Vd @ 10.1V, OK.
That Drain is connected to the next Source and is the output voltage. OK.
Pair B
Vgs -3.5V, OK.
Vds -10.1V, OK.
Vd 0Vdc. OK. This is Audio Ground and by definition is the 0.0mVdc reference
BTW, the Drain PIN is some 30mm away from the Audio Ground and there will be a small volts drop along that trace. Expect Vd ~0.1mVdc to 0.2mVdc with a non hotrodded DCB1. A hotrodded version will show a higher volts drop along the trace.
Does that adequately explain the working half? Can you follow what I have said so far? Mark those voltages and voltage drops on a big copy of the schematic. Look at what is happening around the two big mosFETs.
Please confirm your understanding, or ask further questions. Then we can look at the misbehaving -ve side.
Similarly matching and/or precise selection of LED Vf is not required. You have 5.46V and 5.48V for the two groups. That would give a 20mV difference in the +ve and -ve output voltages. That 20mV "error" is completely swamped by other +ve and -ve mismatches in semiconductors.
For a 10Vdc output your Vs on pair A @ ~16.6Vdc is a bit low. I wonder what transformer you have used? Did you measure the open circuit voltage of your transformer before adopting it for the build? How much lower will it go if mains voltage drops to minimum limit? I have ~+-22Vdc as my input voltage from a 15+15Vac transformer. A 12+12Vac transformer kills the performance of the CCS when main's voltage goes low.
The +ve voltages look about right.
16.7Vdc input to 16.7Vs.
Vgs ~-3.5V, OK.
Vds ~-6.6V, low but OK.
Vd @ 10.1V, OK.
That Drain is connected to the next Source and is the output voltage. OK.
Pair B
Vgs -3.5V, OK.
Vds -10.1V, OK.
Vd 0Vdc. OK. This is Audio Ground and by definition is the 0.0mVdc reference
BTW, the Drain PIN is some 30mm away from the Audio Ground and there will be a small volts drop along that trace. Expect Vd ~0.1mVdc to 0.2mVdc with a non hotrodded DCB1. A hotrodded version will show a higher volts drop along the trace.
Does that adequately explain the working half? Can you follow what I have said so far? Mark those voltages and voltage drops on a big copy of the schematic. Look at what is happening around the two big mosFETs.
Please confirm your understanding, or ask further questions. Then we can look at the misbehaving -ve side.
Last edited:
There's lots of Andrew* on this Forum.
I'll assume you meant me and thank you for the complement.
Fritz,
Wake up, it's past dawn on the east coast !!!!
CA should be up by now.
I'll assume you meant me and thank you for the complement.
Fritz,
Wake up, it's past dawn on the east coast !!!!
CA should be up by now.
Thanks Andrew,
Yes that all seems to make sense and I can see the interconnections looking at the PCB layout .
Is it accurate to say the the shunt regulator pairs a voltage supply with a controlled shunt to ground to increase the current flow in the devices for a better/quiter voltage regulation?
Yes Im using a 12x12 transformer but will switch to a 15 by 15. Currently I have the trans hooked to a variac so I can turn the voltages up for testing if helpful.
Yes that all seems to make sense and I can see the interconnections looking at the PCB layout .
Is it accurate to say the the shunt regulator pairs a voltage supply with a controlled shunt to ground to increase the current flow in the devices for a better/quiter voltage regulation?
Yes Im using a 12x12 transformer but will switch to a 15 by 15. Currently I have the trans hooked to a variac so I can turn the voltages up for testing if helpful.
The circuit needs a substantially constant voltage across the CCS resistor/s to maintain CCS operation. As the input voltage falls, the Vdrop across the resistor/s does indeed hold constant. Then when the Vdrop decreases with a further decrease in input voltage then the input voltage is too low. The minimum input voltage is that which does not allow a decrease in the resistor Vdrop. Using your Variac you can identify the input voltage required to maintain effective CCS operation. From there you can determine the minimum mains voltage required.
I've tried to turn my DCB1 into hotrodded version by changing the 3 paraleled 68R resistors with one Caddock Mp930 10R and i have some nasty behaviour: the output voltages are ok (about 10.3/-10.8), the voltage across every 10R is about 1.9V but one IRF (the one close to the buffer part)on the positive rail becomes extremely hot compared to the other one.
I know i have small heatsinks but on the negative rails the corresponding IRF stays a lot cooler.
What to do, where to check ?
I know i have small heatsinks but on the negative rails the corresponding IRF stays a lot cooler.
What to do, where to check ?
Check the corresponding rail on oscilloscope to see if there is oscillation. Although there has not been any such reported history it may be. Are there any extra wires you used for those resistors? Do a sanity check that its gate stopper is the designated value also.
I am not allowed to edit the quote of your post !
So I will repeat it here - edited:
A Series FET = CCS
Vg -13.1
Vs -16.6
Vd -13.1
B Shunt FET
Vg -9.8
Vs -13.1
Vd 0
CCS fet
Vgs = +3.5Vgs. OK
Vds = +3.5Vds, something not right.
Shunt FET
Vgs = +3.3Vgs, OK
Vds = +13.1Vds, OK
The series FET can operate with a low Vds, but this circuit does not use it that way.
Mark those voltages on your big schematic. Where does it look like it is going wrong? Is there short g to d? Measure Vdg at the FET PINs.
Last edited:
My problem has been solved after rotating resistors between rails also they became mounten mirrored. This surprised me but everything is equally hot now but not to much.
I know it sounds stupid but have this resistors a correct way of mounting them in circuit ? On Mouser they are called current sense, it's something related ?
I know it sounds stupid but have this resistors a correct way of mounting them in circuit ? On Mouser they are called current sense, it's something related ?