Fortunately it was a bad joint. One of the legs didn't get enough juice and when I tried to measure, completely broke it. Not it works fine.
Now one side measures 3.4, other 3.7Vgs, meaning a bit more amp on that side, but it is not that much.
Strangely the shunt worked just fine for weeks with that half dead joint.
![]()

I built the DCB1 today with a Teabag full kit, LED strings are matched at 5.45V and 9.0V respectively, got the following measurements:
V+ = 9.61V
V- = -9.60V
Across CCS resistor (2 x 56R in parallel, to be hotrodded later), V=1.8V for -ve side and 1.9V for +ve side
DC offset at output: 1.5mV (left) and 0.8mV (right)
Are the above numbers OK? The DC offset seems a bit high, especially the left channel, and I would like to know whether it can be lowered by adjusting the voltage or CCS current of the 2 regulators.
V+ = 9.61V
V- = -9.60V
Across CCS resistor (2 x 56R in parallel, to be hotrodded later), V=1.8V for -ve side and 1.9V for +ve side
DC offset at output: 1.5mV (left) and 0.8mV (right)
Are the above numbers OK? The DC offset seems a bit high, especially the left channel, and I would like to know whether it can be lowered by adjusting the voltage or CCS current of the 2 regulators.
Attachments
+1.5mVdc or -1.5mVdc and +0.8mVdc or -0.8mVdc?
+1.5mVdc and +0.8mVdc
That says the upper device is running at more than Idss, i.e +ve Vgs.
You need to swap upper with lower to get the upper device running at less than Idss.
You need to swap upper with lower to get the upper device running at less than Idss.
1.5mVdc that is blocked by the DC blocking capacitor fitted in a subsequent stage will have absolutely no deleterious effect on sound quality nor the life of the speaker.
If you have DC coupling through out the DCB1 and all subsequent equipment, then live in hope.
If you have DC coupling through out the DCB1 and all subsequent equipment, then live in hope.
That says the upper device is running at more than Idss, i.e +ve Vgs.
You need to swap upper with lower to get the upper device running at less than Idss.
Thanks Andrew. Those 2SK170BLs come as matched quads with idss label on them, perfectly matched within 0.02mA so I didn't measure them again before soldering into the board.
With your comments, at the risk of destroying 4 precious Toshiba JFETs, I de-soldered them to measure the idss again, turned out they're not as closely matched as I thought (9.93mA, 9.94mA, 9.98mA, 10.04mA respectively, probably since the room temperature is different). I solder them back in based on your comment and L-channel offset is now -1.0mV whereas right is +0.9mV.
I feel more comfortable with the DC offset now, thanks again.
I'm also planning to buid this hot rodded B1 pre. Now I'm using a kit from velleman (with better parts)
pls take a look at my personal web blog.....
users.telenet.be/offie
I'm in need of more inputs and don't need a tone control.
The Velleman kit isn't bad at all. (when using good opamps). But ...I would like to swap this with a B1 buffer. Input sensitivity of the main amplifier is 1Vrms. So no worries about that.
Speakers are B&W DCM7's
So I'm looking for matched and unmatched pairs of jfet's..but i'm living in europe (Belgium) Postage fees are more expensive than the components.
Any one living nearby?
pls take a look at my personal web blog.....
users.telenet.be/offie
I'm in need of more inputs and don't need a tone control.
The Velleman kit isn't bad at all. (when using good opamps). But ...I would like to swap this with a B1 buffer. Input sensitivity of the main amplifier is 1Vrms. So no worries about that.
Speakers are B&W DCM7's
So I'm looking for matched and unmatched pairs of jfet's..but i'm living in europe (Belgium) Postage fees are more expensive than the components.
Any one living nearby?
Last edited:
😱 It's me again! I've really enjoyed building this but I seem to be having more problems than most, it's a bit demoralising to read about perfect first time builds but I'll stay positive and get there. It's a good learning experience.
Anyway... I cased my board and connected it up with only one input device (a DAC) to keep it simple. I had an unlistenable ground loop at first until I joined the two separate GND tags on the stereo pot (a Tocos Cosmos - tag 1).
Now the problem is that I only have one channel working. The board still gives the same DC offset readings as before (about 2mV on each channel) so I'd assumed it to be ok. The problem is that I've continuity checked all wires (input and output) from molex to RCA (through the pot for the outputs) and there are no breaks. I'm stumped. Any suggestions for further testing?
The other odd thing is that I tested it with an old integrated amp and the sound comes out the same whatever input I select on the amp.
Thanks
Anyway... I cased my board and connected it up with only one input device (a DAC) to keep it simple. I had an unlistenable ground loop at first until I joined the two separate GND tags on the stereo pot (a Tocos Cosmos - tag 1).
Now the problem is that I only have one channel working. The board still gives the same DC offset readings as before (about 2mV on each channel) so I'd assumed it to be ok. The problem is that I've continuity checked all wires (input and output) from molex to RCA (through the pot for the outputs) and there are no breaks. I'm stumped. Any suggestions for further testing?
The other odd thing is that I tested it with an old integrated amp and the sound comes out the same whatever input I select on the amp.
Thanks
First make sure that the problem is in the preamp. Swap the input plugs; does the same channel stay dead? If not, then the problem is upstream of the preamp. If the same channel is dead, swap the output plugs. If the same channel is dead the problem is downstream of the preamp.
If you verify that the preamp is at fault, then it's time to start measuring things. Check the power supply voltages, and the voltages across the FETs. Make sure the output relay is OK.
I don't follow what you are saying about the integrated amp.
If you verify that the preamp is at fault, then it's time to start measuring things. Check the power supply voltages, and the voltages across the FETs. Make sure the output relay is OK.
I don't follow what you are saying about the integrated amp.
Hello, thanks for the reply. The integrated amp is behaving as it should now - it was playing the input from the DCB1 plugged into AUX no matter what source I selected.
I swapped the inputs and outputs and the 'dead' channel swaps around so it's definitely the DCB1 - also the DAC plugged into the integrated amp is fine. I'm using that amp to test the DCB1 - so that I can limit what reaches the speakers until I know the buffer is working fine.
The channel isn't actually dead, it's just very quiet - about a tenth of the volume of the other one.
The only tests I know how to do are these (with the results):
Vdrop across CCS (1.63v and 1.73v)
Vout at the power supply (+9.86v and -9.89v)
DC offset (left 1.5mV and right 2.5mV)
The relay clicks a second or two after the unit is switched off. I'm afraid I don't know how to check that or the voltage across the FETs but I'll get googling...
Thanks again.
I swapped the inputs and outputs and the 'dead' channel swaps around so it's definitely the DCB1 - also the DAC plugged into the integrated amp is fine. I'm using that amp to test the DCB1 - so that I can limit what reaches the speakers until I know the buffer is working fine.
The channel isn't actually dead, it's just very quiet - about a tenth of the volume of the other one.
The only tests I know how to do are these (with the results):
Vdrop across CCS (1.63v and 1.73v)
Vout at the power supply (+9.86v and -9.89v)
DC offset (left 1.5mV and right 2.5mV)
The relay clicks a second or two after the unit is switched off. I'm afraid I don't know how to check that or the voltage across the FETs but I'll get googling...
Thanks again.
Use 300Hz sinewave signal and trace it along all route nodes from input to output. If not having generator and scope, use a free software gen and your DMM on AC mode.
I'm sorry Salas but I have no idea what that means. I'm trying to learn but it's slow going - where can I read about how to do that? If that's the only way to fault check it then I'm pretty disheartened - my project that I've enjoyed so much will have to go into a drawer or the bin 🙁
Don't lose heart. Maybe that JFET pair you touched up was problematic from the start or some socket/wire connection is not done well.
What I meant is by feeding sinewave signal from a PC or laptop or smartphone etc from a free signal generator app, your DMM set to measure AC can detect it. So when you will probe from input towards pot in, pot out, 220 Ohm input, 220 Ohm output, relay output, on the bad channel, you will be able to pinpoint where the signal stops and fix it. Use red DMM probe on every point of the signal path while keeping black probe connected on any handy ground point.
What I meant is by feeding sinewave signal from a PC or laptop or smartphone etc from a free signal generator app, your DMM set to measure AC can detect it. So when you will probe from input towards pot in, pot out, 220 Ohm input, 220 Ohm output, relay output, on the bad channel, you will be able to pinpoint where the signal stops and fix it. Use red DMM probe on every point of the signal path while keeping black probe connected on any handy ground point.
Ok, thanks for the support, it'll be a while before I give up - I've just had a day of it following the ground loop... and so on. I was hoping to be sat here now listening to music with a pint of Madness ale!
I've isolated it as much as I can - I checked the pot and fully open it was reading 19+ kOhm on both sides so that should be ok. It's the left channel on the board (yep, the one that I touched up the other day).
So the signal goes into the RCA input? I then put the DMM black lead on GND and follow the components? Maybe easier for me if I work back from the pot until I get a big AC change (increase)?
I've isolated it as much as I can - I checked the pot and fully open it was reading 19+ kOhm on both sides so that should be ok. It's the left channel on the board (yep, the one that I touched up the other day).
So the signal goes into the RCA input? I then put the DMM black lead on GND and follow the components? Maybe easier for me if I work back from the pot until I get a big AC change (increase)?
Also make sure you did not ground the signal when you rearranged the pot connections. Try ground its metal body instead.
So the signal goes into the RCA input? I then put the DMM black lead on GND and follow the components? Maybe easier for me if I work back from the pot until I get a big AC change (increase)?
Yes. Its possible there is a lifted ground somewhere for instance. Because you had signal before even if there was a buzz, correct?
Yes, that's right. I think you're onto something re the grounding. I put a wire link between the two GND tags on the pot to get rid of the hum but one channel is much louder than the other regardless of how I plug the RCA input in - it's as if the signal is being 'mixed'. I'll try to post a picture of the pot wiring, it might help others (as in how not to do it if that's the problem).
Thanks again!
Thanks again!
- Home
- Source & Line
- Analog Line Level
- Salas hotrodded blue DCB1 build