Maybe also check that the chassis power connector for the Meanwell is not turned 180 degree as this would reverse the voltage and +V would be grounded?
Yes,that white wire is shorting out the thermistor because of the way the Meanwell plug is wired.We have discussed that previously.I agree with Adason,you don't need it.However that doesn't address your issue unless there is a stray wire strand from there to the other pins on the connector..
Last edited:
Pass DIY Addict
Joined 2000
Paid Member
Another potential problem spot is the power connector - there is not much clearance between the positive supply and the ground wire for soldering those wires that run to the filter board.
Folks:
Thanks for the quick feedback!
I removed the white wire connecting the PS board to one of the two unused pins on the power inlet. I then removed the ground wire connecting the L channel FE board to the PS board and the ground and V+ wires connecting the L channel OS board to the PS board, isolating (a) the L channel FE and OS boards together from (b) the R channel FE and OS boards and the PS board. All of the input and output connections to the rear panel were also disconnected.
Naturally, the ground wire on the L channel FE board to the ground wire on the L channel OS board measures 0R. When measuring either ground wire to the L channel V+ wire, the meter starts at a few ohms and within 4-5 seconds climbs to about 1k (and continues to slowly rise). The same thing happens when measuring the R channel V+ connection to any ground connection on the R channel FE board, R channel OS board or the PS board. Both channels essentially measure the same.
R7 has still not been added to either FE board.
Your thoughts?
Regards,
Scott
Thanks for the quick feedback!
I removed the white wire connecting the PS board to one of the two unused pins on the power inlet. I then removed the ground wire connecting the L channel FE board to the PS board and the ground and V+ wires connecting the L channel OS board to the PS board, isolating (a) the L channel FE and OS boards together from (b) the R channel FE and OS boards and the PS board. All of the input and output connections to the rear panel were also disconnected.
Naturally, the ground wire on the L channel FE board to the ground wire on the L channel OS board measures 0R. When measuring either ground wire to the L channel V+ wire, the meter starts at a few ohms and within 4-5 seconds climbs to about 1k (and continues to slowly rise). The same thing happens when measuring the R channel V+ connection to any ground connection on the R channel FE board, R channel OS board or the PS board. Both channels essentially measure the same.
R7 has still not been added to either FE board.
Your thoughts?
Regards,
Scott
ZM & Co.:
Since the R channel FE and OS boards were already connected to the PS board, I tested the R channel. There was no change -- the voltage between V+ and ground rapidly bounced around in the 0.5 to 3.3 VDC range, the same as that channel and the L channel before the L channel was disconnected from the PS board.
I searched for the meaning of the Meanwell's dimly lit, rapidly blinking LED but couldn't find anything. Might that be helpful?
Eric:
My connections at each of the boards are clean -- no stray wires. And I put a little heatshrink at each of the connection tabs on the power inlet. There's no short-circuit there.
Regards,
Scott
Since the R channel FE and OS boards were already connected to the PS board, I tested the R channel. There was no change -- the voltage between V+ and ground rapidly bounced around in the 0.5 to 3.3 VDC range, the same as that channel and the L channel before the L channel was disconnected from the PS board.
I searched for the meaning of the Meanwell's dimly lit, rapidly blinking LED but couldn't find anything. Might that be helpful?
Eric:
My connections at each of the boards are clean -- no stray wires. And I put a little heatshrink at each of the connection tabs on the power inlet. There's no short-circuit there.
Regards,
Scott
Your SMPS brick output voltage on test is OK 36V ? Maybe something inside psu input connector ?
have you any way of testing either PSU alone (load of effective 18R, but must be at least 100W) , or testing amp with other PSU
24Vdc will do the job, even 19V laptop brick
24Vdc will do the job, even 19V laptop brick
The switching supplies I work on will many times "thrash" like you describe when trying to drive a short or near short. Is the DIN connector installed upside down? Sorry if someone already asked. I didn't see a picture showing the rear of the amp and the orientation of the connector. But I agree with these guys, it would be nice to test the SMPS and amplifier separate from each other to narrow it down at least.
I would plug linear power supply to it, carefully, with fuses to protect in case of short. Left and right channel separately. That should sort things out.
Disconnect both channels from the PS board and repeat the switch voltage test.
What happens then?
Is the key slot for the power plug on the top or the bottom when installed? Can you take a pic of the back panel?
What happens then?
Is the key slot for the power plug on the top or the bottom when installed? Can you take a pic of the back panel?
Last edited:
Reporting that my VFET is operating swimmingly every day. I checked the bias this past evening after the amp played quiet background music for about 16 hours. Both channels still were at 20V. I nudged one down from 20.09 to 20.04 to match the other.
I let the amp open up a couple nights ago and it was sublime. I don’t think I will need any more gain at this point. However I may need a different phono stage with just a little more gain in case I want to play quiet records loud. Not critical at this point.
Once my work calms down the seismic gravity of having and enjoying this amp may fully hit me.
I let the amp open up a couple nights ago and it was sublime. I don’t think I will need any more gain at this point. However I may need a different phono stage with just a little more gain in case I want to play quiet records loud. Not critical at this point.
Once my work calms down the seismic gravity of having and enjoying this amp may fully hit me.
The optional front end cards called Bulwark, Marauder, and Dreadnought, have a front end gain which is set by a ratio of two resistors in the negative feedback loop. Insightful builders who really know what they're doing, could possibly make a modification (unsupported!) to that ratio and, thereby, achieve a different front end gain.
On the other hand, neither the original Nelson Pass (2 x Toshiba JFET) front end card, nor the Scourge card, have NFB divider resistors. So they cannot execute this particular stunt.
On the other hand, neither the original Nelson Pass (2 x Toshiba JFET) front end card, nor the Scourge card, have NFB divider resistors. So they cannot execute this particular stunt.
Some of those optional front end cards also have a few dB more gain, even without making any unsupported modifications. That could be enough to make a difference.
Not really a huge amount, just an extra 1.73dB of gain. 20 * log10(6.1 / 5.0) = 1.7272 dB
Gives the OPA552 and the discrete OTA, some extra stability margin. Feels good too.
Gives the OPA552 and the discrete OTA, some extra stability margin. Feels good too.
This amp; where do I start; so much detail coming through with a lovely wide and deep soundstage and space around voices and instruments. Just had a listen to several LP's and am one happy bunny😀. Thank you again Nelson.
Marra,
Congratulations on getting your DIY Sony VFET amplifier up and running!
What speakers and preamp are you using?
Congratulations on getting your DIY Sony VFET amplifier up and running!
What speakers and preamp are you using?