As a late builder, I notice in post 3705 Calvin mentions the idle current through Q1 etc. and Q2 etc is roundabout 3.5mA for each transistor in the input section. Apparently, it is important to use the right current when matching for hFE. As I understand the schematic for matching bipolars on page 2 of the assembly guide is set for 1 mA. Would it be possible to set the resistors of the matching schematic for 3,5 mA idle current, and still have a sensible reading on a multimeter?
correlation between different currents at measurement and operation is pretty good with these transistors, so IMHO it will not matter much. Low current will help to not change the temperature during measurement, as Andrew mentions - this can lead to big errors. Best would be to plug the transistor in and wait until the reading settles.
Of course, if you want to be utterly precise, then use 3.5mA. I would predict that a lot of any difference found would be swept away by the emitter resistors,anyway..... I did my selection work with the hfe measurement option of a cheap multimeter (don't even know what the current was) and it worked beautifully.
just my two cents
alfred
Of course, if you want to be utterly precise, then use 3.5mA. I would predict that a lot of any difference found would be swept away by the emitter resistors,anyway..... I did my selection work with the hfe measurement option of a cheap multimeter (don't even know what the current was) and it worked beautifully.
just my two cents
alfred
Finally finished building my second Paradise board and I'm having problems with the shunt regulator. The two flanking led banks all light up. However the bank of six at the back glow at different intensities, the two inner leds are much brighter than the four, two pairs, outer ones.
I'm getting about 0.5v on the regulator output with 26vdc input. The heat sink is becoming reasonably warm...Just no output.
I'll test around the regulator board at the tps, but any pointers would be very useful.
Thank you.
I'm getting about 0.5v on the regulator output with 26vdc input. The heat sink is becoming reasonably warm...Just no output.
I'll test around the regulator board at the tps, but any pointers would be very useful.
Thank you.
Simon,
I have the R2 version.
I've not mixed up the 327 or 337s. I've checked all the resistor markings, check diode polarity.
I did check that none of the power devices were shorted to the heatsink, that was about 2 two years ago and those clips are so tough to put on I would surprised that they would have moved...
I've checked all the voltages around the board using published regulator check cct and it all checks out...even upto the IRF610 and IRF9610 devices.
Interestingly I did notice that on initial switch on all the leds under the heatsink glow the same, then the four outer ones fade, but don't go out.
Initial switch on, LEDS
XXX XXX
Two secs later
xxX Xxx
Where X indicate full brightness and x partial brightness.
The other LEDS glow nicely.
I do have to make a confession...
When I originally wired the board up I swapped the Ov and the -25Vdc feeds the board over by mistake.
I have probably popped a device somewhere, I'm guess on the two centre ICs on the heatsink...But why this would effect the LEDs I'm not sure...Or could I have popped the two diodes D101 and D201?...though I would get no voltages on the front of the regulator at all...
Simon,
I have the R2 version.
I've not mixed up the 327 or 337s. I've checked all the resistor markings, check diode polarity.
I did check that none of the power devices were shorted to the heatsink, that was about 2 two years ago and those clips are so tough to put on I would surprised that they would have moved...
I've checked all the voltages around the board using published regulator check cct and it all checks out...even upto the IRF610 and IRF9610 devices.
Interestingly I did notice that on initial switch on all the leds under the heatsink glow the same, then the four outer ones fade, but don't go out.
Initial switch on, LEDS
XXX XXX
Two secs later
xxX Xxx
Where X indicate full brightness and x partial brightness.
The other LEDS glow nicely.
I do have to make a confession...
When I originally wired the board up I swapped the Ov and the -25Vdc feeds the board over by mistake.
Or could I have popped the two diodes D101 and D201?...though I would get no voltages on the front of the regulator at all...
And then like a moment of foolishness I noticed that I was taking voltage readings from the pads that are the feeds to the phono amp section....
Doh!!!!!!
Oh look, 18.48Vdc and -18.48Vdc on the regulator output....what an IDIOT!!!
Regulator all working...
I'd forgotten how hot that heatsink gets..wow!
Simon,
When I originally wired the board up I swapped the Ov and the -25Vdc feeds the board over by mistake.
...
Electrolytic capacitors do not like that at all maybe one is shorthed out
Those board changes and J310 replacements in the PSU
I have J109s in positions Q107 and Q207. Is the replacement with J113s a simple fit, that is not leg twisting? I have looked over their datasheets and it looks the case of a simple remove and re-fit.
Also, just to check that R108 and R208 each need replacing with 1k ohm to complete the swap?
I know stupid questions, but I want to make doubly sure as I've been away from these thread for a long while.
And in case I have missed anything from this thread, in terms of updates and changes to the R2 Paradise boards, these four are what I have found so far:
Ta!
About some more reasons for a (not to bad) (but still) noisy PSU. Salas has found that it is possible that the PSU generates 20 up to 50 mV of white noise. This level of PSU contamination is not affecting the RIAA but we want perfection.
After some discussion, between Salas and me, it was thought that the J310 may be the cause of the problem. Back to the simulator, and the datasheet then, what I find is this, the J310 is used as a CCS of 1mA, for this it needs a Rgs of (about) 3Kohm, that implies a Vgs of 3Volts. The Vgs of the HexFet is (about) 4Volts, in this application. That leave about 1Volt margin on Vds of the J310, given production tolerances and specification spread between manufacturers, this may be not enough.
In case where the Vds margin is to low, the J310 will not be a constant current source, in these cases the PSU will not function optimal, and a higher than normal noise level will be present.
Now how to solve this? Easy! Use a NJF that shows a higher margin. For instance, the 2SK170 needs a Vgs off (about, according to simulation data) 360mVolt, and that leaves a margin of 3.6Volts. That would be enough, but the 2SK170 is something like unobtainium. That is where I started my hunt for a replacement NJF. If should be cheap, leave lots of margin, and easy to get.
From my graphs, the best, commercially available and cheap NJF is the J113, good choices are 2SK170, BF256A, BF245A, BF244A, J113, U309 and J309. If you want to propose others, then let me know, if applicable, I will add them to the graph
I have J109s in positions Q107 and Q207. Is the replacement with J113s a simple fit, that is not leg twisting? I have looked over their datasheets and it looks the case of a simple remove and re-fit.
Also, just to check that R108 and R208 each need replacing with 1k ohm to complete the swap?
I know stupid questions, but I want to make doubly sure as I've been away from these thread for a long while.
And in case I have missed anything from this thread, in terms of updates and changes to the R2 Paradise boards, these four are what I have found so far:
Four caps in the PSU to stop the oscillations - done
Calvin's boards - pending
Q107/Q207 and R108 and R208 replacement - pending
70Mhz stopper cap in amp - not needed.
Calvin's boards - pending
Q107/Q207 and R108 and R208 replacement - pending
70Mhz stopper cap in amp - not needed.
Ta!
Well I've built a pair of Calvin's boards and only the lights under the shunt heatainks light up...on both boards...
I've checked for shorts and I can't find any...
Without the boards all the psu least glow as usual.
I've replaced the 10R with 8R2.
I'm using MJE350Gs
I can't see anything wrong at all...
Sent from my SM-G900H using Tapatalk
I've checked for shorts and I can't find any...
Without the boards all the psu least glow as usual.
I've replaced the 10R with 8R2.
I'm using MJE350Gs
I can't see anything wrong at all...
Sent from my SM-G900H using Tapatalk
Did you try the boards fully without Calvins? If not then check for shorts on the gain stage/RIAA.
Do all the LED's come on with the psu connected to the rest of the board but without the Calvins fitted? If not check for shorts.
Did you remove the old output stage before fitting the Calvins?
Have you wired them up correctly? Awfully tempting to wire the Calvin OUT to the PSU GND...
A nice hi-res pic usually helps.
Do all the LED's come on with the psu connected to the rest of the board but without the Calvins fitted? If not check for shorts.
Did you remove the old output stage before fitting the Calvins?
Have you wired them up correctly? Awfully tempting to wire the Calvin OUT to the PSU GND...
A nice hi-res pic usually helps.
