diyAudio group buy Paradise R3 phono stage boards
- Thread starter Markus S
- Start date
Update on my builds - I got the Peak DCA and re-measured all of the BC327/337, including removing all of them from my second and third pair of boards. The result? My other meters did a decent job of matching the 327's and were close to the Peak DCA, the 337's were miles off - between 15-20% higher than I measured with my other meters. Looking at my original results, at a guess the BC327's in my pair of "working" boards have a HFE of around 440, the 337 around 500, which explains my offset issues.
I have a very nice spread of around 200 x 327's around 420-440, but not enough 337's to match them up with so am waiting on more 337's to arrive.
If anyone hasn't started building theirs yet, make sure you have a device that you know mesures HFE accurately....
As an aside, when initially testing the boards I was using a self-built LM317/337 supply and the reg's were working fine with no oscillation. Despite what the designer says I will at least try using a gyrator when building my second Paradise to see If that makes an improvement. The on-board regs are supposedly "high-bandwidth" but then so is a gyrator. The reg's are supposedly based on Salas's regs, and they benefit from a gyrator in front of them so I think it is at least worth a try....
I have a very nice spread of around 200 x 327's around 420-440, but not enough 337's to match them up with so am waiting on more 337's to arrive.
If anyone hasn't started building theirs yet, make sure you have a device that you know mesures HFE accurately....
As an aside, when initially testing the boards I was using a self-built LM317/337 supply and the reg's were working fine with no oscillation. Despite what the designer says I will at least try using a gyrator when building my second Paradise to see If that makes an improvement. The on-board regs are supposedly "high-bandwidth" but then so is a gyrator. The reg's are supposedly based on Salas's regs, and they benefit from a gyrator in front of them so I think it is at least worth a try....
337alant
Negatively Biased
I just arranged to buy a full kit of parts (Paradise/Calvin/PSU) from a gentleman on diyAudio. Two quick questions:
1: The guy I am buying from already checked his transistors and did not get a matching set in his kit, does anyone have any leftovers I can buy at cost?
2: I don't have a meter that can measure HFE, any suggestions of something inexpensive and accurate?
1: The guy I am buying from already checked his transistors and did not get a matching set in his kit, does anyone have any leftovers I can buy at cost?
2: I don't have a meter that can measure HFE, any suggestions of something inexpensive and accurate?
Thanks for the kind offer, but I am in the Netherlands now. I found a new DC55 on ebay for a reasonable price, from what I have read this seems the best way to measure the transistors. I will follow your advice from earlier in the thread and order 200 of each to get started, measure these, and see how I get on before making a giant order.
I have arranged to buy boards from a chap called Ulrich (username ui2). They will be here in a couple of weeks because he is about to leave on holiday.
I have arranged to buy boards from a chap called Ulrich (username ui2). They will be here in a couple of weeks because he is about to leave on holiday.
Sorry, another question 
I have some guidance from the chap who is sending me boards on measuring HFE:
He also provided a couple of sketches of some circuits that might accomplish these things here and here.
Does anyone have any thoughts? The circuits (if I can decipher the writing) can be easily breadboarded, but on the other hand, I think it might be easier and faster to just use a Peak meter.
Cheers,
Edward
I have some guidance from the chap who is sending me boards on measuring HFE:
He also provided a couple of sketches of some circuits that might accomplish these things here and here.
Does anyone have any thoughts? The circuits (if I can decipher the writing
) can be easily breadboarded, but on the other hand, I think it might be easier and faster to just use a Peak meter.Cheers,
Edward
PigletsDad
My intelligence test came back negative.
Without knowing the circuit in detail, usually matching Vbe at the operating point is more relevant to current sharing than matching Hfe.
For any bipolar transistor, the slope (how rapidly Ic changes with Vbe) is always the same. This is always exponential, over very wide ranges of Ic, until saturation sets in. The exponent is 26mV at room temperature and is given by fundamental constants. So the only thing you need to match in a voltage controlled circuit is Vbe.
For any bipolar transistor, the slope (how rapidly Ic changes with Vbe) is always the same. This is always exponential, over very wide ranges of Ic, until saturation sets in. The exponent is 26mV at room temperature and is given by fundamental constants. So the only thing you need to match in a voltage controlled circuit is Vbe.
337alant
Negatively Biased
The problem I noted during matching of 3000+ transistors is that the 327 and 337 have different temperature coefficients and that drove me nuts trying to get a match I ended up with a temp controlled room at 22deg and got them all to match eventually
But you have to consider at what temp is your build is running at? because if that is 30 or 40 deg than your 337 / 2227 wil no longer be matching , and that all depends on which transformers you chose and the enclosure heat sinking and cooling for the shunt regs, which appear to oscillate at the drop of a hat BTW
To be perfectly honest Im totally pi**ed of by the amount of components I have now had to bodge on to the bottom of these boards to stop them oscillating, and I note that the new boards incorporate all of these mods
The original design recommends 24v transformers yet the circuit only requires 18v so you end up having to dissipate the heat from a DC voltage of 34v down to 18v so a considerable amount of heat is wasted inside the amplifier enclosure so I found it beneficial to use 18v transformers now my reg heat sinks just run hand warm so the temp inside may amp enclosure is much more stable and controlled
At this moment in time my phono stage is sound ing good, better than anything I have owned before but it hasnt been easy getting to this point
Alan
But you have to consider at what temp is your build is running at? because if that is 30 or 40 deg than your 337 / 2227 wil no longer be matching , and that all depends on which transformers you chose and the enclosure heat sinking and cooling for the shunt regs, which appear to oscillate at the drop of a hat BTW
To be perfectly honest Im totally pi**ed of by the amount of components I have now had to bodge on to the bottom of these boards to stop them oscillating, and I note that the new boards incorporate all of these mods
The original design recommends 24v transformers yet the circuit only requires 18v so you end up having to dissipate the heat from a DC voltage of 34v down to 18v so a considerable amount of heat is wasted inside the amplifier enclosure so I found it beneficial to use 18v transformers now my reg heat sinks just run hand warm so the temp inside may amp enclosure is much more stable and controlled
At this moment in time my phono stage is sound ing good, better than anything I have owned before but it hasnt been easy getting to this point
Alan
Sounds like lots of fun ahead . I am getting the R3 boards which seem to have solved most of the problems but I am sure there will be many bumps in the road.
The PSU I got (the pre-reg kit) uses 24v PSUs, but I hope with the giant (1H) hammond chokes on each rail it should be a little lower than 34v coming into the paradise regs. I have seen a few people commenting on the problem of matching transistors at a different temperature to what they will experience when in use. Nobody has a solution to this that I have seen.
My oven can go down to 30 degrees, but I think this might be taking things too far
. I am getting the R3 boards which seem to have solved most of the problems but I am sure there will be many bumps in the road.The PSU I got (the pre-reg kit) uses 24v PSUs, but I hope with the giant (1H) hammond chokes on each rail it should be a little lower than 34v coming into the paradise regs. I have seen a few people commenting on the problem of matching transistors at a different temperature to what they will experience when in use. Nobody has a solution to this that I have seen.
My oven can go down to 30 degrees, but I think this might be taking things too far
sq225917
Bit of this, bit of that
You don't have to have the transistors running in a hot box, you can vent the heat from the shunt via chimney where the airflow does not affect the input stage transistors, just partition off the box, or shield them directly from airflow.
Certainly worthwhile matching them in as warm a room as you can tolerate. I managed 24 degrees for my last build and got them matched to way better than 1%, in fact the input stage transistors were all a numerical match across both types and both channels. I've built 9 of these now for myself and various people, I hate to think how many transistors I've matched, must be close to 10,000. (and worth every second).
Certainly worthwhile matching them in as warm a room as you can tolerate. I managed 24 degrees for my last build and got them matched to way better than 1%, in fact the input stage transistors were all a numerical match across both types and both channels. I've built 9 of these now for myself and various people, I hate to think how many transistors I've matched, must be close to 10,000. (and worth every second).