I know ver well what you feel ....
Congrats.
One quid:
Perhaps, did you couple al input BJT in "equipotential-thermal-position"!
.... In the same bed-room & bed & blanket & night-cap ....
It's music (with the same diapason) .......
That's cool (actually hot 🙂). Do not forget to add a few (little) ventilation holes into the 'Little House on the Paradise' 🙂 otherwise it may get to hot inside.
It'll actually be just be an up-n-over, made of 1mm acylic and bent over the top of the transistors leaving 3-5mm all around. Just to stop the draught
So you leave the sides open? then that is good. Do not forget to add a small heat-spreader to thermal connect all transistors (in the house).
Hi,
Received the boards a while ago and have been waiting for an updated BOM. The old one seems to have disappeared.
Received the boards a while ago and have been waiting for an updated BOM. The old one seems to have disappeared.
I'll have to get my tame plastics bender on the job. I'll need to draw it up first
So please post a simple drawing.... Should be easier for us to visualize the concept.
Today I received a new batch of bjt.... finally I will start to build my third paradise 🙂
Today I received a new batch of bjt.... finally I will start to build my third paradise 🙂
Hi!
After some time not being here, I see that the latest discussions turn around the issue of thermo-coupling and the offset in the stages. I missed also the discussion around the J113 in the PSU.
I am now in the testing phase. What a pain is the matching of the transistors! I had to buy another batch of 100+100 of BC 327/337 without getting better results. The hfe is so different between 327 and 337!. If I had read before that someone had to test 6000 transistors in order to use 36, I wouldn't have begun the paradise history. But finally I was in the 10 % as the guide advise. Well, 10% in each mirror stage, not between stages as I wanted - in this case the differences of average hfe is bigger. And between channels, forget about! I'm not happy with that, but I can't have all in the life. The worse is that the bigger mismatch inside one stage occurs just in the input stage. I had the intuition of the necessity of cooling the tested transistors and I blow over the trans for few seconds, but I wasn't smart enough to use a fan. So the measurings wasn't probably very accurate.
Well, enough with the history. I need help, help for me (as the King, BB). Take a look, please, of my voltage readings and tell me what is wrong.
the voltages in the PSU are OK.
A. RIAA: according to the guide, -1.5V, I have something between -1.9 and -2.3 V wandering continuously.
B. pin 6 of the OPA: -8 to -12 V (right ch.) and -3 to -5 V (left), wandering
C. Input: 0.0 - 0.1 mV stable
D. Output: wandering between -250 mV and + 250 mV
I noticed that in each stage there is one transistor warmer than the others, so naturally I thought to a thermo-coupling solution, me too. Is this the cause of those strange readings? Or some oscillations? Or it's all OK? I'll post my thermo-coupling solution one of these days. I just begun to work on it.
Sorry for this long post and thanks for your help
Emil
After some time not being here, I see that the latest discussions turn around the issue of thermo-coupling and the offset in the stages. I missed also the discussion around the J113 in the PSU.
I am now in the testing phase. What a pain is the matching of the transistors! I had to buy another batch of 100+100 of BC 327/337 without getting better results. The hfe is so different between 327 and 337!. If I had read before that someone had to test 6000 transistors in order to use 36, I wouldn't have begun the paradise history. But finally I was in the 10 % as the guide advise. Well, 10% in each mirror stage, not between stages as I wanted - in this case the differences of average hfe is bigger. And between channels, forget about! I'm not happy with that, but I can't have all in the life. The worse is that the bigger mismatch inside one stage occurs just in the input stage. I had the intuition of the necessity of cooling the tested transistors and I blow over the trans for few seconds, but I wasn't smart enough to use a fan. So the measurings wasn't probably very accurate.
Well, enough with the history. I need help, help for me (as the King, BB). Take a look, please, of my voltage readings and tell me what is wrong.
the voltages in the PSU are OK.
A. RIAA: according to the guide, -1.5V, I have something between -1.9 and -2.3 V wandering continuously.
B. pin 6 of the OPA: -8 to -12 V (right ch.) and -3 to -5 V (left), wandering
C. Input: 0.0 - 0.1 mV stable
D. Output: wandering between -250 mV and + 250 mV
I noticed that in each stage there is one transistor warmer than the others, so naturally I thought to a thermo-coupling solution, me too. Is this the cause of those strange readings? Or some oscillations? Or it's all OK? I'll post my thermo-coupling solution one of these days. I just begun to work on it.
Sorry for this long post and thanks for your help
Emil
Hi Emil,
I tested 6000 transistors just to get a very good match overall for 3 stereo sets and very high hFE tightly matched in the input stage for at least 2 of them. I had totally no luck with my first batch and I tried several distributors to achieve this, but
it's not a specified design target. It maybe even called something crazy 🙂 There is local feedback and variations in hFE are partially compensated.
I cannot say much about your readings, possible to be just thermal drift; probably a 100MHz scope will provide more details.
Good luck!
I tested 6000 transistors just to get a very good match overall for 3 stereo sets and very high hFE tightly matched in the input stage for at least 2 of them. I had totally no luck with my first batch and I tried several distributors to achieve this, but
it's not a specified design target. It maybe even called something crazy 🙂 There is local feedback and variations in hFE are partially compensated.
I cannot say much about your readings, possible to be just thermal drift; probably a 100MHz scope will provide more details.
Good luck!
Last edited:
An externally hosted image should be here but it was not working when we last tested it.
Little house for the input stage, to shelter them from the wind...
I hate to be the bearer of bad tidings regarding matching transistors, having just matched all of mine to better than 1% in ever position on both channels. But it makes a huge difference to the sound. My last batch from Farnel in the UK were almost all used, I got more from that 100 than the previous 1500 i purchased. (Diotec 40 grade on ammo tape)
If anyone needs some matches let me know mine are all still numbered so selecting them will be quick and painless.
MOrning (or Good night)
Thanks metallicus for your answer.
As for the proposal of sq225917, I would be interested if you can find again the 7X2 groups of transistors matched all in no more than 5%.
Emil
Thanks metallicus for your answer.
As for the proposal of sq225917, I would be interested if you can find again the 7X2 groups of transistors matched all in no more than 5%.
Emil
Strange that you get so poor match which brand did you buy?
Of the 3000 I have the Diotech show tight Hfe groupings
I still have problem matching PNP to NPN after 4 different orders to get different batches
but I have loads of matches whitinn the Type.
How did you mesure them and expecialy how constant is the ambient temperature
I am prety sure that like Simon is just question of luck to get the right batch
My lot is in storagge at present as I am moving house but as soon as she will alowd I will be posting again in the BCXXX tread.
Do have a look at it post what you have and with a bit of luck some one got your matching batch.
Of the 3000 I have the Diotec show tight Hfe groups
Hi Emil, sorry to hear that. There were quite some people offering swapping transistors and also matched sets here on this thread, so with a little help we might have resolved this..... Anyway, imagine 670 PCBs with 36 transistors each , that gives 24.000 transistors, and with everybody buying 200pcs or more per stereo set, we are talking close to 100.000 transistors.... This is a good statistical sample, and it should definitely be possible to find matches there.
The fundamental problem is that we are trying to have identical specs between NPNs and PNPs, that come from different technologies, so that in itself presents a problem.
Blowing over the transistors is not a reproducible thing. Did you use the little circuit from the assembly guide? Its probably quite time-consuming, but the best would be to leave the transistor in until the reading stabilizes, without moving air around (maybe by putting a plastic cup over the transistor). This gives you the most exact measurement conditions from transistor to transistor.
Anyway, as a minimum the 16 input stage transistors (both stereo input stages) should be matched as good as possible, for the remaining stages its OK if they are matched within themselves. If you need more transistors for achieving that, please PM me.
These voltage ratings are ballpark, sounds like you are in good shape. Did you check for oscillations? Does it amplify?
Also, it would be good to let the phono stage powered for a couple weeks and then re-check the voltage ratings, especially B (the opamp output voltage). This reflects a certain mismatch in the input stage, maybe not so much current gain but leakage current. I never tried, but if you want you could try and adjust the input offset a little bit, to see if the opamp voltage changes, in order to get it closer to zero without causing too much input voltage. But the best would probably be to replace the input stage transistors with better matches.